Developing A Strategy For Red Cell Antigen Typing AND Matching of Blood For Chronic Transfusion

DEVELOPING A STRATEGY
FOR RED CELL ANTIGEN TYPING
AND
MATCHING OF BLOOD
FOR CHRONIC TRANSFUSION
Jennifer Mary Davies
A thesis submitted in partial fulfilment of the requirements of

the Faculty of Health and Life Sciences
University of the West of England, Bristol for the degree of
Professional Doctorate in Biomedical Science
This research program was carried out in collaboration with the

School of Biomedical and Healthcare Sciences
Plymouth University Peninsula Schools of Medicine and
Dentistry, Plymouth
and
The Royal Devon and Exeter NHS Foundation Trust, Exeter
Submission date: July 2018
i
ABSTRACT
Red cell allo and autoantibodies have the potential to cause haemolytic
transfusion reactions and can also create challenges with subsequent
compatibility testing, sourcing compatible blood, delays in the provision of
blood and cost implications. This is particularly relevant in patients who
are dependent on chronic transfusion support; such as those with sickle
cell disease, thalassaemia, haematological malignancies and renal
disease. A retrospective review of chronically transfused patients with
haematological malignancies and renal disease, treated at the Royal
Devon and Exeter NHS Foundation Trust (RD&E), revealed considerably
higher costs associated with provision of blood for patients with red cell
allo/autoantibodies than for non-immunised patients, providing new
evidence to support the cost effectiveness of extended antigen matching
for these groups of patients.
The risk of development of red cell allo/autoantibodies could be reduced
by a type and match strategy for, at least, Rh (CcEe) and K antigens as
well as the standard ABO and RhD match. To support this strategy a
novel technique was validated for high throughput extended automated
red cell antigen serological phenotyping suitable for use in the hospital
transfusion setting.
Red cell genotyping assays can also support a type and match strategy;
however, these assays have not been evaluated for use in the hospital
transfusion service setting in the United Kingdom (UK). A platform for this
ii
technique was evaluated; the results compared well to the serological
assay, and the assay showed some potential for routine use.
A large scale randomised controlled trial to investigate the benefit and
cost effectiveness of a type and match strategy has not been attempted
in the UK, therefore a pilot study was performed in which patients were
randomly assigned to a standard care or intervention (type and match)
group. The pilot study demonstrated the feasibility of a larger scale trial
and provided new evidence supporting the potential to implement a type
and match strategy within the hospital transfusion service, using routine
blood stocks.
Implementation of an extended type and match strategy, utilising the new
assay for extended serological phenotyping and the lessons learned from
the pilot study, has been implemented for patients with haematological
malignancies and renal disease at the RD&E. Patients identified as
potential recipients of chronic transfusion support are serologically
phenotyped for Rh (CcEe) and K antigens prior to transfusion. Changes
have been made to the red cell stock storage area and the laboratory
information management system to support provision of matched blood
for transfusion. The program is being monitored for its effectiveness in
reducing both allo/autoimmunisation rate and the cost of provision of
blood.
iii
Contents
1 INTRODUCTION ............................................................................... 1
1.1 Blood and blood components for transfusion ............................... 1
1.2 Blood component support in chronic transfusion ......................... 4
1.3 Blood group systems and red cell antigens ................................. 9
1.3.1 Blood group systems ............................................................. 9
1.3.2 Biochemistry, genetic basis and function of blood groups
systems............................................................................................ 15
1.3.2.1 ABO system .................................................................. 15
1.3.2.2 Rh system ..................................................................... 18
1.3.2.3 Kell system ................................................................... 22
1.3.2.4 MNS system ................................................................. 25
1.3.2.5 Duffy system ................................................................. 26
1.3.2.6 Kidd system .................................................................. 29
1.3.2.7 Diego system ................................................................ 30
1.3.2.8 Dombrock system ......................................................... 33
1.3.2.9 Colton system ............................................................... 35
1.3.2.10 Yt system ...................................................................... 37
1.3.2.11 Lutheran system ........................................................... 37
1.4 Antibodies to red cell antigens and their clinical relevance ........ 40
1.4.1 Clinical relevance of red cell antibodies .............................. 40
1.4.2 Laboratory relevance of red cell antibodies ......................... 42
1.4.2.1 Group and antibody screen........................................... 44
1.4.2.2 Antibody Identification ................................................... 45
1.4.2.3 Serological red cell phenotyping ................................... 46
1.4.2.4 Direct Antiglobulin Test ................................................. 47
1.4.2.5 Serological crossmatch ................................................. 47
1.5 Serological red cell phenotyping ................................................ 49
1.5.1 Principles of direct agglutination in tube technology ............ 50
1.5.2 Principles of the Column Agglutination Test (CAT) ............. 52
iv
1.5.3 Principles of automated phenotyping methodology on the
Immucor NEO® analyser ................................................................. 56
1.5.4 Grading reactions in phenotyping techniques ..................... 61
1.6 Red cell antigen genotyping ....................................................... 63
1.6.1 Principles of the BLOODChip blood group genotyping assay .
............................................................................................ 66
1.6.2 Principles of the Luminex xMAP analyser ........................... 71
1.7 Risk of alloimmunisation in chronic transfusion ......................... 73
1.8 Red cell autoantibodies.............................................................. 75
1.9 Extended red cell antigen matching of blood in chronic
transfusion ........................................................................................... 76
1.10 Aims and Hypothesis ............................................................. 80
2 MATERIALS AND METHODS ......................................................... 82
2.1 Overview of the project .............................................................. 82
2.2 Retrospective review of chronically transfused patients ............ 83
2.2.1 Sample size for retrospective review ................................... 83
2.2.2 Patients with haematological malignancies ......................... 83
2.2.3 Patients with renal insufficiency .......................................... 85
2.3 Comparison of manual and automated red cell phenotyping ..... 86
2.3.1 Manual red cell phenotyping ............................................... 87
2.3.2 Automated red cell phenotyping using the NEO analyser ... 93
2.3.3 Extended phenotyping profile result interpretation on the
NEO® analyser .............................................................................. 101
2.3.4 Quality control testing ........................................................ 102
2.4 Blood Group genotyping .......................................................... 107
2.4.1 DNA Extraction .................................................................. 107
2.4.2 IDCOREXT test ................................................................. 108
2.4.2.1 IDCOREXT Workflow ................................................. 108
2.4.2.2 DNA Amplification ....................................................... 109
2.4.2.3 Hybridisation of amplified DNA to allele specific probes
111
2.4.2.4 Fluorescent labelling of hybridised DNA ..................... 113
2.4.3 Luminex xMAP Analyser ................................................... 115
2.4.3.1 Luminex® X-Map system analysis .............................. 115
2.5 Randomised Controlled Trial ................................................... 118
2.6 STATISTICAL ANALYSIS........................................................ 121
2.6.1 Retrospective review of haematology and renal patients .. 121
v
2.6.2 Comparison of the manual and automated red cell
phenotyping results ........................................................................ 122
2.6.3 Comparison of the serological phenotyping and genotyping
results .......................................................................................... 122
2.6.4 Review of the pilot study – randomised controlled trial...... 123
3 RETROSPECTIVE REVIEW RESULTS ........................................ 124
3.1 Introduction .............................................................................. 124
3.2 Patients with haematological malignancies.............................. 126
3.2.1 Antibodies and disease conditions .................................... 127
3.2.2 Antibody Specificities ........................................................ 132
3.2.3 Antibodies and patient factors ........................................... 135
3.2.3.1 Gender and immunisation rates .................................. 135
3.2.3.2 Age and immunisation rates ....................................... 137
3.2.3.3 Immune suppression and incidence of immunisation . 139
3.2.3.4 Number of transfusions and incidence of immunisation ....
.................................................................................... 143
3.2.3.5 Time between transfusions and the incidence of
immunisation .....................................................................................
.................................................................................... 147
3.2.4 Transfusion Reactions....................................................... 149
3.2.5 Cost of additional testing ................................................... 151
3.3 Patients with renal insufficiency ............................................... 153
3.3.1 Antibody Specificities ........................................................ 154
3.3.2 Antibodies and patient factors ........................................... 157
3.3.2.1 Gender and immunisation rate.................................... 157
3.3.2.2 Age and immunisation rate ......................................... 159
3.3.2.3 Immune suppression and immunisation rate .............. 161
3.3.2.4 Number of transfusion episodes and incidence of
immunisation ............................................................................... 163
3.3.2.5 Time between transfusion and the incidence of
immunisation .....................................................................................
.................................................................................... 165
3.3.3 Transfusion Reactions....................................................... 167
3.3.4 Cost of additional testing ................................................... 169
3.4 Comparison of renal patients and haematology patients ......... 171
3.4.1 Gender and the incidence of immunisation ....................... 171
3.4.2 Number of units transfused and immunisation rate ........... 173
vi
........................................................................................ 174
3.4.3 Age at first transfusion and incidence of immunisation...... 175
3.5 Discussion ................................................................................... 177
4 AUTOMATED RED CELL PHENOTYPING ................................... 194
4.1 Introduction .............................................................................. 194
4.2 Results of the automated phenotyping validation .................... 196
4.3 Comparison of test costs for manual and automated serological
phenotyping ....................................................................................... 209
4.4 Discussion ............................................................................... 212
5 BLOOD GROUP GENOTYPING ................................................... 216
5.1 Introduction .............................................................................. 216
5.1.1 Rh system CcEe................................................................ 224
5.1.2 Kell system ........................................................................ 225
5.1.3 Kidd system ....................................................................... 226
5.1.4 Duffy system ..................................................................... 226
5.1.5 MNS system ...................................................................... 227
5.1.6 Diego system .................................................................... 228
5.1.7 Dombrock system.............................................................. 228
5.1.8 Colton system ................................................................... 229
5.1.9 Yt system .......................................................................... 229
5.1.10 Lutheran system ............................................................ 230
5.2 Study design ............................................................................ 233
5.3 Results ..................................................................................... 234
5.3.1 DNA extraction .................................................................. 234
5.3.2 Cost of testing ................................................................... 236
5.3.3 Comparison of phenotyping and genotyping results.......... 238
5.4 Discussion ............................................................................... 246
6 PILOT STUDY-RANDOMISED CONTROLLED TRIAL ................. 252
6.1 Introduction .............................................................................. 252
6.2 Results of the randomised controlled pilot study ...................... 258
6.2.1 Number of patients transfused in each study group .......... 259
6.2.2 Time spent in trial for patients in each study group ........... 260
6.2.3 Comparison of the number of transfusion episodes and units
given for patients in each study group ........................................... 263
6.2.4 Disease conditions in the pilot study patients .................... 266
6.2.5 Matching red cell antigens for transfused patients in the
intervention group .......................................................................... 268
6.3 Discussion ............................................................................... 277
vii
7 COST ANALYSIS .......................................................................... 284
7.1 Introduction .............................................................................. 284
7.2 Automated extended serological phenotyping ......................... 284
7.1 Genotyping Assay .................................................................... 288
7.2 Cost comparison of serological phenotyping and genotyping
techniques ......................................................................................... 288
7.3 Discussion ............................................................................... 291
8 FINAL DISCUSSION AND CONCLUSIONS.................................. 294
9 REFERENCES .............................................................................. 311
10 APPENDICES ................................................................................. 1
10.1 Appendix 1: Raw data for retrospective review of haematology
cohort ................................................................................................. 1
10.2 Appendix 2: Raw data for retrospective review of renal cohort ...
............................................................................................... 47
10.3 Appendix 3: Programmed cut off values for the serological
phenotyping profiles on the NEO analyser .......................................... 88
10.4 Appendix 4: Raw data for genotyping results ......................... 90
10.5 Appendix 5: Patient information leaflet given to the participants
in the pilot study ................................................................................ 290
10.2 Appendix 6: Consent form signed by participants in the pilot
study ............................................................................................. 295
viii
List of figures
Figure 1.1 Structure of the ABO antigens ............................................... 17
Figure 1.2 Diagrammatic representations of the RhD and RhCE antigen
structures ................................................................................................ 21
Figure 1.3 The Kell and Xk heterodimer ................................................. 24
Figure 1.4 The Duffy glycoprotein ........................................................... 28
Figure 1.5 The Kidd glycoprotein ............................................................ 30
Figure 1.6 Diego antigens located on band 3 glycoprotein ..................... 32
Figure 1.7 Dombrock antigens on the ART4 glycoprotein ....................... 34
Figure 1.8 The Colton glycoprotein structure .......................................... 36
Figure 1.9 Structure of the Lutheran glycoprotein ................................... 39
Figure 1.10: Principles and reactions of the direct agglutination assay
tube technique for the determination of red cell antigen status ............... 51
Figure 1.11: Principles of the column agglutination assay ...................... 53
Figure 1.12: Reaction patterns of the column agglutination assay .......... 55
Figure 1.13: Principles of the solid phase phenotyping assay ................ 60
Figure 1.14: Comparison of reaction patterns in phenotyping assays ..... 62
Figure 1.15: The principles of the IDCOREXT Assay .............................. 67
Figure 1.16 Principles of the Luminex xMAP analyser ............................ 72
Figure 2.1: Rh and K antigen phenotyping by column agglutination ....... 89
Figure 2.2: Manual Serological phenotyping for Fy (a and b), Jk (a and b)
and s antigens by column agglutination .................................................. 91
Figure 2.3: Reaction patterns generated by automated red cell antigen
phenotyping for Rh(CCwcEe), K, M and N antigens ................................ 97
ix
Figure 2.4: Reactions patterns generated by automated red cell antigen
typing for Fya, Fyb, Jka, Jkb, S, s and k antigens ................................... 100
Figure 2.5: NHSBT ten cell ID panel profile .......................................... 104
Figure 2.6: NHSBT three cell screen profile product ............................. 105
Figure 2.7: NHSBT titration cell set profile ........................................... 106
Figure 3.1: Correspondence analysis of the association between disease
types and antibody types in the haematology cohort ............................ 131
Figure 3.2: Incidence of antibody specificities within the haematology
cohort .................................................................................................... 134
Figure 3.3: Immunisation rates in males and females within the
haematology cohort .............................................................................. 136
Figure 3.4: Influence of age at first transfusion and allo/autoantibody
development within the haematology cohort ......................................... 138
Figure 3.5: Influence of immune suppression (platelet requirement) on
allo/autoantibody development in the haematology cohort ................... 140
Figure 3.6: Influence of immunosuppression (irradiated requirement) on
allo/autoantibody development in the haematology cohort ................... 142
Figure 3.7: Influence of the number of transfusion episodes on the
development of all/autoantibodies in the haematology cohort .............. 144
Figure 3.8: Influence of the number of red cell units transfused on the
development of allo/autoantibodies in the haematology cohort ............ 146
Figure 3.9: Influence of the frequency of transfusion episodes on the
development of allo/autoantibodies in the haematology cohort ............ 148
Figure 3.10: Incidence and specificity of allo/autoantibodies in the renal
cohort .................................................................................................... 156
x
Figure 3.11: Influence of gender on the development of
allo/autoantibodies in the renal cohort .................................................. 158
Figure 3.12: Influence of age at first recorded transfusion on the
development of allo/autoantibodies in the renal cohort ......................... 160
Figure 3.13: Influence of immunosuppression on the development of
allo/autoantibodies in the renal cohort .................................................. 162
Figure 3.14: Influence of the number of transfusion episodes and number
of units transfused on the development of allo/autoantibodies in the renal
cohort .................................................................................................... 164
Figure 3.15: Influence of the frequency of transfusion episodes on the
development of allo/autoantibodies in the renal cohort ......................... 166
Figure 3.16: Comparison of the influence of gender on the development
of allo/autoantibodies in the haematology cohort versus the renal cohort
.............................................................................................................. 172
Figure 3.17: Comparison of the number of red cell transfusions given in
the haematology and renal cohorts ....................................................... 174
Figure 3.18: Comparison of age at first transfusion and incidence of
immunisation in the haematology and renal cohorts ............................. 176
Figure 4.1: The influence of reagent temperature on the reaction strength
of automated serological phenotyping testing ....................................... 204
Figure 4.2: Influence of age of sample and temperature of reagent on the
reactions strength of automated serological phenotyping ..................... 205
Figure 5.1: DNA analysis report generated by the Luminex XMAP
analyser ................................................................................................ 239
xi
Figure 6.1: Comparison of number of patients transfused in the pilot study
groups ................................................................................................... 259
Figure 6.2: Recruitment rates for pilot study patients ............................ 261
Figure 6.3: Number of red cell units transfused to patients in pilot study
.............................................................................................................. 264
Figure 6.4: Number of transfusion episodes for patients in the pilot study
.............................................................................................................. 265
Figure 6.5: Disease conditions in the pilot study patients ..................... 267
Figure 6.6: Frequencies of Rh phenotypes in patients recruited to the pilot
study ..................................................................................................... 276
xii
List of tables
Table 1.1: World Health Organisation (WHO) classification of lymphoid
neoplasms, myeloid neoplasms and acute leukaemia .............................. 6
Table 1.2: The blood group systems with the gene symbols as recognised
by the HUGO (Human Genome Organisation) Gene Nomenclature
Committee and the ISBT (International Society of Blood Transfusion) ... 11
Table 1.3: Likely clinical significance of red cell alloantibodies ............... 48
Table 2.1: Reagent volumes required to produce the PCR mix using the
IDCOREXT ........................................................................................... 109
Table 2.2 Amplification program details for IDCOREXT ....................... 111
Table 2.3 Thermal cycler program for denaturation and hybridisation .. 113
Table 2.4 Fluorescent labelling mix....................................................... 114
Table 3.1: Allo/autoantibodies and disease conditions ......................... 128
Table 3.2: Red cell antibody type in relation to disease ........................ 130
Table 3.3: Transfusion reactions in the haematology cohort ................. 150
Table 3.4: Cost of additional testing in the haematology cohort ............ 152
Table 3.5: Transfusion reactions in the renal cohort ............................. 168
Table 3.6: Cost of additional testing in the renal cohort ........................ 170
Table 4.1: Consistency of automated screening reactions with
homozygous and heterozygous expression of red cell antigens ........... 198
Table 4.2: Phenotyping anomalies in automated and manual assays .. 201
Table 4.3: The frequency of antigens found in samples selected for
validation of the automated serological phenotyping assays ................ 208
Table 4.4: Breakdown of test costs for the manual serological
phenotyping assay. ............................................................................... 210
xiii
Table 4.5: Breakdown of the test costs for the automated serological
phenotyping assay ................................................................................ 211
Table 5.1: Blood group system alleles detected by the BLOODChip ID
COREXT assay ...................................................................................... 219
Table 5.2: The red cell antigen types determined by the IDCOREXT
assay. ................................................................................................... 231
Table 5.3: Concentration and quantity of DNA extracted from whole blood
samples ................................................................................................ 235
Table 5.4: Cost of Blood Group genotyping .......................................... 237
Table 5.5: Antigen phenotypes by serological analysis and genotyping241
Table 6.1 Optimum stock of red cell units held in RD&E stock ............. 257
Table 6.2: Provision of Rh (CcEe) and K matched blood for transfused
patients in the intervention group .......................................................... 269
Table 6.3: Rh and K types of patients recruited to the pilot study ......... 273
Table 7.1: Cost effectiveness of serological phenotyping and genotyping
techniques ............................................................................................ 290
xiv
PUBLICATIONS
Typing and matching blood for chronically transfused patients – a pilot
study. Davies, J. Transfusion Medicine (2016), 26 (Suppl.2): 35.
Blood Group Genotyping Using the BloodChip ID Core XT: Potential for
Use in the Hospital Transfusion Setting. Davies, J. Transfusion Medicine
(2014), 24 (Suppl. 2), 39.
Developing a strategy for red cell antigen typing and matching of blood
for chronic transfusion. Davies, J. Transfusion Medicine (2013), 23,
(Suppl. 2): 43.
xv
ABBREVIATIONS
AHG Anti-human globulin
ALL Acute lymphocytic leukaemia
AML Acute myeloid leukaemia
ANCOVA Analysis of covariance
APML Acute pro-myelocytic leukaemia
BCSH British Committee for Standards in Haematology
BMT Bone marrow transplant
C3d Complement component 3
CAT Column agglutination test
CD Cluster of differentiation
CI Confidence interval
CLL Chronic lymphocytic leukaemia
CMML Chronic myeloid/monocytic leukaemia
DAT Direct antiglobulin test
DARC Duffy antigen chemokine receptor
DNA Deoxyribonucleic acid
DSTR Delayed serological transfusion reaction
EDTA Ethylenediamine tetraacetic acid
EPO Erythropoietin
ESRD End stage renal disease
xvi
FAB French-American-British
HEA Human erythrocyte antigens
HLA Human leukocyte antigens
hrEPO Human recombinant Erythropoietin
HUGO Human Genome Organisation
IAT Indirect antiglobulin test
IgG Immunoglobulin G
IgM Immunoglobulin M
IPS Integrated Pathology System
ISBT International Society of Blood Transfusion
ITP Idiopathic thrombocytopaenic purpura
JPAC Joint United Kingdom Blood Transfusion and Tissue
Transplantation Services Professional Advisory
Committee
LISS Low ionic strength saline
MALDI-TOF MS Matrix Assisted Laser Desorption/ionisation
Time of Flight Mass Spectrometry
MDS Myelodysplastic syndrome
MPD Myeloproliferative disorder
MPN Myeloproliferative neoplasm
MPS Massively parallel sequencing
xvii
NCABT National Comparative Audit of Blood Transfusion
NEP Nonexofacial polymorphism
NGS Next Generation Sequencing
NHL Non-Hodgkin’s lymphoma
NHS National Health Service
NHSBT National Health Service Blood and Transplant
NICE National Institute for Health and Care Excellence
NIHR National Institute for Health Research
NRES National Research Ethics Service
NTD Not determined
PBS Phosphate buffered saline
PCR Polymerase chain reaction
PRV Polycythaemia rubra vera
QC Quality control
RCT Randomised controlled trial
SAPE Streptavidin R-phycoerythrin
SCD Sickle cell disease
SD Standard deviation
SNP Single nucleotide polymorphism
SSP-PCR Sequence specific primer – polymerase chain
reaction
xviii
UK United Kingdom
VAT Value Added Tax
VWD Von Willebrand’s disease
WW1 World War 1
xix
ACKNOWLEDGEMENTS
I would like to express my gratitude to the following people who
have supported me in the completion of this project:
Dr Ruth Morse and Dr Paul White at the University of the West
of England. Dr Morse for her help and support as my supervisor
and Dr White for his assistance with the statistical elements of
the work.
Dr Neil Avent and Dr Tracey Madgett at the University of
Plymouth for their support with the blood group genotyping
testing.
Dr Paul Kerr at the Royal Devon and Exeter NHS Foundation
Trust for his support and assistance with the pilot study.
Stacey Flay at the Royal Devon and Exeter NHS Foundation Trust
for recruiting the patients to the pilot study.
xx
The Exeter Leukaemia Fund at the Royal Devon and Exeter
NHS Foundation Trust for donating funding to support the
purchase of the blood group genotyping assay kit.
The Molecular Genetics Department at the Royal Devon and
Exeter NHS Foundation Trust for DNA extraction work.
John Herbert for creating the diagrams of the red cell antigen
structures in section 1.3.2.
xxi
1 INTRODUCTION
1.1 Blood and blood components for transfusion
The collection and storage of blood for transfusion was introduced during
the First World War (WW1) in 1917 (Robertson, 1918; Makins, 1922).
Blood was collected from “lightly wounded” soldiers and mixed with a
preservative/anticoagulant solution of sodium citrate with glucose,
allowing storage for up to 14 days. This process became known as blood
banking and it gave the army the ability to transfuse close to the front line
with lifesaving effects (Makins, 1922). The birth of the modern blood
bank, with blood collected from volunteer donors from the general
population is attributed to Percy Lane Oliver, who, in 1921, received a
request for an emergency donor and within five years had recruited 400
volunteers to a donor panel (Starr, 1999).
In the interval between the First and Second World Wars, methods for the
separation of whole blood into its component parts (red cells and
plasma), led to the clinical use of plasma for treating severe infections by
Strumia and his group in 1927 (Strumia and McGraw, 1941). The
introduction of plastic bags for the storage of blood in 1950 (Walter and
Murphy, 1952), rather than the glass bottles as originally used, facilitated
further separation of the donated blood. This enabled the separation of
whole blood into the blood components (including fresh frozen plasma,
platelet concentrates, and cryoprecipitate) available today. The
availability of these blood components enables targeted therapy for
1
specific deficiencies of blood constituents in patients, caused by inherited
or acquired conditions.
Blood and blood components are used in a wide variety of medical and
surgical conditions and their life saving qualities are without question. In
the United Kingdom (UK) more than 50% of the red cell components
collected, processed and distributed by the National Health Service Blood
and Transplant (NHSBT) are transfused for non-surgical conditions
(JPAC, 2014). A national audit of blood use in medical patients conducted
in 2011 in the UK (NCABT, 2011) revealed that the majority of red cells
were transfused to patients with haematological diagnoses (28%), despite
the audit restricting the number of patients with those conditions that were
included within the review. Other recipients of red cells were patients with
acute or chronic gastrointestinal haemorrhage (17%), non-haematological
cancer (15%), patients in intensive care units (7%) and patients with renal
failure (4%). A later audit in 2014 (NCABT, 2014) revealed that medical
usage accounted for 67.4% of all red cell units used in the UK, with
haematological and non-haematological (including renal failure) anaemia
being the largest user (54.5%) Other blood usage was seen in
gastrointestinal bleeding (12%), obstetrics and gynaecology (6%),
cardiothoracic surgery (6%), gastrointestinal surgery (4%), trauma (4%)
and orthopaedics (4%). Thus, a large proportion of the approximately two
million red cell components collected annually in the UK are given to
patients requiring chronic transfusion support, namely those with
2
haematological malignancies, non-haematological cancers and renal
failure.
At the Royal Devon and Exeter NHS Foundation Trust (RD&E) the red
cell usage is broken down as follows; medical conditions (78%), surgery
(21%) and obstetrics/gynaecology (1%). The majority of red cell usage is
for patients with haematological (41%) and non-haematological anaemia
(27%), followed by gastrointestinal bleed (8%) and trauma (8%). The
RD&E clinical haematology department has an inpatient and outpatient
service providing care for patients with haematological conditions, and
also performs autologous peripheral blood stem cell transplants for
patients with conditions such as myeloma and lymphoma. A
comprehensive day case, inpatient and outpatient renal service is
provided at the main hospital site and in five satellite hospital units. The
service provides treatment for patients with acute (usually reversible) and
chronic (usually irreversible) kidney failure, and those with kidney
transplant. The fact that the majority of red cells are transfused to patients
with haematological or renal anaemia provided the rationale for the focus
on these cohorts within the research work. The non-emergency nature of
blood transfusion in chronic anaemia, compared to that in gastrointestinal
haemorrhage and trauma, would, in theory, allow for antigen typing and
extended matching of blood for these patients.
3
1.2 Blood component support in chronic transfusion
Cancer is the one of the most common causes of mortality in the
developed world (Cartwright et al., 1999; Ferlay et al., 2007; Office for
National Statistics, 2012). In 2012 lymphoid cancer, a haematological
malignancy, accounted for 2.6% of the death rate in males (Office for
National Statistics, 2012). In addition to the substantial mortality rate for
this type of cancer, and other haematological malignancies, treatment
costs can also be high, including prolonged hospitalisation,
chemotherapy, drugs and other specialised support such as blood
component therapy (Yu et al., 2007).
The haematological malignancies are a diverse group which were
originally divided into the leukaemias/myelodysplasias, coded by the
French-American-British (FAB) classification (Bennet et al., 1976) and the
Morphology Immunology Cytogenetic classification (Van den Berghe,
1988), and the lymphomas, which historically were subjected to a
multitude of classifications (Gerrard-Marchant et al., 1974; Lukes and
Butler, 1986; Harris et al., 1994; Chan et al., 1994; Chan et al., 1995).
They were classified according to lineage by the World Health
Organization Clinical Advisory Committee (Swerdlow et al., 2008) in an
attempt to present a classification system that is recognised worldwide.
The classifications have been more recently revised in 2016 for lymphoid
neoplasms (Swerdlow et al., 2016) and myeloid neoplasms and acute
leukaemia (Arber et al., 2016). A simplified version of the World Health
4
Organisation (WHO) classification of lymphoid neoplasms, myeloid
neoplasms and acute leukaemias is shown in table 1.1.
5
Table 1.1: World Health Organisation (WHO) classification of lymphoid neoplasms, myeloid neoplasms and acute
leukaemia
The table shows a simplified version of the WHO classification of lymphoid neoplasms (Swerdlow et al., 2016) myeloid neoplasms and
acute leukaemia (Arber et al., 2016). For the purposes of this study the table only includes the major subtypes (in bold type) and
categories seen in the patients included within the retrospective review (chapter 3) and the pilot study (chapter 6). The full
classifications can be found using the hyperlinks for lymphoid neoplasms and myeloid neoplasms and acute leukaemia.
Lymphoid neoplasms Myeloid neoplasms Acute leukaemia
Chronic lymphocytic leukaemia (CLL) Myelodysplastic/myeloproliferative Acute myeloid leukaemia (AML)
neoplasms (MDS/MPS)
Hairy cell leukaemia (HCL) Chronic myelomonocytic leukaemia Acute lymphoblastic leukaemia (ALL) (B-
(CMML) cell or T-cell)
Waldenström macroglobulinemia Chronic myeloid leukaemia (CML)
Mantle cell lymphoma (MCL)
Diffuse large B-cell lymphoma (DLBCL)
Hodgkin lymphoma
Plasma cell myeloma
6
Morbidity and mortality rates have historically been high in patients with
haematological malignancies, with mortality rates reaching 75.6% in
certain countries (Grais, 1962). In England and Wales during the years
1945 to 1957, mortality rates were found to range between 19.8% for
chronic myeloid leukaemia to 59% for acute leukaemia (Court-Brown and
Doll, 1959). Recent advances in the treatment of patients with
haematological malignancies, using chemotherapy or high dose
chemotherapy combined with bone marrow or stem cell transplantation,
have increased the survival rate in this group of patients (Champlin and
Gale, 1987), with complete remission rates of more than 80% seen in
young adult patients with acute lymphocytic leukaemia (Dombret and
Gardin, 2016). However, the intense treatment regimens directed at the
progenitor cells within the bone marrow result in anaemia and
thrombocytopenia, which, if untreated, are also life threatening conditions,
consequently the patient’s survival becomes dependent on blood
component support (NICE 2003).
Patients with acute or chronic kidney disease may also suffer from
anaemia caused by the disease (Brugnara and Eckardt, 2011) and be
dependent on blood transfusion for periods of time. Anaemia is common
in patients with kidney disease, which is caused by decreased synthesis
of erythropoietin, a hormone produced by the kidneys that stimulates the
bone marrow to make red blood cells (Erslev, 1953; Jacobson et al.,
1957). The identification and treatment of anaemia in this group of
patients is important because anaemia has the potential to cause adverse
7
effects including reduced oxygen utilisation, increased cardiac output and
left ventricular hypertrophy, increase in the progression of the disease
and reduction in immune responsiveness (NICE 2011). A study in
diabetic nephropathy, the leading cause of end stage renal disease
(ESRD), found that even mild anaemia increased the risk of progression
to ESRD and that anaemia was an independent risk factor for disease
progression (Mohanram et al., 2004). For patients with ESRD a common
treatment option is haemodialysis, however, this process can cause blood
loss during the processing stage within the haemodialysis machine,
exacerbating the extent of anaemia (NIH, 2016). The prevalence of
anaemia in chronic kidney disease in the UK is approximately 12% (NICE
2011). A review in the United States estimated that anaemia in patients
with kidney disease ranged from 2% in the early stages to over 70% in
ESRD (Astor et al., 2002).
The Canadian Hemodialysis Morbidity Study found that for patients with
ESRD on haemodialysis the probability of surviving 12 months without
receiving a blood transfusion was 47.2% for males and 27.5% for females
(Churchill et al., 1992). Although blood transfusion therapy in patients
with renal disease has dramatically reduced over the last 20 years since
the introduction of human recombinant erythropoietin (hrEPO) therapy,
blood transfusion is still a treatment option if clinically indicated (NICE
2011; Mikhail et al., 2010) especially during haemodialysis. In addition
approximately 10% of patients become resistant to the effects of hrEPO
(Van der Putten et al., 2008) for various reasons including iron deficiency,
8
infection/inflammation, severe hyperparathyroidism and inadequate
dialysis (Drüeke, 2001; Priyadarshi and Shapiro, 2006; Ribeiro et al.,
2013).
1.3 Blood group systems and red cell antigens
1.3.1 Blood group systems
The term blood group refers to variations and polymorphisms in the
blood, related to red cell surface antigens (Daniels, 2013). The most well-
known system is the ABO system and the most well-known antigen is the
RhD, due to their use in the blood donor setting. Variations in the
antigens on the red cell surface can only be called a blood group if they
are defined by an antibody.
The ABO blood group system and corresponding antibodies were first
discovered by Landsteiner in 1900 using direct agglutination techniques
(Landsteiner, 1900 cited in Daniels 2013). However, not all blood group
antibodies are able to directly agglutinate red cells, and it was not until
the development of the indirect antiglobulin test (IAT) in 1945 (Coombs et
al., 1945a; Coombs et al., 1945b) that other, so-called “incomplete” red
cell antibodies could be identified, and their corresponding blood group
systems and antigens named. New DNA based technologies have
facilitated the identification of new blood group systems and allowed
mapping of the genes controlling the expression of the antigens within the
system to the chromosomal location. As such, the number of recognised
9
blood group systems is ever increasing. A standard nomenclature for
these blood groups and their antigens was first developed by the
International Society of Blood Transfusion (ISBT) Working Party on Red
Cell Immunogenetics and Blood Group Terminology in 1980. Each blood
group system is assigned a number, the system name which, for the well-
known systems may be the historically assigned name, the system
symbol and the symbol for each gene within the system (table 1-2). To
facilitate international standardisation of gene symbols, each gene
symbol for the blood group families are recognised and approved by the
Human Genome Organisation (HUGO). Application of this nomenclature
on an international scale ensures that publications including information
on blood group systems have standardised terminology, facilitating
literature searches for researchers. In order for a red cell antigen to be
named as a blood group system it is required to fulfil the criteria set out
by the ISBT; it must be defined by a corresponding human alloantibody, it
must be an inherited character, the gene encoding the antigen must have
been identified and sequenced and its chromosomal location must be
known (Daniels et al., 2004).
An explanation of the structure, molecular basis and function of blood
group systems relevant to this study are given in section 1.3.2. This
includes blood group antigens detected using the serological phenotyping
assay (chapter 4) and/or the genotyping assay (chapter 5) described in
this study. As such, the section will focus on the following blood group
systems; ABO, Rh, Kell, Kidd, Duffy, MNS, Diego, Dombrock, Colton, Yt
and Lutheran.
10
Table 1.2: The blood group systems with the gene symbols as recognised by the HUGO (Human Genome
Organisation) Gene Nomenclature Committee and the ISBT (International Society of Blood Transfusion)
Table of the blood group systems with the gene symbols as recognised by the HUGO Gene Nomenclature Committee (www.genenames.org
updated October 2012) and the ISBT. Approval by the HUGO Gene Nomenclature Committee ensures that each gene symbol is unique thus
facilitating electronic data retrieval from publications and databases. The nature of the encoded protein is included; if unknown this is denoted in the
table. Key:N/A = not applicable
No. System name System No. of Gene symbol(s) Chromosomal CD number Nature of the encoded protein
symbol antigens location
001 ABO ABO 4 ABO 9q34.2 N/A Galactosyltransferase and N-
acetylgalactoseaminyltransferase
002 MNS MNS 48 GYPA, GYPB, 4q31.21 CD235 Unknown

GYPE
003 P1PK P1PK 3 A4GALT 22q13.2 N/A Galactosyltransferase
004 Rh RH 55 RHD, RHCE 1p36.11 CD240 Unknown
005 Lutheran LU 22 LU 19q13.32 CD239 Adhesion molecule
006 Kell KEL 36 KEL 7q34 CD238 Metalloendopeptidase?
11
(continued)
(continued) (continued) symbol antigens (continued) location (continued) (continued)
(continued) (continued)
007 Lewis LE 6 GUT3 19q13.3 N/A Fucosyltransferase
008 Duffy FY 5 DARC 1q23.2 CD234 Chemokine receptor
009 Kidd JK 3 SLC14A1 18q12.3 N/A Urea transporter
010 Diego DI 22 SLC4A1 17q21.31 CD233 Anion exchanger
011 Yt YT 3 ACHE 7q22.1 N/A Acetylcholinesterase
012 Xg XG 2 XG, MIC2 Xp22.33 CD99 (MIC2 Adhesion molecule
product)
013 Scianna SC 7 ERMAP 1p34.2 N/A Adhesion molecule?
014 Dombrock DO 7 ART4 12p12.3 CD297 ADP ribosyltransferase
015 Colton CO 4 AQP1 7p14.3 N/A Water channel
016 Landsteiner- LW 3 ICAM4 19p13.2 CD242 Adhesion molecule
Wiener
017 Chido/Rodgers CH/RG 9 C4A, C4B 6p21.3 N/A Complement factor
018 H H 1 FUT1 19q13.33 CD173 Fucosyltransferase
019 Kx XK 1 XK Xp21.1 N/A Membrane transport protein?
12
(continued) (continued) symbol antigens (continued) location (continued) (continued) (continued)
020 Gerbich GE 11 GYPC 2q14.3 CD236 Cytoskeletal element
021 Cromer CROM 19 CD55 1q32.2 CD55 Complement cascade regulator
022 Knops KN 9 CR1 1q32.2 CD35 Cell surface receptor
023 Indian IN 5 CD44 11p13 CD44 Adhesion molecule
024 Ok OK 3 BSG 19p13.3 CD147 Plasma membrane protein
025 Raph RAPH 1 CD151 11p15.5 CD151 Adhesion molecule
026 John Milton JMH 6 SEMA7A 15q24.1 CD108 Adhesion and signalling molecule?
Hagen
027 I I 1 GCNT2 6p24.2 N/A N-acetylglucosaminyltransferase
028 Globoside GLOB 1 B3GALT3 3q26.1 N/A N-acetyl galactosaminyltransferase
029 Gill GIL 1 AQP3 9p13.3 N/A Water channel
030 Rh-associated RHAG 4 RHAG 6p21-qter CD241 Ammonium and carbon dioxide
glycoprotein transporter
031 FORS FORS 1 GBGT1 9q34.13 N/A Unknown
13
(continued) (continued) symbol antigens (continued) location (continued) (continued) (continued)
032 JR JR 1 ABCG2 4q22 N/A Unknown
033 LAN LAN 1 ABCB6 2q36 N/A Unknown
034 Vel VEL 1 SMIM1 1p36.32 N/A Unknown
035 CD59 CD59 2 CD59 11p14-p13 CD59 Unknown
036 Augustine AUG 2 SCL29A1 6p21.1 N/A Unknown
14
1.3.2 Biochemistry, genetic basis and function of blood groups
systems
Blood group antigens are either protein determinants, which represent the
primary products of blood group system genes, or carbohydrate
determinants on glycoproteins or glycolipids, in which the products of the
genes controlling antigen expression are glycosyltransferase enzymes
(Daniels, 2013). The following subsections explain the biochemistry,
molecular basis and function of the blood group systems and antigens
important to this study. Unless otherwise referenced, the information in
sections 1.3.2.1 to 1.3.2.11 is taken from the Human Blood Groups
reference book by Daniels (2013).
1.3.2.1 ABO system
The ABO blood group system comprises of four antigens, the H antigen
encoded by the FUT1 locus on chromosome 19 and the ABO antigens
encoded by the ABO gene on chromosome 9. The H allele produces
α1,2-L fucosyltransferase at the FUT1 locus, this enzyme catalyses the
transfer of L-fucose to its acceptor substrate, D-Galactose, resulting in
the H determinant. The ABO blood group type is determined by alleles at
the ABO locus. The product of the A allele is α1,3-N-acetyl-D-
galactosaminyltransferase, which attaches N-acetylgalactosamine to the
H determinant and the B allele produces α1,5-D-galactosyltransferase,
which attaches D-Galactose to the H determinant. The third allele at the
locus, O, does not produce an active enzyme and so the H determinant
15
remains unchanged. ABO antigens are expressed on carbohydrate
chains on glycoproteins, mainly the poly-N-acetyllactosaminyl N-glycans
of the band 3 or band 4.5 proteins. Expression of the antigens is
dependent on the presence of specific monosaccharides attached to a
precursor disaccharide at the non-reducing end of a carbohydrate chain,
shown in simple diagrammatic form in figure 1.1. The carbohydrate
chains are built up by the sequential addition of monosaccharides
catalysed by a specific glycoslytransferase. It is these
glycoslytransferases that represent the primary products of the ABO
genes.
The function of the ABO antigens is unknown; however, the carbohydrate
structures which carry the ABO red cell antigens form part of the
glycocalyx, an extra-cellular matrix acting to protect the cell from damage
or invasion by pathogenic micro-organisms (Viitala and Järnefelt, 1985;
Kościelak, 2012). The ABH determinants on the red cells are carried,
mainly, on the anion exchange protein, band 3, and band 4.5 which is a
glucose transporter.
16
Figure 1.1 Structure of the ABO antigens
Red cell membrane
The diagram shows the immunodominant sugars in the ABO blood group
system. The L-Fucose sugar denoting the blood group O is, more
correctly, known as the H determinant, if this sugar is not present then A
or B antigens cannot be expressed, even if the genes are present. This
results in the rare H-deficient (Bombay) phenotype. Expression of A and
B antigens is determined by enzymes produced by alleles at the ABO
locus which catalyse the attachment of the relevant sugar determinants to
the H structure. (Adapted from Diamed, cited in Immunobase 1,
http://www.bio-rad.com/en-uk/product/immunobase?ID=LO52IK5Y7).
17
1.3.2.2 Rh system
The Rh blood group system is a highly complex system, with 55 antigens
encoded by two homologous, closely linked genes on chromosome 1
(ISBT 004 RHCE alleles v4.0_20180208). The RHD gene produces the
RhD antigen, and the RHCE gene produces the Cc and Ee antigens. C
and c, E and e represent two pairs of antithetical antigens, with
polymorphisms controlled by RHCE. The C/c and E/e phenotypes arise
as a result of a single amino acid substitution; serine to proline at position
103 for C/c and proline to alanine at position 226 for E/e.
Each antigen has been allocated a number, for example, RH1 (D), RH2
(C), RH3 (E), RH4 (c), RH5 (e) and RH8 (Cw). The alleles are inherited as
haplotypes, often referred to using the Fisher-Race or Wiener
terminology. The Fisher-Race terminology was based on the premise that
the Rh phenotype was controlled by three closely linked loci on each
chromosome of a homologous pair, each of which has its own set of
alleles; Dd, Cc and Ee (the d antigen does not exist but is used in this
terminology to denote the absence of the D antigen). They postulated that
the loci were so closely linked that the three genes on chromosome were
always inherited together (Fisher, cited in Race, 1944). Wiener,
conversely, theorised that there was only one gene complex with eight
alleles, Ro, R1, R2, Rz, r, r', r", ry, controlling the Rh phenotype (Wiener,
1943). Both theories are too simplistic to explain the Rh system antigens
and resultant phenotypes but they are commonly used for the
interpretation of serological reactions within hospital transfusion services.
18
The Rh system antigens (RhD, C, c, E and e) are carried on hydrophobic,
non-glycosylated proteins which cross the red cell membrane 12 times
(Daniels, 2013)(figure 1.2). An associated protein, the RhAG glycoprotein
forms complexes with the Rh protein within the cell membrane. The exact
functions of the Rh and RhAG proteins are not yet known, but there is
some evidence to suggest that the RhAG is involved in the transport of
ammonium ions and carbon dioxide (Marini et al., 1997; Endeward et al.,
2008). The RhD negative phenotype is a result of the total absence of the
RhD polypeptide from the red cell membrane, in Caucasian populations
this is usually as a result of homozygosity of a deletion of RHD. In black
Africans, however, the D negative phenotype is often caused by the
presence of an inactive RHD called the RHD pseudogene, or a hybrid
gene RHD-CE-Ds. The hybrid genes do not produce a D antigen, but
probably produce an abnormal C antigen, and are associated with the
VS+V- phenotype (Daniels, 2013). There are many variant antigens
within the Rh system, as described by Daniels (2013), which create
challenges for serological phenotyping and genotyping assays. The most
common variants in the UK involve the D antigen, recommended practice
in hospital laboratories is to test the patient cells against an IgM
monoclonal anti-D reagent that does not detect the DVI variant (BCSH
2012). This practice is recommended as individuals with the DVI
phenotype are likely to produce anti-D if transfused with D positive red
cells, whereas other D variants are at less risk of developing anti-D.
Variants to other Rh antigens are rare, particularly in Caucasian
19
populations, and so testing is not included in serological phenotyping
assays. However, assays for some of these variants are included in
genotyping assays (see chapter 5). The Rhnull phenotype, in which neither
RhD nor CcEe antigens are expressed on the red cell surface, is
extremely rare. There are two types of Rhnull; the amorph type, which is a
result of inactivating mutations in the RHCE and deletion of the RHD, or
the regulator type, in which the Rh genes are normal but there are
inactivating mutations in the RHAG gene.
20
Figure 1.2 Diagrammatic representations of the RhD and RhCE
antigen structures
C/c E/e
Ser103Pro Pro226Ala
Red cell membrane
The RhD and RhCE proteins span the red cell membrane and are similar,
differing by only 32 to 35 amino acids, the sections of the protein above
and below the cell membrane are denoted by dotted lines on the diagram.
RhD is composed of many different epitopes (highlighted in the oval
above the cell membrane). The RhD variant phenotype occurs when one
or more of these epitopes are missing. The RhCE protein carries the Cc
and Ee antigens, two pairs of antithetical antigens that differ by a single
amino acid change. At position 103 C antigen expression is result of
serine and c antigen expression by proline, and at position 226 E results
from proline and e from alanine. (Diagram adapted from Westhoff, 2007)
21
1.3.2.3 Kell system
The Kell system comprises of 36 antigens encoded by a single gene
(KEL) on chromosome 7 ((ISBT 006) blood group alleles v4.0 160701).
The system includes seven sets of antigens with allelic relationships: K
and k; Kpa, Kpb and Kpc; Jsa and Jsb; K11 and K17; KEL14 and KEL24;
KEL25 and KEL28; KEL31 and KEL38. The antithetical antigens arise as
a result of single amino acid changes at different positions on the Kell
glycoprotein. The Kell antigens are located on CD238, a red cell
transmembrane glycoprotein, which is closely associated with another
protein, Xk (see figure 1.3). The Kell-Xk heterodimer is part of the red cell
membrane complex that contains band 3, the Rh complex and
glycophorin C.
The prevalence of the commonly encountered Kell antigens differs by
ethnic group: for example, K positive is more common in Caucasian
samples and less often seen in African-Americans; Kp(a+) phenotype is
almost always found in whites; and Js(a+) is almost exclusively found in
individuals of African ethnicity (Westhoff and Shaz, 2013). Similar to the
RhD antigen, the K antigen is strongly immunogenic and the antibody is
clinically significant, hence it is recommended in the UK to select K
negative red cells for transfusion in females of child bearing potential
(BCSH 2012). The Kell-null phenotype (K0), in which none of the Kell
antigens are expressed on the red cell, is a result of homozygosity for
KEL inactivating mutations. In the Kmod phenotype, which also results
from KEL mutations, the antigens are present but expressed weakly.
22
The function of the Kell glycoprotein is unclear, although suggestions
have been made regarding possible physiological functions in heart, red
cell ion transport, neovascularisation in tumours and motor function (Zhu
et al., 2009) and a role in erythropoiesis (Belhacene et al., 1998).
23
Figure 1.3 The Kell and Xk heterodimer
a b
Js /Js
K/k
Xk
Cell membrane
The Xk protein and the Kell glycoprotein are linked by a disulphide bond.
The diagram shows the approximate locations of the antithetical antigens,
K/k, Jsa/Jsb and Kpa/Kpb on the Kell glycoprotein. The antithetical
antigens arise as a result of single amino acid changes in the
glycoprotein; K/k change at position 578 (threonine to methionine),
Jsa/Jsb at position 597 (leucine to proline) and Kpb/Kpa at position 281
(arginine to tryptophan). The Kx system consists of one antigen which is
encoded by the X-linked gene, XK, and located on the Xk protein. The
glycoprotein structure also carries a pentameric zinc-binding motif,
HELLH. Picture adapted from Slideshare
(https://image.slidesharecdn.com/kellbloodgroupsystem-150908110628-
lva1-app6892/95/kell-blood-group-system-10-638.jpg?cb=1441710657).
24
1.3.2.4 MNS system
The MNS system is another highly complex blood group system, with 48
antigens included in the system, encoded by three genes (GYPA, GYPB
and GYPE) located on chromosome 4 (MNS (ISBT 002) blood group
alleles v4.1 170119). M and N determinants are carried on glycophorin A
(GPA), the major red cell sialic acid-rich glycoprotein. The M and N
antigen status of red cells is defined by an amino acid sequence
polymorphism at positions 1 and 5 on Glycophorin A. The M antigen has
serine at position 1 and glycine at position 5, whereas, the N antigen has
leucine at position 1 and glycine at position 5. In addition, the first 26
amino acids of the Glycophorin B protein on the red cell surface are
identical to those of the N antigen on Glycophorin A (Daniels, 2013). This
gives rise to an antigen denoted “N” to distinguish it from the N antigen on
Glycophorin A, this antigen is present on almost all human red cells,
irrespective of their MN status.. Rarely, individuals may lack the GPA,
and cannot express M and N antigens, this is known as the En(a-)
phenotype.
The S and s antigens are located on Glycophorin B (GPB); the S antigen
has methionine at position 29, whereas the s antigen has threonine.
Individuals deficient in GPB do not express the S, s or the high frequency
U antigens. The S-s- phenotype is very rare in Europeans but is noted
with a higher frequency in African populations.
25
Mur is a low incidence antigen, part of the, now obsolete, Miltenberger
series, which arises as a result of a replacement of a small section of
GYPB with a 55 base pair section of GYPB. This gives rise to a hybrid
glycophorin GP (B-A-B) expressing the Mur antigen. Although the Mur
phenotype is rare in white populations, it is relatively common in Asia and
is important in terms of clinical significance (see section 1.4.1), hence
testing for the antigen is a useful addition to genotyping assays.
GPA and GPB cross the cell membrane once, and probably exist in their
monomeric and dimeric forms. GPA is closely associated with band 3 and
is part of the band 3/Rh macrocomplex. Glycophorins have a long,
glycosylated extracellular domain, which carries a negative charge. It is
thought that the function of the glycophorins may involve prevention of
aggregation of red cells, protection against mechanical damage and
microbial attack. It is known that individuals with the En(a-) phenotype
and those with S-s- have red cells that are less susceptible to malarial
invasion.
1.3.2.5 Duffy system
The Duffy blood group system consists of five antigens encoded by a
single gene (DARC) located on chromosome 1 ((ISBT 008) blood group
alleles v4.1 160816). The system includes the antithetical antigens Fya
and Fyb, which are polymorphic in both black and white populations, and
the high frequency antigens Fy3, Fy5 and Fy6. The Fy(a-b-) phenotype is
common in the African population as it confers resistance to the malarial
26
parasite, Plasmodium vivax. This phenotype is caused by homozygosity
for a mutation within an erythroid-specific, GATA-1 transcription-factor
binding site upstream of the coding region of the Duffy gene. The Duffy
antigens are located on a glycoprotein which acts as a chemokine
receptor (Pruenster and Rot, 2006) (see figure 1.4), hence it is often
referred to as the Duffy antigen receptor for chemokines (DARC). DARC
is N-glycoslyated and appears to be a member of the band 3, Rh, Kell/Xk
macrocomplex. The GATA mutation prevents the expression of the Duffy
glycoprotein on the red cells. The Fya/Fyb polymorphism is a result of an
amino acid change from glycine to aspartic acid at position 42. Fy3 and
Fy6 are high frequency antigens expressed on red cells of all Duffy
phenotypes and Fy5 is similar to Fy3, with the exception that it is not
expressed on Fy(a-b-) and Rhnull cells. The Fyx phenotype, which is
encoded by the FY*X allele results in an amino acid substitution in the
intracellular domain of the glycoprotein. This phenotype, which is
relatively common in white populations, behaves as a weak expression of
the Fyb antigen. It is also associated with reduced expression of Fy3, Fy5
and Fy6, suggesting that it is a product of reduced levels of DARC in the
membrane.
27
Figure 1.4 The Duffy glycoprotein
N-terminal domain
C-terminal domain
A diagrammatic illustration of the Duffy glycoprotein showing how the
structure spans the red cell membrane seven times, the cylindrical
shapes represent the area of the glycoprotein that passes through the cell
membrane. The Fya/Fyb polymorphism is located at position 42 near the
N terminal. Fyx is expressed in the intracellular component at position 89
and is thought to be associated with reduced levels of DARC. The
approximate positions of the Fy3 and Fy6 antigens are shown on the
diagram.
Adapted from Genome Research
(http://genome.cshlp.org/content/17/5/577/F3.expansion).
28
1.3.2.6 Kidd system
The Kidd blood group system comprises of three antigens encoded by
the SLC14A1 gene on chromosome 18 ((ISBT 009) blood group alleles
v4.0 160705). The antigens include the antithetical pair, Jka and Jkb and
the high incidence antigen Jk3. The Kidd antigens are carried on a
glycoprotein that spans the red cell membrane 10 times (figure 1.5). The
antithetical antigens, Jka and Jkb, are differentiated by a single amino acid
change on the fourth extracellular loop of the glycoprotein (aspartic acid
to asparagine at position 280).
The null phenotype, Jk(a-b-), is rare in most populations, with the
exception of the Polynesians. It is caused by homozygosity, or compound
heterozygosity for a variety of inactivating mutations in the SLC14A1
gene. Another rare cause of the Jk null phenotype, seen in a Japanese
family was considered to be the result of a dominant inhibitor gene that is
not inherited at the JK locus.
The Kidd glycoprotein is known to be a urea transporter (Sidoux-Walter et
al., 1999). It is very similar to another urea transporter, UT-A (Olives et
al., 1996), and it is thought that compensating activity by this urea
transporter is one of the reasons why the Kidd null phenotype is not
associated with any clinical problems.
29
Figure 1.5 The Kidd glycoprotein
a b
Jk /Jk
Cell membrane
Diagrammatic representation of the Kidd glycoprotein which spans the
red cell membrane 10 times. The Jka and Jkb antigen site is located at
position 280 (demonstrated by the filled circle on the diagram); the
difference between the two antigens is the result of a single amino acid
change (aspartic acid and asparagine). The Jk3 phenotype is expressed
when either Jka and/or Jkb are present.
(Adapted from: https://ars.els-cdn.com/content/image/1-s2.0-
S0065242316300518-f06-04-9780128046869.jpg)
1.3.2.7 Diego system
The Diego system comprises of 22 antigens encoded by a single gene
(SLC4A1) located on chromosome 17 ((ISBT 010) blood group alleles
v3.0 160630). The system includes three pairs of antithetical antigens
(Dia and Dib, Wra and Wrb, Wu and DISK) and 16 low frequency antigens.
30
The Diego antigens are expressed on the red cell membrane glycoprotein
band 3, Dia and Dib, and Wra and Wrb are the result of single amino acid
substitutions, leucine/proline and glutamic acid/lysine, at their respective
positions 854 and 658. The Wu/DISK polymorphism is a result of a single
amino acid substitution, alanine/glycine, at position 565 (see figure 1.6).
The remaining low frequency antigens in the Diego system are also the
result of a single amino acid substitutions.
Band 3 is an intrinsic membrane glycoprotein acting as a structural
protein and also an anion exchanger (Tanner, 1993). Band 3 plays a vital
role in the red cell membrane which is why the Diego null phenotype
(band 3 deficiency) has not been reported.
31
Figure 1.6 Diego antigens located on band 3 glycoprotein
Wu/DISK
Position 565
Wra/Wrb Dia/Dib
Diagrammatic representation of the band 3 glycoprotein with associated
Diego antigens, including the antithetical antigens Wra/Wrb, Dia/Dib and
Wu/DISK antigens at positions 658, 854 and 565 respectively, expressed
as a result of single amino acid substitutions. The approximate positions
of some of the other low frequency antigens in the system are also shown
on the diagram.
(Adapted from:
http://www.bloodjournal.org/content/bloodjournal/92/12/4836/F2.large.
jpg?sso-checked=true )
32
1.3.2.8 Dombrock system
The Dombrock system includes 7 antigens ((ISBT 014) blood group
alleles v4.0 160623) encoded by a single gene (ART4) located on
chromosome 12. The system includes the antithetical antigens Doa and
Dob, and the high frequency antigens Gya, Hy, Joa, DOYA, DOMR and
DOLG. Doa and Dob are expressed as a result of a single amino acid
substitution (asparagine to aspartic acid) at position 265. Expression of
the remaining high frequency antigens are also the result of single amino
acid substitutions, with the exception of Gya. The Gya negative phenotype
can result from a variety of different mutations in the gene, and includes
the Donull phenotype in which the red cell lacks all of the Dombrock
antigens.
The Dombrock antigens are located on the ART4 glycoprotein, which is
attached to the red cell membrane through glycosylphosphatidylinositol
(see figure 1.7). The exact function of the ART4 glycoprotein is not
known, although it does belong to a family of adenosine diphosphate
(ADP)-ribosyltransferases responsible for catalysing the transfer of ADP-
ribose from nicotinamide adenine to a protein substrate.
33
Figure 1.7 Dombrock antigens on the ART4 glycoprotein
Cell membrane
Diagrammatic representation of the structure (shown here as a line) that
expresses the antigens of the Dombrock system. The Doa/Dob
polymorphism arises as a result of a single amino acid substitution at
position 265 (asparagine to aspartic acid). The ART4 glycoprotein is
attached to the red cell membrane via the GPI anchor which is inserted
into the red cell membrane.
(Adapted from: https://clinicalgate.com/human-blood-group-antigens-and-
antibodies/)
34
1.3.2.9 Colton system
The Colton blood group system includes four antigens encoded by the
AQP1 gene located on chromosome 7 ((ISBT 015) blood group alleles
v3.0 160623). The system includes the antithetical antigens Coa and Cob,
and two high frequency antigens Co3 and Co4. The Colton polymorphism
is represented by a single amino acid change at position 45; alanine
gives rise to Coa and valine to Cob. The Co3 antigen is present on all cells
with the exception of the Conull phenotype Co(a-b-). The Co4 antigen is
represented by glutamine at position 47.
The glycoprotein structure that carries the Colton antigens acts as a
water channel (King et al., 2004 (a)). The structure spans the red cell
membrane six times (see figure 1.8), and includes three extracellular
loops and two cytoplasmic loops. One extracellular loop and one
cytoplasmic loop pass into the membrane to create the channel which
allows water molecules to pass.
35
Figure 1.8 The Colton glycoprotein structure
Model of the glycoprotein structure that carries with Colton antigens,
which spans the red cell membrane six times, as represented by the
cylindrical structures in the diagram. The extracellular and cytoplasmic
loops that create the water channel are represented by B and E. The two
antithetical antigens Coa and Cob, which result from a single amino acid
change, are expressed at position 45, the approximate position of the
antigens on the glycoprotein are demonstrated in the diagram.
(Adapted from:
https://www.sciencedirect.com/science/article/pii/S1246782001001422)
36
1.3.2.10 Yt system
The Yt system consists of two antigens, Yta and Ytb, encoded by the
ACHE gene located on chromosome 7 ((ISBT 011) blood group alleles
v5.0 180207). The antigens Yta and Ytb are antithetical, and are
represented by a single amino acid change at position 353 (histidine to
asparagine). A third antigen, YTEG, a high frequency antigen, was
reported in 2017 (Laundy et al., 2017). An inherited Ytnull phenotype, Yt(a-
b-) has not been reported.
The Yt antigens are located on acetylcholinesterase (AChE), a GPI linked
glycoprotein, similar to the Dombrock antigens (see figure 1.7). The exact
nature of the red cell AChE is not known, but AChE is known to play an
essential role in neurotransmission. This essential role may explain why
the null phenotype is not seen.
1.3.2.11 Lutheran system
The Lutheran system has 22 antigens ((ISBT 005) blood group alleles
v4.1 170106) and is encoded by the LU gene on chromosome 19. The
system includes four pairs of antithetical antigens: Lua and Lub, Lu6 and
Lu9, Lu8 and Lu14 and Aua and Aub. These antithetical antigens are all
the products of single amino acid changes; Lua/Lub (histidine to arginine
at position 77), Lu6/Lu9 (phenylalanine to serine at position 275),
Lu8/Lu14 (lysine to methionine at position 204) and Aua/Aub (alanine to
threonine at position 539). The null phenotype, Lu(a-b-) exists as a result
37
of homozygosity for an inactivated Lutheran gene. Red cells of individuals
with the null phenotype lack all Lutheran antigens.
Lutheran antigens are located on glycoprotein structures which are
members of the immunoglobulin superfamily, a large family of receptors
and adhesion molecules (Isacke and Horton, 2000). The Lutheran
antigens are associated with two glycoproteins, which act as ligands for
laminin, an extracellular matrix glycoprotein (see figure 1.9).
38
Figure 1.9 Structure of the Lutheran glycoprotein
Red cell membrane
Diagrammatic representation of the Lutheran glycoproteins, which consist
of five extracellular domains, with a flexible linker and laminin binding site
between domains 2 and 3. Each domain is composed of two β-sheets
stabilised by a disulphide bond and consists of approximately 100 amino
acids. The position of the laminin binding site and the approximate
locations of the four antithetical pairs of antigens (which result from single
amino acid substitutions) on the immunoglobulin superfamily domains are
shown on the diagram.
(Adapted from: http://www.bloodjournal.org/content/89/11/4219)
39
1.4 Antibodies to red cell antigens and their clinical relevance
1.4.1 Clinical relevance of red cell antibodies
One of the effects of the transfusion of allogeneic blood in any group of
patients and in particular with chronic blood transfusion, is the
development of red cell alloantibodies, produced in response to antigenic
differences between the recipient and the donor red cells. As the blood
group antigens are extremely polymorphic it is almost inevitable that there
will be antigenic differences between the donor (transfused) red cells and
the recipient red cells. This can lead to the recognition of these foreign
antigens by the recipient immune system and the production of
alloantibodies.
Alloimmunisation has the potential to cause transfusion reactions in
individuals if they are subsequently transfused with allogeneic red cells
bearing the corresponding antigen. Immune mediated transfusion
reactions can be acute or delayed. Acute reactions are defined by the
Serious Hazards of Transfusion (SHOT) group as fever and other
symptoms/signs of haemolysis within 24 hours of transfusion, whereas
delayed reactions may be seen more than 24 hours after transfusion
(SHOT 2016). Transfusion reactions can result in deleterious health
effects including haemolysis, jaundice, fever, renal failure and shock, and
may be fatal in some cases (Beauregard and Blajchman, 1994; SHOT
2016). Red cell antibodies most commonly implicated in transfusion
reactions include; Anti-D, -C, -c, -E, -e, (Rh system) -K, -k, (Kell system) -
Jka, -Jkb, (Kidd system) -M, -S, -s, -U, (MNS system) -Fya, -Fyb (Duffy
40
system) and Wra (Diego system) (BCSH 2013). Antibodies to antigens
within the Dombrock, Colton, and Yt systems have been implicated in
haemolytic transfusion reactions, and antibodies to Lutheran antigens
may cause mild or delayed transfusion reaction (Daniels, 2013).
Alloimmunisation also has the potential to cause haemolytic disease of
the fetus and newborn (HDFN), if there is an antigenic difference between
the maternal red cells and the fetal red cells. Maternal red cell antibodies,
which may have developed as a result of transfusion or via transfer of
fetal cells to the maternal circulation, may cross the placenta and cause
destruction of the fetal red cells. Certain antibodies, notably anti-D, anti-c
and anti-K can cause severe fetal symptoms, including anaemia,
jaundice, hydrops and stillbirth (RCOG, 2014).
Red cell alloantibodies are typically IgM or IgG in nature and are usually
produced in response to a foreign red cell antigen stimulus. Others, such
as the IgM antibodies directed against the well-known A and B antigens,
are believed to be naturally occurring as there is no requirement for red
cell antigen exposure. Instead they are probably made in response to
bacterial antigens that are similar in structure to red cell antigens
(Springer et al., 1959a; Springer et al., 1959b; Springer and Horton,
1969). Some red cell antibodies, such as the IgM type anti-A and anti-B,
are capable of causing intravascular lysis of the red cell by their ability to
activate the complement cascade through the formation of the membrane
attack complex, and the subsequent release of anaphylatoxins leading to
41
acute systemic effects (Murphy et al., 2013). Other red cell alloantibodies,
normally IgG type, can cause extravascular destruction of foreign red
cells, in the spleen, by interacting with the Fcγ receptors on the cells of
the mononuclear phagocytic system (Murphy et al., 2013).
Red cell alloantibodies may be transient in nature, this is particularly well
known for alloantibodies such as anti-Jka and anti-Jkb. Non-persistent
antibody detection has been reported in several studies (Ramsey and
Larson, 1988; Rosse et al., 1990; Reverberi, 2008). A similar
phenomenon has also been reported for autoantibodies (Young et al.,
2004). The reasons given for this phenomenon may include; lack of
antigenic stimulus resulting in decrease of antibody production over time,
low levels of antibody production that are at the limit of the antibody
detection assays and may only be detectable on some occasions,
improvements in the sensitivity of antibody detection assays over time
and inter-batch variation within the same assay technique.
1.4.2 Laboratory relevance of red cell antibodies
In addition to having the potential to cause haemolytic transfusion
reactions, the presence of red cell alloantibodies in patient blood samples
also brings laboratory related issues. They increase the workload and
cost pressures on the hospital transfusion laboratory services by
introducing requirements for specialised testing techniques and sourcing
antigen matched blood for transfusion, which may cause delays in blood
provision. Standard laboratory testing, within the hospital transfusion
service, prior to the transfusion of red cells, includes ABO and RhD typing
42
of the patient red cells and a screening technique for the detection of any
atypical red cell antibodies in the patient’s plasma, commonly referred to
as a group and antibody screen. Additional testing, if atypical red cells
antibodies are detected, includes identification of the antibody specificity,
serological red cell phenotyping, direct agglutination test (DAT) and
serological crossmatch of donor red cells to determine compatibility. At
the RD&E all of these assays are performed in the hospital transfusion
laboratory. If antibody specificity or compatibility cannot be determined
using the techniques available in the hospital transfusion laboratory,
referral to a reference laboratory at the NHSBT may be required. A
simplified flow chart for pre-transfusion testing in the hospital transfusion
laboratory is shown below:
Group and
antibody
screen
No atypical Atypical antibodies

antibodies detected
detected
Red cells issued Antibody Serological

DAT
with no further identification phenotype
testing
Red cells issued following

compatibility testing by
serological crossmatch
43
1.4.2.1 Group and antibody screen
Prior to any blood transfusion episode, a blood sample must be taken
from the patient for blood grouping (ABO and RhD) and antibody
screening. Blood grouping and antibody screening may be performed
using manual or automated techniques. ABO and RhD typing utilises
direct agglutination techniques with patient red cells using commercial
anti-sera. Antibody screening is performed using indirect antiglobulin
techniques using patient plasma against commercial antibody screening
cells with a known antigen profile. For the purposes of detection of
clinically significant red cell antibodies (that are known to be capable of
causing haemolysis of transfused red cells) in patient samples presented
for testing within the hospital transfusion setting the BCSH (2013)
recommend that the screening cell set employed contains the following
common antigens:
C, c, Cw, D, E, e, K, k, Fya, Fyb, Jka, Jkb, S, s, M, N, P1, Lea and Leb.
If no red cell alloantibodies are detected in the patient’s plasma there is
no requirement for further laboratory testing. Blood for transfusion is
selected that is ABO and RhD compatible for the patient and may be
issued electronically, with no serological compatibility testing between the
donor cells and recipient plasma, in accordance with BCSH guidelines
(BCSH 2013).
At the RD&E red cells are selected for transfusion dependent patients
(excluding SCD) in accordance with the BCSH (2013) recommendations;
44
ABO/D matched. The BCSH guidance states that additional typing and
matching for Rh and K should be a local decision. At the RD&E it has
been recommended that red cells for these patients should be K
negative, but this was not policy and there were no processes in place to
ensure that this occured.
To reduce the risk of the development of red cell antibodies known to
cause Haemolytic Disease of the Fetus and Newborn, in some countries,
including France, red cell transfusions for females with child-bearing
potential may be matched for Rh (CcEe) and K antigens (French
Transfusion guidelines cited in Moncharmont et al., 2015). In the UK it is
recommended that K negative red cell units are selected for this group of
patients, but without the requirement for K typing the patient (BCSH
2013). At the RD&E it is recommended that red cell units are selected
that are also negative for Rh c and E antigens wherever possible.
However, phenotyping has not been advocated in our institution as
transfusion is uncommon in this group of patients and suitable red cells
can easily be sourced from routine stocks.
1.4.2.2 Antibody Identification
If red cell alloantibodies are detected in the antibody screening technique
further laboratory assays are required in order to ascertain the specificity
of the alloantibody. Antibody identification techniques employ the same
principles as the antibody screening techniques but using commercial
cells with an extended range of known antigen profiles. Whereas a
45
screening cell set is commonly composed of two or three red cell
reagents, an antibody identification set may contain more than 10, the
extended range enabling determination of the antibody specificity, rather
than just the presence of a red cell alloantibody. As with antibody
screening methodology, antibody identification may be performed
manually or on an automated system, column agglutination and solid
phase technology being commonly employed.
To avoid potential haemolytic transfusion reactions the BCSH (2013)
recommend that, for patients whose plasma has been shown to contain
clinically significant antibodies as identified in antibody screening and
identification techniques, either antigen negative blood, or unselected but
crossmatch compatible (by IAT) blood should be selected for transfusion.
As a general rule if the patient has an antibody that has been reported to
cause haemolytic transfusion reactions then antigen negative blood must
be selected, whereas the presence of an antibody with no reported cases
of transfusion reactions needs only to be crossmatch compatible (table 1-
3) (BCSH, 2013).
1.4.2.3 Serological red cell phenotyping
Serological phenotyping is the identification of red cell antigens present
on the patient red cells, other than ABO/D, using commercially available
antisera. Currently available methods for serological phenotyping include
direct and indirect agglutination either by tube techniques (see section
1.6.1), or column agglutination techniques based on the principles of
46
Lapierre (1990) (section 1.6.2). Additionally, solid phase tests, based on
the principles of Plapp (1984) and Juji (1972) (section 1.6.3), in which red
blood cells are bound to a solid support such as a microtitre plate well,
can be used for the identification of red cell antigens. Serological red cell
phenotyping is explained in more detail in section 1.5.
1.4.2.4 Direct Antiglobulin Test
Direct Antiglobulin Testing (DAT) may also be required; this test detects
the presence of antibodies or complement fractions on the surface of the
patient’s red cells. Methodology commonly employed in hospital
transfusion laboratories includes column agglutination tests or solid
phase technology, which may be performed manually or on automated
systems.
1.4.2.5 Serological crossmatch
If atypical red cell antibodies are present, compatibility testing, known as
crossmatching, of each unit of blood must be performed. Testing must be
performed by IAT, either column agglutination or solid phase may be
used to ensure compatibility of the selected red cell units for transfusion
(BCSH, 2013). The technique may be performed manually or by
automated technology.
47
Table 1.3: Likely clinical significance of red cell alloantibodies
Likely clinical significance of red cell antibodies reported in this study and
recommendations for the selection of blood for patients with their
presence. Taken from the BCSH guidelines for pre-transfusion
compatibility procedures in blood transfusion laboratories (BCSH 2013)
and Daniels (2013).
System Antibody Likely clinical Recommendation
significance in
transfusion
ABO Anti-A1 No IAT crossmatch
compatible
H Anti-H (in A1 and A1B No IAT crossmatch
patients) compatible
Rh Anti-D,-C,-c,-E,-e,-f Yes Antigen negative
Rh Cw No IAT crossmatch
compatible
Kell Anti-K,-k, Yes Antigen negative
Kell Anti-Kpa No IAT crossmatch

compatible
Kidd Anti-Jka,-Jkb Yes Antigen negative
MNS Anti-M (active at 37oC) Yes Antigen negative
MNS Anti-M (not active at No IAT crossmatch

37oC) compatible
MNS Anti-N, -Mur, -Hut No IAT crossmatch
compatible
MNS Anti-S,-s,-U Yes Antigen negative
Duffy Anti-Fya,-Fyb Yes Antigen negative
P Anti-P1 No IAT crossmatch

compatible
Lewis Anti-Lea,-Leb,-Le-a+b No IAT crossmatch
compatible
Lutheran Anti-Lua No IAT crossmatch
compatible
Diego Anti-Wra Yes IAT Crossmatch
compatible
Yt Anti- Yta, -Ytb Possible IAT crossmatch, least
incompatible for anti-
Yta
Chido/Rodge Anti-Chido-Rodgers No IAT Crossmatch
rs compatible
Knops Anti-Knops McCoy No IAT Crossmatch
compatible
Others Autoantibodies No ABO/D Rh (CcEe)
and K matched
48
1.5 Serological red cell phenotyping
Positive identification of a red cell alloantibody in a patient’s plasma also
includes testing the patient’s red cell antigen type. This is performed by
serological phenotyping using a combination of monoclonal and
polyclonal reagent antisera, utilising direct or indirect agglutination
techniques. Identification of alloantibody specificity can be supported if
the patient’s red cells are negative for the corresponding antigen (BCSH,
2013). Methods for serological phenotyping within the hospital transfusion
setting include direct agglutination in tube technology, column
agglutination and microtitre plates and indirect agglutination using column
agglutination and solid phase methods. However, unlike blood grouping,
antibody screening and antibody identification, serological phenotyping
methods are currently, generally performed manually, and may employ a
combination of the available techniques. Automation of these techniques
has been limited by the availability of appropriate reagents and testing
profiles that could be used on the high throughput blood grouping
analysers commonly used by hospital transfusion services in the UK.
However, manual techniques are labour intensive and are prone to errors
during sample preparation, testing, result interpretation and result
reporting stages. It is for this reason that extended phenotyping, typing
for the presence of more than the antigen implicated by the antibody
identification, has not been common practice, even though the
information gained from the extended phenotype could be used to select
blood that might protect the patient from further alloimmunisation. In
addition, extended serological phenotyping could be used for prophylactic
49
antigen matching of blood to prevent alloimmunisation for patients with
conditions that might require chronic transfusion in the future.
1.5.1 Principles of direct agglutination in tube technology
Direct agglutination is a well-recognised technique (as described by Issitt
and Anstee, 1998) for detecting the presence of antibodies or antigens
and is used across many applications in clinical medicine. Some
commercially available monoclonal antisera are capable of agglutinating
red cells in test tubes to determine the presence of specific blood group
antigens on red cells. In these techniques, an aliquot of a suspension of
patient red cells, with unknown antigen status, is placed in a test tube
along with the antisera. A short incubation period allows the antisera to
combine with the red cell antigens, if present, on the patient red cells.
Centrifugation of the test tube then forces the red cells together, allowing
the antisera to crosslink between red cells, causing agglutinates to form.
In the absence of the corresponding antigens on the patient red cells,
there will be no antigen-antibody reaction and no agglutinates will form
following centrifugation. The presence of agglutinates is determined
macroscopically by gently shaking the test tube after the centrifugation
stage, and the agglutination strength may be graded as shown in figure
1.10. This technique is performed manually and the macroscopic reading
of the results is subjective and dependent on the experience of the
operator. As a result of these factors the technique is labour intensive and
prone to the known errors of manual testing (Stainsby et al., 2006) in both
preparation and the result determination.
50
Figure 1.10: Principles and reactions of the direct agglutination
assay tube technique for the determination of red cell antigen status
Reactions are
graded from
strongly positive
(4+) through to
negative (0)
4+ 3+ 2+ 1+ +/- 0
In direct agglutination assays patient red cell suspension is mixed with the
relevant anti-serum and incubated within the test tube. Centrifugation forces
the red cells together allowing antigen-antibody complexes to form if antigen
is present on the patient red cells; reaction patterns are graded from strongly
positive (4+) to weakly positive (+/)-, as indicated in the figure. A negative
reaction is seen if the patient red cells do not carry the antigen and therefore
do not form complexes.
(Picture courtesy of Bio-Rad, cited in Immunobase 1 – Diamed,
51
1.5.2 Principles of the Column Agglutination Test (CAT)
The column agglutination test is performed in specially designed cards
which employ the use of either antisera within the column designed to
capture red cells with the corresponding antigen (direct agglutination), or
via agglutination of red cells through indirect agglutination techniques.
The techniques may be performed manually or using automated
methods. At the RD&E lack of automation for these tests means that they
must be performed and read manually, thus introducing the risk of error.
Phenotyping using monoclonal antisera, for Rh (CcEe) and K antigens
using the Diamed technique (Bio-Rad Laboratories Inc.), is performed on
a single card containing six separate reaction columns. Each column
contains a specific antiserum (anti-C, anti-c, anti-E, anti-e or anti-K) within
a gel matrix and a negative control well is included to confirm that any
positive reactions are due to the presence of the corresponding red cell
antigen on the patient’s cells and not false positive reactions due to auto-
agglutination. A red cell suspension containing patient cells of unknown
antigen status is added to the top of the column, the card is then
centrifuged which forces the patient red cells through the gel matrix
containing the specific antisera. If the patient red cells carry the
corresponding antigen, antigen-antibody complexes form, trapping the
cells in the column and demonstrating a positive result. Conversely, cells
lacking the corresponding antigen will not form antigen-antibody
complexes and will be forced to the bottom of the column denoting a
negative reaction (see figure 1.11). Results can be read macroscopically
if performed manually or this can be performed via an automated system.
52
Figure 1.11: Principles of the column agglutination assay
Patient cell suspension

is added to the column
Red cells carrying the

antigen are trapped in the
gel matrix
Anti-serum is contained in
the gel matrix
Red cells lacking the
antigen travel to the
bottom of the column
Reaction strengths are graded

from negative (-) to strongly
positive (++++)
A diagrammatic representation of the principle of the column agglutination
technique for the determination of red cell antigen status. The patient cell
suspension is added to the column and centrifugation forces the red cells
through the gel matrix containing the anti-serum. Red cells positive for the
antigen form antigen-antibody complexes and are trapped in the gel matrix.
Cells that do not express the antigen travel to the bottom of the column during
centrifugation. See section 1.5.4 for information on grading of results.
53
Phenotyping techniques using column agglutination may also be
performed via indirect agglutination. Generally this is required when using
polyclonal antisera, which are incapable of forming antigen-antibody
complexes that can bridge between red cells forming agglutinates
directly. In these techniques, the use of antihuman globulin (antibodies to
human antibodies, usually of goat or murine extraction) is employed to
link red cells coated with the specific antiserum. An aliquot of a
suspension of patient red cells of unknown antigen status is incubated
with the specific antiserum in an incubation chamber at the top of the
column. An incubation period performed at 37 oC, simulating an in vivo
situation, allows antigen-antibody complexes to form if the corresponding
antigen is present on the patient red cells. A centrifugation stage then
forces the red cells through the gel matrix containing antihuman globulin
(AHG), if antigen-antibody complexes are present on the patient red cells
they are trapped by the AHG within the column. In the absence of
antigen-antibody complexes the red cells will travel to the bottom of the
column, denoting a negative result (see figure 1.12). The results can be
read macroscopically or via automated technology.
54
Figure 1.12: Reaction patterns of the column agglutination assay
Patient plasma is incubated

with the reagent red cells to
allow antigen-antibody
complexes to form
Anti-Human Globulin is
contained in the gel matrix
Reactions are given a numerical Positive reactions are seen if the

grading to denote the variations in patient’s plasma contains red cell
strength of positive reactions antibodies
A diagrammatic representation of the results seen with IAT column
agglutination technique for the determination of red cell antigen status
Patient red cells are incubated with the commercial anti-sera to allow
antigen-antibody complexes to form in the incubation well. Centrifugation
then forces the red cell/antiserum mix through the gel matrix containing
Anti-Human Globulin (AHG). If the patient red cells do not carry the
antigen being tested then the unsensitised red cells will fall to bottom of
the column. Positive reactions are denoted by patient red cells trapped in
the gel. Positive reactions are graded from strongly positive (4+) through
to weakly positive (+/-).
55
1.5.3 Principles of automated phenotyping methodology on the
Immucor NEO® analyser
Development of high throughput automated methods for extended red
cell phenotyping would not only introduce a technique for reducing
manual testing (with its inherent risks) for patients who had developed a
red cell alloantibody, but also facilitate extended phenotyping for patients
prior to transfusion. This has the potential to allow for more specific
matching of red cell antigens on the donor blood for patients likely to
require chronic transfusion support, thus potentially reducing the risk of
alloimmunisation. Column agglutination technology may be automated,
currently available platforms that could support this type of automated
phenotyping include; the Bio-Rad IH500 and IH1000 (Bio-Rad
Laboratories Ltd., Watford, UK), the Grifols Erytra and Erytra Eflexis
(Grifols UK Ltd., Cambridge, UK) and the Ortho AutoVue ® Innova System
(Ortho Clinical Diagnostics, Buckinghamshire, UK).
The Immucor NEO Iris and ECHO Lumena automated analysers (IBG
Immucor Ltd., West Midlands, UK) offer serological phenotyping using
direct agglutination techniques in microtitre plates and solid phase
methodology. The Immucor NEO analyser is currently utilised in the
RD&E transfusion laboratory for routine ABO/D typing, antibody
screening and antibody identification, but has not been developed for
automated serological phenotyping. This study details the development
and validation of a novel, high throughput method for extended
56
serological red cell phenotyping using the Immucor NEO® analyser
system.
The automated technique using the Immucor NEO® analysers employs
two distinct methods for the determination of the presence of red cell
antigens; direct agglutination and solid phase technology. In the direct
agglutination test the patient red cells are mixed with the specific
antiserum in a well of the microtitre plate, an incubation phase allows the
red cells to settle together permitting cross-linking of the IgM molecules in
the antiserum with the corresponding antigen on the red cells. This effect
is increased during a centrifugation phase enabling a quicker reaction to
occur. Direct agglutination of red cells with a particular reagent indicates
the presence of the corresponding antigen. The microtitre plate is then
gently shaken by the analyser, during this process red cells that have not
been agglutinated by the antiserum will break apart and form a loose
layer of cells in suspension, and cells that have agglutinated may stay
together forming a clump, reactions similar to those seen in direct
agglutination in tubes. The analyser captures images of these reactions
and the differing patterns are then converted into reactions grades (which
vary from negative (no agglutinates), through to strongly positive
(indicated by a tight button of cells at the bottom of the well) using in-built
computerised algorithms.
Solid phase (Capture-R) assays are based on the procedure of Plapp
(Plapp et al., 1984) and Juji (Juji et al., 1972) and are used in conjunction
57
with polyclonal antisera that are not capable of direct agglutination. The
individual wells of the microwell strips are coated with an affinity isolated
goat anti-murine antibody and a murine anti-RBC antibody; these are
used to support the formation of a red cell monolayer, created from the
patient sample by the automated analyser, to the microwell surface.
Reagent antiserum is then dispensed into the well, followed by low ionic
strength saline (LISS) and an incubation phase. Lowering the ionic
strength acts to increase the rate of formation of antigen-antibody
complexes, and can reduce the incubation time needed for the
complexes to form (cited in Rolih et al., 1985). During the incubation
period the antigens carried on the immobilised patient red cells are bound
to the specific IgG antibodies in the antiserum. Following incubation,
residual unbound immunoglobulins are removed from the wells by an
automated washing step and anti-human IgG coated Capture-R Ready
indicator red cells are added. A centrifugation step allows the indicator
red cells to come into contact with any antibodies bound onto the surface
of the patient’s immobilised red cells. If the patient’s cells are positive for
the antigen being identified then IgG-anti-IgG complexes will form
between the indicator cells and the sensitised patient red cells. The
indicator cells will then adhere to the surface of the immobilised patient
red cells forming a second layer of cells. If the patient cells are negative
for the antigen being determined then the indicator cells will not bind to
the immobilised cells and, during the centrifugation phase, will form a
pellet at the bottom of the well (see figure 1.13). The analyser image
58
capture system converts the reactions seen in the wells into reaction
grades denoting the presence or absence of the specific red cell antigen.
59
Figure 1.13: Principles of the solid phase phenotyping assay
Test red cells are fixed to the

microtitre well
Specific red cell antigen antisera is
added to the red cell monolayer
LISS acts as a potentiator

The reagent antiserum binds to the
antigens on the test cells if the
antigen is present (denoted by a ^) WASH
Indicator cells coated with anti-IgG will

Indicator red cells bind to any anti-sera captured on the
red cell monolayer
CENTRIFUGE
SIDE VIEW
Unbound Indicator cells will be

forced into a button by the
centrifugation stage, interpreted as
a negative reaction by the analyser
TOP VIEW software
STRONG POSITIVE WEAK POSITIVE NEGATIVE
Bound indicator cells remain on

the red cell monolayer and are
interpreted as a positive reaction
Diagrammatic view of the solid phase technology utilised by the IBG NEO®
analyser system Red cells are bound to the microwell surface by the analyser, a
potentiator LISS (low ionic strength saline) is added which increases the rate of
formation of antigen-antibody complexes. Antibodies to red cell antigens are
“captured” on the microwell during incubation, unbound immunoglobulins are
then rinsed from the wells in a wash phase. Capture R anti-IgG Indicator red
cells are added, which “sandwich” the “captured” antibodies, making them
visible. A centrifugation phase then brings the indicator red cells in contact with
antibodies bound to the reagent red cell membranes. (Diagram courtesy of
Caryn Conway, Immucor, personal communication).
60
1.5.4 Grading reactions in phenotyping techniques
The reaction strengths in the different methods for serological red cell
antigen phenotyping can be graded from negative (-) through to strongly
positive (4+); a comparison of the grading reactions seen in Capture-R
Select, tube agglutination and gel column agglutination techniques is
shown in figure 1.14. The grading of the reaction strength is important as
weak, or mixed field, reactions may be indicative of the presence of
transfused red cells, or, in the case of polyclonal antisera, the presence of
a positive DAT (BCSH, 2013). In either case the true phenotype of the
patient cannot be determined until the cause of the weak or mixed field
reaction has been investigated. Grading of reaction strengths for tests
performed manually is subjective and dependent on the experience of the
individual performing the test, thus some operators may misclassify a
weak, false positive, reaction as a positive reaction and report an
incorrect result without further investigation (SHOT, 2014). Automation of
phenotyping methods, as for the column agglutination and solid phase
assays, removes this subjectivity and allows a more standardised
approach to grading, thus reducing the risk of misreading a false positive
result due to the presence of transfused cells or a positive DAT (Stainsby
et al., 2006).
61
Figure 1.14: Comparison of reaction patterns in phenotyping assays
A comparison of grading reaction between Capture R Select (IBG NEO
analyser), haemagglutination performed in a tube (direct agglutination)
and column agglutination (Gel/CAT). Grading ranges from 0 (negative),
no reactions between the red cells and the plasma, through 1+ (weakly
positive reactions) to 4+ (strongly positive reaction) Note: weak (+/-)
reactions shown in previous figures are not demonstrated.
(Picture courtesy of Caryn Conway, Immucor – personal
correspondence).
62
1.6 Red cell antigen genotyping
Serological red cell phenotyping in patients with a recent history of blood
transfusion is complicated by the presence of donor red cells within the
circulation. In these situations it is not possible to determine whether
positive reactions seen in serological phenotyping tests are due to the
presence of red cell antigens on the patient (recipient) cells, or on those
of the transfused donor cells. In these circumstances the red cell
phenotype may be predicted by genotyping the patient’s DNA. Genetic
testing of patient DNA to determine the red cell antigen genotype, and
thus predict the phenotype, is possible post transfusion, because DNA is
extracted from the patient’s white blood cells, which are uncontaminated
by transfused blood containing only red cells. However, this is an
expensive, labour intensive, specialised test that, currently, does not lend
itself to the hospital setting; the DNA based approaches that are available
include the BLOODChip (Progenika Biopharma SA) (Avent et al., 2007),
sequence-specific primer-polymerase chain reaction (SSP-PCR) (Renoud
et al., 2006), BeadChip (Hashmi et al., 2005; Hashmi et al., 2007), PCR-
restriction fragment length polymorphism GenomeLab SNPstream
(Denomme and Van Oene, 2005; Montpetit et al., 2006) and the matrix-
assisted laser desorption/ionisation time-of-flight mass spectrometry
(MALDI-TOF MS) (Gassner et al., 2013). The availability of some of
these techniques on high throughput platforms, such as the BLOODChip
(using the Luminex xMAP analyser) and the MALDI-TOF MS, are making
these options more attractive to the hospital transfusion service.
63
As discussed in section 1.3, the molecular basis of most blood group
antigens is now known. The majority of blood group polymorphisms are
caused by single nucleotide polymorphisms (SNPs) in the genes that
dictate the blood group antigen; this may be the protein itself where the
gene codes for the blood protein antigen, or where gene codes for a
glycosyltransferase which then catalyses the addition of
monosaccharide(s) to the blood group oligosaccharide (Avent, 2009).
Commercial blood group genotyping techniques use polymerase chain
reaction (PCR) to amplify DNA obtained from a whole blood sample,
using target specific primers. The amplified products can then be
identified using probes corresponding to allelic pairs of blood group
SNPs. Scanning systems are employed to detect the allele specific
probes and specially designed software systems can interpret these
results into genotypes and predicted phenotypes.
High throughput, automated genotyping techniques have been available
now for more than a decade but have yet to be adopted into routine use
in the UK, either in the NHSBT for donor typing, or in the hospital
transfusion service for patient typing. One of the reasons for this slow
uptake is the cost of genotyping; new equipment would be required as the
assays could not be added to the current equipment performing
serological testing. In addition, manual steps in the PCR processing
phase remain a limiting factor in the implementation of the technique into
a busy hospital transfusion laboratory setting. A common criticism of
genotyping is that the genotype does not always reflect the phenotype
64
(Avent, 2009). This may be due to the null phenotypes seen in some
blood group systems, such as, Rh, ABO, Jk and Fy, or due to a new,
previously unknown allele. The length of time for genotype testing is also
considered restrictive, particularly within the hospital transfusion service
where provision of matched blood may be required urgently. The
BLOODChip IDCOREXT assay has been demonstrated to have a
processing time of less than 4 hours, which is shorter than that for other
systems including PCR-SSP and BeadChip (Finning et al., 2016).
Wagner and co-workers (Wagner et al.; 2017) have attempted to address
this by devising a novel method using PCR amplification without the need
for DNA extraction. This group have reported that, using the new method
results can be available within 40 minutes of receipt of the blood sample,
which may make the technique applicable to the hospital transfusion
setting.
This study aimed to compare the results from blood group genotyping
using one of the currently commercially available genotyping assays,
BLOODChip, with those found by traditional serological methods for a
variety of blood group antigens. Additionally, the study aimed to assess
the feasibility of implementing such a DNA based approach in the routine
hospital laboratory, including a cost comparison of the two different
approaches to red cell antigen typing and their advantages and
disadvantages.
65
1.6.1 Principles of the BLOODChip blood group genotyping assay
The BLOODChip IDCOREXT (Progenika Biopharma, Derio, Spain) uses
a microarray system for the determination of polymorphisms within the
Rh, Kell, Kidd, Duffy, MNS, Diego, Dombrock, Colton, Yt (previously
Cartwright) and Lutheran blood group systems (see section 5.1).
Extracted DNA is amplified using a PCR mix containing HotStarTaq
polymerase, primers, biotinylated nucleotides and a buffer solution. The
amplification products are then hybridised to allele specific
oligonucleotide probes attached to microspheres, and then labelled with
fluorescence-conjugated streptavidin-phycoerythrin (SAPE) (figure 1.15).
The microspheres contain two distinct dyes, red and infrared, which emit
light in different regions of the spectrum. The combination of dyes in
different concentrations results in 100 unique fluorescent colour tones,
each correlating to a different spectral address defining a bead class
(Goldman et al., 2015). Quantification of the amounts of labelled PCR
products hybridised to the beads is then performed using the Luminex
xMAP analyser system (Luminex Corporation, Austin, Texas, USA).
66
Figure 1.15: The principles of the IDCOREXT Assay
Diagrammatic representation of the amplification, hybridisation and labelling phases in the IDCOREXT assay,
Step 1: The amplification phase separates the DNA strands allowing the target specific primers to bind to the DNA. The addition of the HotStarTaq
polymerase allows new DNA strands to be produced. The cycle is repeated 40 times to produce sufficient DNA for analysis (picture courtesy of Dr Ruth
Morse, University of the West of England, personal communication).
A: Double stranded DNA extracted from patient sample
o
B: The DNA strands are denatured at 96 C to create single
stranded DNA
C: The primers contained in the IDCOREXT assay kit then
bind to the DNA strand when the temperature is lowered to

o
50 C
D: The HotStarTaq polymerase is then added
67
E: The HotStarTaq polymerase then synthesises
nucleotides from the end of the primer to create a new
stand of DNA
F: The temperature is then raised again to denature the
DNA.
The entire process (A-F) is repeated 40 times to amplify the
the target DNA
68
Step 2: The amplified DNA is then hybridised with allele specific probes attached to beads in suspension (picture courtesy of Louise Redfern, Grifols,
personal communication).
Probe specific for blood group gene of interest
Biotinylated PCR product
Bead with allele specific probe attached
Step 3: A fluorescent marker (SAPE) is then added to the bead suspension. This marker will then attach to the DNA if present on the allele specific probe
forming a dye-DNA conjugate. Excitation of the dye-DNA conjugate by the Luminex analyser then allows the bead specific emission to be quantified by the
Luminex xMAP analyser. (Picture courtesy of Louise Redfern, Grifols, personal communication).
Streptavidin R-Phycoerythrin Conjugate (SAPE)

attached
69
Step 4: The sample is analysed on the Luminex xMAP analyser system. The allele specific probes are attached to beads of different combinations of red and
infrared dyes which allow the Luminex xMAP analyser to distinguish one bead set from another and hence identify the presence of specific alleles (picture
courtesy of Luminex (figure courtesy of Lawrence Rentoul, Merck Millipore, personal communication)).
70
1.6.2 Principles of the Luminex xMAP analyser
The Luminex xMAP analyser system uses a multiplexing system to detect
multiple nucleic acid sequences in a single reaction. In a process similar
to that of a flow cytometer, the analyser directs the beads into a cuvette
and past red and green lasers. The beads are sorted into their different
classes, in accordance with the dye concentration, by the red laser,
allowing the analyser to distinguish one bead class from another. The
green laser excites the SAPE allowing the analyser to identify the
presence of a specific allele hybridised to an allele specific probe.
Fluorescence and side scatter are measured and the information is
reported as median fluorescent intensity (MFI) (see figure 1.16). The
BLOODChip analysis software extracts this information from the analyser
and converts this into a report containing the genotypes and predicted
phenotypes. The system can test up to 48 samples in a single batch. The
manufacturer recommends that a negative and positive control is
included with each batch.
71
Figure 1.16 Principles of the Luminex xMAP analyser
The Luminex xMAP® technology uses fluidics, optics, and digital signal
processing combined with the BLOODChip IDCOREXT proprietary
microsphere technology to deliver multiplexed assay capabilities. The
sample is moved through the instrument and reaches the cuvette. As the
beads enter the cuvette, the sheath fluid moves at 90ul/sec and forms a
core around the beads which move at 1ul/sec. This is known as
hydrodynamic focusing. The red laser excites the red and infrared dyes of
each of the beads, the analyser collects this information to determine
where the bead falls on the bead map, and side scatter is also measured
to determine if the bead is 5.6μm. A Double Discriminator is gated to rid
the data of any outliers such as small particulates and doublet beads. The
green laser excites the reporter dye, SAPE, and the information is
interpreted to determine the analyte being tested. (Picture reproduced
courtesy of Louise Redfern, Grifols, personal communication).
Beads with attached allele

specific probes are
directed through the
cuvette
The red laser excites the

dye within the bead and
the green laser excites the
SAPE
72
1.7 Risk of alloimmunisation in chronic transfusion
The risk of alloimmunisation is dependent on the immunogenicity of the
red cell antigen, hence it has long been advocated that all blood
transfusions should be matched for the highly immunogenic RhD antigen
(Woodrow and Donohoe, 1968) as well as the standard ABO antigens
discovered by Landsteiner (1900 cited in Daniels 2013). Other red cell
antigens which may evoke an immune response, and cause a transfusion
reaction, in antigen negative transfusion recipients include those of the
Rh (CcEe), K, Fy, Jk, MN and Ss blood group systems. However, these
are only weakly immunogenic, immunising only 2-6% of transfusion
recipients (Seyfried and Walewska, 1990; Heddle et al., 1995; Hoeltge et
al., 1995) compared to the RhD antigen which has an alloimmunisation
rate of 30-90% (Freda et al., 1967; Gunson et al., 1970; Cook, 1971;
Frohn et al., 2003). The RhD antigen is particularly immunogenic as RhD
negative individuals lack the entire RhD protein from the red cell
membrane (Daniels, 2013).
Studies have suggested that alloimmunisation may also be influenced by
other factors such as; underlying disease, gender, age, inflammatory
status and the frequency and number of transfusions (Coles et al., 1981;
Blumberg et al., 1983; Blumberg et al., 1984; Fluit et al., 1990;
Hendrickson et al., 2006). Alloimmunisation rates in patients with
haematological malignancies vary. A retrospective study undertaken in
the Netherlands (Schonewille et al., 1999) found an overall
alloimmunisation rate of 9% in patients with malignant myeloproliferative
73
and lymphoproliferative diseases whereas another study by Blumberg
and co-workers (1983) showed rates of 16% in patients with
myelogenous leukaemias and 11% in aplastic anaemia but no
alloimmunisation in patients with lymphocytic leukaemias. Another study
demonstrated a relatively high alloimmunisation rate amongst patients
with myelodysplastic syndrome MDS; 36% of patients given more than 30
units of blood, and 64% given in excess of 100 units were found to have
developed alloantibodies (Milic et al., 2011).
More recent studies have been published regarding the alloimmunisation
rates of patients with non-haematological malignancies or renal
insufficiency. Alloimmunisation rates of 13.8% have been reported for
patients with SCD (Master et al., 2016) and 7.98% for those with
thalassamia (Abdelrazik et al., 2016). An interesting study by Türkmen
and co-workers (2016) reported an alloimmunisation rate of 0.12% in a
cohort of neonates and children up to the age of 3 years. In contrast, their
control group, comprising of adult patients aged 45 years and older, had
an immunisation rate of 3.55% after a median of 5 red cell transfusions,
rising to 10.24% after 40 red cell units. A review of the published
literature regarding alloimmunisation rates in SCD by Zheng and Maitta
(2016) revealed that the United States had a higher immunisation rate
(22.3%) compared to other regions (16.5%) including South America, the
Caribbean, Middle East, Africa and Europe. They concluded that this was
probably a result of the limited ethnicity of the donor pool, where Black
donors represent a minority. In populations where donors and recipients
74
are ethnically similar, such as Uganda and Jamaica, the alloimmunisation
rate is much lower (Olujohungbe et al., 2001). Differential expression of
red cell antigens are noted between Caucasian and Black populations for
many of the blood group systems, most notably for Rh (Daniels, 2013).
There is a paucity of recent publications on the alloimmunisation rates in
the UK for chronically transfused patients with haematological
malignancies or renal insufficiency, providing a rationale for this research.
1.8 Red cell autoantibodies
Patients with haematological malignancies may also make antibodies to
red cell antigens on their own cells (Timura et al., 2000; Solal-Celigny et
al., 1984) and this can also create challenges with the provision of blood
by the hospital transfusion services. The production of red blood cell
autoantibodies has been shown to be associated with blood transfusion
and the subsequent production of alloantibodies by the recipient (Young
et al., 2004). This can sometimes cause haemolytic episodes (Young et
al., 2004) but probably occurs more commonly with no ill effects in the
patient (Garratty, 2004). It has been suggested that the production of
autoantibodies in a transfused patient may be due to the presence of
donor lymphocytes in the transfused blood that produce an alloantibody
to the recipient red cells, thus mimicking an autoantibody (Ishikura et al.,
1993). In the United Kingdom (UK) there is currently no reported data on
75
the occurrence of red cell autoantibodies associated with
alloimmunisation following blood transfusion in patients with renal disease
or haematological malignancies and whether the presence of these
antibodies has an effect on the transfusion requirements of the patients
remains unknown.
1.9 Extended red cell antigen matching of blood in chronic
transfusion
There is no current evidence on the rates of alloimmunisation and
autoimmunisation in patients with chronic transfusion requirements as a
result of haematological malignancies or renal failure in the UK and
whether there are risk factors within this heterogeneous group of patients
that could be used to predict those at greatest risk of producing allo
and/or autoantibodies during the course of their treatment.
It has been advocated by some workers (Michail-Merianou et al., 1987;
Rosse et al., 1990) that certain groups of chronically transfused patients,
for example those with sickle cell disease (SCD) and thalassaemia,
should receive blood that is matched for more than the standard ABO and
RhD antigens. In the UK the British Committee for Standards in
Haematology (BCSH) also endorse this practice for patients with SCD
and thalassaemia (BCSH, 2013) but it is not standard practice for other
groups of chronically transfused patients, such as those with
haematological malignancies. This has mainly been due to the cost
76
implications of performing the tests, the perceived difficulties in providing
matched blood for large numbers of patients and the lack of evidence of
the benefits of such an approach.
Blood transfusions in patients with renal disease and haematological
malignancies account for a substantial proportion of the total red cells
transfused within the hospital setting. At the RD&E this usage is
approximately 30% of the total red cells transfused. There are some
historical studies (Fluit et al., 1990; Shirey et al., 2002; Schonewille et al.,
2009; Kadar, 2010; Schonewille et al., 1999; Domen and Ramirez, 1988;
Blumberg et al., 1984) that have investigated the occurrence of red cell
antibodies in these groups of patients and the potential benefits and cost
effectiveness of strategies to reduce immunisation. However, the advent
of improvements in high throughput automated analysers in the hospital
transfusion service, allowing rapid automated red cell phenotyping, along
with novel genotyping platforms and increased red cell antigen typing of
donor red cells has now made this topic relevant to review.
This research aimed to review the current extent of allo and/or
autoimmunisation in chronically transfused patients attending the RD&E,
and review the effect that extending matching of blood for transfusion
from ABO/D to additional antigens might have on the immunisation rate. It
aimed to validate a new technique for extended serological phenotyping
using the IBG NEO analyser and investigate the feasibility of introducing
a genotyping assay into a hospital transfusion laboratory. It aimed to
77
review the cost effectiveness of implementing serological phenotyping or
genotyping assays to support a type and match program. In addition,
ethical approval was obtained from the National Research Ethics Service
(NRES) Committee South West - Cornwall & Plymouth (MREC
No.12/SW/0251 R&D Study No. 13017733), to recruit patients attending
the haematology clinic at the RD&E to a pilot study, a small randomised
controlled trial. Patients recruited to the study were randomly assigned to
one of two groups; patients assigned to group 1 received standard care
with regard to blood transfusions (ABO and RhD compatible blood) and
group 2 patients were to be assigned blood that was matched for Rh (C,
e, E and e) and K in addition to the standard ABO and RhD. The pilot
study was designed to investigate the feasibility of the provision of blood
matched for Rh (CcEe) and K antigens within this cohort of patients as
well as reviewing recruitment and drop-out rates. A more extensive
randomised controlled trial has been performed by Sanquin in the
Netherlands (Schonewille et al, 2016); this multicentre study aimed to
review the incidence of alloimmunisation in a general transfusion
population using a multi-centre, parallel group randomised trial.
Randomisation of study participants into one of two study arms; ABO-D
matched only, or extended match (including c, C, K, Fya, Jka and S
antigens) demonstrated a 64% reduction in alloimmunisation risk in
surgical patients. The pilot study undertaken in our institution was
designed to investigate the feasibility of performing a larger multicentre
study in the UK. The results of the pilot study could then be used to
design a larger randomised control trial with the aim of investigating the
78
benefits and cost effectiveness of a strategy for preventive extended
antigen matching for patients with haematological malignancies. Neither a
pilot study, nor a randomised control trial such as this, has yet been
attempted in the UK.
A type and match program for chronically transfused patients has the
potential to reduce the risk of alloimmunisation, with the subsequent
patient safety benefits of reduction of risk of transfusion reaction or HDFN
(see section 1.4). It is this hypothesis that underpins this research project.
79
1.10 Aims and Hypothesis
 The hypothesis underpinning this research is that allo and
autoimmunisation in chronically transfused patients can be
reduced by the provision of blood subjected to extended red cell
phenotyping and/or genotyping.
 The research aims to test the hypothesis are:
o To evaluate the frequency of allo- and autoantibodies in two
cohorts of patients with chronic transfusion requirements
using retrospective patient data from patient cohorts
transfused at the RD&E (Haematology and Renal patients)
o To identify any risk factors for the development of allo- and
autoantibodies in these two cohorts of patients with chronic
transfusion requirements using the retrospective patient
data
o To evaluate the use of novel fully automated, high
throughput, red cell serological phenotyping profiles on the
NEO® analyser (IBG Immucor) automated analyser
o To evaluate and compare the cost effectiveness of a
strategy of red cell phenotyping and matching blood for
transfusion in patients treated for haematological
malignancies and those treated for renal insufficiency using
information obtained from the retrospective patient data.
o To evaluate the use of the BLOODChip IDCOREXT blood
group genotyping kit and compare the predicted phenotypes
obtained from genotypes with those obtained by the
80
automated serological phenotyping profiles on the NEO®
analyser.
o To evaluate the results of a pilot study recruiting patients
with haematological malignancies investigating the effects
of prospectively typing and matching red cell antigens for
this cohort of patients. This pilot study is designed to assist
with the protocol for a larger randomised controlled trial.
81
2 MATERIALS AND METHODS
2.1 Overview of the project
An overview of the work undertaken for this research is show in the form
of a flow chart:
Retrospective review of chronically transfused patients treated at the RD&E to establish:

 Rate of allo/autoimmunisation
 Risk factors for allo/autoimmunisation
 Costs associated with allo/autoimmunisation
 Potential benefit of a type and match strategy
Validation of a fully automated high throughput serological phenotyping assay to support a type and
match strategy:
 Comparison of phenotype results with current manual methods
 Cost analysis
 Suitability for routine use to support type and match strategy
Validation of the BLOODChip IDCOREXT blood group genotyping assay:

 Comparison of IDCOREXT predicted phenotype with the serological phenotype
 Feasibility of technique for use in hospital transfusion service
 Cost analysis
 Suitability for routine use to support type and match strategy
Pilot study – randomised controlled trial:

 Feasibility of introducing a type and match stategy in a hospital transfusion service
 Identification of changes required in service provision to support the strategy
Cost analysis:
 Analysis of the costs associated with allo/autoimmunisation in chronically transfused patients
and methods for high throughput phenotyping/genotyping
 Review of cost effectiveness of a type and match strategy
82
2.2 Retrospective review of chronically transfused patients
2.2.1 Sample size for retrospective review
Two cohorts of transfusion dependent patients were selected for inclusion
in the retrospective review; patients with haematological malignancies
and patients with renal insufficiency. Sample size for the retrospective
review was calculated with the aim of achieving a 95% confidence level
with the minimum margin of error (<5%), whilst taking into account the
amount of historic information available in the computer records and the
timeframe of the study. The decision was taken to include a minimum of
800 patients in each study cohort, giving a margin of error of 3.4%.
2.2.2 Patients with haematological malignancies
Data were collected for 1107 randomly selected patients admitted for red
cell transfusions within the haematology ward and haematology day case
unit at the RD&E between 1995 and 2012. Patients were identified using
an in-house enquiry package designed to extract the hospital number and
name of patients who had crossmatch tests recorded in the pathology
computer system (Integrated Pathology System (IPS)) and had the
location codes of YAR (haematology in-patient ward) or YARD
(haematology day case). Patients who had received at least one unit of
blood on one occasion were included in the database. Ethical approval
was not considered appropriate for this retrospective audit as agreed at
the Cornwall and Plymouth Research Ethics Committee on the 5th
September 2012 (MREC No.12/SW/0251 R&D Study No. 13017733).
83
Blood samples from patients attending the haematology ward and day
case unit at the RD&E during this time period were routinely screened for
atypical red cell antibodies, using a selected set of reagent red cells for
antibody detection, as part of the pre-compatibility testing. From 1995 to
2000 antibody detection was undertaken manually using the low ionic
strength solution (LISS) gel cards (column agglutination technology;
Diamed UK). From the years 2000 to 2008 automated antibody screening
was performed by the IBG Multisampler (IBG Immucor) robotic system
using low ionic strength solution gel test cards (Diamed UK). From 2008
to the present day, antibody screening has been undertaken using
capture technology (solid phase technology) on the IBG Galileo/NEO and
Echo fully automated analysers (IBG Immucor). Over the study period,
antibody identification was performed using a variety of techniques
including column agglutination (Diamed UK) and automated Capture
Ready ID (IBG Immucor). Laboratory processes for antibody identification
also included the DAT by column agglutination (Diamed UK) and, if
appropriate, the use of column agglutination with monospecific anti-
human globulin (AHG) reagents to identify the type of antibody and/or
complement fraction coating the red cells.
Test costs were calculated using an in-house spreadsheet calculator
(Microsoft Excel) taking into account reagent costs, equipment
maintenance, consumables, quality control and staff time incurred during
the financial year 2011-2012. Total test costs for the additional testing
84
performed on patient samples during the retrospective review period were
calculated using current tests costs, although it is accepted that this is an
approximation of the actual costs of the tests performed and is subject to
error due to variations such as different testing techniques and inflation.
Data were collected on the gender, age at first transfusion, red cell
transfusion history, platelet transfusion history, requirement for irradiated
blood components, allo/autoantibody status and additional testing
requirements for 1107 patients admitted for red cell transfusions within
the haematology ward and day case unit at the RD&E between 1995 and
2012.
2.2.3 Patients with renal insufficiency
Data were collected for 877 randomly selected patients admitted for red
cell transfusions within the renal ward and haemodialysis unit at the
RD&E between 1995 and 2013. Patient data were collected as described
for haematology patients in section 2.2.2, with the exception that the
location codes of SID (renal in-patient ward) or CRE (haemodialysis unit)
were used in the computer enquiry. Patients who had received at least
one unit of blood on one occasion were included. Ethical approval was
not considered appropriate for this retrospective audit as agreed at the
Cornwall and Plymouth Research Ethics Committee on the 5th
September 2012 (MREC No.12/SW/0251 R&D Study No. 13017733).
85
Screening of patients’ samples for atypical red cell antibodies during this
time period was as for patients admitted to the haematology wards. Data
were collected on the gender, age at first transfusion, red cell transfusion
history, renal transplant history, allo/autoantibody status and additional
testing requirements for 877 patients admitted for red cell transfusions
within the renal ward and haemodialysis unit at the RD&E between 1995
and 2013.
2.3 Comparison of manual and automated red cell
phenotyping
One hundred patient samples, to be discarded, were selected for
comparative extended serological red cell antigen phenotyping in the
current manual systems and novel automated method. The BCSH
guidelines for validation and qualification of new techniques in the
laboratory setting (BCSH, 2012) recommend that a predetermined
number of samples are tested in duplicate with the current system,
quoting “2 weeks or 250 samples whichever occurs first” as a target
value. At the time of the validation approximately 140 samples per year
were phenotyped for Rh (CcEe) and K antigens, and less typed for Jk,
Fy, Ss and MN antigens. It was, therefore, considered that 100 samples
would be an acceptable size for comparison testing, which would test all
the reagents and provide the sensitivity required to give confidence in the
performance of the technique. All samples were collected into EDTA
anticoagulant and stored at 2-8oC until tested. Patients who had records
86
on the hospital computer system, of blood transfusion in the three months
prior to the collection of the sample were excluded from analysis due to
the interference caused by circulating transfused red cells in these
serological techniques. Ethical approval for this stage of the project was
not required as it was considered to be service development. Samples
had been taken for routine blood group and antibody screening and no
additional samples were required.
2.3.1 Manual red cell phenotyping
Manual red cell phenotyping for Rh (CcEe) and K antigens was
performed by column agglutination using DiaClon Rh-Subgroups and K
(Catalogue number: 50110, Bio-Rad, Watford, UK) microgel card for 100
patient samples. The microgel card contains monoclonal antibodies (anti-
C (cell line MS-24), anti-c (cell line MS-33), anti-E (cell line MS-260), anti-
e (cell lines MS-16, MS-21 and MS-63) and anti-K (cell line MS-56))
within each gel matrix and also includes a negative control well.
Method:
A 5% suspension of patient red cells was prepared by mixing 0.5mL of
ID-Diluent 2, a modified, low ionic strength solution (catalogue number
05761, Bio-Rad) with 25 µL of packed cells, in accordance with
manufacturer’s instructions.
Patient red cell suspension (10 µL) was then added to each well of the
microgel and the card was then centrifuged in an ID-Centrifuge 12 S II
(Bio-Rad) at 85xg for 10 minutes.
87
The results were read macroscopically, with positive reactions denoted by
red cell agglutinates visible in the gel matrix and negative reactions by a
button of red cells at the bottom of the column (see figure 2.1).
88
Figure 2.1: Rh and K antigen phenotyping by column agglutination
Positive reactions are

seen if the antigen is
present on the patient
red cells
Negative reactions are

seen if the antigen is not
present on the patient
red cells
Specific antiserum is
contained within the gel
matrix
Example of a patient red cell phenotype performed using column agglutination
methodology (Bio-Rad). Each column within the card contains a specific antiserum, a
suspension of the patient red cells is added to each column, and centrifugation then
forces the patient red cells through the gel matrix containing the antiserum. If the patient
red cells have the corresponding antigen then the cells are trapped in the gel matrix, if
the red cell antigen is absent then the patient cells will travel to the bottom of the
column. The patient phenotype in the example shown here is C+, c+, E-, e+, K-; the
negative reaction in the control column validates the results and confirms the absence of
any false positive reactions caused by auto-agglutination.
89
Manual red cell phenotyping for the Fy (a and b), Jk (a and b), s and k
antigens was performed using polyclonal reagents supplied by Lorne
(Lorne Laboratories Limited, Early, UK) in an IAT. The following reagents
were used: monoclonal anti-Fya (cell line DG-FYA-02 code 774002) anti-s
(cell line P3YAN3 code 771002) and anti-Fyb (code 317002), anti-Jka
(code 323002), anti-Jkb (code 324002), anti-k (code 320002) prepared
from pooled human sera.
Method:
A 0.9% suspension of patient red cells was prepared by mixing 10 µL of
packed red cells with 1mL of ID-Diluent 2 (Bio-Rad).
Patient red cell suspension (50 µL) was then added to each well of a
LISS Coombs Anti-IgG microgel card (Catalogue number: 50540, anti-
IgG, rabbit, Bio-Rad) followed by 25 µL of the appropriate anti-sera.
A control well was included on the microgel card by the addition of 50 µL
of patient red cell suspension to one of the wells within the Coombs Anti-
IgG microgel card.
The microgel was incubated for 15 minutes in a DiaMed-ID Incubator
37oC (Bio-Rad) and then centrifuged in an ID-Centrifuge 12 S II (Bio-Rad)
at 85 x g for 10 minutes.
The results were then read macroscopically, a positive reaction in the
control well invalidated the results (see figure 2.2).
90
Figure 2.2: Manual Serological phenotyping for Fy (a and b), Jk (a
and b) and s antigens by column agglutination
Incubation well
Positive reactions are seen if the

antigen is present on the patient red
cells
In the manual column agglutination Indirect Antiglobulin Test (Bio-Rad)

patient red cells in suspension are incubated with specific antisera at
37oC to allow antigen-antibody complexes to form. Centrifugation then
forces the red cells through a gel containing anti-human globulin. Red
cells with antibodies attached are then caught by the anti-human globulin
within the gel. Negative reactions are denoted by a button of patient red
cells at the bottom of the column and positive reactions by red cell
agglutinates trapped within the gel matrix. The patient shown in this figure
has the phenotype Fy (a-b+), Jk (a+b-) and s+, the negative reaction in
the control well indicates that the reactions with the antisera are valid and
not due to the presence of a positive Direct Antiglobulin Test.
91
Manual red cell phenotyping for the M, N, S and C w antigens was
performed using monoclonal reagents manufactured by Millipore
(Millipore UK Ltd, Watford, UK). The following antisera were utilised; Anti-
M (cell line LM110/140), anti-N (cell line BO3), anti-S (cell line MS-94)
and anti-Cw (cell line MS-110). A 3-5% red cell suspension was prepared
by mixing 0.5mL of Phosphate Buffered Saline (PBS 0.9% pH7.0, IBG
Immucor) with 75 µL of the patient packed cells. The same cell
suspension was used for all the tube techniques described below.
Method:
N antigen status was determined by adding 75 µL of cell suspension and
75 µL of anti-N to an appropriately labelled test tube.
The mixture was agitated gently and then centrifuged in an Immufuge II
(Baxter Healthcare Ltd, Newbury, UK) for 1 minute at approximately 200-
300 x g.
The results were then read by gently agitating the tube to dislodge the red
cells and examining macroscopically for agglutination.
M antigen status was determined by adding 75 µL of red cell suspension
and 75 µL of anti-M to an appropriately labelled test tube.
This was mixed well and incubated at room temperature for 5 minutes.
The tube was then centrifuged at approximately 1000 x g for 20 seconds.
cells and examining macroscopically for agglutination.
92
S antigen status was determined by adding 75 µL of patient red cell
suspension and 75 µL of anti-S to an appropriately labelled test tube.
This was mixed and then centrifuged at approximately 1000 x g for 20
seconds.
cells and examining macroscopically for agglutination. Cw antigen status
was determined by the same methodology as the S antigen status.
All negative and weakly positive reactions for M, N, S and Cw antigens
were then incubated at room temperature for a further five minutes before
being centrifuged (Immufuge II; Baxter Healthcare Ltd) for 1 minute at
approximately 200-300 x g) again for repeat macroscopic reading.
2.3.2 Automated red cell phenotyping using the NEO analyser
Automated red cell phenotyping was performed using the fully automated
Galileo-NEO® analyser (IBG Immucor). This analyser is used within the
transfusion laboratory at the RD&E for routine ABO/D typing, antibody
screening and antibody identification techniques. It also has the capacity
for automated DAT testing and crossmatching, although these tests have
yet to be validated on the analyser at the RD&E. Reagents for red cell
antigen typing are also supplied by IBG Immucor, which are CE marked
for tube, slide and automated techniques, however, these are not used
for automated testing within the UK (personal communication from
Berndine Kokkelink, IBG Immucor). Provision of extended, automated
93
serological phenotyping profiles was considered to be a novel way of
introducing high throughput phenotyping technology within the hospital
transfusion service, which could then be adopted by other centres using
the same platforms. The assay would enable identification of a range of
red cell antigens in a single, automated test profile, which is not yet
available in the UK.
A profile to determine the Rh (CcEeCw), K, M and N antigen status was
set up on the NEO® analyser. Immuclone® IgM (IBG Immucor)
monoclonal reagents suitable for use on the IBG NEO® analyser system
in 96 well U-bottom microplates were utilised for this profile. The
monoclonal reagents used were; anti-C (derived from cell line MS-24),
anti-c (derived from cell line MS-33), anti-E (derived from cell lines MS-80
and MS-258), anti-e (derived from cell lines MS-16, MS-21 and MS-63),
anti-K (derived from cell line MS-56), anti-M (clone M-11H2), anti-N
(clone 1422-C7) and anti-Cw (derived from human cell line MS-110). The
reagents were loaded into a single reagent rack, enabling a simple and
fast method for reagent presentation to the analyser. The patient primary
blood sample was centrifuged in a Labofuge400 centrifuge for 11 minutes
at 2,250 x g prior to placing on the NEO® analyser.
Method:
The automated process enables the analyser to produce a red cell
suspension from the patient primary sample tube; 10μL of red cells taken
94
from the primary sample are mixed with 314μL of Diluent (IBG Immucor
kit, ref: 0066058).
Aliquots (40μL) of this cell suspension are then added to the first eight
wells of the 96 well U-bottom microplate in a vertical pattern.
An aliquot (10μL of anti-Cw, -C, -c, -E, -e, -K or 20μL of anti-M, -N) of the
monoclonal reagent is then added to the appropriate well of the microtitre
plate in the order shown in figure 2.3.
An incubation phase performed at 20oC for 10 minutes allows the
antigen-antibody complexes to form, if the antigens are present.
Centrifugation of the plate (30 seconds at 80 x g followed by 30 seconds
at 170 x g) enhances the antigen-antibody reaction by forcing the red
cells closer together allowing the antisera to bind red cells together as
agglutinates.
Automated shaking of the plate for 60 seconds then disperses red cells
that have not been agglutinated, leaving a loose layer of red cells within
the microwell. Cells that had been agglutinated by the formation of
antigen-antibody complexes remained as a tight button of cells at the
bottom of the microwell.
The automated reader within the analyser is then able to interpret the
reaction patterns of the red cells, recognising a positive reaction (tight
button of cells at the bottom of the microwell) from a negative reaction
(loose layer of cells around the microwell). The pattern of cells detected
95
by the automated reader is then converted into a numerical reaction
strength (between 0 and 99) and a corresponding reaction grade
(between 0 and 4; figure 2.3). The user is able to view both reaction
strengths and grades; the availability of the reaction grade allows some
comparability with manual techniques that may also be employed in the
laboratory.
96
Figure 2.3: Reaction patterns generated by automated red cell
antigen phenotyping for Rh(CCwcEe), K, M and N antigens
Format of the results and reaction patterns generated by the automated
monoclonal profile on the NEO®. The left hand side of the screen shows
the laboratory number associated with the patient sample being tested
and the antigen status of the patient red cells. The right hand side of the
screen shows the image of the reaction in the well captured by the
analyser camera, the reaction strength generated by the analyser and the
grade of reaction. A negative reaction is equivalent to reaction strength of
23 or below, equivocal and positive reactions are variable and dependent
on the antisera used. The “well image” column shows a photographic
image of the agglutination of the cells (positive reaction) or the loose layer
of non-agglutinated cells denoting a negative reaction. (Picture courtesy
of Caryn Conway, IBG Immucor, personal communication).
97
Automated red cell phenotyping for Fy(a and b), Jk(a and b), S, s and k
antigens was performed using commercial polyclonal antisera prepared
from pools of human sera (IBG Immucor) in Capture-R Select microwell
strips (IBG Immucor ref: 006446). The reagents are loaded in a single
reagent rack, as for the monoclonal reagents. The patient primary blood
sample was centrifuged in a Labofuge400 centrifuge for 11 minutes at
2,250 x g prior to placing on the NEO® analyser.
Method:
The automated process enables the analyser to produce a red cell
suspension from the patient primary sample tube; 4μL of red cells from
the sample tube are mixed with 270μL of Diluent (IBG Immucor kit).
Aliquots (40μL) of this cell suspension are then added to the first eight
wells of the 96 well of the Capture-R Select microwell strips in a vertical
pattern. The Capture-R Select microwell is supplied as a ready to use kit
and is pre-coated with an affinity isolated goat anti-murine and an anti-
RBC antibody to immobilise red blood cells to the microwell surface.
The analyser performs a shake phase for 15 seconds to enable the
patient red cells to bind to the Capture-R Select coated microwell.
A double centrifugation stage (60 seconds at 150 x g followed by 60
seconds at 500 x g) then enables the formation of the red cell monolayer.
An aliquot (25μL) of the polyclonal reagent was then added to the
appropriate well of the microtitre plate along with 35μL of negative control
reagent (IBG Immucor) in the order shown in figure 2.4.
98
An incubation phase performed at 39oC (enabling a 37oC temperature
within the reaction well) for 20 minutes allows the antigen-antibody
complexes to form, if the antigens are present on the patient’s red cells.
The analyser then completes a wash cycle to remove any unbound
antibodies. Immucor Capture-R Ready Indicator Red Cells (a suspension
of red blood cells coated with murine monoclonal anti-human IgG
molecules, 50μL) were then added to the microwell.
These cells are captured by the antigen-antibody complexes in the
presence of the antigen, but remain free in suspension in the absence of
the antigen. A double centrifugation stage (30 seconds at 80 x g followed
by 30 seconds at 170 x g) followed by a shake phase and further
centrifugation (20 seconds at 15 x g) then forces free cells to the bottom
of the microwell but bound cells remained as a monolayer of cells around
the microwell.
Automated reading of the microwell plate by the analyser allowed the
distinction between the monolayer of cells, indicating a positive reaction,
or a button of cells in the bottom of the microwell indicating a negative
reaction (figure 2.4).
99
Figure 2.4: Reactions patterns generated by automated red cell
antigen typing for Fya, Fyb, Jka, Jkb, S, s and k antigens
Screenshot showing the reaction patterns generated by the NEO

analyser. In the Capture R Select solid phase test positive reactions are
denoted by a layer of patient red cells coating the bottom of the well and
negative reactions are denoted by a tight button of cells, this is in contrast
to the reaction patterns seen in the direct agglutination testing using the
monoclonal reagents (as shown in figure 2.3). The serological phenotype
of the patient sample is shown in the top left hand side of the screen.
(Picture courtesy of Caryn Conway, Immucor, personal communication).
100
2.3.3 Extended phenotyping profile result interpretation on the
NEO® analyser
A monoclonal antisera reagent profile (anti-C, Anti-c, Anti-E, anti-e, anti-
K, anti-Cw, anti-M and anti-N) was created on the IBG Immucor Galileo-
NEO® analyser (section 2.3.2, figure 2.3) along with a polyclonal antisera
reagent profile (anti-Fya, anti-Fyb, anti-Jka, anti-Jkb, anti-S, anti-s, anti-k
and auto control; all reagents from IBG Immucor, figure 2.4). The Galileo-
NEO® analyser provides the test result in three different formats; a “+” or
“-“ result, a numerical value (0 to 99) and a reaction strength value (0, 1+,
2+, 3+ and 4+). The numerical value has programmed cut-off points for
generating the reaction strength values, which vary according to the
antisera (see appendix 3). These cut-off points have been validated by
the supplier and cannot be influenced by the laboratory staff operating the
analyser. A numerical value of 23 for the monoclonal profile and 20 for
the polyclonal profile, or below generated by the analyser is set to
correspond to a negative result and borderline values between negative
and the set positive value (1+ reaction) are expressed as “not
determined” (NTD). The analyser returns an “X” in the testing well if it
detects an insufficient volume; this could be due to insufficient sample or
insufficient antisera and acts as an internal quality control system to
minimise the risk of false negative results.
The antisera used for the automated testing were all CE marked and
validated for use in automated techniques, with the exception of the anti-
N (immuClone) which was CE marked only for use in tube and slide tests.
101
This validation also intended to review the effectiveness of this antiserum
for use in the automated technique on the Galileo-NEO®.
2.3.4 Quality control testing
The NHSBT ten cell ID panel profile (0.8% cell suspension) in Cellstab
preservative (item code: PR143, NHSBT, UK) with known red cell
phenotypes (figure 2.5) was used to determine the accuracy of the
automated phenotyping profiles prior to commencing comparative work
with patient samples. The panel was tested on four occasions over four
weeks to determine any intra- and inter-test variation. The panel
contained a mixture of cells with homozygous and heterozygous
expression of red cell antigens. The NHSBT three cell screen profile
product (3% cell suspension) in Alsevers preservative (item code: PR122,
NHSBT) as shown in figure 2.6), selected to include all the relevant
antigens including homozygous and heterozygous expression of red cell
antigens, was tested over five separate occasions to challenge the
stability and accuracy of the antisera over an extended period. The
NHSBT titration cell set profile (0.8% cell suspension in Cellstab, item
code: PR407), including a two cell set with homozygous and
heterozygous expression of red cell antigens (figure 2.7), was tested on
two separate occasions.
The purpose of quality control for red cell antigen phenotyping is to detect
the weakest expression of the antigen, hence the requirement for
examples of heterozygous expression of antigens. Red cells that are
102
heterozygous express both the antigen and its antithetical counterpart, for
example Jk(a+b+). Whereas, homozygous expression is the presence of
one antigen but not the antithetical antigen, for example Jk(a+b-). There
are no clear recommendations from the BCSH (BCSH 2013) on the
requirements for quality control for phenotyping reagents, only that
suitable positive and negative controls should be included. In the
absence of any commercial quality control for phenotyping reagents, with
the exception of RhCcEe and K, the NHSBT cells were selected as they
included heterozygous and homozygous expression of the majority of
antigens and, thus, could provide both the positive and negative controls
required.
103
Figure 2.5: NHSBT ten cell ID panel profile
An example of an NHSBT ten cell ID panel. Panels are available in a variety of preservative media; the ID Panel in CellStab was used for quality
control of the automated extended phenotyping technique. Antigen profiles of the red cells in the panel are denoted by a positive (+) or negative (-)
in the table. The panel sheet supplied by the NHSBT (shown in this example) also allows for result entry in the blank section on the right hand side
and for documentation of a DAT result at the bottom of the table. This sheet can then be retained as a worksheet within the laboratory as part patient
result record. (Picture reproduced courtesy of NHSBT, taken from the reagent product profiles http://hospital.blood.co.uk/diagnostic-
services/reagents/product-profiles/).
104
Figure 2.6: NHSBT three cell screen profile product
An example of an NHSBT three cell screen profile set. The product is available in different preservative media suitable for use in a
variety of antibody screening techniques; the product in Alsevers (3% cell suspension) was used for quality control of the extended
automated phenotyping technique. (Picture reproduced courtesy of NHSBT, taken from the reagent product profiles
http://hospital.blood.co.uk/diagnostic-services/reagents/product-profiles/).
105
Figure 2.7: NHSBT titration cell set profile
Example of an NHSBT two cell titration set profile. This product is available for use in techniques designed to determine the strength of a red cell
antibody by titration methods. It was selected for use as a quality control product in the validation of the extended automated phenotyping technique
as it has a selection of red cells that are heterozygous for some of the red cell antigens which were used for determination of the performance of the
automated technique with red cells of a weaker expression than homozygous cells. (Picture reproduced courtesy of NHSBT, taken from the reagent
product profiles http://hospital.blood.co.uk/diagnostic-services/reagents/product-profiles/).
106
2.4 Blood Group genotyping
2.4.1 DNA Extraction
DNA was obtained from whole blood samples taken from 47 patients
attending the haematology clinic for routine blood samples and who
consented to have an additional sample taken for genetic analysis
(ethical approval for this was obtained from the NRES Committee South
West - Cornwall & Plymouth (MREC No. 12/SW/0251, R&D Study No.
1301733)). The BLOODChip IDCOREXT kit funded for this research
included 48 tests, 47 patient assays and one negative quality control
sample. The sample size for this study group was determined by the
funding obtained. It was not possible within the funding of the research to
include replicates to investigate the reproducibility of the results.
DNA was extracted from the whole blood samples in EDTA anticoagulant
(2.7ml) using the Autopure LS (Qiagen, Hilden, Germany), a fully
automated, high throughput analyser that uses PureGene chemistry
(Qiagen). This method gently removes contaminants and inhibitors to
enable purification of high molecular weight (100-200kb) DNA that can be
stored for analysis.
The analyser uses a red cell lysis solution (Qiagen, 15ml), part of the
Autopure LS whole blood extraction kit (used according to manufacturer’s
instructions) to remove the red cells from the whole blood sample and
then uses centrifugation (3000 x g for 5 min) to produce a pellet of white
107
blood cells. The white cells are then dispersed and lysed using a cell lysis
solution (Qiagen Autopure LS whole blood extraction kit used according
to manufacturer’s instructions, 5ml) and the proteins precipitated (2.2ml
Autopure protein precipitation solution, Qiagen) by mixing vigorously
(rapid tube inversions for 5 minutes). The supernatant containing the
DNA is then dispensed into tubes containing isopropanol (Autopure 100%
isopropanol 6.5ml). These tubes are rotated and centrifuged (rotated 50
times, centrifuged at 3000 x g for 5 min) to precipitate and pellet the DNA,
the isopropanol (entire volume) is then discarded and the DNA treated
with 70% ethanol (Autopure 70% ethanol, 6.5ml), centrifuged (3000 x g
for 5 min) and drained to evaporate any remaining alcohol. A DNA
hydration solution (Autopure DNA hydration solution, 300µl) is then
dispensed into the tubes to hydrate the DNA. Concentration and purity
were measured using the Nanodrop ND8000 spectrophotometer
(measured in ng/µl and ratios to indicate the purity of the samples). The
samples were then stored at -20°C until required for testing.
2.4.2 IDCOREXT test
2.4.2.1 IDCOREXT Workflow
The IDCOREXT protocol consists of four steps:
 Amplification
 Hybridisation/Labelling
 Data acquisition on Luminex
 Data analysis
108
2.4.2.2 DNA Amplification
DNA preparation was undertaken in a laminar flow fume hood (Astec –
Bioquell UK Ltd, Andover, UK). The IDCOREXT PCR master mix,
containing the allele specific primers (product ref: 1020120100, Progenika
Biopharma S.A), was brought to room temperature, vortexed for 10
seconds and then spun for 10 seconds in a micro-centrifuge (13000 xg)
before use. The HotStarTaq® DNA polymerase (Qiagen) was kept frozen
until ready for use and then immediately prior to use the tube was flicked
gently and spun down in a micro-centrifuge for 10 seconds. A PCR
reaction mix was prepared using the IDCOREXT PCR master mix and
the HotStarTaq® DNA polymerase in accordance with the quantities
(volumes in μL) shown in table 2.1.
Table 2.1: Reagent volumes required to produce the PCR mix using
the IDCOREXT
Reagent volumes required to produce the PCR mix (volumes in μL) using
the IDCOREXT PCR master mix and the HotStarTaq® DNA polymerase
(Qiagen). Volumes are given for sample batch sizes of 1, 8, 24 and 48.
Number of 1 8 24 48
samples
IDCOREXT 22.5 180 540 1080
PCR master
mix
HotStarTaq 0.5 4 12 24
DNA
Polymerase
(5 U/μL)
109
The 47 patient samples were run in three batches for ease of sample
preparation. A 96 well reaction plate (MicroAmp® Optical 96 well reaction
plate, Applied Biosystems®) (ThermoFisher Scientific, Cramlington, UK)
was used for the amplification phase. The PCR reaction mix was
vortexed and spun rapidly for 10 seconds in a micro-centrifuge, 20μL of
this mix was then dispensed into one well of the PCR reaction plate for
each sample. The frozen DNA samples were brought to room
temperature, vortexed and spun rapidly. DNA sample, 5μL,
(concentration >10ng/μL, A260/A280 ratio in the range 1.6-1.95, as
recommended by the package insert) was then added to the
corresponding wells and the resulting suspension mixed gently by
pipetting up and down.
The plate was then sealed with adhesive film (Microamp® Clear adhesive
film, Applied Biosystems®). The plate was then placed in the thermal
cycler block (Veriti® Applied Biosystems®) with a PCR compression pad
on the top. An amplification program was applied to the samples as
shown in table 2.2, which resulted in the production of the biotinylated
PCR product (Biotin labelled dCTP, included within the amplification kit).
110
Table 2.2 Amplification program details for IDCOREXT
Program details for the amplification stages; the denaturation, annealing
and extension phases are linked for 40 cycles and then the plates were
kept in the thermal cycler at 4°C until ready for the hybridisation phase.
Temperature Time Cycles

Polymerase 95°C 15 mins 1
Activation
Denaturation 95°C 30 sec 40
Annealing 60°C 30 sec 40
Extension 72°C 80 sec 40
Final extension 72°C 7 min 1
Hold 4°C ∞ 1
2.4.2.3 Hybridisation of amplified DNA to allele specific probes
The ID-COREXT Beads Master Mix (product ref: 1020120200, Progenika
Biopharma S.A.), a solution containing a mix of beads with allele specific
probes attached, was brought to room temperature and vortexed
vigorously for 30 seconds before use. The beads mix include probes
specific for the following alleles:
RHCE blood group: RHCE*ce; RHCE*Ce, RHCE*cE; RHCE*CE,
RHCE*CeCW, RHCE*ceCW, RHCE*CECW, RHCE*ceAR, RHCE*CeFV,
RHCE*CeVG , RHCE*cEFM, RHCE*ce[712G] , RHCE*ce[733G] ,
HCE*ce[733G,1006T] , RHCE*CE-D[2, 5,7]-CE,
RHCE*cE[697G,712G,733G], RHD*r’s-RHCE*ce[733G,1006T].
KELL blood group: KEL*K_KPB_JSB, KEL*k_KPB_JSB,
KEL*k_KPA_JSB, KEL*k_KPB_JSA.
111
Kidd blood group: JK*A, JK*B, JK*B_null(871C), JK*B_null(IVS5-1a).
Duffy blood group: FY*A; FY*B, FY*B_GATA, FY*B[265T]_FY*X
MNS blood group: GYPA*M, GYPA*N, GYPB*s, GYPB*S, GYPB*Mur,
GYPB*deletion, GYPB*S_, null(230T) , GYPB*S_null(IVS5+5t)
Diego blood group: DI*A, DI*B.
Dombrock blood group: DO*A, DO*B, DO*B_HY-, DO*A_JOA
Colton blood group: CO*A, CO*B
Cartwright blood group: YT*A, YT*B
Lutheran blood group: LU*A, LU*B
This process was repeated before dispensing for each sample to ensure
that the beads remained fully dispersed within the mix. ID-COREXT
Beads Master Mix (46µL) was dispensed into relevant wells of a 96 well
Thermowell P polycarbonate clear PCR plate (Corning, New York, USA).
The PCR plate containing amplified DNA from patient samples was
removed from the thermal cycler and 4μL of the PCR product was
dispensed into the corresponding well containing the hybridisation mix;
the resulting suspension was gently mixed by pipetting up and down.
The plate was then covered with a non-adhesive sealing film (Bio-Rad
MSA-5001 film) and placed in the thermal cycler with silicone pads. A
program of denaturation and hybridisation was then applied to allow
amplified DNA to attach to the specific alleles on the beads (table 2.3).
112
Table 2.3 Thermal cycler program for denaturation and hybridisation
PCR amplicons from patient samples were hybridised to beads with allele
specific probes was performed by denaturation at 95°C, followed by a
hybridization phase at 52°C, the samples were then held prior to
fluorescent labelling.
Temperature Time
Denaturation 95°C 2 min
Hybridisation 52°C 30min
Hold 52°C ∞
2.4.2.4 Fluorescent labelling of hybridised DNA
The SAPE (Streptavidin R-Phycoerythrin Conjugate) mixture containing
the fluorescent marker (product ref: 000020100, Progenika Biopharma
S.A.) and SAPE Dilution Buffer (product ref: 000020200, Progenika
Biopharma S.A.) were brought to room temperature and the SAPE
vortexed and spun rapidly. A labelling mix was then prepared using the
SAPE Dilution Buffer and the SAPE reagent (volumes μL) as shown in
table 2.4. The labelling mix was then vortexed and protected from the
light by covering with aluminium foil.
113
Table 2.4 Fluorescent labelling mix
The hybridised DNA attached to the allele specific beads was then
labelled with a fluorescent mix for detection and quantification by the
Luminex analyser. The labelling mix using the SAPE Dilution Buffer and
the SAPE reagent (volumes μL) for different batch sizes of patient
samples are shown in the table below.
Number of 1 8 24 48
samples
SAPE 87 696 2088 4176
Dilution
Buffer
SAPE 4.6 36.8 110.4 220.8
At the 52°C hold step in the thermal cycler program (shown in table 2-3)
the lid was opened, and the silicone pads and sealing film carefully
removed. Labelling mix (80μL) was then dispensed into each well of the
hybridisation plate and gently mixed by pipetting up and down. The
samples were then immediately analysed on the Luminex X-Map®
system. Fluorescent labelling is stated here in volumes rather than final
concentrations, as stated in the IDCOREXT package insert, and in
accordance with confidential, proprietary information (kit product ref:
1020120034, Progenika Biopharma S.A.).
114
2.4.3 Luminex xMAP Analyser
2.4.3.1 Luminex® X-Map system analysis
The Luminex Gold heater block was placed on the plate holder and the
temperature was set to 52°C to correlate with the hold temperature of the
thermal cycler as rapid cooling of the samples could result in non-specific
hybridisation (it is a recommendation of the Luminex assay process that
the temperature is maintained through the process). This was done one
hour in advance of the labelling phase to enable the heater block to reach
the correct temperature.
Prior to analysis, the Luminex analyser was calibrated using the following
reagents (used as per manufacturer’s instructions, Luminex®
xPONENT® 3.1 Quick Guide, Luminex Corporation)
xMAP Classification Calibrator Microspheres (CAL 1), polystyrene
microspheres internally labelled with fluorescent dyes and intended to
normalise the kittings for the classification channels (CL1 and CL2).
xMAP Reporter Calibrator Microspheres (CAL 2), polystyrene
normalise the analyser reporter channel (RP1)
xMAP Classification Control Microspheres (CON1), polystyrene
115
verify the calibration and optical integrity for the classification channels
and the doublet discriminator channel.
xMAP Reporter Control Microspheres (CON2), polystyrene microspheres
internally labelled with fluorescent dyes and intended to verify the
calibration and optical integrity for the reporter channel.
Batches for the sample runs were created on the Luminex® xPONENT®
3.1 software (in accordance with the manufacturer’s instructions) using
the Create a New Batch from an existing Protocol option in the
batches tab. The corresponding protocol (TLMX_ID-COREXT) was then
selected. The batch run was named and the sample identification
numbers were manually input into the corresponding positions on the
plate layout.
A negative control was run with the sample batch, as per manufacturer’s
recommendation. Due to the limited funding for the research it was not
possible to include a positive control as well. The negative control was
performed by setting up a test reaction with molecular biology grade
water known to be free of any DNA contamination.
The Eject icon was clicked to eject the plate holder, and the hybridisation
plate was placed into the plate holder. The Retract icon was then
selected to place the plate holder into the analyser ready for analysis.
The analysis process was then started, once the batch analysis had been
116
completed the data was exported as a comma separated values (csv) file
and saved into a folder labelled with the batch name.
The IDCOREXT analysis software was then used to convert the raw data
into a report including the predicted phenotype. The analyser raw data
includes internal quality control designed to identify any invalid results
and prevent reporting of potentially incorrect genotypes and phenotypes.
A “green” alert denotes a valid analysis for genotype and predicted
phenotype. Invalid results are denoted by the following alerts:
 Indeterminate genotype – the software is not able to assign a
result for one or more polymorphisms.
 Low signal intensity – the signal intensity does not fulfil the
acceptable threshold for one or more polymorphisms.
 Low bead count – the bead count does not reach the acceptable
threshold for one or more polymorphisms.
 Unknown genotype(s) and predicted phenotype(s) – a particular
combination of SNPs is not included in the software. The genotype
has not been previously described as associated to a phenotype.
In this circumstance an alternative method should be used to
confirm the genotype and serology used to predict the phenotype.
 Invalid test – the sample analysis result is not acceptable.
 Invalid reading – the raw data cannot be analysed by the software.
In the event of invalid results, a report will not be generated and the
sample(s) must be re-tested. The raw data obtained from the analysis
can be seen in appendix 4.
117
2.5 Randomised Controlled Trial
A total number of 60 patients were recruited to the pilot study, selected
from patients attending the haematology out-patient clinic. The number of
participants was calculated to provide 80% power to detect an
acceptance rate of 85% or lower if the true acceptance rate in the
population is 95% using a one-sided binomial test of size α = 0.05 (as
determined with Dr Paul White (University of the West of England) and in
accordance with Moore et al., 2011).
The eligibility criteria for recruitment was that the patient was an adult
capable of giving informed consent for inclusion into the trial, in
accordance with the Mental Capacity Act 2005, and that there was no
record of historic red cell antibodies on the patient’s electronic records.
The pilot study was approved by the Cornwall and Plymouth Research
Ethics Committee on the 5th September 2012 (MREC No.12/SW/0251
R&D Study No. 13017733) and patient information leaflets were provided
for all patients approached for recruitment into the trial. Recruitment
commenced on the 9th November 2012 and the required number of
recruits was reached by the 16th July 2014. Patients were randomly
assigned to one of two groups; the standard care group, who had a
standard blood group performed for ABO and RhD type and would
receive blood matched only for ABO and RhD groups or to the
intervention group, who, in addition to the standard care, also had their
blood typed for Rh (CcEe) and K antigens and were flagged on the
pathology computer system to receive blood matched for the additional
118
antigen types. Provision was made for the intervention group patients to
receive standard blood for transfusion should there be an emergency
situation that did not allow for additional time for the acquisition or
matching of blood with the extra antigens. Laboratory staff was requested
to create a comment on the pathology computer system (visible to
laboratory staff only) in the event that matched blood could not be
provided from the routine blood stocks in the time frame required by the
clinical team attending the patient.
Patients were assigned to the standard care or intervention group using a
modified block randomisation technique to ensure that the groups were
comparable in size and the total number of patients recruited amounted
to 60. The two groups were allocated a number (1 for the standard care
and 2 for the intervention); each number was registered on 30 separate
pieces of paper which were then sealed into 60 individual envelopes.
Upon recruitment of each patient an envelope was selected at random by
a healthcare member of staff not related to the research and the
researcher was then informed of the allocated group of the patient. If the
patient was allocated to the intervention group the researcher then
requested that the patient blood group sample was typed for Rh (CcEe)
and K antigens, in addition to the standard ABO and RhD. A clinical note
was then added to the patient electronic record on the pathology
computer system stating the patient Rh (CcEe) and K antigen type and
the antigen requirements of blood that would be required in the event of a
transfusion.
119
The patients remained in the trial until it was closed on the 31st July 2015,
at this point data were collected for each patient which included:
 Number of patients transfused in each study group
 Recruitment rates
 Underlying disease condition of the patient
 Number of units of blood transfused during the trial period
 Number of transfusion episodes
 Provision of matched red cell units for patients in the intervention
group
 Development of any red cell allo/autoantibodies
 Length of time that the patient had been included in the trial.
 Rh (CcEe) and K antigen type of the patient
 Frequencies of Rh phenotypes
This data was collated on an excel file and subjected to statistical
analysis in order to investigate any significant differences between the
number of units of blood transfused during the trial period, the number of
transfusion episodes, development of any red cell allo/autoantibodies and
the length of time that the patient had been included in the trial.
Ethnicity of the patients was not included in the data collection, primarily
because this data is not recorded in the pathology computer system as
part of the patient demographics. It was not felt necessary to request this
information at recruitment as part of the purpose of the pilot study was to
review the availability of matched blood from routine stocks held in the
120
transfusion laboratory. The region served by the RD&E does not have an
ethnically diverse population; the Exeter census profile (2011) reported
80.49% of the population to be white British. The results of the pilot study
could then be used to review the ease of providing matched blood from
the current stocks, and to direct any changes that may be required to
blood ordering and stock storage.
2.6 STATISTICAL ANALYSIS
2.6.1 Retrospective review of haematology and renal patients
Patient information obtained from the retrospective review of
haematology and renal transfusions, excluding any personal identifiers,
was input into an excel spreadsheet and analysed using SPSS version 19
(see appendices 1 and 2). Fisher’s Exact Test was used to examine
associations between allo/autoimmunisation and categorical factors
(gender, immunosuppression) and the independent samples t-test was
used to examine any associations on numeric scale variables (age at first
transfusion, number of units transfused, number of transfusion episodes).
Correspondence analysis was used to investigate any associations
between red cell antibody type (alloantibody only, allo- and autoantibody
in combination or autoantibody only) with disease conditions within the
haematology cohort. Correspondence analysis is a multivariate graphical
statistical technique designed to depict associations in cross-tabulation
121
tables (Hirschfeld, 1935). The process is a decomposition of the classic
chi-squared test of association into two orthogonal components which are
plotted against one another; points which are close on the plot show
positive associations. The process is a non-parametric exploratory
depiction of potential associations and is not based on the ideas of
statistical significance. The technique allows an exploration of
relationships between multiple categorical variables at the same time,
removing the need to conduct separate chi-square tests for each pair of
variables (Sourial et al, 2010).
2.6.2 Comparison of the manual and automated red cell
phenotyping results
Red cell antigen typing results obtained using serological automated
testing were directly compared to those obtained by current manual
techniques. Red cell antigen results obtained by serological automated
testing using the NHSBT commercial red cell panels were compared to
the known antigen status as recorded on the antigen profile sheets
supplied with the panels (see figures 2.5 – 2.7).
2.6.3 Comparison of the serological phenotyping and genotyping
results
Red cell antigen types obtained using the automated serological
phenotyping techniques were directly compared to the antigen type
122
predicted by the genotyping technique, with result types being identified
as positive or negative.
2.6.4 Review of the pilot study – randomised controlled trial
Patient information obtained from the pilot study involving haematology
patients, excluding any personal identifiers, was input into an excel
spreadsheet and analysed using SPSS version 19. Fisher’s Exact Test
was used to examine any differences in the number of patients in the
standard care group that received a transfusion compared to the number
in the intervention group. The t-test was used to investigate any
significant differences between the two groups with regard to numeric
scale variables, time spent in trial, number of units transfused and
number of transfusion episodes.
123
3 RETROSPECTIVE REVIEW RESULTS
3.1 Introduction
This study reports on a retrospective review of over 1000 patients
admitted for at least one blood transfusion into the haematology inpatient
ward and outpatient day case ward, and over 800 patients transfused as
a result of renal insufficiency in a National Health Service (NHS)
Foundation Trust hospital. The data collected for each patient included
the total number of units of blood transfused, number of transfusion
episodes, age at first documented transfusion, gender, diagnosis,
intensity of treatment regime and extent of immunosuppression. A
tabulated summary of the raw data obtained for the retrospective review
can be seen in appendix 1 (haematology cohort) and appendix 2 (renal
cohort).
Concomitant platelet therapy and the requirement for irradiated blood
components were used as surrogate markers for severe
immunosuppression in the haematology patients, as historical studies
have suggested this may result in a tolerance to alloimmunisation
(Schonewille et al., 1999; Holohan et al., 1981; Dutcher et al., 1981;
Asfour et al., 2004; Ramsey et al., 1989). Renal transplant was used as a
marker for severe immunosuppression in patients transfused on the renal
wards. Renal transplant patients are treated with life-long or long-term
immunosuppressive drugs aimed at reducing the risk of acute graft
rejection, often using multiple immunosuppressive agents (Lee and
Gabardi, 2012). Immune suppression is achieved either by depleting the
124
number of lymphocytes, diverting the lymphocyte traffic or by blocking the
lymphocyte response pathway (Hallorhan, 2004). Patients receive
induction therapy of high dose immunosuppressives at the time of the
transplant to prevent acute rejection (Norman and Turka, 2001;
Hallorhan, 2004; Nashan, 2005; Terasaki and Mizutani, 2006; Cornell et
al., 2008; Gabardi et al., 2011) followed by a lower maintenance dose
using a combination of drugs (Gaston, 2001; Zhang, 2013). Suppression
of the immune system in renal transplant to reduce the risk of production
of HLA alloantibodies could also, in theory, reduce the risk of production
of red cell antibodies in transfused renal transplant patients.
The aim of this review was to identify potential risk factors and/or
protective factors related to the production of allo and/or autoantibodies
within these groups of chronically transfused patients. This information
could then be used to identify an allo/autoantibody immunisation risk
group that would benefit from extended red cell phenotyping and
matching of blood for transfusion. Risk factors for alloimmunisation, such
as the number of red cell transfusions (Zalpuri et al., 2012) and the
gender of the recipient (Hoeltge et al., 1995) have already been
suggested and, in addition, it is also known that patients who develop one
alloantibody are more likely to produce additional antibodies following
subsequent red cell transfusion (Schonewille et al., 1999; Fluit et al.,
1990). The identification of a high risk group destined for chronic red cell
transfusion therapy would help to target the implementation of a strategy
for pre-transfusion extended red cell antigen phenotyping and matching
125
of blood for transfusion in this group of patients. Such a strategy could
have the potential to reduce the risk of alloimmunisation, in turn reducing
the burden of additional testing for the hospital transfusion service and
the risk of transfusion reactions in the patients.
3.2 Patients with haematological malignancies
A total of 1107 records of patients admitted to the haematology wards for
transfusion were reviewed, 1006 patients for anaemia secondary to
haematological disorders, 54 with solid organ cancer and 47 for anaemia
secondary to other conditions (see table 3-1). The 1107 patients
received a total of 31,890 units of red cells during 14,013 clinical
episodes, the average number of units transfused was 29 (SD 42.1). The
average number of transfusion episodes was 13 (SD 17.1) and the age of
the patients at first recorded transfusion ranged from 13 to 99 with an
average of 72 years (SD 14.7).
Of the 1107 patients, 256 (23%) either presented with, or developed, allo
and/or autoantibodies to red cell antigens. Of the original 1107 patients
sixty patients (5.4%) presented with allo and/or autoantibodies at the first
recorded test, of these, 21 (1.9%) presented with an identifiable red cell
alloantibody only. Eleven (18.3%) of the 60 patients who presented with
red cell allo and/or autoantibodies developed additional allo and/or
autoantibodies following transfusion.
126
A total of 1086 patients presented with no detectable red cell
alloantibodies, 140 (12.9%) developed one or more alloantibodies post
transfusion. The average number of units given prior to the production of
allo/autoantibodies was 15.46 (maximum =270, minimum = 2, average =
5.5, SD = 6.35).
3.2.1 Antibodies and disease conditions
Allo and/or autoantibodies were most commonly seen in patients with
haemolytic anaemia (78%), aplastic anaemia (57%), Waldenström
macroglobulinaemia (42%) and chronic lymphocytic leukaemia (39%).
Within the disease groups chronic myeloid / monocytic leukaemia
(CMML), myelodysplastic syndrome (MDS) and myeloproliferative
disorder (MPD), 30% of patients were found to have allo/autoantibodies.
In addition 30% of patients transfused for less common haematological
conditions (shown as “other” in table 3-1), such as amyloidosis,
sideroblastic anaemia and essential thrombocythaemia, were found to
have allo/autoantibodies. Allo/autoantibodies were less commonly seen
in patients with acute lymphocytic leukaemia (0%), Hodgkin’s lymphoma
(7%) and non-Hodgkin’s Lymphoma (16%). Table 3-1 shows the number
of patients with allo and/or autoantibodies in relation to disease condition.
It is important to note that in some of the groups the percentages quoted
may be high, or low, because the patient numbers are small.
127
Table 3.1: Allo/autoantibodies and disease conditions
A retrospective review of patients transfused within the haematology cohort (n=1107)
was performed to investigate any associations between the risk of development of
allo/autoantibodies and the disease status of the patient. This table shows the total
number of patients reviewed in each disease group, along with the number of patients in
that group found to have allo/autoantibodies and the relative percentage.
Disease Total Number of patients % with

number of with allo/autoantibodies
patients allo/autoantibodies
Acute lymphocytic 14 0 0
leukaemia
Acute myeloid leukaemia 145 35 24
Anaemia 26 7 27
Aplastic anaemia 14 9 57
Chronic 20 6 30
myeloid/monocytic
Leukaemia
Chronic lymphocytic 57 22 39
leukaemia
Haemolytic anaemia 9 7 78
Hodgkin’s lymphoma 29 2 7
Lymphoma 61 13 21
Myelodysplastic 151 45 30
syndrome
Myeloma 187 33 18
Myeloproliferative 23 7 30
disorders
Non-Hodgkin’s 258 41 16
lymphoma
Solid organ cancer 54 10 19
Other 47 14 30
Waldenström 12 5 42
macroglobulinaemia
128
Further analysis of the incidence of alloantibodies only, alloantibodies in
combination with autoantibodies, and autoantibodies only, was performed
in some of the disease conditions (Table 3-2). Correspondence analysis
was performed to investigate any associations between disease status
and antibody type (Figure 3.1). This demonstrated that haemolytic
anaemia, MDS, and MPD show an association with allo and autoantibody
in combination, whereas AML, CLL and aplastic anaemia are associated
with autoantibody production only. Four (57.1%) of the seven immunised
patients with haemolytic anaemia had alloantibodies in combination with
autoantibodies. One patient with a non-autoimmune haemolytic anaemia
developed anti-Fya and anti-Chido-Rogers. Twelve (54.5%) of the 22
immunised patients with CLL had autoantibodies only, 9 (40.9%)
developed autoantibodies in combination with alloantibodies and one
patient developed an alloantibody only. Twenty six (57.8%) of the
immunised patients with MDS had alloantibodies in combination with
autoantibodies, 11 (24.4%) had alloantibodies only and eight (17.8%) had
autoantibodies only.
129
Table 3.2: Red cell antibody type in relation to disease
Analysis of red cell antibody type in relation to disease within the immunised patient group (n=256) in the haematology
patient cohort (n=1107) was performed for a selection of disease groups. Antibody types are grouped into alloantibody only,
allo and autoantibody in combination and autoantibody only. The total number of immunised patients in each disease group
is represented, along with the total numbers of patients within each disease group with certain antibody types and the
representative percentages.
Total number of
Disease Alloantibody only Allo and autoantibody Autoantibody only
immunised patients
MDS 45 11 (24.4%) 26 (57.8%) 8 (17.8%)
AML 35 10 (28.6%) 11 (31.4%) 14 (40.0%)
Myeloma 33 13 (39.4%) 13 (39.4%) 7 (21.2%)
CLL 22 1 (4.5%) 9 (40.9%) 12 (54.5%)
Lymphoma 13 4 (30.8%) 5 (38.5%) 4 (30.8%)
Solid organ cancer 10 7 (70.0%) 3 (30.3%) 0 (0.0%)
Aplastic anaemia 9 2 (22.2%) 3 (33.3%) 4 (44.4%)
Haemolytic anaemia 7 1 (14.3%) 4 (57.1%) 2 (28.6%)
Myeloproliferative
7 1 (14.3%) 4 (57.1%) 2 (28.6%)
disorders
CML/CMML 6 0 (0.0%) 3 (50.0%) 3 (50.0%)
Waldenström
5 2 (40.0%) 1 (20.0%) 2 (40.0%)
macroglobulinaemia
130
Figure 3.1: Correspondence analysis of the association between
disease types and antibody types in the Haematology cohort
Correspondence analysis showing clusters of disease types associated

with antibody type in the immunised patients within the haematology
cohort (n=256). The blue squares represent the antibody type
(alloantibody only, autoantibody only and allo and autoantibody in
combination) and the red circles represent the disease type. The
components (1 and 2) and the X and Y axes are derived from a process
known as eigenvalue decomposition of the chi-square test of association.
The “map” produced by this process will demonstrate any relationships
between the variables by the proximity of the plotted points; haemolytic
anaemia, MDS, and MPD show an association with allo and autoantibody
in combination, whereas AML, CLL and aplastic anaemia are associated
with autoantibody production only. Analysis performed by Dr Paul White,
University of the West of England.
131
3.2.2 Antibody Specificities
Of the 256 red cell antigen immunised patients, 70 (27.3%) developed
alloantibodies only, 101 (39.4%) developed alloantibodies in combination
with autoantibodies and 85 (33.2%) developed autoantibodies only. The
incidence of antibody specificities is presented in figure 3.2. A total of 242
alloantibodies were detected, of these 71 (29.3%) could not be identified.
Of the 171 identifiable alloantibodies 114 (66.7%) were to Rh (D, C, c, E,
e) or Kell antigens. The most common specificity was anti-E (21.5%),
followed by anti-K (11.6%) and anti-C (7.4%). Two hundred and four
autoantibodies were detected, 18 (8.8%) had definable specificities within
this group auto-anti-e was the most common (61%). Autoantibody
specificity was either confirmed at the reference laboratory (NHSBT) (11
patients) or assigned in the transfusion laboratory based on antibody
identification and phenotype results (7 patients). Antibodies to variant
antigens were not excluded for any of these patients. Of the 256 red cell
antigen immunised patients, 101 (39.2%) had alloantibodies in
combination with autoantibodies. Of these 101 patients, 80 (79.2%)
developed the autoantibody after the alloantibody, 8 (7.9%) developed
the alloantibody after the autoantibody, 9 (8.9%) presented with
alloantibody and autoantibody in combination, and 4 (3.9%) developed
allo and autoantibody in combination following transfusion. Analysis of the
specificities of patients with identifiable alloantibodies revealed 60
patients with single specificity (61.6%), 28 with two alloantibodies (25%),
11 with three antibodies (9.8%), three with four antibody specificities
132
(2.7%) and one patient who developed five separate alloantibodies
(0.9%).
Patients (n=186) who presented with, or developed autoantibodies had
records detailing the type of antibody or complement fraction on the
surface of their red cells. Within this group 140/186 (75.3%) were found to
have IgG only, 17/186 (9.1%) were found to have C3d only and 29/186
(15.6%) were found to have a combination of IgG and C3d.
Autoantibodies of a transient nature could be demonstrated in 117 of the
186 patients (62.9%).
133
Figure 3.2: Incidence of antibody specificities within the
haematology cohort
The incidence and specificity of allo- and autoantibodies within the immunised
group of the haematology patient cohort (n=256). The specificity of the antibody
is shown on the Y-axis, alloantibodies are denoted by the antigen specificity and
autoantibodies are denoted separately. Autoantibodies showing no specificity
were most commonly seen in this cohort of patients (n=186/256), within the
group of autoantibodies demonstrating a red cell antigen specificity, auto anti-e
was the most common (n=11/18). Alloantibodies to antigens within the Rh and
Kell blood group systems were the most common (n=114/171). (AHG = Anti-
human globulin).
134
3.2.3 Antibodies and patient factors
3.2.3.1 Gender and immunisation rates
Of the 1107 patient records reviewed in the haematology cohort, 532
were female and 575 were male. Analysis (Figure 3.3a) of the
allo/autoantibody immunisation rate in males (122/575, 21.22%) and
females (134/532, 25.19%) showed no statistical significant difference
(Fisher’s Exact Test P=0.1340). In addition, no significant difference was
seen in the rates in males (81/575, 14.10%) and females (90/532,
16.92%) with alloantibodies only (Fisher’s Exact Test P=0.1310) (Figure
3.3b).
135
Figure 3.3: Immunisation rates in males and females within the Haematology cohort
A B
700 700
600 600
Number of patients
Number of patients
500 500
400 Patients with 400
allo/autoimmunisation
300 300 Alloimmunised
200 Patients without 200 Non immunised
100 allo/auto immunisation 100
0 0
Female Male Female Male
Gender Gender
Graphs showing the allo/autoimmunisation rates in males and females in the haematology cohort (n=1107). The number of patients who did not
demonstrate allo/autoimmunisation is denoted by the blue bar and the number with red cell allo/autoimmunisation is denoted by the red bar.
Statistical analysis of the allo/autoantibody immunisation rate in males (122/575= 21.22%) and females (134/532 =25.19%) showed no significant
difference (Fisher’s Exact Test P=0.1340) Figure A. In addition there was no significant difference seen in the rates in males (81/575=14.10%) and
females (90/532=16.92%) with alloantibodies only (Fisher’s Exact Test P=0.1310) Figure B.
136
3.2.3.2 Age and immunisation rates
The age of patients receiving blood transfusions in the haematology
cohort ranged from 13 to 99 years (n=1107). The mean age at the first
recorded transfusion for patients with allo/autoimmunisation was 69.18
(range 13-95 years, standard deviation (SD) 14.62) and the mean age for
those patients without allo/autoantibodies was 69 years (range 18-99, SD
14.73). There was no statistically significant difference between the mean
age at first recorded transfusion and the incidence of allo/autoantibody
immunisation (t-test P=0.8637). Further review of patients to investigate
any relationship between age at first recorded transfusion and the
development of alloantibodies only revealed a mean age at first recorded
transfusion of 69.82 (SD 14.22) for those with alloimmunisation and a
mean age of 69 (SD 14.73) for the non-immunised patients (Figures 3.4A
and 3.4B respectively). Again, no statistical significance was
demonstrated (t-test P= 0.5042)
137
Figure 3.4: Influence of age at first transfusion and allo/autoantibody development within the haematology cohort
A B
120 120
100 100
Age in years at first recorded
Age in years at first recorded

80 80
transfusion
transfusion
60 60
40 40
20 20
0 0
Immunised Non Immunised Alloantibody only Non Immunised
Box-and-whiskers plots showing the minimum and maximum age at first transfusion (black line), the first quartile, median (second quartile) and third quartile. This
figure shows the age at first recorded transfusion in patients in the haematology cohort (n=1107) with or without allo/autoantibodies (figure A) and with or without
alloantibodies only (figure B). There was no significant difference between the mean age at first recorded transfusion and the incidence of allo/autoantibody
immunization (mean 69.18 years, SD14.62) compared to the non-immunised group (mean = 69 years, SD14.73) (t-test P=0.8637); Figure A. Figure B shows the
mean age at first recorded transfusion in patients with alloantibodies only (mean 69.82, SD 14.22) and those non-immunised (mean = 69, SD 14.73), no significant
difference was seen (t-test P= 0.5042).
138
3.2.3.3 Immune suppression and incidence of immunisation
Concomitant platelet support and the requirement for irradiated blood
were used as surrogate markers for immune suppression in order to
investigate any relationship between immune suppression and the
incidence of immunisation to red cell antigens. In the haematology cohort
(n=1107) 534 patients were recorded as having received concomitant
platelet therapy, of these 123 (23.0%) patients had a recorded allo/auto
antibody. The incidence of allo/autoantibodies in patients not receiving
any platelet support (n=573) was 23.2%. Therefore, concomitant platelet
therapy appeared to have no influence on the rate of allo/autoantibody
formation (Fisher’s Exact Test P=1.00). Further analysis of patients with
alloantibodies only revealed that 79 (14.8%) of the 534 patients receiving
platelets were recorded as having alloantibodies and 92 (16.0%) had no
alloantibodies, also showing no statistically significant difference
(P=0.6750) (figures 3.5a and b respectively).
139
Figure 3.5: Influence of immune suppression (platelet requirement) on allo/autoantibody development in the
haematology cohort
A B
700 700
600 600
500
Number of patients
500
Number of patients
400 400
Allo/auto immunised alloimmunised
300 300
Non immunised non immunised
200 200
100 100
0 0
No Yes No Yes
Platelet Therapy Platelet Therapy
Concomitant platelet therapy was used as an indicator of immune suppression. In the haematology cohort (n=1107) 534 patients were recorded as
having received concomitant platelet therapy, of these 123 (23.0%) patients had a recorded allo/auto antibody. The incidence of allo/autoantibodies
in patients not receiving any platelet support (n=573) was 23.2%. Therefore, concomitant platelet therapy had no influence on the rate of
allo/autoantibody formation (Fisher’s Exact Test P=1.00) or on the rate of alloimmunisation only (P=0.6750) (Figures A and B respectively).
140
A total of 362 patients in the haematology cohort were given irradiated
blood components, 91/362 (25.1%) of these patients developed an
allo/autoantibody and 165/745 (22.2%) of patients not requiring irradiated
blood components had allo/autoantibodies. Analysis of patients with
alloantibodies only demonstrated an incidence 14.6% (53/362) in patients
receiving irradiated components and 15.6% (116/745) in patients not
receiving irradiated components. Therefore the requirement for irradiated
blood components appeared to have no influence on the rate of
allo/autoantibody formation (Fisher’s Exact Test P=0.2876) or on the rate
of alloimmunisation (P= 0.6750) (Figures 3.6a and b).
141
Figure 3.6: Influence of immunosuppression (irradiated requirement) on allo/autoantibody development in the
haematology cohort
A B
800 800
700 700
600 600
Number of patients
Number of patients
500 500
400 Allo/autoimmunised 400
Alloimmunised
300 Non immunised 300
Non immunised
200 200
100 100
0 0
No Yes No Yes
Irradiated requirement Irradiated requirement
The allo/autoimmunisation rate in patients with a requirement for irradiated blood components was 25.1% (91/362) compared to 22.2% (165/745)
not requiring irradiated blood components (figure A). Analysis of patients with alloantibodies only demonstrated an incidence 14.6% (53/362) of
alloantibodies in patients receiving irradiated components and 15.6% (116/745) in patients not receiving irradiated components (Figures B).
Therefore the requirement for irradiated blood components appeared to have no influence on the rate of allo/autoantibody formation (Fisher’s Exact
Test P=0.2876) or on the rate of alloimmunisation (P= 0.6750).
142
3.2.3.4 Number of transfusions and incidence of immunisation
In the haematology cohort (n=1107) the average number of transfusion
episodes recorded over the review period for all patients was 13 (range 1-
155) and the average number of units given to the patients was 29 (range
1-382).
In the immunised group (n=256) the average number of transfusion
episodes was 20.3 (SD 23.68), compared to an average of 10.36 (SD
13.33) episodes in the non-immunised group (n=851), demonstrating a
statistically significant difference in the number of transfusion episodes
and the incidence of allo/autoimmunisation (t-test P=<0.0001) (figure
3.7a). Further analysis of immunised patients showed a significant
difference in the number of transfusion episodes between patients with
alloantibodies only (average 21.57, SD 24.98) and those with allo and
autoantibodies (average 45.95, SD 57.45 (t-test P= 0.0007), as shown in
figure 3.7b. In addition, a significant difference was also demonstrated in
the number of transfusion episodes in patients with allo +/- autoantibodies
(average 23.04, SD 24.93) compared to those with autoantibodies only
(average 13.37, SD 16.95) (t-test P= 0.0014), as shown in figure 3.7c.
143
Figure 3.7: Influence of the number of transfusion episodes on the development of all/autoantibodies in the
haematology cohort
A B C
180 180 450

Number of transfusion episodes

160 400

160
140 140 350
120 120 300
100 100 250
80 80 200
60 60 150
40 40 100
20 20 50
0 0 0
Immunised Non Immunised Alloantibody only Allo autoantibody autoantibody onlyAllo/autoantibody
Box-and-whiskers plots showing the minimum and maximum numbers (black line) of transfusion episodes (figure A) (the first quartile, median
(second quartile) and third quartile are denoted by the coloured blocks) in immunised and non-immunised patients in the haematology cohort
(n=1107). There was a statistically significant difference (t-test P=<0.0001) in the number of transfusion episodes in the immunised group (n=256,
average number of episodes 20.3, SD 23.68) compared to the non-immunised group (n=851 average10.36, SD 13.33) (figure A). A significant
difference (t-test P= 0.0007), was also seen in the number of transfusion episodes in patients with alloantibodies only (average 21.57, SD 24.98)
compared to those with allo and autoantibodies (average 45.95, SD 57.45) as shown in figure B and also between those with allo+/-autoantibody
(average 23.04, SD 24.93) and those with an autoantibody only (average 13.37, SD 16.95) (t-test P= 0.0014) as shown in figure C.
144
Analysis of the number of red cell units transfused within the haematology
cohort (n=1107) during the review period revealed that the number of
units transfused ranged from 1 to 382, with an average of 29 (SD 42.1).
The average number of red cell units given to patients in the immunised
group (n=256) was 46.99 (SD 59.65) compared to 23.43 (SD 32.04) in
the non-immunised group (n=851), demonstrating a statistically significant
difference (t-test P=<0.0001), as shown in figure 3.8a. This difference
was also seen in the number of units given in patients with allo +/-
autoantibodies (n=171, average number of units given 54.42, SD 65.40)
compared to those with autoantibodies only (n=85, average number of
units given 32.72, SD 42.82) (t-test P= 0.0059), as shown in Figure 3.8b.
This demonstrates that the risk of developing an allo +/- autoantibody
increases as the number of transfusion episodes and number of red cell
units given increases. However, the incidence of autoantibodies alone
does not appear to be influenced by the number of transfusion episodes
or number of units given.
145
Figure 3.8: Influence of the number of red cell units transfused on the development of allo/autoantibodies in the
haematology cohort
A B
450 450
400 400
Number of units transfused

350 350
300 300
250 250
200 200
150 150
100 100
50 50
0 0
Immunised Non Immunised Allo Autoantibodies Autoantibodies only
Box-and-whiskers plots showing the minimum and maximum numbers (black line) of red cell units transfused and numbers of red cell units transfused (the first
quartile, median (second quartile) and third quartile are denoted by the coloured blocks) in immunised and non-immunised patients in the haematology cohort
(n=1107). The average number of red cell units given to patients in the immunised group (n=256) was 46.99 (SD59.65) compared to 23.43 (SD 32.04) in the non-
immunised group (n=851), demonstrating a statistically significant difference (t-test P=<0.0001) (figure A). The number of units given in patients with allo +/-
autoantibodies (n=171, average number of units given 54.42, SD 65.40) compared to those with autoantibodies only (n=85, average number of units given 32.72, SD
42.82) (t-test P= 0.0059), as shown in Figure B.
146
3.2.3.5 Time between transfusions and the incidence of
immunisation
Analysis was performed on the time (in days) between transfusion
episodes for patients within the haematology cohort (n=1107). The time
between transfusion episodes for all patients ranged from 0 to 2882 days,
with an average time of 107.39 (SD 220.42) days. The average time
between transfusion episodes for the immunised patients was 113.3 (SD
192.42) days (n=256, range 0 to 1629 days) compared to an average of
102.5 (SD 228.22) days for non-immunised patients (n=851, range 0 to
2882.5 days). There was no significant difference seen in the average
time between transfusion episodes in the two groups (t-test P=0.4925), as
shown in figure 3.9. This suggests that the frequency at which the
recipients are exposed to foreign red cell antigens does not influence the
development of allo/autoantibodies.
147
Figure 3.9: Influence of the frequency of transfusion episodes on the development of allo/autoantibodies in the
haematology cohort
3000
Time between transfusion episodes (days)

2500
2000
1500
1000
500
0
Immunised Non Immunised
Box-and-whiskers plot showing the minimum and maximum (black line) number of days between transfusion episodes (the first quartile, median
(second quartile) and third quartile are denoted by the coloured blocks) in the haematology cohort. The average time between transfusion episodes
for the immunised patients was 113.3 (SD 192.42) days (n=256, range 0 to 1629 days) compared to an average of 102.5 (SD 228.22) days for non-
immunised patients (n=851, range 0 to 2882.5 days). There was no significant difference seen in the average time between transfusion episodes in
the two groups (t-test P=0.4925).
148
3.2.4 Transfusion Reactions
Of the 1107 patient records reviewed within the haematology cohort, 24
(2.17%) were investigated for a suspected transfusion reaction, giving an
overall transfusion reaction incidence of 1 in 1328.75 units. In the
immunised group the incidence of investigation of transfusion reactions
was 11/256 (4.3%) compared to an incidence of 13/851 (1.5%) in the
non-immunised group. An approximate three fold difference was seen in
the rate of suspected transfusion reactions in the allo/auto immunised
group compared to the non-immunised group (see table 3-3). Analysis
using the Fisher’s Exact Test shows that the difference between the
groups is statistically significant (P=0.0128).
Patients included within the immunised group for review of suspected
transfusion reactions were those with known red cell allo/autoantibodies
being transfused with antigen negative and/or crossmatch compatible
blood, not patients who had suffered a transfusion reaction as a result of
development of antibodies subsequent to a transfusion. Therefore the
difference between the immunised and the non-immunised group is
suggestive that even when allo/antibodies have been detected and blood
compatible by current in vitro methods is provided the likelihood of a
transfusion reaction is higher if the patient has known allo/autoantibodies.
149
Table 3.3: Transfusion reactions in the haematology cohort
The number of suspected transfusion reactions reported in the
haematology cohort (n=1107) for immunised (patients with
allo/autoantibodies) patients was 11/256 and those reported in non-
immunised patients was 13/851. Fisher’s Exact Test shows that the
difference between the groups is statistically significant (P=0.0128). The
risk of transfusion reaction was higher in the immunised group despite the
fact that this group was transfused with blood that was deemed
compatible by in vitro methods.
Total Number of %
number of investigated
patients transfusion
reactions
Non- 851 13 1.5
immunised
Immunised 256 11 4.3
Total 1107 24
150
3.2.5 Cost of additional testing
A total of 6914 serological tests, in addition to the standard ABO and Rh
D group and antibody screen, were performed for the patients in the
haematology cohort (n=1107) at a total cost of £50,262.42. A total of 358
serological tests were performed for the non-immunised patients (n=851)
compared to 6556 additional tests for the immunised group (n=256). A
breakdown of the additional serological tests performed and the costs of
the additional testing are shown in table 3-4. Test costs were calculated,
using an in-house spreadsheet calculator (Microsoft Excel), taking into
account reagent costs, equipment maintenance, consumables, quality
control and staff time incurred during the financial year 2011-2012. Total
test costs for the additional testing performed on patient samples during
the retrospective review period were calculated using these current costs,
although it is accepted that this is not representative of the actual costs of
the tests performed due to changes in costs of staff salaries over the time
period reviewed and changes in test procedures, reagents and analysers
used.
151
Table 3.4: Cost of additional testing in the haematology cohort
The number and cost of additional serological testing performed in
immunised (256/1107) and non-immunised patients (851/1107) within the
haematology cohort. The total cost of additional testing for patients with
allo/autoantibodies was approximately 32 fold greater than that for those
without antibodies.
Immunised group Non immunised group
Test Number of Cost Number of Additional

tests tests Cost
Antibody 2694 £19,181.28 0 0
identification
panel
Direct 2174 £9,152.54 355 £1,494.55
Antiglobulin
Test
Monospecific 1279 £7,098.54 2 £11.10
Coombs test
NHSBT RCI 249 £12,474.90 0 0
referral
Rh CcEe type 98 £426.30 0 0
Red cell 50 £209 1 £4.18

phenotype
Autoabsorption 3 £44.88 0 0
Cold agglutinin 9 £165.15 0 0

screen
Total 6556 £48,752.59 358 £1,509.83
152
3.3 Patients with renal insufficiency
A total of 877 blood transfusion records of patients admitted to the renal
wards for transfusion were reviewed. The 877 patients received a total of
11822 units of red cells during 5238 clinical episodes, the average
number of units transfused was 13.5 (SD 18.36) with a range of 1 to 246.
The average number of transfusion episodes was 5.98 (SD 7.37) with a
range of 1 to 67. The age of the patients at first recorded transfusion
ranged from 15 to 95 with an average of 68.14 years (SD 14.42).
Of the 877 patients in the renal cohort, 134 (15.3%) either presented with,
or developed, allo and/or autoantibodies to red cell antigens. Of the
original 877 patients, 41 patients (4.7%) presented with allo and/or
autoantibodies at the first recorded test, 27 (3.1%) of these presented
with a red cell alloantibody only. Five (12.2%) of the 41 patients who
presented with red cell allo and/or autoantibodies developed additional
allo and/or autoantibodies following transfusion.
A total of 850 patients presented with no detectable red cell
alloantibodies, 33 (3.9%) of these developed one or more alloantibodies
post transfusion
The average number of units given prior to the production of
alloantibodies was 9.66 (range 0 - 143, SD = 16.79). The average
number of units given prior to the production of autoantibodies was 11.46
(range 0 - 91, SD = 18.02).
153
3.3.1 Antibody Specificities
Of the 134 red cell antigen immunised patients in the renal cohort, 60
(44.78%) developed alloantibodies only, 55 (41.04%) developed
alloantibodies in combination with autoantibodies and 19 (14.2%)
developed autoantibodies only. The incidence of antibody specificities is
presented in figure 3.10. A total of 147 alloantibodies were detected, of
these 42 (28.6%) could not be identified. Of the 112 identifiable
alloantibodies 75 (67.0%) were specific to Rh (D, C, c, E, e) or Kell
antigens. The most common specificity was anti-E (16.3%), followed by
anti-K (11.6%) and anti-C (11.6%).
A total of 66 autoantibodies were detected, 7/66 (10.6%) had definable
specificities with auto-anti-e being the most common (3/7, 42.9%) (figure
3.10). A total of 34 (25.4%) of the 134 red cell antigen immunised patients
had alloantibodies in combination with autoantibodies. Of these patients
six (17.6%) developed the autoantibody after the alloantibody, none
developed the alloantibody after the autoantibody, 26 (76.5%) developed
the alloantibody and autoantibody in combination following transfusion
and two presented with an allo/autoantibody combination (5.9%).
Analysis of the specificities of patients with identifiable alloantibodies
revealed 38 patients with single specificity (60.3%), 18 with two
alloantibodies (28.6%), 4 with three antibodies (6.34%) and three with
four antibody specificities (4.76%). Sixty six patients who presented with,
or developed autoantibodies had records detailing the type of antibody or
154
complement fraction on the surface of their red cells. Sixty one (61/66,
92.4%) were found to have IgG only, two (2/66, 3.0%) were found to have
C3d only and two (2/66, 3.0%) were found to have a combination of IgG
and C3d. One patient did not have a recorded autoantibody type.
155
Figure 3.10: Incidence and specificity of allo/autoantibodies in the
renal cohort
The incidence and specificity of allo- and autoantibodies within the renal patient cohort (n= 877).
The specificity of the antibody is shown on the Y-axis, alloantibodies are denoted by the antigen
specificity and autoantibodies are denoted separately. Autoantibodies with no definable specificity
were most commonly detected (n=58), along with alloantibodies reactive in Indirect Anti Human
Globulin (AHG) tests with no definable specificity (n=42). The most commonly detected
alloantibody was anti-E (n= 24), followed by anti-C (n=17) and anti-K (n=17).
156
3.3.2 Antibodies and patient factors
3.3.2.1 Gender and immunisation rate
Of the 877 records reviewed in the renal patient cohort, 339 (38.7%) were
female and 537 (61.2%) were male. The allo/autoantibody immunisation
rate in males was 11.7% (63/537), compared to 20.9% in females
(71/339) demonstrating a significant difference (Fisher’s Exact Test
P=0.0003) and indicating that females are significantly more likely to
develop allo/autoantibodies than males in this group of patients (figure
3.11).
157
Figure 3.11: Influence of gender on the development of
allo/autoantibodies in the renal cohort
600
500
Number of
patients
400
300 immunised
200
non-immunised
100
0
Female Male
Gender
Incidence of immunisation rates in males and females in the renal cohort
(n=877). The incidence of immunisation in male patients was 11.7%
(63/537) compared to an incidence rate of 20.9% (71/339) in female
patients. A significant difference (Fisher’s Exact Test P=0.0003) was seen
between allo/autoimmunisation rates in males and females in this group
of patients.
158
3.3.2.2 Age and immunisation rate
In the renal cohort (n=877) the age of the patients at first recorded
transfusion ranged from 15 to 95 years with an average of 68.14 years
(SD14.42). Within the immunised group (n=134) the average age at first
transfusion was 66.58 years (range 24 – 90, SD 14.70) and within the
non-immunised group the average age was 68.42 years (range 15 – 95,
SD 14.36). There was no significant difference between the mean age (in
years) at first recorded transfusion in the patients with allo/autoantibodies
and those without (t-test P=0.4599) (figure 3.12), indicating that age at
first transfusion does not influence the incidence of immunisation.
159
Figure 3.12: Influence of age at first recorded transfusion on the
development of allo/autoantibodies in the renal cohort
100
90
80
Age of patients (years)
70
60
50
40
30
20
10
0
Box-and-whiskers plot showing the range of age at first transfusion (black line) in the
immunised and non-immunised patients within the renal cohort (n=877). The first and
third quartiles and the median (second quartile) are denoted by the red and green
blocks. The average age at first transfusion in the immunised group was 66.58 years
(range 24 – 90, SD 14.70) and in the non-immunised group the average age was 68.42
years (range 15 – 95, SD 14.36). No significant difference was seen between the mean
age (in years) at first recorded transfusion in the patients with allo/autoantibodies and
those without (t-test P=0.4599)
160
3.3.2.3 Immune suppression and immunisation rate
Receipt of a renal transplant was used as a surrogate marker of immune
suppression, within the renal cohort, in order to investigate any
relationship between immunosuppression and immunisation rate. Forty
seven patients within this group (n=877) had records indicating that they
had received a renal transplant. An immunisation rate of 2.1% (7/47) was
seen in the patients with a record of renal transplant compared to an
immunisation rate of 15.3% (127/830) in the non-transplanted patients.
Statistical analysis of the groups showed that there is a statistical
significance between these rates indicating that immune suppression in
this group of patients may have a protective effect against immunisation
to red cell antigens (Fisher’s Exact Test P=0.0029) as shown in figure
3.13.
161
Figure 3.13: Influence of immunosuppression on the development of
allo/autoantibodies in the renal cohort
1000
800
Number of patients
600
400
Immunised
200
non-immunised
0
Yes No
Transplant
Comparison of the incidence of immunisation in renal patients (n=877) with and without
recorded renal transplant, used as a surrogate marker of immune suppression. An
immunisation rate of 2.1% (7/47) was seen in the patients with a record of renal
transplant compared to an immunisation rate of 15.3% (127/830) in the non-transplanted
patients. The difference in these rates is statistically significant (Fisher’s Exact Test
P=0.0029)
162
3.3.2.4 Number of transfusion episodes and incidence of
immunisation
The number of transfusion episodes and the number of red cell units
transfused was reviewed for the immunised and the non-immunised
patients in the renal cohort of patients. The average number of
transfusion episodes across the patients in this group (n=877) was 5.98
(range 1-67, SD 7.38). Within the immunised group (n=134), the average
number of transfusion episodes was 9.69 (range 1-58, SD 11.53),
compared to an average of 5.30 (range 1-67, SD 6.12) for the non-
immunised group, demonstrating a significant difference in the number of
transfusion episodes and incidence of allo/autoimmunisation (t-test P=
<0.0001) (figure 3.14a).
The average number of red cell units transfused to all patients in the renal
cohort (n=877) was 13.50 (range 1-246, SD 18.36). Within the immunised
group (n=134), the average number of units was 21.85 (range 2-143, SD
25.72), compared to an average of 11.97 (range 1-246, SD 16.26) in the
non-immunised group (n=743), again demonstrating a significant
difference (t-test P= <0.0001) (figure 3.14b)
163
Figure 3.14: Influence of the number of transfusion episodes and number of units transfused on the development
of allo/autoantibodies in the renal cohort
A B
80 300
70

250
60
200
50
40 150
30
100
20
50
10
0 0
Immunised Non Immunised Immunised Non Immunised
Box-and-whiskers plot showing the range (black line) of transfusion episodes and number of units given in the immunised and non-immunised patients within the
renal cohort (n=877). The first and third quartiles and the median (second quartile) are denoted by the red and green blocks. The average number of transfusion
episodes in the immunised group (n=134, average=9.69, SD 11.53) was significantly different to that seen in the non-immunised group (n=743, average=5.3,
SD6.12) (figure A). Similarly, the average number of units transfused in the immunised group (n=134, average=21.85, SD25.72) was significantly different to that
seen in the non-immunised group (n=743, average=11.97, SD 16.26) (figure B)
164
3.3.2.5 Time between transfusion and the incidence of immunisation
Time between transfusions, in days, was reviewed for the patients within
the renal cohort (n=877) to investigate any relationship between the
incidence of immunisation and the time frames for transfusion of blood
units. For all patients in this group (n=877), the time (days) between
transfusion ranged from 0 to 2120.5 days, with an average of 136.64 (SD
218.09). Within the immunised group (n=134) the average time between
transfusions was 118.6 days (range 0-931.9, SD 157.65) and within the
non-immunised group (n=743) the average was 139.5 days (range 0-
2120.5, SD 227.30), demonstrating no significant difference between the
two groups (t-test P=0.3076) (figure 3.15). This would suggest that the
frequency (time between transfusion) of exposure to foreign red cell
antigens does not influence the incidence of immunisation in this group of
patients.
165
Figure 3.15: Influence of the frequency of transfusion episodes on
the development of allo/autoantibodies in the renal cohort
2500
Time between transfusion episodes (days)
2000
1500
1000
500
0
Box-and-whiskers plot showing the range of time in days (black line) between transfusion
episodes in the immunised and non-immunised patients within the renal cohort (n=877). The first
and third quartiles and median (second quartile) are denoted by the red and green blocks. The
average time between transfusion for the immunised group (118.6 days, range 0-931.9, SD
157.65) was not significantly different to that seen in the non-immunised group (139.5 days, range
0-2120.5, SD 227.30) (t-test P=0.3076)
166
3.3.3 Transfusion Reactions
Of the 877 patient records reviewed within the renal cohort, 5 patients
(0.6%) were investigated for a suspected transfusion reaction, giving an
overall transfusion reaction incidence of 1 in 2364.4 units. An incidence of
2.2% of patients were investigated for a suspected transfusion reaction in
the immunised group (n=134, 3/134), compared to an incidence of 0.3%
(n=743, 2/743) in the non-immunised group. An approximate eight fold
difference was seen in the rate of suspected transfusion reactions in the
allo/auto immunised group compared to the non-immunised group (see
table 3-5). The difference between the suspected transfusion reaction
rates between the two groups showed a significant difference (t-test
P=0.0276), although it should be noted that the number of patients in the
group that experienced reactions was very small. Patients included within
the immunised group for review of suspected transfusion reactions were
those with known red cell allo/autoantibodies being transfused with
antigen negative and/or crossmatch compatible blood, not patients who
had suffered a transfusion reaction as a result of development of
antibodies subsequent to a transfusion, therefore the difference between
the immunised and the non-immunised group is suggestive that even
when allo/antibodies have been detected and blood compatible by current
in vitro methods is provided the likelihood of a transfusion reaction is
higher if the patient has known allo/autoantibodies.
167
Table 3.5: Transfusion reactions in the renal cohort
Within the renal cohort (n=877), immunised patients (n=134) showed an
incidence of 2.24% suspected transfusion reaction compared to just 0.3%
in the non-immunised cohort (n=743). An approximate eight fold
difference was seen in the rate of suspected transfusion reactions in the
allo/auto immunised group compared to the non-immunised group. The
risk of transfusion reaction was higher in the immunised group despite the
fact that this group was transfused with blood that was deemed
compatible by in vitro methods.
Total Number of %
number of investigated
patients transfusion
reactions
Non- 743 2 0.27
immunised
Immunised 134 3 2.24
Total 877 5
168
3.3.4 Cost of additional testing
A review was performed of the additional laboratory testing undertaken
for the patients within the renal cohort (n=877). A total of 1919 serological
tests, in addition to the standard ABO and Rh D group and antibody
screen were performed for the 877 patients at a total cost of £14,604.59.
A total of 25 additional tests costing £155.53 were performed for the non-
immunised patients (n=743) compared to 1894 additional tests and a cost
of £13,094.76 for the immunised group (n=134). A breakdown of the
additional serological tests performed and the costs of the additional
testing are shown in table 3-6. The cost of testing for patients with
allo/autoantibodies was approximately nine times greater than that for the
non-immunised group, compared to the 32 fold difference seen within the
haematology immunised and non-immunised cohort. Test costs were
calculated, using an in-house spreadsheet calculator (Microsoft Excel),
taking into account reagent costs, equipment maintenance, consumables,
quality control and staff time incurred during the financial year 2011-2012.
Total test costs for the additional testing performed on patient samples
during the retrospective review period were calculated using these
current tests costs, although it is accepted that, as discussed previously
(section 3.2.5), there is a lack of accuracy due to estimates based on
current costings.
169
Table 3.6: Cost of additional testing in the renal cohort
The table shows the additional laboratory tests performed for patients
within the renal cohort (n=877), the test type, number of tests performed
and cost of testing is shown for the immunised and non-immunised
groups. The costs for the additional laboratory tests required for the
patients within the renal cohort were substantially higher for the
immunised group (n=134) than those for the non-immunised group
(n=743).
Immunised group Non-immunised group
Test Number of Cost Number of Additional

tests tests Cost
Antibody 962 £6,849.44 1 £7.12
identification
panel
Direct 588 £2,475.48 21 £88.41
Antiglobulin
Test
Monospecific 221 £1,226.55 1 £5.55
Coombs test
NHSBT RCI 44 £2,204.40 1 £50.10
referral
Rh CcEe type 51 £221.85 1 £4.35
Red cell 28 £117.04 0 0

phenotype
Total 1894 £13,094.76 25 £155.53
170
3.4 Comparison of renal patients and haematology patients
3.4.1 Gender and the incidence of immunisation
Statistical analysis of the numbers of male and female patients with and
without allo/autoantibodies treated on the haematology wards showed no
significant differences, indicating that the sex of the patient had no
influence on the development of allo/autoantibodies. This difference,
however, was not replicated in the patients transfused on the renal wards.
In order to further investigate this anomaly the rates of allo/autoantibody
production in each of the sexes across the two disciplines were subjected
to statistical analysis using the Fisher’s Exact Test.
An immunisation rate of 21.2% was seen in the male haematology cohort
(n=575) compared to a rate of 11.7% in the male renal cohort (n=537).
Female patients showed an immunisation rate of 25.2% in the
haematology cohort (n=532) and a rate of 20.9% in the renal cohort
(n=340). A statistically significant difference was seen in the rates of
allo/autoantibody production for males (Figure 3.16a) in the two cohorts
(Fisher’s Exact Test, P=<0.0001), but this was not seen for females
(Fisher’s Exact Test, P=0.1640) (Figure 3.16b). This suggests that
females are equally likely to develop red cell all/autoantibodies whether
they are transfused for haematological malignancies or renal insufficiency
whereas males are less likely to produce red cell allo/autoantibodies if
they are transfused for renal insufficiency than they are if transfused for
haematological malignancies.
171
Figure 3.16: Comparison of the influence of gender on the development of allo/autoantibodies in the haematology
cohort versus the renal cohort
A B
700 600
Number of female patients

600
Number of male patients
500
500
400
400 Allo/auto Allo/auto
300
300 immunised Yes immunised Yes
Allo/auto 200 Allo/auto
200
immunised No immunised No
100 100
0 0
Haem Renal Haem Renal
Discipline Discipline
The rates of red cell allo/autoantibody production in male patients transfused on the haematology wards (haem) (n= 575, incidence of immunisation 21.2%) and the
renal wards (n=537, incidence of immunisation 11.7%) are shown in figure 3.16a. There is a statistically significant difference in allo/autoantibody development
between the two groups (Fisher’s Exact Test, P=<0.0001). The immunisation rates in females within the haematology cohort (n=532, incidence 25.2%) compared to
the rate in the renal cohort (n=339, incidence 20.9%) is shown in figure 3.16b. No significant difference in immunisation rates was seen in these groups.
172
3.4.2 Number of units transfused and immunisation rate
Patients within the haematology cohort (n=1107) were transfused with an
average of 28.8 units (SD 42.10) over the review period, compared to an
average of 13.5 (SD 18.36) for the renal patients (n=877). The average
number of red cell units transfused to immunised haematology patients
(n=256) was 46.99 (SD 59.65), compared to an average of 21.85 (SD
25.72) units transfused to immunised renal patients (n=134). Non-
immunised patients in the haematology cohort (n=851) received an
average of 23.43 (SD 32.01) units of red cells compared to an average of
11.97 (SD 16.26) transfused to the non-immunised renal patients
(n=743). Analysis of the number of units transfused to patients on the
haematology wards compared to those on the renal wards demonstrated
that patients transfused for haematological malignancies received
significantly more units of blood than those transfused for renal
insufficiency irrespective of whether they had red cell allo/autoantibodies
(figure 3.17a, t-test P value P=<0.0001) or not (figure 3.17b, t-test P value
P=<0.0001).
173
Figure 3.17: Comparison of the number of red cell transfusions given in the haematology and renal cohorts
A: Immunised patients B: Non-immunised patients

450 400
400 350

350
300
300
250
250
200
200
150
150
100
100
50 50
0 0
Box-and-whiskers plots showing the range of number of transfusions given (black line) in the immunised and non-immunised patients within the
haematology cohort (n=1107) and renal cohort (n=877). The first and third quartiles and the median (second quartile) are denoted by the red and
green blocks. Immunised haematology patients received an average of 46.99 units of red cells (SD 59.65) compared to an average of 21.85 (SD
25.72) in the immunised renal patients (figure 3.17a). In comparison, non-immunised haematology patients received an average of 23.43 units (SD
32.01) compared to an average of 11.97 units (SD 16.26) in the renal patients (figure 3.17b). Patients transfused for haematological malignancies
received significantly more units of blood than those transfused for renal insufficiency irrespective of whether they had red cell allo/autoantibodies
(Figure 3.17a, t-test P value P=<0.0001) or not (Figure 3.17b, t-test P value P=<0.0001).
174
3.4.3 Age at first transfusion and incidence of immunisation
The average age of patients at first transfusion in the haematology cohort
(n=1107) was compared to that of patients within the renal cohort (n=877)
for immunised and non-immunised patients. Non-immunised patients
within the haematology cohort (n=851) were transfused at an average
age of 69.18 (SD 14.7) years compared to an average of 68 years (SD
14.4) in the renal cohort (n=743). For immunised patients the average
age at first transfusion was 69 (SD 14.7) years for haematology patients
(n=256) and 67 (SD 14.7) for renal patients (n=134). No statistically
significant difference was seen in age at first transfusion within the
immunised patients (figure 3.18a t-test P value = 0.2007) or the non-
immunised patients (figure 3.18b t-test P value = 0.1067).
175
Figure 3.18: Comparison of age at first transfusion and incidence of immunisation in the haematology and renal cohorts
A Immunised patients B Non-immunised patients
100 110
90 100
90
Age at first transfusion

Age at first transfusion
80
70 80
70
60
60
50
50
40 40
30 30
20 20
10 10
0 0
Box-and-whiskers plots showing the range (black line) of number of transfusions given in the immunised and non-immunised patients within the haematology
(haem) cohort (n=1107) and renal cohort (n=877). The first and third quartiles and the median (second quartile) are denoted by the red and green blocks.
Immunised haematology patients received the first transfusion at an average age of 69 years (SD 14.7) compared to an average age of 67 (SD 14.7) in the
immunised renal patients (figure A). In comparison, non-immunised haematology patients received the first transfusion at an average age of 69.18 (SD 14.7)
compared to an average age of 68 years (SD 14.4) in the renal patients (figure B). No statistically significant differences were seen in the age at first
transfusion for either the immunised (Figure A, t-test P value P=0.207) or the non-immunised patients (Figure B, t-test P value P=0.1067).
176
3.5 Discussion
It is well accepted in the UK that patients with SCD and thalassaemia
should have extended red cell antigen typing prior to transfusion, and
should receive blood that is matched for at least Rh (CcEe) and Kell
antigens (BCSH, 2013). This strategy has been implemented in response
to the wealth of evidence that this group of patients have a relatively high
alloimmunisation rate (Coles et al., 1981; Davies et al., 1986; Michail-
Merianou et al., 1987; Rosse et al., 1990; Spanos et al., 1990; Hmida et
al., 1994; Tahhan et al., 1994; Moreira et al., 1996). Patients with SCD
and thalassaemia require frequent transfusion support over a long period
of time; the frequency of transfusion makes red cell antigen phenotyping
difficult if left until after the production of a red cell alloantibody when
transfused cells are present in the circulation. The majority of
alloantibodies produced by these patients were found to be Rh or Kell
system related (Davies et al., 1986; Rosse et al., 1990), hence the
recommendation to match for these antigens prophylactically. More
recent studies have noted that patients with SCD have high rates of
alloimmunisation despite strategies to match for Rh and Kell antigens
(Chou et al., 2013; Woldie et al., 2015). This may be related to ethnic
differences between donor and recipient populations, altered RH alleles
when the donor and recipient populations are matched, or patients
receiving transfusions in other centres where matching does not occur.
This information has reignited the debate regarding the benefits and cost
effectiveness of a prophylactic matching strategy for these chronically
177
transfused patients. It has also questioned whether the introduction of
genotyping, particularly for RH, would improve matching and reduce the
alloimmunisation rate.
Patients with haematological malignancies or renal failure are also likely
to require frequent blood transfusion support over an extended period,
often several years, and so it is reasonable to consider the
implementation of such a strategy for typing and matching blood for
transfusion in these groups of patients as well. There has been some
debate in the literature regarding typing and matching of blood in
chronically transfused patient groups, other than those with SCD and
thalassaemia. Initially publications reported that prophylactic matching
was either unnecessary or not cost effective (Blumberg et al., 1984;
Domen and Ramirez, 1988; Schonewille et al., 1999). However, later
studies supported this strategy in certain selected groups (Fluit et al.,
1990; Shirey et al., 2002; Schonewille et al., 2009; Kadar, 2010). A study
by Alves and co-workers (2012) found a relatively high rate of
alloimmunisation in patients transfused for surgical or clinical
emergencies. Their results suggest that a type and match strategy could
also be beneficial for non-chronically transfused patients.
To continue the debate, this retrospective study of patients who received
blood transfusions in a UK haematology ward and a renal ward setting,
collected and analysed data on the allo and auto-immunisation rates,
potential risk factors and cost of additional testing, providing new
178
evidence to support a type and match strategy. The retrospective review
found an overall alloimmunisation rate in patients transfused on the
haematology wards of 12.9%, which is not dissimilar to the 9% found by
Schonewille et al., (1999) and 11.8% by Fluit et al., (1990) in similar
groups of patients. In this study alloantibodies were most commonly seen
in patients with aplastic anaemia (35.7%), MDS (24.5%), solid organ
cancers (18.5%), CLL (17.5%), AML (14.5%) and myeloma (13.9%).
Other workers have found lower rates, 11% in aplastic anaemia
(Blumberg et al., 1983) and 15% (Sanz et al., 2013) in MDS/CMML.
Higher rates were seen in MDS by Milic and co-workers (2011), who
found 36% alloimmunisation rate in patients who received more than 30
units of blood and 64% in those who received more than 100 units. These
differences may be explained by factors such as, the number of patients
studied, the sensitivity of the antibody screening techniques and the
transfusion policies in place. Some studies have indicated that patients
with lymphocytic leukaemias do not develop alloantibodies (Blumberg et
al., 1983; Blumberg et al., 1984; Fluit et al., 1990; Seyfried and Waleska,
1990), however, both this study and that of Schonewille et al., (1999)
have found alloimmune responses in these patients. Correspondence
analysis of the data found in this study revealed that haemolytic anaemia,
MDS and MPD showed an association with the development of allo and
autoantibody, whereas CLL, AML and aplastic anaemia appeared to be
associated with the development of autoantibodies only.
179
In the cohort of 877 patients transfused due to renal failure, 15.3% were
found to have allo and/or autoantibodies, with 13% having red cell
alloantibodies. This is slightly higher than the rates seen in a similar study
by da Silva and co-workers (2013) who reported 7.4% of 393 patients
with chronic kidney disease on a transplant waiting list having red cell
alloantibodies. Sangole and Chaudhari (2013) found red cell
alloantibodies in 25% of patients with chronic renal failure who had
experienced transfusion reactions; however, this study involved a very
small number of chronically transfused patients.
Age appeared to have no influence on the alloimmunisation rate in either
the haematology patients or the renal patients, a fact supported by
Schonewille et al., (1999) and Sanz and co-workers (2013), but not by
Seyfried and Waleska (1990) who reported that the probability of
alloimmunisation is a quadratic function of age. A large study of patients
and donors in the Unites States of America found the alloimmunisation
increased with age for certain antibodies (anti-K, -Kpa, -Fya, -D, -C and –
E) and decreased with age for others (anti-Lea, -Leb, -M and -P1)
presumably due to the inclusion of increasing numbers of women
immunised through pregnancy and both sexes immunised through
transfusion (Winters et al., 2001).
This retrospective review reported only on adult patients, the youngest
patients included were 13 years for the haematological cohort and 15
years for the renal cohort. Other studies have reported much lower rates
180
of alloimmunisation in paediatric patients transfused as a result of SCD or
thalassaemia (Spanos et al., 1990; Aygun et al., 2002), with statistically
significant differences seen in the rate of alloimmunisation in patients
transfused before the age of three, compared to those transfused after
three years of age (Spanos et al., 1990). In a study of transfused patients
awaiting kidney transplantation, which included a patient population with
an age range of 4 to 77 years, no differences were seen in the mean age
of the immunised group compared to the non-immunised group (da Silva
et al., 2013). It would seem reasonable to assume that in a non-selective
population the incidence of red cell alloantibodies increases with age but
in transfused population this finding is not apparent after three years of
age.
In the haematological cohort, gender appeared to have no influence on
the alloimmunisation rate, a fact supported by other studies in similar
groups of patients (Blumberg et al., 1983; Redman et al., 1996
Schonewille et al., 1999; Sanz et al., 2013). Within the renal cohort
gender did appear to have an influence on the alloimmunisation rate, with
females being more likely to produce alloantibodies. This association has
also been reported by other groups in less selective cohorts of patients
and/or blood donors (Hoeltge et al., 1995; Winters et al., 2001; Bauer et
al., 2007), and is generally believed that the greater frequency of red cell
antibodies seen in females is due to pregnancy related alloimmunisation.
The study of renal patients by da Silva and co-workers (2013) suggested
that the determining factor for the differences in alloimmunisation rates in
181
males and females was not only the gestations, but also due to greater
numbers of transfusions given to females (average 7 +/- 3) compared to
males (average 3 +/- 2.6). However, in this study of renal patients, no
statistically significant difference was seen between the number of red
cells transfused between males and females (average 13.51 and 13.43
respectively t-test P value 0.9499) or the number of transfusion episodes
(average 5.90 and 6.01 respectively, t-test P value 0.8296). Further
analysis of the red cell immunisation rates in males and females between
the haematological cohort and the renal cohort in this study found that
that there was no significant difference in the rates for females (t-test P
value 0.1640) but that there was a highly significant difference in the red
cell immunisation rate for males (t-test P value 0.0001). This shows that
male patients transfused for renal failure are much less likely to produce
red cell antibodies than males transfused for haematological
malignancies, suggesting that the disease status itself may put male
haematology patients on an equal footing to females with regard to the
development of red cell allo and/or autoantibodies. This apparent equality
in red cell allo/autoimmunisation could be related to the immune
dysregulation found in certain haematological malignancies (Barrett et al.,
2000; Maciejewski et al., 2007; Alfinito et al., 2010). Alternatively, it could
be related purely to the fact that patients with haematological
malignancies received, on average in this study, significantly more red
cell transfusions over the course of their treatment than the renal patients
(46.99 units vs 12.0 units t-test P value 0.0001 for immunised patients
182
respectively and 23.43 units vs 8.0 units t-test P value 0.0001 for non-
immunised patients).
The current study found that the requirement for concomitant platelet
therapy and irradiated blood components, used here as surrogate
markers for intensive chemotherapy and immune suppression, did not
influence alloimmunisation rates, while other workers have reported that
immunosuppressed patients are less likely to produce antibodies
(Ramsey et al., 1989; Schonewille et al., 1999; Asfour et al., 2004; Zalpuri
et al., 2014). It is possible that the surrogate markers used in this study
did not identify patients whose immune suppression was of a similar state
to those reported in the other studies. Schonewille and co-workers (1999)
reported occurrence of antibodies in patients receiving intensive
chemotherapy, whereas the patients studied by Asfour and co-workers
(2004) had received bone marrow and/or peripheral blood stem cell
transplant, and Ramsey and co-workers (1989) studied patients who
received liver and heart transplants. The comprehensive two-centre case-
reference study by Zalpuri et al. (2014) demonstrated that exposure to
immunosuppressives was associated with a considerably lower risk of
alloimmunisation. Using a case controlled study they demonstrated that,
patients taking only corticosteroids, patients taking only other
immunosuppressants, and those taking both, had a lower
alloimmunisation risk (adjusted Risk Ratios of 0.70 (95% CI, 0.42-1.16);
0.51 (95% CI, 0.04-7.10) and 0.19 (95% CI, 0.07-0.53) respectively) than
that of patients not taking any immunosuppressants (Zalpuri et al., 2014).
183
The surrogate markers used in this study were chosen because they
were believed to be adequate markers for patients undergoing the type of
intensive chemotherapy which would substantially suppress the immune
system. The difference seen in this study and that of Zalpuri and co-
workers (2014) may be due to the fact that in this study it was not known
if the immunosuppressive therapy was given before or after the
alloimmunisation, which may have influenced the results. In addition,
there is evidence to suggest that the transfusion of platelets can, in fact,
increase the risk of red cell alloantibody development by inducing an
inflammatory response in the recipient from infused cytokines (Yazer et
al., 2009).
Review of the transfused renal patients showed an apparent relationship
between immune suppression as a result of renal transplantation and
allo/autoimmunisation, suggesting that immune suppression may have a
protective effect against immunisation to red cell antigens in this group of
patients. However, the number of patients identified as having received
renal transplants in this study was small and it was not possible to
ascertain whether the patients had produced allo/autoantibodies prior to
or subsequent to the transplant.
The current data supports that of other studies (Sarnaik et al., 1986; Fluit
et al., 1990; Hoeltge et al., 1995; Zalpuri et al., 2012), that the
development of alloantibodies is related to the number of red cell
184
transfusions given; in this study both the total number of units and the
number of transfusion episodes were related to the risk of antibody
formation in both cohorts of patients. Antibodies were formed after an
average of 15.46 units had been transfused in the haematology cohort, a
figure similar to that found by Schonewille et al., (1999) and after an
average of 9.66 units in the renal cohort.
In this study a substantial proportion of alloantibodies could not be
identified; 29.3% in the haematology cohort and 28.6% in the renal
cohort. This is similar to the figures found by Schonewille et al., (1999),
Redman et al., (1996) and da Silva et al., (2013) who found 22.5%,
17.6% and 34.5%, respectively. The majority of patients formed a single
antibody specificity (61.6% in the haematology cohort and 60.3% in the
renal cohort) with 43 haematology patients and 25 renal patients forming
multiple specificities, although no patients formed more than five antibody
specificities, comparable with results found by Winters et al., (2001) in an
American population.
As in other studies (Fluit et al., 1990; Hmida et al., 1994; Redman et al.,
1996; Schonewille et al., 1999; Sanz et al., 2013; da Silva et al., 2013)
the majority of antibodies identified were Rh and Kell related, with anti-E
being the most common in both patients transfused for haematological
malignancies and those transfused for renal insufficiency. Almost all red
cell units received from the NHSBT in our institution have their Rh (CcEe)
and K antigen status documented on the unit, making these antigens the
185
easiest to match with phenotyped patients. A strategy to type and match
Rh (CcEe) and K antigens in chronically transfused patients with
haematological malignancies and renal insufficiency has the potential to
eliminate alloantibodies to these antigens and reduce the additional cost
burdens currently seen with alloimmunised patients.
The analysis of additional testing and the associated costs for the
patients included in our retrospective review showed that immunised
patients in the haematology cohort, in general, required approximately 19
times the number of additional serological tests than the non-immunised
patients, and that the increase in costs required were 32 fold. In the renal
cohort, the additional testing for the immunised group was approximately
75 times more than that for the non-immunised group, and the increase in
cost was approximately nine fold. The differences in the costs between
the groups is, in part, due to the higher number of some additional tests
(in particular the DAT) being performed for haematology patients as part
of routine screening for certain diseases, and regardless of immunisation
to red cell antigens. It is likely that these figures would lead to an
underestimation of projected figures, of both test numbers and costs, if
used to predict cost effectiveness of a type and match program. This is
because changes in transfusion policy throughout the study period meant
that, in contrast to current pre-transfusion compatibility guidelines (BCSH,
2013), tests for antibody identification, confirmatory antigen status and
the use of elution techniques for patients with autoantibodies were
regularly omitted from the screening process and so have not been
186
included in the costs. The costs of crossmatching red cell units for the
immunised and non-immunised group were not included in this study, but
would, obviously, impact on the financial burden to the hospital service as
well. Non-immunised patients may have red cells issued for transfusion
by electronic issue, whereas immunised patients always require
serological crossmatching (BCSH, 2013) with the ensuing cost
implication.
A less well documented, and not completely understood, complication of
blood transfusion is the development of red cell autoantibodies, which are
often found in combination with alloantibodies (Shirey et al., 2002; Young
et al., 2004; Ahrens et al., 2007) in transfused patients. The frequency of
auto and alloimmunisation in combination varied in these studies from
28% to 40%, not dissimilar to the figure found in this study for
haematology patients (39.4% of immunised patients) and renal patients in
this study (41.0% of immunised patients). In the majority of cases, in this
study, production of the autoantibody was subsequent to production of
the alloantibody, in accordance with results found by Sanz and co-
workers (2013), and also, in this study the autoantibodies appeared to be
clinically insignificant. Unfortunately, throughout the study period, the
transfusion policy within our institution did not require the use of an eluate
kit to determine the specificity of an autoantibody appearing post
transfusion, and so no records were available to confirm that it was not
indicative of a delayed or acute transfusion reaction. A substantial
proportion (75.3% of haematology patients and 92.4% of renal patients)
187
of the patients with a positive DAT had IgG demonstrable on the surface
of their red cells; elution of these antibodies would be a useful piece of
further work to investigate any potential specificities. However, the
presence of the autoantibodies did not appear to increase the transfusion
requirements of the patients, suggesting that they were not of clinical
significance. In this study, 158 patients treated on the haematology wards
with autoantibodies had follow up testing and, of these, 123 (77.9%) of
the autoantibodies were transient, a phenomenon also described by
Young and co-workers (2004). This would seem to support the theory of
Ahrens and co-workers (2007) that the alloantibody is initially of low
affinity (compared to the matured antibody) and so it cross reacts with
other structures on the red cell membrane as well as the alloantigen. A
similar phenomenon has been demonstrated in an animal model by
Weigle (1965), who demonstrated that autoimmunity could be induced in
rabbits by the injection of heterologous or chemically altered homologous
thyroglobulin. Another theory that has been put forward to explain
transfusion-induced autoantibodies is the nonexofacial polymorphism
(NEP) hypothesis. This hypothesis is based on the theory that differences
in red cell antigenic structures are present both on the surface of the red
cell and within the cytoplasmic or transmembrane domains of the blood
group molecule (Zimring et al., 2007). NEPs have been shown to exist for
some blood group systems including Duffy (Neote et al., 1994; Mallinson
et al., 1995; Olsson et al., 1998; Parasol et al., 1998), Kidd and Kell
(Zimring et al., 2007). It has been demonstrated that autoreactive B cells
specific for red cell antigens exist naturally even in the presence of a
188
functional tolerance mechanism (Murakami and Honjo 1996), and that
they are prevented from maturing by the thymic deletion of autoreactive
CD4+ T helper cells. Zimring and co-workers (2007) postulated that
transfusion of red cells that have the same external red cell antigens, but
different NEPs, could lead to non-deletion of the autoreactive T helper
cells, leaving them available to assist the autoreactive B cells, ultimately
leading to red cell autoantibody production. Another suggestion that has
been put forward is that the appearance of red cell autoantibodies is due
to the presence of lymphocytes from previous transfusions which have
survived and produce alloantibodies that attach to the recipient red cell,
thus mimicking an autoantibody (Petz and Garratty, 1980; Garratty, 1996;
Petz and Garratty, 2004).
The relatively high level of autoantibodies seen in the haematology cohort
of this study, with or without associated alloantibodies, is further
complicated by the original pathology of the patients within the study
group. Patients with lymphoproliferative disorders are prone to the
production of red cell autoantibodies (Kipps and Carson, 1993; Timura et
al., 2000) due to immune dysregulation resulting from the underlying
disease. Similarly, patients with MDS have also been shown to produce
autoantibodies to red cell antigens (Sokol et al., 1989; Novaretti et al.,
2001) as do patients with CML (Steegmann et al., 2003). In this study the
highest proportion of autoantibodies were found, not unexpectedly, in
patients with haemolytic anaemia (85.7%), followed by aplastic anaemia
(50.0%), CLL (36.8%), CML/CMML (30%), MDS (21.9%), AML (17.2%)
189
and myeloma (10.6%), compared to only 5.5% of patients with solid
organ cancer. The levels of autoantibodies present in patients with MDS
in this study is higher than that found by others; Sokol and co-workers
found just 7% (Sokol et al., 1989) with other groups finding 3.7% (Martin-
Vega et al., 1988) and 8.2% (Mufti et al., 1986), however the current data
is similar to the 21.6% found by Seitanides and co-workers (Seitanides et
al., 1988). The higher levels in this study may be due to differences in
sensitivity of screening techniques and/or the smaller number of patients
included. Red cell autoantibodies in patients with renal disease are not
so well documented. They may be related to the immune dysregulation
seen in ESRD, which is exacerbated by dialysis, and can result in
inflammation caused by immunoactivation (Deschamps-Latscha and
Chatenoud 1996; Kato et al., 2008). The relatively high rates of
autoantibodies and alloantibodies in combination, in both the
haematology and renal cohorts in this study, is in contrast to the rate
(3.8%) seen in a non-selected population of red cell immunised patients
and blood donors (Winters et al., 2001), which supports the theories that
non-clinically significant red cell autoantibodies in combination with
alloantibodies may be related to the disease condition.
Throughout the study period, no patients in either cohort were
documented as having suffered haemolytic transfusion reactions, either
acute or delayed. A small percentage (2.17% of haematology patients
and 0.57% of renal patients) of patients were investigated for suspected
transfusion reactions, slightly higher than that found in cancer patients
190
(0.3%) by other workers (Huh and Lichtiger, 1987). This difference may
be explained by improvements, since 1987, in the clinical observations of
transfused patients and changes in the laboratory investigations of
transfusion reactions. A large proportion of the reactions seen by Huh
and Lichtiger (1987) were febrile non-haemolytic reactions, presumably
caused by transfused white blood cells. Since November 1999, all
allogeneic red cell units in the UK have been leucocyte depleted, the UK
specification requires that more than 90% of manufactured leucocyte
depleted blood components should have less than 1 x 10 6 leucocytes and
more than 99% should have less than 5 x 106 leucocytes (JPAC, 2013).
Leucocyte depletion of red cells has been demonstrated to reduce the
incidence of febrile non-haemolytic transfusion reactions (King et al, 2004
(b)) hence, relatively small numbers of this type of reaction would have
been seen in this study. Review of the 11 haematology patients with allo
and/or autoantibodies, investigated for suspected transfusion reactions in
this study, revealed that none of these patients had developed a newly
positive DAT or newly developed alloantibody after the transfusion,
although no eluates were performed to confirm this. Delayed Serological
Transfusion Reactions (DSTR), defined by the detection of a positive
DAT and a new alloantibody specificity post transfusion, were reported as
being a frequent finding in multiply transfused patients (Ness et al., 1990),
although generally benign in nature and not causing clinical haemolysis,
however, this was not seen in our patients. Introduction of the use of
eluate testing as part of the laboratory investigation of positive DATs post
transfusion, as recommended in the BCSH guidelines (BCSH, 2013),
191
may increase the number of regularly transfused patients diagnosed with
DSTR in our patient population.
In conclusion, this study found that patients chronically transfused for
haematological or renal indications are at risk of alloimmunisation. The
antibodies developed by the patients are mainly to the Rh and Kell
antigens, which could, potentially, be reduced by a type and match
strategy. There are considerable costs associated with the provision of
blood in the presence of alloantibodies, and an increased risk of
transfusion reaction for the patients. Blood units supplied by the NHSBT
are typed for Rh and Kell antigens, and the information is visible on the
component label, facilitating blood selection in a type and match strategy.
However, implementation of a strategy for type and match would need a
suitable method, within the hospital transfusion laboratory, to type patient
samples for the corresponding antigens. The method would need to
support a high throughput, and have a high level of automation to reduce
the risk of error. Hospital transfusion laboratories currently use high
throughput automation for routine blood group and antibody screening, a
novel method for serological phenotyping on one of these platforms is
explored in chapter 4. Alternatively, commercial assays are available for
blood group genotyping, which could also be suitable for a hospital
transfusion laboratory. The suitability of one of these assays is explored
in chapter 5. Although red cell units for transfusion supplied by NHSBT
are typed for Rh and Kell antigens, provision of matched units from the
stocks held in a hospital transfusion laboratory could be challenging. The
192
challenges for the laboratory provision of blood are explored in the pilot
study, detailed in chapter 6. The high throughput assay selected to
support a type and match strategy would also need to be cost effective.
The costs for serological phenotyping and genotyping assays are
explored in the relevant sections and compared in chapter 7.
193
4 AUTOMATED RED CELL PHENOTYPING
4.1 Introduction
Recent developments in the use of automated serological phenotyping
techniques in the NHSBT facilities in the UK have meant that the majority
of red cell units supplied to hospital transfusion services are now typed
and labelled not only for ABO and RhD, but also for Rh (CcEe) and Kell
antigens. The BCSH, however, do not recommend that hospital
transfusion services in the UK routinely type patients for antigens other
than ABO and RhD prior to transfusion (BCSH 2013). This is mainly due
to the lack of automated techniques available for performing these tests
within the hospital transfusion service setting and the risks associated
with performing these tests manually on a large scale. Development of a
system to perform automated red cell antigen phenotyping profiles for
pre-transfusion patients would enable hospital transfusion services to
prophylactically select and match blood for, at least, Rh (CcEe) and Kell
antigens and possibly other antigens from routine blood stocks.
The NEO® blood grouping analyser system (IBG Immucor) was
implemented in the RD&E in 2011 for blood grouping, antibody screening
and identification. The analyser also has the potential to perform rapid
automated extended red cell phenotyping profiles including Rh (CcEeCw),
K, Fy (a and b), Jk (a and b), M, N, S, s and k antigens. Although this
technology has been in widespread use throughout the UK for many
years for blood grouping, antibody screening and antibody identification,
profiles for automated serological phenotyping have not been developed
194
or implemented for routine use. Introducing serological phenotyping to the
system offers an opportunity for hospital transfusion services to access
large scale, cost effective extended red cell serological phenotyping. As
part of this study, development work was undertaken with the supplier to
create the assay profiles on the automated system. This involved
determination of the antisera to be used, setting the assay protocols on
the analyser software and creating a report format. The study aimed to
validate the antisera for automated use using commercially available red
cells with known antigen types. In addition, comparability of manual red
cell phenotyping, using a combination of techniques, and automated red
cell phenotyping using the NEO® blood grouping analyser (IBG Immucor)
for patient samples, was performed in accordance with BCSH guidelines
for validation (BCSH, 2012).
The study also reviewed the comparability of cost between manual and
automated red cell phenotyping, with the aim of providing a projected cost
for the implementation of a strategy for targeted extended red cell
phenotyping and matching of blood for transfusion in patients with
haematological malignancies or renal insufficiency.
195
4.2 Results of the automated phenotyping validation
Automated extended red cell antigen testing on the IBG NEO® using the
monoclonal reagents (anti-C, anti-c, anti-E, anti-e, anti-K, anti-M and anti-
N) with the commercial NHSBT ten cell profile (0.8% cell suspension in
Cellstab), three cell profile (3% cell suspension in Alsevers) and two cell
titration profile (0.8% cell suspension in Cellstab) over multiple occasions
(ten cell profile n=4, three cell profile n=5 and two cell profile n=2) showed
no failures and no discrepancies with the known red cell antigen types
reported in the product insert. Reaction values on the IBG NEO® were
consistently strongly positive (4+) with both homozygous and
heterozygous expression of red cell antigens.
Automated extended red cell antigen testing on the IBG NEO® with the
polyclonal reagents (anti-Fya, anti-Fyb, anti-Jka, anti-Jkb, anti-S, anti-s and
anti-k) with NHSBT ten cell profiles, three cell profile and two cell titration
profile over multiple occasions (ten cell profile n=4, three cell profile n=5
and two cell profile n=2) showed one invalidated result with the anti-k
reagent and one weak and undefinable (NTD) result during Jk a testing.
Following repeat testing, the invalidated result was positive and the weak
positive result was negative, giving expected results. There were no
discrepancies found with the known red cell antigen types reported in the
product insert. Reaction values on the IBG NEO® were consistently
strongly positive (4+) with homozygous and heterozygous expression of
red cell antigens. One cell, on one occasion, produced a 3+ result with
the anti-k reagent. The consistency of the reactions with homozygous and
196
heterozygous expression of red cell antigens is shown in table 4-1 using
the NHSBT titration two cell profile as an example.
197
Table 4.1: Consistency of automated screening reactions with homozygous and heterozygous expression of red cell
antigens
The automated profiles were tested on two occasions using the NHSBT titration two cell profile to review the consistency of the reaction
grades. The titration panel includes red cells demonstrating both homozygous (shown in green) and heterozygous (shown in red) expressions
of red cell antigens. The antigen profiles of the cells are shown in bold type and the reaction grades are shown as negative (-) or the numerical
value (graded 1- 4 by the analyser). On both testing occasions results for homozygous and heterozygous antigen expression were strongly
positive (3 or 4 strength reactions).
C c E e Cw K M N S s Fya Fyb Jka Jkb k

Cell 1 + - + + - - - + - + - + + + +
antigen
profile
Cell1 – 4 - 4 4 - - - 4 - 4 - 4 3 4 4
test 1
Cell 1 – 4 - 4 4 - - - 4 - 4 - 4 3 4 4
test 2
Cell 2 + + + + - + + + + + + + + + +
antigen
profile
Cell2 – 4 4 4 4 - 3 4 4 4 4 4 4 4 4 4
test 1
Cell2 – 4 4 4 4 - 3 4 4 4 4 4 4 4 4 4
test 2
198
One hundred patient samples were selected for automated extended red
cell antigen serological phenotyping on the IBG NEO® and comparative
testing in the manual system. A mixed field reaction, demonstrating the
presence of a dual population of cells with, and without, the E antigen,
was seen with the anti-E reagent for one patient in both the automated
and the manual testing. This type of reaction suggests interference from
red cells from a recent blood transfusion. Following discussion with the
patient it was confirmed that a transfusion has been performed at another
hospital.
A weak reaction was seen with the anti-Fya reagent in both the
automated (1+ reaction strength) and the manual technique for one
patient. This sample also produced a weakly positive (2+) result with the
anti-S reagent in the automated profile; manual testing with anti-S
produced a negative result. Weakly positive results are suggestive of
interference by red cell autoantibodies, which should be identified by the
inclusion of a control test using patient cells analysed against the patient
plasma or a commercial neutral reagent. A DAT on the patient sample
confirmed the presence of a weak autoantibody.
A weak reaction (2+) was seen with the anti-Fyb reagent in automated
testing in one sample, repeat testing of this sample showed a negative
result, in accordance with the manual testing results (table 4.2). This was
considered to be a false positive reaction generated by the analyser
system. The cut off values for the anti-Fyb assay between the top end of
199
the “not determined” result (40.998) and the lower end of the 2+ positive
result (50.999) are relatively tight (see appendix 3). There are a number
of sample and/or analytical factors that may affect the value of a reaction
within the system, these include; insufficient centrifugation of the sample,
reagent deterioration and particles of dirt on the imaging system. These
weak reactions seen in the initial testing with commercial cells, and again
during patient testing were important in terms of recognition of false
positives. In the hospital transfusion setting it is extremely important to
reduce the risk of false positive results as these could lead to transfusion
of incorrectly matched red cell units. The information obtained from the
validation was used to determine the cut off values for reporting
serological phenotyping results. It was concluded that only strongly
positive values of 3+ and 4+ could be confidently reported as positive,
and any results between “not determined” and 2+ would be considered
invalid results requiring further investigation.
200
Table 4.2: Phenotyping anomalies in automated and manual assays
Anomalous results in serological phenotyping assays were seen for three patients; patient 1 demonstrated a mixed field reaction (MF) (shown
in green) with the anti-E reagent in manual testing and weak positive in the automated assay, patient 2 reacted weakly positive with the anti-Fya
reagent in both the automated and manual assays, this patient also produced a weakly positive reaction with the anti-S reagent in the
automated profile (shown in red). Patient 3 reacted weakly with the anti-Fyb reagent in the automated assay (shown in blue). The validation
determined that reaction strengths between 1 and 2 are suggestive of false positive results and should be invalidated.
Red cell antigen
E Fya Fyb Jka Jkb S s k
Patient 1 manual MF + + + + + + +
assay
Patient 1 NTD 4 4 4 4 4 4 4
automated assay
Patient 2 manual + Weak + + + - + +
assay positive
Patient 2 4 1 4 4 4 2 4 4
automated assay
Patient 3 manual - + - + + + - +
assay
Patient 3 - 4 2 4 4 4 - 4
automated assay
(primary test)
Patient 3 N/A 4 - 4 4 4 - 3
automated assay
(repeat test)
201
One discrepancy was seen with N typing for one patient, this sample
repeatedly produced a negative result in automated testing using the
immuClone anti-N and a positive result in manual testing using the
Millipore anti-N. Confirmatory testing at the Red Cell
Immunohaematology Laboratory at Filton NHSBT supported the positive
result seen in the manual testing assay, and testing of the patient sample
using the immuClone anti-N in a manual tube technique also gave a
positive result. The discrepancy was thought possibly to be due to
stability issues relating to sample storage time prior to testing. In addition,
the temperature of the reagents at testing was also considered as a
potential source of the weakened reactions (personal communication
from Berndine Kokkelink, IBG Immucor). The performance of certain
Immucor reagents, in particular LISS, are affected by temperature, and it
is noted in the manufacturer’s instructions that they should be allowed to
come to 18-20oC before use. This was not noted in the instructions for the
phenotyping antisera. However, the anti-N reagent had not been
validated by the supplier for automated use, and so it was reasonable to
exclude temperature as a cause of the discrepancy. A further 12 N
positive samples (10 M+N+ and two M-N+) were selected for stability
testing and reagent temperature testing. These samples were repeatedly
tested over a 10 day period; each day the samples were tested with
reagents taken straight from refrigerated storage and again after the
reagents had been allowed to stand at room temperature for at least 4
hours. With the exception of one N heterozygote all other M+N+ samples
tested gave negative results with the anti-N reagent at day 7 (figures 4.1a
202
and b). Using the manual technique, all nine of these samples were still
strongly positive with anti-N (Millipore) at day 8. The reaction strength
with anti-N (immuClone) in the automated technique for the nine samples
clearly demonstrated a reduction in strength over the testing period with
reagents taken directly from refrigerated storage and those held at room
temperature prior to testing (figure 4.2). However, investigation of the
potential effect of temperature of the reagents on the reaction strength
showed no significant difference (figure 4.2) on each day of testing and
testing was discontinued on day 8.
203
Figure 4.1: The influence of reagent temperature on the reaction strength of automated serological phenotyping
testing
A B
Sample 1
100 100 Sample 1
M+N+
Sample 2 M+N+
Reaction strength reported by analyser assay
90 90 Sample 2
Reaction strength reported by analyser assay

M+N+
Sample 3 M+N+
80 80 Sample 3
M+N+
Sample 4 M+N+
70 70 Sample 4
M-N+ M-N+
Sample 5 Sample 5
60 M+N+ 60
M+N+
Sample 6 Sample 6
50 M+N+ 50 M+N+
Sample 7 Sample 7
40 M+N+ 40 M+N+
Sample 8 Sample 8
30 M+N+ 30 M+N+
Sample 9 Sample 9
20 M+N+ 20 M+N+
Sample 10 Sample 10
10 M+N+ 10 M+N+
Sample 11 Sample 11
0 M-N+ M-N+
Sample12
0
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Day 7
Day 8
Day 9
Day 10
Sample12
Day 1
Day 2
Day 3
Day 4
Day 5
Day 6
Day 7
Day 8
M+N+ M+N+
Reaction strengths of the automated serological phenotyping assay using monoclonal reagents over time with M+N+ and M-N+ cells using reagents
taken from refrigerated storage (A) or used at room temperature (B). The numerical value calculated by the automated analyser is shown on the Y
axis and the day of testing on the X axis. The majority of M+N+ cells reach the negative cut off value of 23 by day 7 regardless of the temperature of
the reagents.
204
Figure 4.2: Influence of age of sample and temperature of reagent on the reactions strength of automated
serological phenotyping
A
90
80
Analyser derived reaction
70
60
50
value
MONOPM
40
MONO
30
20
10
0
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
B
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
MONOPM 82 84 77 51 46.5 36.5 25 26
MONO 83 82.5 80 59 38.5 30.5 23.5 32.5
P value 0.8287 0.785 0.6345 0.3974 0.4607 0.5867 0.8932 0.5711
This figure (A) shows the comparison between the average numerical value for each of the nine M+N+ samples over time for serological phenotyping using reagents
directly from refrigerated storage (MONO) and those held at room temperature prior to testing (MONOPM). The numerical value calculated by the automated
analyser is shown on the Y axis and the day of testing on the X axis. T-test values for each paired results shown in the table (B) did not demonstrate any statistically
significant difference, suggesting that the temperature of the reagent during use doe not influence the reaction strength of the result.
205
Manual testing for Rh (CcEe) and K, Cw, M, N took approximately 25
minutes per patient, which included 10 minutes of staff time preparing the
tests and reading the results. Manual testing for Fya, Fyb, Jka, Jkb, S, s
and k required approximately 35 minutes per patient which included 10
minutes of staff time preparing the tests and reading the results. To
reduce the risk of error, best practice is to test one patient at a time and
include positive and negative reagent cell controls for each anti-serum
used (BCSH, 2013).
Automated testing incurred approximately 2 minutes of staff time to load
the reagents onto the analyser; this is independent of the number of
patient samples tested. Preparation of samples for testing took
approximately 2 minutes for 12 samples. The analyser is capable of
processing three patient samples for monoclonal profiles in 28 minutes
and three patient samples for polyclonal profiles in 50 minutes. The
analyser is capable of processing single samples and samples in batches
of various sizes.
Overall, the repeat rate for automated extended red cell antigen
phenotyping was 1/100 for monoclonal antibodies (anti-C, Anti-c, Anti-E,
anti-e, anti-K, anti-M and anti-N) and 3/100 for polyclonal antibodies (anti-
Fya, anti-Fyb, anti-Jka, anti-Jkb, anti-S, anti-s and anti-k). The repeat rate
for manual testing was 1/100 for polyclonal antisera (anti-Fya, anti-Fyb,
anti-Jka, anti-Jkb, anti-s and anti-k). No repeat testing was required for
manual tests performed using monoclonal antisera.
206
The frequency of red cell antigens in the tested patient samples (table 4-
3) showed percentages comparable with that reported in the literature for
Caucasian populations (Issit and Anstee, 1998; Dean 2005; Daniels and
Bromilow, 2010) demonstrating that the selection of samples for testing
was representative of the mainly Caucasian population tested in our
institution.
207
Table 4.3: The frequency of antigens found in samples selected for validation of the automated serological
phenotyping assays
The frequency of antigens present in the sample population selected for validation of the automated serological phenotyping assay was
representative of a mainly Caucasian population when compared to others studies performed in similar populations (Issitt and Anstee, 1998; Dean,
2005; Daniels and Bromilow, 2010;). Antigenic frequencies not stated by the reporters are denoted by NS in the table.
w a b a b
Antigen C c E e K C Fy Fy Jk Jk S s M N k
% positive in this 68.7 80.8 37.4 94.9 15.2 2.02 65.7 81.8 81.8 70.7 53.5 86.7 81.9 77.8 99
study
Dean, 2005 68 80 29 98 10 NS 66 83 77 74 55 89 78 72 99
Daniels and 68 81 29 98 9 NS NS NS 76 72 NS NS NS NS NS
Bromilow, 2010
Issitt and Anstee, 70 80 30 98 9 1 66 83 76.9 72.5 57 88 78 72 99.8
1998
208
4.3 Comparison of test costs for manual and automated
serological phenotyping
Test costs were calculated for the manual and automated serological
phenotyping assays. Staff costs were calculated at Band 5 (Agenda for
Change, NHS pay scale 2015) pay point 20 and reagent costs were
calculated at supplier cost 2014-2015 (excluding VAT). Costs for the
relevant equipment maintenance were included, based on the proportion
of tests processed on the equipment per year, and internal quality control
costs were also included. Manual serological phenotyping assays
performed in column agglutination tests and tube tests required
approximately five minutes of staff time to prepare patient sample
solutions, set up the test and quality control, and approximately five
minutes for result interpretation and recording (table 4-4). Automated
phenotyping took approximately 2 minutes for sample preparation and
reagent preparation, with no time required for result interpretation and
recording (table 4-5). Quality control costs for the manual phenotyping
assays were based on the use of NHSBT two cell titration profile cells
used with each assay; quality control for the automated phenotyping
assay was based on the use of the NHSBT three cell profile in Alsevers,
used on a monthly basis. Test costs for automated red cell antigen
profiles processed on the NEO® analyser were considerably less than
those for manual typing, mainly due to the reduction of staff time required
for the process.
209
Table 4.4: Breakdown of test costs for the manual serological phenotyping assay.
Cost per test was calculated for extended phenotyping performed in manual assays, for the antigens shown in the first column. The table shows the
breakdown of the cost per test of the anti-serum, quality control (QC), consumables, diluent, equipment and staff time. The cost for equipment is
allocated as a proportion of the total tests performed on the system. The total cost for an extended phenotype performed using manual assays was
£50.46.
Antigen Anti-serum cost QC cost Consumables (£) Diluent 2 (£) Equipment (£) Staff time Total (£)
per test (£) (£) (£)
Fya 0.28 0.043 1.49 0.11 0.22 2.35 4.50
Fyb 0.19 0.043 1.49 0.11 0.22 2.35 4.40
s 0.31 0.043 1.49 0.11 0.22 2.35 4.53
S 1.81 0.043 0.46 NA 0.22 2.35 4.88
Jka 0.73 0.043 1.49 0.11 0.22 2.35 4.94
Jkb 0.37 0.043 1.49 0.11 0.22 2.35 4.59
k 0.24 0.043 1.49 0.11 0.22 2.35 4.45
M 0.76 0.043 0.46 NA 0.22 2.35 3.84
N 0.76 0.043 0.46 NA 0.22 2.35 3.84
Cw 0.94 0.043 0.46 NA 0.22 2.35 4.01
Rh (CcEe) and K NA 0.033 3.78 0.11 0.22 2.35 6.49
Total 50.46
210
Table 4.5: Breakdown of the test costs for the automated serological phenotyping assay
Cost per test was calculated for extended phenotyping performed in manual assays, for the antigens shown in the first column. The
table shows the breakdown of the cost per test of the anti-serum, quality control (QC), consumables, diluent, equipment and staff
time. The cost for equipment is allocated as a proportion of the total tests performed on the system. The total cost for an extended
phenotype performed by the automated method was £19.55.
Antigen QC cost (£) Consumables (£) Liss and Equipment Staff time (£) Total (£)
Indicator cells (£)
Fya 0.05 0.24 0.028 0.66 0.07 1.40
Fyb 0.05 0.24 0.028 0.66 0.07 1.55
s 0.05 0.24 0.028 0.66 0.07 1.42
S 0.05 0.24 0.028 0.66 0.07 1.42
Jka 0.05 0.24 0.028 0.66 0.07 1.42
Jkb 0.05 0.24 0.028 0.66 0.07 1.42
k 0.05 0.24 0.028 0.66 0.07 1.51
M 0.05 0.07 0.028 0.66 0.07 1.18
N 0.05 0.01 0.028 0.66 0.07 1.09
Cw 0.05 0.01 0 0.66 0.07 1.19
C 0.05 0.01 0 0.66 0.07 0.97
c 0.05 0.01 0 0.66 0.07 0.89
E 0.05 0.01 0 0.66 0.07 0.89
e 0.05 0.01 0 0.66 0.07 0.98
K 0.05 0.01 0 0.66 0.07 1.15
Neg control 0.05 0.24 0.028 0.66 0.07 1.09
Total 19.55
211
4.4 Discussion
The automated phenotyping profiles were demonstrated to be simple,
robust and rapid assays for the serological testing of Rh (CcEe), K, Cw,
M, N, Fya, Fyb, Jka, Jkb, S, s and k antigens, suitable for use in routine
hospital transfusion service laboratories. Weakly positive reactions in the
automated assay (reactions strengths between 1+ and 2+) were
suggestive of false positive results and so the assay was amended to
exclude these reactions strengths from being reported as positive.
Only one discrepancy was identified during comparative testing of patient
samples; N antigen testing of one patient sample gave repeatedly
positive results in the manual test and negative results in the automated
test. It was initially thought that this may have been caused by the
automated reagent having cross reactivity with the “N” antigen. The anti-
N reagent used in manual typing for the comparative study was a murine
monoclonal IgG reagent (Millipore (UK) Ltd), which is stated not to react
with the “N” antigen, when used by the recommended tube technique.
The anti-N used in the automated typing was also a murine monoclonal
reagent (immuClone), licensed for use in tube and slide techniques and
was also stated not to have “N” reactivity. Consequently, the discrepancy
between the Millipore and immuClone anti-N reagent reactivity with this
one patient sample was discounted as being the result of cross reactivity
with the “N” antigen in the manual tube technique, since both suppliers
clearly state that there is no “N” reactivity with their reagents. Anecdotal
212
evidence suggested that the strength of reactivity of the anti-N reagent in
the automated profile (verbal correspondence) may be dependent on the
age of the sample at testing and/or the temperature of the reagents. The
manufacturer of the immuClone anti-N used in the automated technique
(Immucor Gamma) stated in the product insert that Ethylenediamine
Tetraacetic Acid (EDTA) samples may be tested up to 14 days from
collection, or until the expiry of the anticoagulant, if stored at 2 – 8oC. This
particular sample was first tested at day 11, then at days 18 and 43.
Review of the reaction strengths with the anti-N reagent revealed that the
reactivity reduced from a score of 22 at day 11, 17 at day 18 and 14 at
day 43, all giving negative qualitative results but noticeably reducing with
age of sample. A stability study was then initiated to investigate this
theory. It was felt that if stability of sample was an issue with anti-N
testing, then this was most likely to affect heterozygous expression of the
N antigen (M+N+ patients). Consequently, the stability study included 10
samples showing heterozygous expression (M+N+) and two samples
showing homozygous expression (M+N-). These samples were tested
daily for up to 10 days to determine a cut-off point for sample testing. The
stability study demonstrated that, as a general rule, the reaction strength
with N heterozygous cells would reduce to a negative value by day 7, and
that to achieve the acceptable 3+ or 4+ reaction strength samples should
be tested within three days of collection. In addition, it was also
suggested (verbal correspondence) that the temperature of the reagents
during testing may affect the reaction strengths. To test this effect, the
stability study was also designed to compare the reaction strengths of the
213
anti-N using reagents taken directly from refrigerated storage 2 – 8oC,
with the reaction strengths using reagents that had reached room
temperature. The comparison of the reaction strength between the two
groups showed no significant difference, indicating that temperature of
the reagents does not affect the strength of the reaction.
The reason for the reduction in reaction strength of the anti-N
(immuClone) is not clear, and would appear not to affect all N
heterozygous cells, since one M+N+ cell remained strongly positive at
day 10. The fact that the nine M+N+ cells, that gave negative results in
the automated testing at day 10, still remained strongly N positive when
tested with the anti-N (Millipore) reagent in manual testing, suggests that
the N antigen itself is not affected by storage. Both the ImmuClone and
the Millipore reagents are murine monoclonal reagents, and are from
different clone lines. However, this difference was not thought to affect
the reaction strength, since the ImmuClone reagent was found to give a
strong positive reaction in the tube technique with the original discrepant
sample. The reduction in reaction strength, seen with the automated
technique over time, is possibly due to the centrifugal forces applied by
the automated analyser, or the forces applied during the shaking step
prior to result analysis. Although this validation demonstrated that the
immuClone anti-N reagent is suitable for use in automated techniques,
samples must be tested within 3 days of collection to minimise the risk of
false negative results.
214
Automated assays are considerably cheaper than manual assays, mainly
due to the reduction in staff time required for preparing and reading the
assays. In addition, reactions in the automated assay are graded by the
analyser in accordance with a computer generated algorithm, in contrast
to the macroscopic, human eye, reading of reactions strengths in the
manual assay which may be subject to discrepancies between individual
staff. Thus, the automated assay has the potential to standardise results
and reduce the risk of mis-interpretation.
This validation confirmed the fact that serological phenotyping assays are
prone to false positive results in the presence of transfused cells and
positive DAT. As reported here, measures can be implemented to reduce
the risk of reporting false negative results in the presence of a positive
DAT, but challenges will inevitably remain in the presence of circulating,
transfused red cells. Genotyping assays overcome the limitations of the
serological assays by testing genetic material extracted from white blood
cells in the patient sample. The red cell phenotype is then predicted from
the genotype. Thus, genotyping assays are useful when attempting to
antigen type regularly transfused patients and those with positive DAT.
The following chapter aims to assess the feasibility of DNA based assays
for use in a hospital transfusion service setting, and to compare
phenotypes predicted by genotyping with those found in the automated
serological assay.
215
5 BLOOD GROUP GENOTYPING
5.1 Introduction
For patients undergoing multiple transfusions, serological phenotyping
can be complicated by the presence of transfused cells within the
circulation (BCSH, 2013). A definitive determination of the red cell
phenotype of the patient is not possible whilst donor cells are still present
in the circulation, and this makes it difficult to determine whether any
antibodies detected are alloantibodies (reacting against the donor cells)
or autoantibodies (reacting against the patient’s own cells). This, in turn,
leads to difficulties with selection of appropriate antigen matched blood
for subsequent transfusions. Serological phenotyping can also be further
complicated by the presence of antibodies bound to antigens on the
surface of the red cells, which can lead to false positive results, if using
polyclonal antisera. The benefits of genotyping techniques have already
been shown in the determination of blood groups in multiply-transfused
patients (Guelsin et al., 2010), as red cells (transfused or recipient) do not
carry DNA, only the DNA from recipient white cells is analysed. Blood
components in the UK are leucocyte depleted, the small amounts of white
cell DNA transferred to the recipient during transfusion does not interfere
with the genotyping results, and even if transfusions of non-white cell
depleted units are given this does not appear to interfere (Rozman et al.,
2000).
216
Advances in gene cloning and sequencing of blood group genes have
determined the molecular basis for almost all of the clinically significant
blood group polymorphisms (Daniels, 2013). The majority of genetically
defined blood group antigens are the result of single-nucleotide
polymorphism (SNP) (Castilho and Junior, 2004) which give rise to the
amino acid changes described in section 1.3.2, and determine the
antigenic phenotype. It is this knowledge that enables the use of DNA
based assays, which detect these SNPs, to ascertain the allelic variations
and predict the blood group antigens expressed on the surface of the red
cells. Information regarding the blood group systems, genotype and
allelic variations that make up the antigens of the various blood group
systems can be found in the ISBT blood group alleles database and the
Erythrogene database (Möller et al., 2016).
DNA techniques have also been shown to have a high concordance rate
with serological phenotyping techniques in blood donor and patient
samples (Karpasitou et al., 2008), although limitations are seen with null
and/or weak phenotypes, which may be detected only by serological
techniques (Drago et al., 2010). Knowledge of the molecular basis of null
phenotypes and weakened expression of antigens on the red cells allows
the inclusion of assays into the red cell antigen genotyping profiles
designed to detect the mutations silencing the expression of the antigen.
This allows the genotyping results to predict the red cell antigen
phenotype with a greater level of accuracy, as the presence of the gene
does not always lead to the presence of the antigen on the surface of the
217
red cell. The IDCOREXT assay, evaluated in this study, includes probes
designed to detect the SNPs responsible for null phenotypes in the Kidd,
Duffy and MNS blood group systems, thus reducing the potential for
incorrect prediction of the presence of the red cell antigens. This study
aimed to evaluate the applicability of this assay in a routine hospital
transfusion service setting, as yet untested in the UK.
The following subsections will focus on the genetic basis of the blood
group antigens predicted by the BLOODChip IDCOREXT assay, table 5.1
contains a summary of the blood group systems, genes, alleles,
nucleotide changes associated with the allelic variations and the
predicted amino acid changes included in the assay. The antigenic
phenotypes predicted by the alleles assayed by the IDCOREXT test are
shown in table 5-2. The IDCOREXT test (Progenika Biopharma S.A.) can
be used to type allelic variants which determine the blood group antigens
tested in the serological phenotyping section of this study (Rh (CcEe), Cw,
Kell (K and k), Kidd (Jka and Jkb), Duffy (Fya and Fyb), M, N, S and s),
and also includes those of other blood group systems (Diego, Dombrock,
Colton, Yt and Lutheran systems). Information regarding the
biochemistry, genetic basis and function of the blood group systems are
detailed in section 1.3.2.
218
Table 5.1: Blood group system alleles detected by the BLOODChip ID COREXT assay
Blood group systems in the Erythrogene database and ISBT Blood Group Alleles showing the system name, gene, allele, nucleotide change and predicted amino
acid change for the blood group antigen phenotypes predicted by the BLOODChip IDCOREXT assay Key: NA = not available in ISBT database
System name Gene Allele (ISBT) Nucleotide change Predicted amino acid change
Rh RHCE RHCE*01 (RHCE*ce) Wild type reference Wild type reference
RHCE*02 (RHCE*Ce) 48G>C 178C>T 203A>G Trp16Cys Leu60Ile Asn68ser

307C>T Pro103Ser
RHCE*03 (RHCE*cE) 676G>C Ala226Pro
RHCE*04 (RHCE*CE) 48G>C 150C>T 178C>A Trp16Cys Leu60Ile Asn68ser

201A>G 203A>G 307C>T Pro103Ser Ala226Pro
676G>C
RHCE*CeCW 122A>G Gln41Arg
RHCE*ceCW NA NA
RHCE*CECW NA NA
RHCE*ceAR 48G>C 712A>G 733C>G Trp16Cys Met238Val Leu245Val

787A>G 800T>A 916A>G Arg263Gly Met267Lys Ile306Val
RHCE*CeFV 667G>T 697C>G 712A>G Val223Phe Gil233Glu Met238Val
RHCE*CeVG 712A>G 733C>G 787A >G

Met238Val Leu245Val Arg263Gly
219
System name Gene (cont.) Allele (ISBT) (cont.) Nucleotide change Predicted amino acid change
(cont.) (cont.)
(cont.)
Rh RHCE*cEFM 697C>G 712A>G Gln233Glu Met238Val
RHCE*ce[712G] 712A>G Met238Val
RHCE*ce[733G] 733C>G Leu245Val
RHCE*ce[733G,1006T] 733C>G 1006G>T Leu245Val Gly336Cys
RHCE*CE-D[2, 5, 7]-CE 307 676 712 733 1006† NA
RHCE*cE[697G,712G,733G] 697C>G 712A>G 733C>G Gln233Glu Met238Val Leu245Val
RHD*r’s-RHCE*ce[733G,1006T] IVS3+3100G 733C>G Leu245Val Gly336Cys

1006G>T
Kell KEL KEL*k_KPB_JSB 578C>T Thr193Met
KEL*K_KPB_JSB Wild type reference Wild type reference
KEL*k_KPA_JSB 841C>T Arg281Trp
KEL*k_KPB_JSA 1790T>C Leu597Pro
220
System name Gene (cont.) Allele (ISBT) (cont.) Nucleotide change (cont.) Predicted amino acid change
(cont.) (cont.)
Kidd SLC14A1 JK*02 (JK*B) Wild type reference Wild type reference
JK*01(JK*A) 838A>G Asn280Asp
JK*B_null(871C) 871T>C Ser291Pro
JK*B_null(IVS5-1a) 1G>A in intron 5 NA
Duffy DARC FY*A; 125A>G Asp42Gly
FY*B Wild type reference Wild type reference
FY*B_GATA -67T>C NA
FY*B[265T]_FY*X 265C>T Arg89Cys
221
(cont.) (cont.)
MNS GYPA GYPA*M Wild type reference Wild type reference
GYPB GYPA*N 59C>T 71G>A 72T>G Ser20Leu Gly24Glu
GYPE
GYPB*s Wild type reference Wild type reference
GYPB*S 143C>T Thr48Met
GYPB*Mur GYP(B1-136-Bψ137-204- NA
A205-229-B230-366).
GYPB*deletion Deletion of whole gene NA
GYPB*S_null(230T) 230C>T Thr77Met
GYPB*S_null(IVS5+5t) 270+5G>T Alternative splicing
222
(cont.) (cont.)
Diego SLC4A1 DI*A 2561C>T Pro854Leu
DI*B Wildtype reference Wild type reference
Dombrock ART4 DO*A Wild type reference Wild type reference
DO*B 793A>G Asn265Asp
DO*B_HY 232G>T Gly108Val
DO*A_JO 350C>Y Thr117Ile
Colton AQP1 CO*A Wild type reference Wild type reference
CO*B 134C>T Ala45Val
Yt (Cartwright) ACHE YT*A Wild type reference Wild type reference
YT*B 1057C>A His353Asn
Lutheran LU LU*A 230G>A Arg77His
LU*B Wildtype reference Wild type reference
223
5.1.1 Rh system CcEe
RHCE encoding the various Rh (CcEe) polymorphisms is located on the
short arm of chromosome 1 (Avent et al., 1990; Cherif-Zahar et al., 1990).
The IDCOREXT assay includes probes for the common alleles encoding
the C, c, E, e and Cw antigens. The alleles at the locus include RHCE*01
(RHCE*ce), in which “c” is represented by 307C in exon 2 and “e” which
is associated with 676G in exon 5. RHCE*C is represented by 307T in
exon 2 and RHCE*E by 676C in exon 5. All the RHCE amplicons
included within the IDCOREXT kit are gene-specific and so no
amplification of RHD is expected. The CW antigen is produced as a result
of a single nucleotide change in exon 1 (122A>G) and is usually
associated with RHCE*Ce (Mouro et al., 1995). Rarely, Cw is associated
with RHCE*ce and RHCE*CE, the IDCOREXT assays includes probes
designed to detect the nucleotide change associated with all of these
alleles. Rhnull phenotypes are identified as a result of no amplifications of
exons 5 and 7 of RHCE.
Additional alleles assayed within the IDCOREXT test in the Rh system
include those that code for the rare V and VS antigens. VS is associated
with a single nucleotide change 733C>G in exon 5 of RHCE*ce (Daniels,
2013). The IDCOREXT assay includes probes designed to detect the
VS+V+ phenotype and the VS+V- phenotype (733C>G and 1006G>T).
224
The hrS and hrB antigens are associated with partial e (Shapiro, 1960;
Shapiro et al., 1972). These antigens are important as they may be
lacking in some black populations, individuals may type serologically as
e+ but make apparent anti-e. Accurate identification of individuals with
these variant antigens is of relevance in the transfusion of patients with
SCD (Chou and Westhoff, 2009). Allelic probes designed to detect the
presence of these SNPs representing these phenotypes are also included
in the IDCOREXT.
5.1.2 Kell system
KEL encoding the Kell antigens is located on chromosome 7 (7q33)
(Zelinski et al., 1991; Purohit et al., 1992; Lee et al., 1993; Murphy et al.,
1993). The K/k polymorphism arises as a result of a nucleotide change,
578C>T, within exon 10 of the KEL gene, resulting in the KEL*K (ISBT
K*01.01) and KEL*k (ISBT K*02) alleles. Additional alleles within the Kell
system include those that code for the antithetical antigens Kp a and Kpb
and Jsa and Jsb. Kpa/Kpb arises as a result of 841C>T substitution in
exon 8 and Jsa/Jsb from a 1790T>C change in exon 17.
The Kell null phenotype can result from a variety of different genetic
events (Reid, 2009) and so it is not feasible within the IDCOREXT assay
to include probes designed to detect the phenotype.
225
5.1.3 Kidd system
The gene encoding the antigens of the Kidd system (SLC14A1) is located
on chromosome 18 (18q11-q12) (Olives et al., 1996). The Jka/Jkb
phenotype arises as a result of a nucleotide change 838G>A in exon 9,
resulting in the JK*A (ISBT JK*01) and JK*B (ISBT JK*02) alleles.
The rare Jk null phenotype, Jk (a-b-), can be caused as a result of one of
two silencing mutations; a splice mutation causing skipping of exon 6,
SLC14A1:c.471-1G>A (Lucien et al., 1998) or a T871C substitution
predicted to disrupt a potential N-glycosylation motif (NSSNSP),
SLC14A1:c.871T>C (Sidoux-Walter et al., 2000). The IDCOREXT assay
contains a probes designed to detect both of these mutations.
5.1.4 Duffy system
DARC (ISBT nomenclature or HUGO gene name ACKR1) encoding the
Duffy system antigens is located on chromosome 1 (1q21-q22) (Mathew
et al., 1994). The Fyb/Fya polymorphism arises as a result of a nucleotide
change, 125A>G, in exon 2.
Tournamille and co-workers (1995) have demonstrated that the Fynull
phenotype, Fy(a-b-), commonly seen in African populations, is a result of
disruption of a GATA motif in the Fy gene promoter (DARC:c.-67T>C). In
addition, a weakened expression of Fyb phenotype, also known to cause
discrepancies between genotyping and phenotyping results, (Murphy et
al., 1997) was later found to be caused by an Arg89Cys substitution
226
(Tournamille et al., 1998), the result of a nucleotide change 265C>T in
exon 2. The IDCOREXT assay includes probes designed to detect both
of these changes.
5.1.5 MNS system
The genes encoding the MNS polymorphisms, GYPA and GYPB, are
located on chromosome 4 (4q31.21) (Kudo and Fukuda, 1990). The third
gene in the family, GYPE, does not encode detectable protein on the red
cell surface, but has been shown to be involved in some gene
rearrangements that do encode hybrid proteins (ISBT, 2017). The M/N
polymorphism arises as a result of nucleotide changes, 59C>T, 71G>A
and 72T>G in exon 2. The S/s polymorphism arises as a result of a single
change, 143T>C in exon 4.
Within the MNS system the alleles assayed include the U antigen, an
antigen associated with the S-s- phenotype (Wiener et al., 1954) and Mia,
a low frequency antigen (Levine et al., 1951). Mur and Mia antigens arise
as a result of hybrid glycophorins (B-A-B). The GYPB*Mur allele detected
by the IDCOREXT assay is a novel sequence derived from the composite
exon; GYPB 5’ pseudoexon 3 and GYPA 3’ exon 3. The nucleotide
change associated with this phenotype is GYP(B1-136-Bψ137-204-A205-
229-B230-366).
The S-s-U- phenotype occurs as a result of Glycophorin B deficiency
(Huang et al., 1987; Rahuel et al., 1991), the IDCOREXT assay includes
227
detection of the GYPB*deletion which gives rise to the S-s-U- phenotype.
The S-s-Uvar (weakened expression of the U antigen) is associated with
a variant Glycophorin B protein that expresses a red cell antigen known
as the He antigen (Huang et al., 1994; Huang et al., 1995; Reid et al.,
1996; Huang et al., 1997; Storry et al., 2003). The S-s- phenotype can
also rise as a result of mutations within GYPB, commonly within intron 5
or caused by a nucleotide change 230C>T substitution. Probes designed
to identify both of these are included in the IDCOREXT assay; GYPB:c.
230C>T and GYPB:c. 270+5G>T.
5.1.6 Diego system
The Diego system antigens include three pairs of antithetical antigens; Dia
and Dib, Wra and Wrb, Wu and DISK and encoded by the gene SLC4A1
on chromosome 7 (7q12-q21) (Schofield et al., 1994). The Dia/Dib
polymorphism arises as a result of a nucleotide change 2561T>C in exon
19. The IDCOREXT test includes an assay for the alleles that encode the
Dia and Dib antigens, DI*A (ISBT DI*01) and DI*B (ISBT DI*02).
5.1.7 Dombrock system
The Dombrock system antigens are encoded by the gene ART4 gene
located on chromosome 12 (12p13.2-12.3) (Eiberg and Mohr, 1996;
Mauthe et al., 2000); the assay tests for alleles that encode four of these
antigens (Doa, Dob, Hy and Joa). The Doa/Dob polymorphism arises from
a single nucleotide change, 793A>G in exon 2, resulting in the DO*A
228
(ISBT DO*01) and DO*B (ISBT DO*02) alleles. Hy is represented by
323G in exon 2 of DO*B and Joa by 350C in exon 2 of DO*A.
5.1.8 Colton system
The Colton blood group system consists of four antigens (Coa, Cob, Co3
and Co4), the gene encoding these antigens, AQP1, is located on
chromosome 7 (7p14) (Moon et al., 1993).. The Coa/Cob polymorphism
arise from a single nucleotide change 134C>T in exon 1. The IDCOREXT
assay includes probes designed to identify the alleles, CO*A and CO*B,
responsible for the expression of the Coa and Cob phenotypes.
Although Conull phenotypes exist, and some are represented by single
nucleotide changes, they are rare and are not included in the IDCOREXT
assay.
5.1.9 Yt system
The antigens of the Yt (previously known as Cartwright) system are
encoded by the gene, ACHE, located on chromosome 7 (7q22) (Getman
et al., 1992; Ehrlich et al., 1992). The Yta/Ytb polymorphism results from
the nucleotide change, 1057C>A, in exon 2, giving rise to the YT*A (ISBT
YT*01) and YT*B (ISBT YT*03) alleles. The allele responsible for the high
frequency YTEG antigen, YT*03, is the result of a nucleotide changes,
266G>A in exon 2.
229
5.1.10 Lutheran system
The Lutheran blood group system includes 22 antigens; however the
IDCOREXT assay includes allelic tests for just two of these antigens (Lu a
and Lub). The gene that encodes the antigens of the Lutheran blood
group, LU, system is located on chromosome 19 (19q12-q13) (Parsons et
al., 1997; El Nemer et al., 1997). The Lua/Lub polymorphism results from
a single nucleotide change, 230A>G, in exon 3, resulting in the alleles,
LU*A (ISBT LU*01) and LU*B (ISBT LU*02). The Lunull phenotype is
extremely rare and, as such, identification is not included in the
IDCOREXT assay.
230
Table 5.2: The red cell antigen types determined by the IDCOREXT assay.
The IDCOREXT assay uses PCR amplification to obtain target sequences for alleles encoding red cell antigens from human blood samples. The
human blood group systems and the alleles assayed by the test are shown in the table below, also shown are the red cell antigen types as denoted
by their International Society of Blood Transfusion (ISBT) phenotype name.
Blood Group System Alleles Assayed Antigens

(ISBT Phenotype)
Rh RHCE*ce; C (RH:2)
RHCE*Ce E (RH:3)
RHCE*cE; c (RH:4)
RHCE*CE e (RH:5)
RHCE*CeCW CW (RH:8)
RHCE*ceCW V (RH:10)
RHCE*CECW hrS (RH:19)
RHCE*ceAR VS (RH:20)
RHCE*CeFV hrB (RH:31)
RHCE*CeVG
RHCE*cEFM
RHCE*ce[712G]
RHCE*ce[733G]
RHCE*ce[733G,1006T]
RHCE*CE-D[2, 5, 7]-CE
RHCE*cE[697G,712G,733G]
RHD*r’s-RHCE*ce[733G,1006T]
Kell KEL*K_KPB_JSB K (KEL:1)

KEL*k_KPB_JSB k (KEL:2)
KEL*k_KPA_JSB Kpa (KEL:3)
KEL*k_KPB_JSA Kpb (KEL:4)
Jsa (KEL:6)
Jsb (KEL:7)
231
Blood Group System Alleles Assayed Antigens (continued)
(ISBT Phenotype)
Kidd JK*A Jka (JK:1)

JK*B Jkb (JK:2)
JK*B_null(871C)
JK*B_null(IVS5-1a)
Duffy FY*A; Fya (FY:1)

FY*B Fyb (FY:2)
FY*B_GATA
FY*B[265T]_FY*X
MNS GYPA*M S (MNS:3)

GYPA*N s (MNS:4)
U (MNS:5)
GYPB*s Mia (MNS:7)
GYPB*S M (MNS:1)
GYPB*Mur N (MNS:2)
GYPB*deletion
GYPB*S_null(230T)
GYPB*S_null(IVS5+5t)
Diego DI*A Dia (DI:1)

DI*B Dib (DI:2)
Dombrock DO*A Doa (DO:1)

DO*B Dob (DO:2)
DO*B_HY Hy (DO:4)
DO*A_JO Joa (DO:5)
Colton CO*A Coa (CO:1)

CO*B Cob (CO:2)
Cartwright YT*A Yta (YT:1)

YT*B Ytb (YT:2)
Lutheran LU*A Lua (LU:1)

LU*B Lua (LU:2)
232
5.2 Study design
The IDCOREXT test allows a simultaneous multiplex reaction in a single well
avoiding the need for running multiple separate methods in parallel. The
results obtained from the Luminex xMAP technology are then processed using
the IDCOREXT analysis software, converting the alleles detected into a
genotype result which is then, in turn, used to predict the phenotype for each
of the allelic variants interrogated by the test (see section 1.7).
This study aimed to evaluate the IDCOREXT test for potential use as a high
throughput genotyping method suitable for the hospital transfusion setting. In
addition, this study aimed to compare the red cell phenotyping results
obtained from blood samples taken from a selection of the patients consented
from the Haematology Clinic and tested on the automated phenotyping
platform, with the genotyping results obtained by the IDCOREXT test using
DNA extracted from the same patient samples.
233
5.3 Results
5.3.1 DNA extraction
The concentration of DNA extracted from the whole blood samples (n=47)
ranged from 6.03ng/μL to 812 ng/μL, the average concentration was
184.75 ng/μL (SD 180.63) (table 5-3). The DNA concentration
recommended for use in the IDCOREXT assay was between 10 ng/μL
and 80 ng/μL, hence, one sample was below the recommended level.
234
Table 5.3: Concentration and quantity of DNA extracted from whole
blood samples
DNA was extracted from whole blood samples taken from 47 patients for
genotyping using the IDCOREXT assay. The concentration of DNA ranged from
6.03ng/μL to 812 ng/μL.
Patient number DNA concentration (ng/μL) DNA quantity
1 45.74 100
2 26.17 100
3 104.8 300
4 25.58 300
5 195.1 300
6 175 300
7 300.4 300
8 6.03 300
9 271.7 300
10 812 300
11 266.1 300
12 137.7 300
13 797 300
14 45.97 300
15 80.03 100
16 64.55 300
17 28.13 100
18 397 300
19 82.78 300
20 28.78 300
21 437.6 300
22 411 300
23 182.7 300
24 25.12 300
25 130.7 300
26 96.24 300
27 215.6 300
28 95.69 300
29 24.3 300
30 185.9 300
31 111.2 300
32 71.69 300
33 300.1 300
34 121.3 300
35 129.9 300
36 297.2 300
37 349.4 300
38 516.2 300
39 168.7 300
40 47.32 300
41 108.4 300
42 134.8 300
43 43.33 300
44 124.3 300
45 258.1 300
46 76.06 300
47 130 300
235
5.3.2 Cost of testing
Test costs were calculated for the red cell antigen genotyping system.
Staff costs were calculated at Band 5 (Agenda for Change, NHS pay
scale 2015) pay point 20 and reagent costs were calculated at supplier
cost 2014-2015 (excluding VAT). Costs for the relevant equipment
maintenance cost per patient test were estimated, based on the number
of patients registered as attending the Haematology or Renal wards for
transfusion at the RD&E per year, and assuming that every patient would
require a blood group genotype. True cost per test for the analyser
maintenance could not be calculated as the system is not in use in our
institution; however, estimated costs were calculated based on the
supplier quotation. However, this may lead to over representation of the
cost as the analyser system can also be used by other specialities, such
as microbiology and immunology, therefore not restricted to blood group
genotyping and, thus increasing the total throughput per year. Internal
quality control costs were also included in the cost of the consumables.
The total cost per patient test was estimated to be £73.66 (table 5-4).
236
Table 5.4: Cost of Blood Group genotyping
The cost per test was estimated for blood group genotyping using the IDCOREXT test. All costs shown in this table are cost
per patient sample tested. Staff time costs were estimated assuming that the testing and reporting of results would be
performed by a band 5 Biomedical Scientist. Equipment maintenance costs were based on an assumption that all patients
admitted for transfusion on the haematology and renal wards would require a blood group genotype test (n=1674 for the
financial year 2014-2015).
DNA IDCOREXT + Consumables (£) Equipment Staff time (£) Total cost per
extraction (£) kit (£) (£) patient sample (£)
8.95 45.00 1.50 0.87 17.34 73.66
237
5.3.3 Comparison of phenotyping and genotyping results
DNA extracted from 47 patients was analysed for the presence of known
blood group SNPs on the Luminex XMAP analyser using the IDCOREXT
kit. Whole blood samples from each of these 47 patients had previously
been analysed for serological phenotype using the automated
phenotyping method (IBG Immucor) to allow comparison of the
phenotyping result with the predicted phenotype generated by the
genotyping result. The IDCOREXT contains assays for 49 different alleles
that are used by the software to predict 37 red cell phenotypes. The
report produced by the software details the blood group system, the
alleles assayed within that system, along with the genotype result, the
corresponding red cell antigen with their ISBT phenotype notation and the
predicted phenotype (figure 5.2). The predicted phenotype, as interpreted
by the software, was then used for direct comparison with the phenotypes
reported by the automated serological phenotyping assay. The raw data
obtained from the IDCOREXT assay are detailed in Appendix 4.
238
Figure 5.1: DNA analysis report generated by the Luminex XMAP
analyser
An example of the report generated by the Luminex XMAP analyser for each sample of
DNA processed. The alleles assayed for each blood group system are shown, along
with the genotype result, the corresponding red cell antigen (ISBT Phenotype) and the
predicted phenotype.
239
No failures of testing were encountered and there were no discrepancies
seen in the serological phenotyping and genotyping for Rh (CcEeCw), K,
k, N, Fy (a and b), Jk (a and b), S and s. One discrepancy was seen in
the M antigen status for one patient (EX1300853), this patient typed as
M+ in serological phenotyping assays but produced an M- genotype
(discussed further in section 5.4). One patient (EX1312673) returned a
“not determined” result for the S antigen in the automated serological
assay. A repeat sample could not be obtained from the patient within the
time frame of the study, due to the death of the patient, but further review
of the serological assay reaction value (21) showed the result to be just
above a negative value (20.998 value for the cut off high) (see result
interpretation cut off values in appendix 3). Therefore, the serological
phenotyping result was suggestive of a negative result, in agreement with
the genotyping predicted phenotype result. Full results of the comparison
of the genotyping assay and the automated phenotyping assay can be
seen in Table 5-5.
Comparison of serological phenotype and genotype for some antigens
predicted by the IDCOREXT assay could not be performed as there were
no serological phenotypes performed, these include the antigens V, hrS,
Vs and hrB of the Rh system, Kpa, Kpb, Jsa and Jsb of the Kell system, U
and Mia of the MNS system, and antigens of the Diego, Dombrock,
Colton, Cartwright and Lutheran system.
240
Table 5.5: Antigen phenotypes by serological analysis and genotyping
Results obtained by serological phenotyping for the 47 selected patient samples are shown in black type and those predicted by genotyping are
shown in red type.
A single discrepancy was seen for the M antigen type for patient EX1300853 which is highlighted in yellow.
S antigen status for patient EX1312673 was not determined by serological phenotyping methods due to undetermined consistently weak positive
reactions (highlighted in yellow)
Key: P = positive result

N = negative result
ND = not determined
Red cell antigens assayed

Sample No C c E e Cw K Fya Fyb Jka Jkb S s M N k
EX1305814 P P N P N P P P P N N P P N P
P P N P N P P P P N N P P N P
EX1305816 N P P P N N N P N P P P P N P
N P P P N N N P N P P P P N P
EX1305815 N P P N N N P N P N N P N P P
N P P N N N P N P N N P N P P
EX1305817 N P P P N N P N P P P P P P P
N P P P N N P N P P P P P P P
EX1300853 N P P N N N P P P P N P P P P
N P P N N N P P P P N P N P P
241
Red cell antigens assayed (continued)
EX1300854 N P P P N N P P P P P P P N P
N P P P N N P P P P P P P N P
EX1303174 N P P N N N P P N P N P P P P
N P P N N N P P N P N P P P P
EX1306850 P N N P N N P N P P P P P P P
P N N P N N P N P P P P P P P
EX1307336 P P P P N N N P P N P N P P P
P P P P N N N P P N P N P P P
EX1307335 P N N P N N N P P P P P P P P
P N N P N N N P P P P P P P P
EX1307657 P N N P N N P P P P N P P N P
P N N P N N P P P P N P P N P
EX1307898 N P N P N N N P P N N P P P P
N P N P N N N P P N N P P P P
EX1307899 N P P P N P P P P P P P P P P
N P P P N P P P P P P P P P P
EX1308240 N P N P N N P P P N N P N P P
N P N P N N P P P N N P N P P
EX1308368 P P N P N N P N P P N P P N P
P P N P N N P N P P N P P N P
EX1308578 N P P P N N P P P P N P P P P
N P P P N N P P P P N P P P P
EX1308762 P N N P N N N P P N P P P N P
P N N P N N N P P N P P P N P
242
EX1308909 P P P P N N P P P N N P P P P
P P P P N N P P P N N P P P P
EX1308908 N P N P N N P N P N P P P P P
N P N P N N P N P N P P P P P
EX1308959 P P N P N N N P N P N P P P P
P P N P N N N P N P N P P P P
EX1311437 P N N P P N N P N P N P P N P
P N N P P N N P N P N P P N P
EX1311438 N P P P N N P P P P P N P N P
N P P P N N P P P P P N P N P
EX1311988 P P N P N P N P P P P P P P P
P P N P N P N P P P P P P P P
EX1311987 P P P P N N N P P P P P P P P
P P P P N N N P P P P P P P P
EX1311986 P P P P N N P P P N P P P P P
P P P P N N P P P N P P P P P
EX1311989 P P N P N N P P P P P P P N P
P P N P N N P P P P P P P N P
EX1312296 P P N P N N P N P N N P P N P
P P N P N N P N P N N P P N P
EX1312295 P P N P N N N P N P P P P N P
P P N P N N N P N P P P P N P
EX1312294 N P N P N N P P P P N P P N P
N P N P N N P P P P N P P N P
243
EX1312297 P P N P N P N P P N N P P P P
P P N P N P N P P N N P P P P
EX1312311 P P N P N N N P N P N P N P P
P P N P N N N P N P N P N P P
EX1312312 P N N P N N N P P P N P P P P
P N N P N N N P P P N P P P P
EX1312453 N P P P N N N P P P P N P N P
N P P P N N N P P P P N P N P
EX1312674 P N N P N N P N P P N P P P P
P N N P N N P N P P N P P P P
EX1312673 P P N P N N N P P P ND P P P P
P P N P N N N P P P N P P P P
EX1312675 P N N P P N N P N P N P N P P
P N N P P N N P N P N P N P P
EX1312967 N P P P N N N P N P P P P P P
N P P P N N N P N P P P P P P
EX1313154 P P N P N N P P N P P N P N P
P P N P N N P P N P P N P N P
EX1313155 N P P P N N P N N P N P N P P
N P P P N N P N N P N P N P P
EX1313515 N P P N N N N P P N P P N P P
N P P N N N N P P N P P N P P
EX1313516 P P N P P N P P P N N P P N P
P P N P P N P P P N N P P N P
244
EX1313820 P P N P N N N P N P N P P P P
P P N P N N N P N P N P P P P
EX1400171 P P N P N N N P P P N P P P P
P P N P N N N P P P N P P P P
EX1400887 P P N P N N N P N P N P N P P
P P N P N N N P N P N P N P P
EX1401612 P P P P N N N P P P P N P N P
P P P P N N N P P P P N P N P
EX1401613 P P N P N P P P P N P P P N P
P P N P N N P P P N P P P N P
EX1401737 P P P P N N P P N P P P P P P
P P P P N N P P N P P P P P P
245
5.4 Discussion
The IDCOREXT assay was demonstrated to be a robust assay showing
good comparison with the automated serological phenotyping assay and
a low failure rate. A discrepancy in testing was seen with the M antigen
type in one patient, unfortunately it was not possible within the time scale
and funding of this project to ascertain the cause of this discrepancy. In
addition, repeat serological testing was not possible due to the death of
the patient. It is unlikely that the discrepancy was caused by human
and/or methodological error, as this would have resulted in discrepancies
involving multiple antigens. The serological M type for this patient was
confirmed as positive by a reference centre (NHSBT, Filton), in
accordance with the serological phenotype determined by automated
testing at the RD&E. Therefore the discrepancy was unlikely to be due to
any issues affecting the performance of the anti-M used in the automated
phenotyping assay. It is possible that the allele coding for the M antigen,
in this particular patient, may have an alteration such that it is not
detected by the BLOODChip IDCOREXT assay. Sequencing of the DNA
could potentially have revealed if this was the case, however, this was not
possible within the funding of the project. This type of discrepancy could
result in a false negative result being reported. This would not cause
clinically significant issues for patients within a type and match strategy
as they would be transfused with antigen negative red cells. However,
within the blood donor service this would have implications as the donor
units would be erroneously reported as antigen negative and could elicit
an immune response in a patient.
246
In a study by McBean and co-workers (2014b) it was noted that, although
the microarray techniques are a useful tool for the determination of blood
group genotypes, difficulties can arise as a result of a new or rare allele
not incorporated in the assay, or a complex genetic variant. In these
cases further analysis by sequence-based typing in specialist reference
laboratories may elucidate the molecular basis of the unusual blood
group phenotype (Seltsam and Doescher, 2009). McBean’s laboratory
experience suggests that up to 5-10% of complex serological cases
cannot be resolved by microarray genotyping, and of these, 1-2% cannot
be resolved by sequencing (McBean et al., 2014b).
Genotyping assays can also generate false positive results, where there
is an apparently normal allele but no corresponding antigen on the red
cells. This can be caused by silencing genes outside of the assay target
area, as seen in “null phenotypes”. The BLOODChip IDCOREXT assay
contains makers for well-known silencing genes, but hitherto unknown
ones may be missed. False positive results in genotyping assays cause
clinical issues within the transfusion service rather than the blood donor
service, in contrast to the false negative results. A false positive result for
a transfusion recipient in a type and match program would lead to
selection of inappropriate red cells. Discrepancies between genotyping
and phenotyping results may also arise if not all the alleles are known,
and many molecular events can give rise to discrepancies. The current,
commercially available, genotyping assays are restricted to testing for
known SNPs, meaning that new variants will not be recognised.
247
Other studies have seen higher genotype-phenotype discrepancy rates
than seen in this study, McBean and co-workers (2014a) found a
discrepancy rate of 10.4% when comparing phenotyping with genotyping
using three different genotyping platforms, the BLOODChip (Progenika),
BeadChip (BioArray Solutions Ltd) and the HemoCarta Blood Group
Typing Panel (Sequenom Inc.). However, it should be noted that the
discrepancies between genotyping and phenotyping results seen in the
McBean study and also in other studies (Drago et al., 2009; Drago et al.,
2010) were due to mutations causing weakened antigen expression or
null phenotypes. The inclusion of markers for identification of null
phenotypes and silencing genes within the commercial genotyping kits
will reduce the risk of incorrect prediction of antigen phenotypes.
A further limitation of the commercially available blood group genotyping
assays is that they may be designed for a specific population (McBean et
al., 2014b). McBean and co-workers report that the BLOODChip assay
was designed for a European population and the BeadChip assay was
designed for the African/America population. Neither of these assays,
they report, are designed for the large proportion of Asian ancestry seen
in the Australian population.
The development of massively parallel sequencing (MPS) as an
alternative approach to blood group genotyping has overcome some of
the limitations of the commercially available microarray kits. MPS is not
248
restricted to the identification of blood group systems with known SNPs
and, thus, allows new variant blood group antigens to be identified. MPS,
or Next Generation Sequencing, can be used to obtain information on the
whole genome, or restricted to whole exome or targeted gene
sequencing. This technique is not currently commonly used for blood
group genotyping but has been applied to resolve the genetic basis of the
Vel antigen (Cvejic et al., 2013) and that of the low frequency orphan
antigen SARA (McBean et al., 2014c). When looking to implement MPS,
ethical issues should also be considered as non-blood group genetic
information is also obtained. Methods for managing this information would
need to be in place to ensure donor and/or patient confidentiality.
One of the advantages commonly quoted for commercially available
blood group genotyping assays is their ability to provide information on
multiple blood group antigens for multiple patients on an automated
platform. In addition, the automated data analysis by the system software
removes the risk of subjective interpretation of the traditional manual
serological assays. The validation of the automated serological assay on
the IBG Immucor NEO® analyser system in this study demonstrates that
serological assays can also be fully automated and provide standardised,
non-subjective, information on a large number of blood group antigens for
multiple patients with minimal manual input. The real advantage of the
genotyping assays is that they can be used to obtain a predicted
phenotype in patients who have been recently transfused and in those
249
who have a positive DAT, areas where serological assays may give
inconsistent or incorrect results.
Although DNA extraction for this study was performed by an automated
system, the genotyping assay required substantial manual input for the
PCR and hybridisation phases. Not only are these manual phases
expensive in terms of labour, they also require specialist skills not
currently used by hospital transfusion scientific staff, and they are at risk
of errors in sample transposition and pipetting which may lead to incorrect
results. Automation of these phases for blood group genotyping assays is
not currently available, and it is this limitation which will hinder the
introduction of the assays into the routine hospital laboratory setting.
The evaluation of the serological and genotyping assays, as described in
this study, have demonstrated robust, accurate systems for high
throughput testing that have potential for use in a hospital transfusion
setting. In particular, the automated serological assay (IBG Immucor)
would appear to be a cost effective method that could be easily
incorporated into a type and match program. To reduce the risk of
inaccurate serological phenotyping results, patient samples would need
to be tested prior to initial transfusion, or in the absence of blood
transfusion in the previous four months. The results of the phenotyping
assay could then be used to provide blood for transfusion that was
matched for Rh (CcEe) and K in addition to the standard ABO/D
matching. The introduction of a “type and match” program would have the
250
potential to reduce the risk of alloimmunisation in chronically transfused
patients. The benefits of a type and match program have not been
evaluated in the UK, and would require a large number of participants
enrolled into a randomised controlled trial (RCT) to adequately address
the research question, given the incidence of alloimmunisation in
chronically transfused patients noted in this study. Prior to initiation of
such an RCT, there are other important questions that need to be
addressed. These include recruitment rates, dropout rates and, more
importantly, the ability of the hospital transfusion laboratories to provide
matched blood from routine stocks. A useful precursor to an RCT is a
pilot study, which can provide answers to some of these questions and
identify any unforeseen issues. A pilot study was initiated at the RD&E,
as part of this research project, designed to provide information that could
be used to lead into a larger RCT, and this is described in chapter 6.
251
6 PILOT STUDY-RANDOMISED CONTROLLED TRIAL
6.1 Introduction
To truly assess whether typing and matching of blood for chronically
transfused patients would reduce the risk of the development of
allo/autoantibodies and realise the estimated cost benefits, a randomised
controlled trial (RCT) is required. Patients would be recruited to either a
standard care group, receiving standard ABO and RhD matched blood or
to an intervention group destined to receive blood additionally matched
for Rh (CcEe) and K antigens. This study has already demonstrated that
the development of allo/autoantibodies occurs in around 15-25% of
chronically transfused patients and that a relatively large number of
transfusions are required before this effect is seen. Therefore any RCT
designed to investigate the potential benefits of typing and matching
blood would need large numbers of patients and would take considerable
time to achieve. Given the number of patients with haematological and/or
renal conditions that would be required for recruitment it is unlikely that
this sort of RCT could be undertaken in one hospital site, but would more
likely require a coordinated multicentre trial.
A feasibility, or pilot, study is a commonly used forerunner to a large RCT,
and is described by the National Institute for Health Research (NIHR) as
a piece of research done before a main study to answer the question “can
this study be done?” In this way they can be used to estimate important
parameters that are needed to design the main study, such as standard
252
deviation of the outcome measure, which is needed in some cases to
estimate sample size required to gain statistical power; number of eligible
patients, willingness of participants to be randomised; willingness of
clinicians to recruit participants; carers or other appropriate participants;
characteristics of the proposed outcome measure. In some cases
feasibility studies might involve designing suitable outcome
measures; follow-up rates, response rates to questionnaires,
adherence/compliance rates; availability of data needed or the usefulness
and limitations of a particular database; time needed to collect and
analyse data (NIHR guidance). Pilot studies do not necessarily need to
be RCTs but the inclusion of a control group and randomisation allows
the review of willingness of the subjects to be randomised which can then
inform on the recruitment requirements of the larger trial.
Pilot studies, because of their limited size, cannot provide any definite
support for any specific therapeutic claims (Loscalzo, 2009), they cannot
be used to evaluate safety, efficacy and effectiveness and neither do they
provide meaningful effect size estimate for planning subsequent studies
due to the imprecision inherent in data from small samples (Leon et al.,
(2011); Friedman et al., 1998; Kraemer et al., 2006), however, the data
obtained can be used to design a larger RCT and they can be useful in
identifying any unexpected harm of the intervention early on in the course
of its development. The ultimate aim of the larger trial should be used to
guide the design of the pilot study and the participants or patients should
be recruited from the same study population (Hinds and Gattuso, 1991),
253
this will enable the researchers to test those aspects of the intended
larger trial and make any amendments in advance of embarking on any
larger RCTs. Recruitment of trial subjects can be slower and more difficult
than researchers imagine and many trials fail to achieve their intended
sample size within the planned time frame (McDonald et al., 2006; Sully
et al., 2013). Assessment of this within the pilot study may prevent issues
developing within the larger RCT. A review of pilot studies published in
seven major medical journals from 2000 to 2001 (Lancaster et al., 2004)
recommended that statistical analysis of pilot studies should be either
mainly descriptive or focus on the estimation of the sample size and that
any results from hypothesis testing should be treated with caution. Arain
and coworkers (2010) revisited this review between the years 2007 and
2008 to assess whether the recommendations made by Lancaster and
his team had produced a change in the practice of reporting pilot studies
but found that pilot studies were still poorly reported with inappropriate
emphasis on hypothesis testing. The review by Arain and coworkers
found that 81% of pilot studies published reported statistics on hypothesis
testing, they also found that only 69% of the studies included a control
group, and although randomisation was noted in 62% of the trials, only
20% of those were blinded randomisation. Teare and coworkers (2014)
recommended that pilot studies include at least 35 participants per group
(total of 70) to allow estimation of the standard deviation for a continuous
outcome and a total of 60 to 100 recruits for trials investigating event
rates. It is recommended that any estimation using confidence intervals
(CI) in pilot studies should use 85% or 75% CI rather than the commonly
254
used 95% CI (Lee et al., 2014). There is no set sample size for a pilot
study. The review by Arain and coworkers (2010) found a median number
of 76 participants, with an interquartile range of 42-216. Generally,
sample sizes of 30, 50 and 70 will provide 48%, 78% and 84% power to
detect an acceptance rate of 85% or lower if the acceptance rate in the
population is 95%, however it has been noted that the sample size is
often related to the funding level rather than the likelihood of achieving
the specific aims (Moore et al., 2011).
With all the benefits and caveats of pilot studies in mind, a small RCT
was designed and performed using participants from the haematology
clinics at the RD&E. It was accepted that, given the time frame in which
the pilot study was required to be completed, and to enable the
participants’ sufficient time to be given blood transfusions, that only 60
patients would be recruited to the trial. The 60 patients were randomised
to the standard care group or to the intervention group using a modified
block randomisation to ensure that there were equal numbers of patients
in each group. Patients approached for recruitment into the trial were
given information leaflets detailing the tests to be performed and the
randomised aspect to the treatment arms (appendix 5). Patients were
given time to read the information and encouraged to ask any questions
regarding the trial. Specific consent forms were signed by each patient
recruited and patients were given the option to drop out of the trial at any
time, should they wish to do so (appendix 6). The pilot study was given
ethical approval from the NRES Committee South West - Cornwall &
255
Plymouth (MREC No. 12/SW/0251, R&D Study No. 1301733). This is the
first time that such a trial has been attempted in the UK.
The aim of the pilot study was to review recruitment rates and dropout
rates for the patients and also to assess the ability to supply matched
blood for transfusion in the intervention group from routine blood stocks
held in a hospital transfusion laboratory. The pilot study was not powered
to investigate any differences in the rates of development of
allo/autoantibodies between the standard care group and the intervention
group. Nor was it powered to provide an estimate of the sample size
required for a larger multicentre randomised controlled trial designed to
investigate any benefits of typing and matching of blood for chronic
transfusion.
The blood stocks held at the RD&E during the time of the pilot study are
shown in table 6.1. The red cell units were held in ABO/D specific trays
within a walk-in cold room. There was no subdivision of ABO/D types for
Rh or Kell antigens. Orders were made to the NHSBT for red cell stocks
on a daily basis, no changes were made to the ordering procedure with
respect to particular Rh antigen types. In order to locate a matching unit
for a pilot study patient, the laboratory staff had to search the physical
stock, there is no search function for this in the laboratory computer
system. Alternatively, appropriate antigen negative blood could be
ordered from the NHSBT. As such, the results of the pilot study could
256
also be used to design effective changes to the way that the stocks are
arranged and ordered to support a type and match program.
Table 6.1 Optimum stock of red cell units held in RD&E stock
The optimum number of red cells and their ABO/D type are shown in the table
below. A small stock of group O and A, K negative, irradiated units are held for
patients with this special requirement. The red cell units are stored in labelled
trays in accordance with their ABO/D type. (Table taken from the RD&E
laboratory standard operating procedure for red cell ordering and stock
control).
Blood group Optimum Red Cell Paediatric Red Cells

Stock
O RhD Pos 30 0
O RhD Neg 10 1
A RhD Pos 30 0
A RhD Neg 12 0
B RhD Pos 6 0
B RhD Neg 4 0
AB RhD Pos 4 0
AB RhD Neg 4 0
O RhD Pos Irradiated K- 4
O RhD Neg Irradiated K- 2
A RhD Pos Irradiated K- 4
257
6.2 Results of the randomised controlled pilot study
A total of 60 patients were recruited to the trial between the 8 th November
2012 and the 16th July 2014. Patients were recruited from the
haematology clinic with the only exclusions being individuals under the
age of 18, those not capable of giving informed consent and those with a
positive antibody screen recorded in the pathology computer system.
Initially, recruitment was restricted to patients with haematological
conditions considered likely to require blood transfusion, such as
myeloma, acute and chronic leukaemias, Hodgkin’s and non-Hodgkin’s
lymphoma, myelodysplastic and myeloproliferative disorders. However,
using these criteria only eight patients were recruited over a seven month
period. In order to recruit a sufficient number of patients in the time scale
relevant to the study, the entry criteria were widened to include any
patient attending the clinic with a confirmed haematological condition.
Only one patient approached declined to enter the trial on the basis that
they were alarmed by the thought of potential blood transfusions. All other
patients approached were satisfied with the information that they were
given regarding the trial and happy to be included.
Initially 30 patients were randomised to Group 1, the standard treatment
group, and 30 to Group 2, the intervention group. One patient in Group 2
was found to have a positive antibody screen on initial testing and so was
excluded from the trial.
258
6.2.1 Number of patients transfused in each study group
A total of 59 patients were included in the trial, of these 19 (32.2%) were
transfused during the trial period. A total of 13 patients received
transfusions in the standard care group, group 1 (n=30), and four patients
were transfused in the intervention group, group 2 (n=29). This
demonstrates that significantly more patients in group 1 received
transfusions than in group 2 (Fisher’s exact test p value = 0.0204) (figure
6.1).
Figure 6.1: Comparison of number of patients transfused in the pilot
study groups
35
30
25
Number of patients
20
Not transfused
15 Transfused
10
0
Group 1 Group 2
Within the standard care group (group 1, n=30) 13 (43.3%) patients were
transfused, compared to 4 (13.8%) patients in the intervention group
(group 2, n=29). Significantly more patients in group 1 were transfused
compared to group 1 (Fisher’s exact test p value = 0.0204).
259
6.2.2 Time spent in trial for patients in each study group
Patients were recruited to the trial between the 8th November 2012 and
the 16th July 2014, the trial was closed and data collected on the 31st July
2015, consequently, in total, the trial ran over 995 days. The recruitment
process lasted 615 days, longer than anticipated due to initial difficulties
with the selection criteria. Over the 21 month recruitment process patients
were accepted onto the trial at an average rate of 2.86 per month,
however, over the first six months of the trial only seven patients were
recruited (average 1.17 per month). At this point it was realised that
insufficient patients would be recruited in the allocated time using the
stringent criteria and so the criteria were relaxed, as a result of this, from
June 2013 to November 2013 a total of 34 patients were recruited
(average 5.67 per month) (figure 6.2).
Patients in group 1 spent an average of 697.6 days (range 396 to 995
days, SD 154.8, n=30) on the trial compared to an average of 583.7 days
(range 380 to 994 days, SD 165.9, n=29) for patients in group 2 (figure
6.3). Of the first 30 patients to be recruited to the trial, 21 (70%) were
randomly assigned to group 1, this unexpectedly high number would
account for the significant difference in length of time in the trial for
patients in each of the groups (t-test p value = 0.0085).
260
Figure 6.2: Recruitment rates for pilot study patients
14
Number of patients recruited
12
10
8
6
4
2
0
Apr-14
Apr-13
Dec-12
May-13
Dec-13
May-14
Jan-13
Mar-13
Aug-13
Jan-14
Mar-14
Jun-13
Oct-13
Feb-14
Jun-14
Nov-12
Jul-13
Nov-13
Jul-14
Feb-13
Sep-13
Month and Year
Patients were recruited to the pilot study over a 21 month period (n=60).
Slow recruitment rates in the initial six month period were thought to be
due to strict selection criteria focussed on patients with disease
conditions considered to be at high risk of transfusion dependence.
Relaxation of the selection criteria and an increase in recruitment rate
was seen from June 2013 onwards, which enabled the recruitment target
to be achieved within the time scale of the project.
261
Figure 6.3: Comparison of the length of time spent on the trial for
patients in the pilot study groups
900
800
700
Time in trial (days)
600
500
400
300
200
100
0
1 2
Pilot study group
Box-and-whiskers plots showing the minimum and maximum numbers
(black line) of time spent in the trial (the first quartile, mean and third
quartile are denoted by the coloured blocks). Patients in group 1
(standard care group, n=30, range 396 to 995 days, SD 154.8) spent
significantly longer on the pilot study trial than patients in group 2
(intervention group, n=29, range 380 to 994 days, SD 165.9). T-test p
value = 0.0085.
262
6.2.3 Comparison of the number of transfusion episodes and units
given for patients in each study group
Within the standard care group (group 1 n=30) 13 patients received blood
transfusions and within the intervention group (group 2, n=29) four
patients received transfusions. The average number of red cell units
given to transfused patients in group 1 over the study period was 27.46
(range 2 to 76, SD 26.99, n=13) and the average number of red cell units
given the transfused patients in group 2 was 22.25 (range 1 to 71, SD
32.71, n=4). There was no statistically significant difference seen in the
number of red cell transfusions given to patients in each study group (t-
test p value = 0.7513) (figure 6.4).
The average number of transfusion episodes within the standard care
group (group 1) was 11.62 (range 1 to 30, SD 10.32, n=13) compared to
an average of 10.75 (range 1 to 34, SD 15.59, n=4) in the intervention
group (group 2), again this demonstrated no statistically significant
difference between the two groups (t-test p value = 0.8971) (figure 6.5).
Statistical analysis of the number of red cell units transfused and the
number of transfusion episodes for the transfused patients in each of the
study groups demonstrated that the patients in each group were well
matched despite the majority of patients recruited during the initial
recruitment phase, when selection for patients likely to require transfusion
was attempted, being allocated to group 1, the standard care group.
263
Figure 6.3: Number of red cell units transfused to patients in pilot
study
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
Group 1 Group 2
(black line) of red cell units transfused (the first quartile, mean and third
quartile are denoted by the coloured blocks) to patients in the pilot study.
Comparison of the number of red cell units given to patients in the
standard care group (group 1, average 27.46, range 2 to 76, SD 26.99,
n=13) and the number given to patients in the intervention group (group
2, average 22.25, range 1 to 71, SD 32.71, n=4) showed no significant
difference (t-test p value = 0.7513).
264
Figure 6.4: Number of transfusion episodes for patients in the pilot
study
40.00
35.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
Group 1 Group 2
(black line) of transfusion episodes (the first quartile, mean and third
quartile are denoted by the coloured blocks) experience by patients in the
pilot study. Comparison of the number of transfusion episodes for
patients in the standard care group (group 1, average 11.62, range 1 to
30, SD 10.32, n=13) and those for patients in the intervention group
(group 2, average 10.75, range 1 to 34, SD 15.59, n=4) did not
demonstrate any statistically significant difference (t-test p value =
0.8971).
265
6.2.4 Disease conditions in the pilot study patients
Patients recruited to the pilot study were suffering from a wide variety of
haematological disease conditions (figure 6.6). The standard care group
(group 1) included a larger number of myeloma and non-Hodgkin’s
lymphoma (NHL) patients compared to the intervention group (group 2).
These disease conditions are considered high risk for requiring
transfusion through the course of treatment and so, as the majority of
patients in this phase were randomly assigned to group 1, this may have
introduced an element of selection bias. This would explain the larger
numbers of patients requiring transfusion within the standard care group.
Other disease conditions expected to require blood transfusions, such as
acute lymphocytic leukaemia (ALL), acute myeloid leukaemia (AML),
acute pro-myelocytic leukaemia (APML), chronic lymphocytic leukaemia
(CLL), chronic myelo-monocytic leukaemia (CMML), Hodgkin’s
lymphoma, lymphoma, myelodysplastic syndrome (MDS) and
myeloproliferative disease (MPD) demonstrated a relatively more even
spread across the two study groups. Relaxation of the selection criteria
meant that conditions unlikely to require transfusion during the course of
the disease, such as idiopathic thrombocytopaenic purpura (ITP),
coagulation disorders, Von Willebrand’s disease (VWD) and
polycythaemia rubra vera (PRV) were also included in the trial groups.
266
Figure 6.5: Disease conditions in the pilot study patients
8
Number of patients in trial
7
6
5
4
3
2
Group 1
1
0 Group 2
AIHA
ALL
AML
APML
Chronic Anaemia
CLL
CML
CMML
Coagulation disorder
ET
Haemochromatosis
Hairy Cell Leukaemia
Hodgkins
ITP
Lymphoma
MDS
MPD
Myelofibrosis
Myeloma
NHL
PRV
VWD
Waldenstroms
A wide variety of haematological disease conditions were seen in the pilot study patients. The number of patients recruited to the standard care group (group 1) for
each disease condition is shown on the graph as blue columns and the number of patients recruited to the intervention group (group 2) is shown by the red column.
Key:
AIHA = autoimmune haemolytic anaemia CML = chronic myeloid leukaemia MPD = myeloproliferative disorder
ALL = acute lymphocytic leukaemia CMML = chronic myelo-monocytic leukaemia NHL = non-Hodgkin’s lymphoma
AML = acute myeloid leukaemia ET = essential thrombocythaemia PRV = polycythaemia rubra vera
APML = acute pro-myelocytic leukaemia ITP = Idiopathic thrombocytopaenia purpura VWD = Von-Willebrand’s disease
CLL = chronic lymphocytic leukaemia MDS = myelodysplastic syndrome
267
6.2.5 Matching red cell antigens for transfused patients in the
intervention group
Patients allocated to the intervention group had Rh (CcEe) and K antigen
types performed using the automated serological method (Immuncor) and
by the manual laboratory method (Bio-Rad). For each of these patients a
note was added to the laboratory computer record stating the Rh (CcEe)
and K antigen type and the corresponding antigen negative requirements
of red cell units for transfusion. Laboratory staff was requested to record,
on the laboratory computer system, a reason if they were unable to
provide matched red cells for the patient for any of the transfusion
episodes.
A total of 89 red cell units were provided for the four transfused patients
in the intervention group. Patient 1 received a total of 10 units over five
transfusion episodes, patient 2 received 71 units over 34 episodes,
patient 3 received seven units over three episodes and patient 4 received
one unit on one occasion (table 6-1).
268
Table 6.2: Provision of Rh (CcEe) and K matched blood for transfused patients in the intervention group
Four patients in the intervention group (patients allocated to receive red cells for transfusion matched with their own Rh
(CcEe) and K types) received transfusions during the trial period. Three patients received red cell units that were matched
with their own Rh (CcEe) and K antigen type (patient numbers 1, 3 and 4). One patient received a total of 71 units of red
cells, of which just 20 (28.2%) were typed and matched to their Rh (CcEe) and K antigen type.
Patient No Rh (CcEe) and K type Number of red cell Number of transfusion % of units matched
units transfused episodes with Rh (CcEe) and K
type
1 R2r (cDE/cde) K neg 10 5 100%
2 R2R2 (cDE/cDE) K neg 71 34 28.2%
3 R1r (CDe/cde) K neg 7 3 100%
4 R1r (CDe/cde) K neg 1 1 100%
269
All four of the transfused patients were found to be negative for the K
antigen, a phenotype shared by over 90% of the Caucasian population
(Daniels, 2013), and as such expected to be seen in around 90% of
donated red cells stored in a hospital blood bank stock. None of the red
cell units transfused to any of the four patients were K positive, giving a
matching rate of 100%.
Two of the four transfused patients (patients 3 and 4) were found to be
positive for the C, c and e antigens and negative for the E antigen, giving
them the mostly likely Rh type R1r (CDe/cde), a phenotype shared by
32.68% of the English population (Daniels, 2013). Patient 3 was provided
with seven units of blood over three transfusion episodes, and patient 4
was provided with a single unit of blood on one occasion, the Rh (CcEe)
type of all of the units matched with the patients Rh (CcEe) phenotype.
Patient 1 was found to be positive for the c, E and e Rh antigens and
negative for the C antigen, giving a most likely Rh type R2r (cDE/cde), a
phenotype shared by 10.97% of the English population (Daniels, 2013).
This patient was provided with 10 units of blood over five transfusion
episodes, all of the units provided carried the same Rh antigens as the
patient.
Patient 2 was found positive for the c and E antigens and negative for the
C and e antigens, giving a most likely Rh type of R2R2 (cDE/cDE), a
phenotype shared with just 1.99% of the English population (Daniels,
270
2013). This patient required a total of 71 units of blood over 34
transfusion episodes during the trial period, the laboratory were only able
to provide 20 units of blood that matched this patient’s Rh (CcEe)
phenotype. On four occasions a reason was given to explain the lack of
matched blood; it was noted that there were no units in stock that
matched. No reason was noted for the remainder of the transfusion
episodes.
The Rh (CcEe) and K types of all the patients recruited to the trial were
elucidated using the automated extended serological assay (table 6-2).
The most likely phenotypes of the patients are shown in both Fisher-Race
(CDE) terminology (Fisher, cited in Race, 1944) and Wiener’s Rh-Hr
(Wiener, 1943). The D antigen status of the patients was determined
using two commercial reagents; NOVACLONE™ anti-D blend (IgM D175-
2 and IgG D145, IE4) and ImmuClone® IgM (derived from cell line RUM
1). The assays were performed using the standard D typing assay on the
NEO® analyser (Immucor).
The R1r (CDe/cde) phenotype was the most common Rh type seen in the
patients recruited to the pilot study (21/59, 35.6%), R2r was seen in nine
patients (15.2%), R1R1 in nine patients (15.2%), R1R2 in eight patients
(13.6%), rr in 7 patients (11.9%) and R2R2 in five patients (8.5%) (Figure
6.7). The R1R1 phenotype is shared by 17.68% of the English population,
R1R2 by 11.87%, and rr by 15.10% (Daniels, 2013), should the pilot
study patients sharing these phenotypes have required a transfusion the
271
provision of blood should not have proved difficult from routine stocks
given the frequency of phenotypes.
272
Table 6.3: Rh and K types of patients recruited to the pilot study
Rh and K antigen status was determined using the automated extended serological assay for all the patients recruited to the pilot
study. In the table below antigen status is denoted as positive (P) or negative (N) for the antigens C, c, E, e and K, assayed using
the extended automated serological phenotyping assay, the D antigen status was determined using two anti-D clones via the
standard automated assay (NEO®analyser, Immucor). The most likely Rh phenotype is shown in both Fisher-Race and Wiener’s
terminologies.
Patient Trial Antigens Probable phenotype

number Group
C c E e K Fisher-Race Wiener
1 1 P P N P P CDe/cde R1r
2 1 N P P P N cDE/cde R2r
3 1 N P P N N cDE/cDE R2R2
5 1 P N N P N CDe/CDe R1R1
6 1 P P P P N CDe/cDE R1R2
9 1 N P N P N cde/cde rr
10 1 N P P P P cDE/cde R2r
15 1 P P N P N CDe/cde R1r
273
Patient Trial Antigens Probable phenotype (continued)
number Group
274
Patient Trial Antigens Probable phenotype (continued)
number Group
275
Figure 6.6: Frequencies of Rh phenotypes in patients recruited to
the pilot study
12
10
Number of patients
6
Group 1
4 Group 2
0
R1r R1R1 R1R2 R2r R2R2 rr
Most likely Rh phenotype (Wiener terminology)
The graph shows the frequencies of the Rh phenotypes of the patients
recruited to the pilot study (group 1, the standard care group (n=30) and
group 2, the intervention group (n=29)) denoted using the Wiener
terminology. The most common phenotype seen was the R1r (35.6%),
other phenotypes were seen with the following frequencies; R2r (15.2%),
R1R1 (15.2%), R1R2 (13.6%), rr (11.9%) with R2R2 (8.5%) being the least
common.
276
6.3 Discussion
The pilot study demonstrated that patients responded well to the
recruitment process and the information that was provided regarding the
trial, with only one patient declining to participate. The recruitment
process itself, however, was slow, particularly when selection was limited
to patients with diagnoses anticipated to require transfusion support.
Recruitment increased when the selection criteria were relaxed to include
patients with any disease condition. However, the number of patients in
each group that actually received a transfusion was low, particularly in the
intervention group where only 13.8% (4/29) of the patients received a
transfusion. Comparison of the number of red cell units transfused and
the number of transfusion episodes for the transfused patients in the
standard care group and the intervention group demonstrated that
patients were well matched in each group. The pilot study was not
powered to investigate any differences in the rate of antibody
development in the groups and it was noted that no patients in either
group developed an allo- or autoantibody as a result of the red cell
transfusions given.
No patients requested to drop out of the trial, suggesting that recruitment
to a larger multi-centre trial would not pose significant issues, however,
two patients died during the course of the trial. When estimating
participant numbers for a multi-centre randomised controlled trial to
277
investigate the benefit of a typing and matching strategy for patients with
haematological conditions or renal insufficient, it should be remembered
that these patients have serious conditions and that the death rate in the
study groups may be high. In addition, the pilot study showed low rates of
transfusion in patients recruited to the trial. This was mainly attributed to a
relaxation of the selection criteria in order to recruit sufficient patients in a
relatively short space of time. In order to undertake a larger RCT,
powered to detect any significant differences in antibody development
rates in patients given standard care compared to those given antigen
matched units, it would be preferable to maintain a stricter selection
criteria based on haematological disease conditions considered most
likely to require transfusion support.
Randomisation is a technique well known and extensively used in clinical
trials to prevent selection bias during the recruitment of subjects to the
trial and also reduce the risk of accidental bias during the data analysis
(Rosenberger and Lachin, 2015). It also allows the use of probability
theory to investigate the likelihood of chance as a source for any
differences seen in the end outcome of the trial. Randomisation is
particularly useful for the investigation of different treatment or
intervention regimes; whether one particular treatment is better than
another, no worse than another or equivalent. Randomisation has the
benefit of ensuring that subjects have an equal chance of being allocated
to one group or another and that the groups are balanced with respect to
known, and unknown, confounding or prognostic variables and are an
278
essential tool for testing the efficacy of a treatment (Suresh, 2011). If
randomisation is not used at the start of a clinical trial then the influence
of imbalance that is introduced needs to be controlled for in the analysis
of the results in order to produce an unbiased result (Kalish and Begg,
1985; Fleiss et al., 2003), using techniques such as analysis of
covariance (ANCOVA) or multivariate ANCOVA. Randomisation is
essential to ensure that the treatment effects are reviewed appropriately.
Clinical trials that been set up with unclear or inadequate randomisation
have been reported to over-estimate the treatment effects by up to 40%
(Schulz and Grimes, 2002a). In order to further reduce the potential for
selection bias during the randomisation process, allocation concealment
can be used, a technique which shields the researcher from the
knowledge of the group to which subject will be allocated.
There are different types of randomisation that can be used for clinical
trials; simple randomisation, block randomisation, stratified randomisation
and covariate adaptive randomisation. Simple randomisation involves a
single sequence of random assignments, which can be achieved by a
process as simple as flipping a coin or rolling a dice. This technique is
useful in clinical trials involving large numbers of subjects (200 or more)
but can result in unequal numbers between the groups in trials involving
smaller numbers of subjects (Altman and Bland, 1999a). In order to
attempt to equalise the number of subjects being allocated to each group,
block randomisation can be used, the total number of subjects required
for the trial is broken down into smaller blocks with pre-determined group
279
assignments which has the effect of keeping the subject numbers similar
at all times (Frane, 1998; Altman and Bland, 1999b). The downside to
block randomisation is that it can introduce some bias as the groups
could then have imbalances as the subjects are recruited, such as gender
or disease type inequalities (Pocock and Simon, 1975) and it can harm
the unpredictability of treatment assignments (Schulz and Grimes,
2002b). Reduction of the bias effects that simple or block randomisation
may introduce within trial groups can be achieved by using stratified
randomisation. This technique requires that any covariates that need to
be controlled within the trial are identified prior to the selection process
and can then be applied to each participant. This approach to
randomisation can prove difficult in clinical trials where subjects are
recruited on an individual basis or if there are numerous covariates that
need to be controlled (Weir and Lees, 2003). In clinical trials involving
small to moderate numbers of subjects, covariate adaptive randomisation
can be used, where recruits are assigned to one of the treatment groups
with reference to any covariates and also assignments of previous
recruits allowing balance of a large number of covariates (Kalish and
Begg, 1985; Hu et al., 2014).
The randomisation technique used to allocate patients to the standard
care or the intervention group in this pilot study resulted in an unexpected
bias towards the standard care group for the first 30 patients recruited.
This meant that patients in the standard care group spent significantly
longer on the trial than patients in the intervention group. To reduce the
280
risk of this occurring in a larger RCT, a block or stratified randomisation
approach should be utilised. Stratified randomisation would allow control
of the important variables, gender and disease condition, whilst ensuring
that patients in each study group have equality in the length of time on
the trial.
This pilot study demonstrated issues with provision of antigen matched
blood from routine stocks held in a hospital transfusion laboratory for
patients with rarer Rh (CcEe) phenotypes. NHSBT blood centres test and
distribute blood donated from a very wide geographical area, as such the
blood stocks held within a hospital transfusion laboratory should
represent the distribution of phenotypes in that general population. Thus,
provision of matched blood for patients with common phenotypes such as
K negative and the Rh type R1r should be relatively easy, whereas
provision of matched blood for the less common Rh phenotype, R 2R2, is
more difficult. This is supported by a study by Klapper and coworkers
(2010) using a virtual type and match strategy which demonstrated that
blood units could be provided from routine stocks for over 90% of
requests if providing K and Rh matches. ABO blood group type will also
impact on the provision of matched blood from routine stock; patients with
blood type O will only be able to received matched blood from the stock
of group O, whereas patients of blood type A are able to receive matched
blood of type A or O, increasing the probability of finding matching blood
within the stocks held, and patients of blood type AB can receive donor
blood of type O, A, B or AB, further increasing the likelihood of finding a
281
match in the stock fridge. Patient 2 in this pilot study was of group O type,
as well as being a less common R2R2 Rh type; this meant that selection
of matched blood was restricted to the stocks of group O RhD positive.
Patients with haematological diseases or renal insufficiency rarely
present with emergency requirement for blood transfusion and so this
should allow laboratory staff time to order blood of the appropriate Rh
(CcEe) and K antigen type for specific patients from the NHSBT. A good
communication system between the haematology/renal clinics and the
laboratory, with regard to patients with transfusion requirement,
adherence to protocols ensuring that transfusion samples are sent to the
laboratory two working days in advance of transfusion, would help
laboratories ensure that they have appropriately matched units of blood in
the stock fridge ready for the patients. Ethnic disparity between the
donor pool and patient population would need to be taken into account
when making changes to the blood stock inventory control and ordering
system to support a type and match strategy. Some Rh types show
disparity, such as R1 which is more common in White than Black
populations, and R0 which is more common in Blacks than Whites
(Daniels, 2013). Ethnic disparity was not considered to impact on the
provision of blood in this pilot study, and no issues were noted with
provision of blood, by the NHSBT, of any particular Rh phenotype.
Improvements in provision of matched blood by the laboratory could be
achieved simply by rearranging the way that the units are stocked and
setting optimum numbers of each phenotype for blood ordering.
282
The results of the retrospective review noted an alloimmunisation rate of
approximately 67% for Rh and Kell alloantibodies for both the renal and
the haematology cohort. The pilot study demonstrates that a type and
match strategy for Rh and Kell antigens could be implemented to reduce
the alloimmunisation rate. Changes would be needed to the blood stock
arrangement and the ordering procedure to ensure that sufficient
numbers of each phenotype were available. There would be no cost
implications for phenotyping the red cell unit, for Rh and Kell, as this is
performed at the NHSBT; however, there are cost implications for typing
the patients which need to be explored (see chapter 7).
283
7 COST ANALYSIS
7.1 Introduction
This study has shown that the implementation of extended automated
serological phenotyping, or genotyping, assays prior to transfusion in
patients with disease conditions likely to require chronic transfusion
support, followed by matching of blood for, at least, Rh (CcEe) and K
antigens, has the potential to reduce the risk of development of red cell
antibodies. However, the cost of implementation of these assays needs to
be assessed and balanced against the potential savings that could be
realised if the development of red cell antibodies were to be prevented.
This study has also shown that both the extended automated serological
phenotyping assay (NEO® analyser Immucor) and the BLOODChip
IDCOREXT genotyping assay are robust and reliable techniques for the
identification of red cell antigens within the hospital transfusion service
setting. The cost effectiveness of both of these assays was assessed
using the data regarding additional laboratory testing collected as part of
the retrospective review. This is the first time that a cost analysis for a
prospective type and match program has been performed in the UK.
7.2 Automated extended serological phenotyping
The total cost of additional laboratory testing incurred by the immunised
patients in the haematology cohort (n=256/1107) was £48,752.59 (see
section 3.2.5). In addition to this cost, each unit of blood for transfusion in
this group of patients required compatibility testing by IAT assay, a test
284
calculated to cost £3.81/unit (RD&E 2014 cost) based on compatibility
testing (crossmatching) by IAT in Column Agglutination (Bio-Rad) by a
Band 5 Biomedical Scientist at pay point 20 (Agenda for Change, NHS
2015). During the retrospective review of the transfusion needs of the
haematology patients, data was not collected on the type of crossmatch
assay performed for each unit transfused for patients with red cell
antibodies and so it is not possible to give accurate costs for compatibility
testing for the immunised group in this patient cohort. However, using the
average number of units transfused to the immunised group (n=256),
46.99 units, and the 2014 cost for compatibility testing (£3.81), a total
cost of £45,711.87 can be estimated for this group for compatibility
testing. This figure is likely to be an overestimate as the majority of
patients in the immunised group developed red cell antibodies following
transfusion and so, initially, provision of blood would not have required
serological crossmatch by IAT. Additional laboratory testing performed for
the immunised patients in the retrospective review amounted to
£48,752.59, this figure, plus the estimated cost of compatibility testing,
gives a total cost of £94,464.46 for laboratory tests performed over the
review period. This gives a cost per head for the immunised haematology
patients (n=256) of approximately £369.00. Additional testing for the non-
immunised patients in the haematology cohort (n= 851) amounted to a
total of £1,509.83. Again, using a cost for compatibility testing based on
the average number of units given to patients in this group (23.43 units)
and the 2014 cost for provision of blood without serological compatibility
testing by IAT (electronic issue, £0.47/unit), it can be estimated that this
285
group incurred a total cost of £9,371.30 for compatibility testing. This
figure, together with the additional testing figure, gives a total of
£10,881.13 for the non-immunised patients, equating to a cost per head
of £12.79/patient. Overall, the cost of provision of blood (additional testing
plus compatibility testing) for the entire haematology cohort (n=1107) can
be estimated at £105,345.59, giving a cost per head of £95.16 per
patient.
Within the renal cohort (n=877) the immunised group (n=134) incurred
additional testing costs of £13,094.76 and the non-immunised group
(n=743) incurred a cost of £155.53. Using the average number of units
transfused to patients in each group (21.85 and 11.97 respectively) and
the cost of provision of blood by IAT crossmatch for the immunised
patients (£3.81/unit) and electronic issue (£0.47/unit) the total cost for
provision of blood for the immunised patients amounted to approximately
£24,250.06 and £4,180.04 for the non-immunised patients. The cost per
head for immunised renal patients can be estimated at £180.97 and for
non-immunised patients at £5.62, the overall cost per head for the entire
renal cohort (n=877) is estimated at £32.41 per patient.
Automated extended serological phenotyping for Rh (CcEe), K, M, N and
Cw antigens cost £9.43 per patient and for Jka, Jkb, Fya, Fyb, S, s and k
antigens the cost amounted to £10.14 per patient, giving a cost per head
of £19.57 per patient. Both figures are considerably less than the current
£95.16 per head for the haematology cohort and £32.41 per head for the
286
renal cohort for blood provision. This suggests that extended automated
serological phenotyping for all haematology or renal patients could be
justified if the patient phenotype could then be used to match blood for
transfusion and reduce the risk of allo and/or auto-immunisation to red
cell antigens. This is particularly relevant for the typing and matching of
Rh (CcEe) and K antigens, for which donor blood is automatically typed.
Automated serological phenotyping for Rh (CcEe) and K antigens only
would amount to £5.97 per patient, this is significantly less than the cost
per head estimated for the immunised haematology and renal patients
(£369.0 and £180.97 respectively) and, furthermore, less than the
estimated cost per head for the non-immunised patients in the
haematology cohort (£12.79) and similar to the cost per head for the non-
immunised renal patients (£5.62).
Manual extended serological phenotyping, at a total cost of £50.46 per
head, could be justified on a cost basis for haematology patients (cost per
head for immunised and non-immunised estimated at £95.16), but not for
renal patients (cost per head for immunised and non-immunised
estimated at £32.41). As for automated serological phenotyping, test cost
for manual Rh (CcEe) and K antigen testing (estimated at £6.49 per
patient) would appear to be a cost effective option if it could be
demonstrated that matching blood for transfusion for these antigens
prevents antibody production. However, the risks of manual testing on
this scale would need to be carefully considered if this approach was
used.
287
7.1 Genotyping Assay
The cost per patient test for the IDCOREXT test was estimated at £73.66.
On a cost basis alone, this technique could be justified for all patients with
haematological conditions prior to initial transfusion, but not for renal
patients. This technique, however, required equipment that is not
currently used in hospital transfusion services, such as extraction hoods,
thermocyclers and the Luminex XMap analyser. The initial set up costs of
introducing these systems into a routine hospital transfusion laboratory
and the staff training required would also need to be taken into account.
Costs for equipment may be kept to a minimum if other disciplines within
the hospital, such as microbiology or immunology, were to share the
equipment, but this may mean that the different aspects of sample
preparation and analysis may take place in geographically separate areas
of the hospital.
7.2 Cost comparison of serological phenotyping and
genotyping techniques
Pre-transfusion phenotyping, in conjunction with a strategy for matching
for the Rh (CcEe) and K antigens, appears to be a cost effective option
when comparing the estimated cost per head for provision of blood for
immunised patients (£369 for haematology patients and £180.97 for renal
patients) with the cost per test for serological phenotyping for the Rh
(CcEe) and K antigens (£5.97 for automated technique) (Table 7-1). The
assumption being that serological phenotype prior to transfusion
combined with a type and match strategy would reduce risk of
288
alloantibody development and associated costs with subsequent
transfusions. Pre-transfusion genotyping appears to be cost effective but
is less easy to justify in clinical practice as described in section 7.1.
289
Table 7.1: Cost effectiveness of serological phenotyping and genotyping techniques
The estimated current costs for the provision of blood for immunised haematology and renal patients per head (standard practice
column), which includes the additional testing required once the patient has developed a red cell allo/autoantibody is considerably
higher than that for pre-transfusion phenotyping or genotyping. Pre-transfusion phenotyping, or genotyping, combined with a
strategy for matching for the Rh (CcEe) and K antigens, both appear to be a cost effective options with the assumption that such a
strategy would prevent the development of antibodies against these antigens.
Standard practice Phenotype for Rh (CcEe) and K antigens Extended genotype

Automated serological Manual serological
Haematology cohort Renal cohort phenotyping phenotyping Genotyping
Non- Non-
Immunised immunised Immunised immunised
£369.00 £12.79 £180.97 £5.62 £5.97 £6.49 £73.66
290
7.3 Discussion
The additional costs incurred for the provision of blood for patients who
have developed atypical red cell allo/auto-antibodies are considerably
higher than the cost for provision of blood for patients without antibodies,
in both the haematology cohort and the renal cohort. There is potential to
reduce this financial burden by the introduction of high throughput
methods of red cell antigen testing and then matching of blood for
transfusion in these chronically transfused patients. Manual and
automated serological red cell antigen phenotyping techniques are useful
methods which can be easily incorporated into routine hospital
transfusion laboratories. Manual techniques have disadvantages in that
they are labour intensive, making them expensive and also introducing
human error risks in the sample preparation, testing, interpretation and
reporting stages. Automated readers, interfacing to laboratory computer
systems and automated interpretation could eliminate some of these risks
but would incur costs for computer upgrades and equipment, if not
already used within the laboratory. Automated serological testing, as
validated in this study, has been shown to be a robust and cost effective
alternative to current manual systems. Interfacing to laboratory computer
systems may incur additional one-off costs, but this would facilitate the
high throughput testing required to antigen type large numbers of patients
by eliminating the need for manual transcription of results.
291
The main disadvantage of serological testing, manual or automated, is
that it cannot be used for recently transfused patients due to circulating
donor cells. This would not be an issue if the hospital transfusion service
chose to adopt the approach of automated serological red cell antigen
typing for all haematology and renal patients prior to initial transfusion,
but would restrict the use of the technique if used only for patients post
development of red cell antibodies. In addition, serological techniques,
manual and automated, using polyclonal anti-sera, such as those used in
the detection of antigens in the Kidd, Duffy, and Ss systems cannot be
used for patients with red cell autoantibodies due to the risk of false
positive results (BCSH, 2013). Again, if used prior to the initial transfusion
for haematology and renal patients this would not pose a significant
issue, but the considerable numbers of patients with autoantibodies seen
in both patient cohorts (186/1107 (16.8%) for haematology patients and
74/877 (8.4%) for renal patients) would restrict the use of the technique if
only implemented post development of antibodies. Genotyping
techniques, however, offer a flexible solution to red cell antigen typing
that can be used for red cell antigen typing using samples from patients
post-transfusion and from those with red cell autoantibodies (Guelsin et
al., 2010).
The potential reduction of costs for additional antibody identification tests
and serological crossmatching could not be realised unless red cell
antigen typing, serological phenotyping or genotyping, was implemented
prior to initial transfusion and the blood provided for transfusion was
292
matched for, at least, the Rh (CcEe) and K antigens. Such a strategy
could significantly reduce the current costs for provision of blood for
patients within the haematology and renal cohorts, if the development of
red cell antibodies to these antigens was prevented. The inclusion of
extended phenotyping techniques, to include red cell antigens within the
Jk, Fy, and Ss systems, would increase the initial costs for phenotyping.
Matching of blood for these antigens is more complex, primarily because
donor blood units are not automatically typed for these antigens but also
because attempting to match at this secondary level reduces the
likelihood of finding matched units within the routine blood stocks held by
the hospital transfusion laboratory.
In addition, other benefits to patients could be realised, such as reduction
in delays in provision of blood and a reduced risk of transfusion reactions.
Individualising patient care in this way also has the potential to improve
the patient journey, particularly for patients with haematological and renal
diseases whose care and treatment can be prolonged and emotionally
challenging.
293
8 FINAL DISCUSSION AND CONCLUSIONS
In order to implement a strategy of red cell antigen typing and matching of
blood for transfusion the hospital transfusion service requires a cost-
effective, robust, high throughput method for extended red cell
phenotyping, or genotyping, of patient samples prior to transfusion.
Current methods in hospital transfusion services for red cell antigen
testing are serological, mainly manual, or, at least, require a selection of
techniques for determination of individual antigen status. This results in a
cumbersome, costly system with all the inherent risks of manual testing
and manual input of results, which is then reserved for patient groups
where there is a national requirement for pre-transfusion testing, namely
the sickle cell disease and thalassaemia groups (BCSH, 2013).
Red cell antigen typing of patients red cells following production of an
alloantibody is generally limited to testing only for the implicated red cell
antigen, requiring a local decision regarding typing and matching for
further antigens (BCSH, 2013). Decisions need to be made in conjunction
with risk assessment, taking into account the limitations with the manual
testing system, costs, and the amount of staff time required to perform
extended typing. This study demonstrated that automated extended
serological red cell antigen typing using the IBG Immucor NEO® analyser
is a robust, cost effective method suitable for high throughput activity
within a hospital transfusion service setting, giving the laboratory access
to a robust assay for extended antigen typing. The method also lends
itself to the addition of an electronic interface from the analyser to the
294
hospital LIM system, removing the requirement for manual input of results
and reducing the risk of error. The automated system has built in quality
control to ensure that sufficient antisera and patient red cells are added to
the reaction mixture, a feature that is difficult to control using a manual
system where such judgments are made by macroscopic reading. Testing
of the automated system demonstrated that strongly positive reactions
are produced for heterozygous and homozygous antigen expression in
both commercial control cells (NHSBT red cell antigen cells) and patient
red cells. Red cell antigen typing using polyclonal antisera in the gel
column agglutination manual techniques employed in this study did not
demonstrate comparable differences in heterozygous and homozygous
antigen expression reaction strengths. It is possible to automate gel
column agglutination tests for red cell antigen typing and it would be
interesting to perform a further comparative study between the IBG
Immucor extended red cell antigen profiles and similar profiles created on
an automated column agglutination analyser. Weakly positive results
demonstrated in this study in the automated system were shown to be
false positives; although the numbers involved were small the specificity
of the automated system could be further improved by instructing the
analyser to invalidate any weak reaction (1+ or 2+ reaction strengths).
The automated system created for testing in this study included an auto
control (patient red cells tested against a non-reactive reagent, designed
to identify non-specific agglutination) as part of the polyclonal antisera
testing profile to minimise the risk of false positive results due to a
295
positive DAT. Only one patient in the study group proved to have a
positive DAT, which was identified due to weakly positive reactions in
both the automated and manual systems, not by the auto control included
in the automated profile. Further testing of a larger number of DAT
positive patients in the automated system is necessary to demonstrate
that inclusion of the auto control will identify DAT positive patients
effectively. In addition, invalidation of weakly positive results will also
reduce the risk of reporting false positive results. The monoclonal reagent
testing profile (anti-C, anti-c, anti-E, anti-e, anti-K, anti-Cw, anti-M and
anti-N) created on the IBG Immucor analyser did not include an auto
control, this was not felt to be a necessary addition as, if this profile was
included as part of pre-transfusion testing, the sample would also be ABO
and D typed by the same analyser, a test profile that already includes an
auto control.
There are currently no commercial quality control samples available
specifically for extended red cell antigen typing profiles in automated
systems. Whole blood quality control cells are available for use in the IBG
Immucor automated systems (WBCorQC), although these were tested
and found to be useful for internal quality control of the automated Rh
(CcEe) and K antisera, they are pooled cells and, as such, cannot be
used for internal quality control of the remaining antisera. The NHSBT 2
cell profile (3% suspension in Alsevers) is marketed for alloantibody
detection but was found to be suitable as an internal quality control
system for the extended red cell antigen typing profiles in this study.
296
However, these cells are not supplied in tubes compatible with use on the
analyser and therefore had to be transferred into suitable tubes,
introducing a potential source of error and removing the potential for any
audit trail and reagent expiry date control on the analyser. To further
improve the use and marketability of the IBG Immucor extended red cell
antigen typing profiles the production of red cell controls, containing cells
with negative red cell antigen expression and, preferably, cells expressing
heterozygous red cell antigens should be explored. Another improvement
to the automated extended red cell antigen phenotyping profiles would be
the replacement of the current polyclonal reagents (anti-Fya, -Fyb, -Jka, -
Jkb, -S, -s and –k) with monoclonal reagents. Not only would this reduce
the cost of testing but it would also allow the use of the reagents on
samples from patients with a positive DAT, currently one of the major
limitations of this assay.
Test costs for automated red cell antigen typing profiles were found to be
considerably cheaper than those for performing the same tests by manual
methods in our institution. In addition, the reduction in staff time required
for performing the automated profiles make this an attractive method for
extended red cell antigen typing in a hospital transfusion service setting.
However, the inclusion of internal quality control material for the
automated system also needs to be taken into account; decisions would
need to be made at a local level as to the frequency of reagent quality
control and the source of the material until such time as commercial cells
become available.
297
If one makes the assumption that typing and matching of blood for
transfusion for Rh (CcEe) and K antigens would eliminate alloantibodies
to Rh C, c, E, e and K, using patient numbers from the retrospective
study patient group (haematology patients), use of such a strategy would
have eliminated the production of Rh and Kell antibodies in 78 of the
1107 patients. Using the average number of antibody identification panels
performed for the retrospective review patients with red cell alloantibodies
(14.25), the total cost of antibody identification panels performed for the
78 patients was £7,913.88. If all 1107 patients included in the
retrospective review of haematology patients had been tested using the
monoclonal reagent testing profile the total cost would have amounted to
£9,232.38. Inclusion of costs for crossmatching (£3.81 per unit) for each
unit transfused for the 78 patients with Rh related or K antibodies, based
on an average of 46.99 units given to immunised patients in this cohort,
gives a total of £13,964.49. The total cost of antibody identification panels
and crossmatching for the 78 patients in the haematology cohort with Rh
related or K antibodies is estimated at £21,878.37, compared to a cost of
£9,232.38 for performing an automated extended phenotyping profile
including Rh (CcEe), K, M, N and Cw antigens for all the haematology
patients in the cohort. Making the assumption that pre-transfusion
phenotyping and matching of blood for Rh (CcEe) and K could prevent
the production of antibodies against these antigens, makes, in theory,
pre-transfusion phenotyping using an automated assay, a cost effective
system for patients with haematological conditions.
298
To realise the cost benefits of introducing a pre-transfusion phenotyping
strategy, particularly tying for Rh (CcEe) and K antigens, for patients with
haematological conditions at most risk of developing an allo/autoantibody
is recommended, this would include patients with haemolytic anaemia,
aplastic anaemia, Waldenström macroglobulinaemia, chronic lymphocytic
leukaemia, chronic myelo-monocytic leukaemia, acute myeloid
leukaemia, myelodysplastic syndrome and myeloproliferative disease,
shown to be high risk for antibody development in this study. The high
proportion of antibodies to Rh and Kell antigens seen in patients in both
the haematology cohort (66.7% of alloantibodies) and in the renal cohort
(67.0%) would support the implementation of a typing and matching of
blood for transfusion strategy for these antigens. Restriction of this
strategy to female patients within the renal cohort could further improve
the cost effectiveness of pre-transfusion antigen typing and matching of
blood, as females in this cohort were seen to have a greater risk of
antibody development than males.
Implementation of pre-transfusion testing for all patients, prior to the
primary transfusion, using the polyclonal reagent testing profile (Jka, Jkb,
Fya, Fyb, S, s, k antigens) is less easy to justify on a cost basis alone,
even if restricted to the patients at most risk of development of
allo/autoantibodies. However, it could be considered a useful addition to
the hospital transfusion service repertoire of tests for patients following
the production of a red cell alloantibody for those patients not on a regular
transfusion regime. It could also be useful, and possibly cost effective, for
299
those hospital transfusion services dealing with more substantial numbers
of sickle cell disease and Thalassaemia patients than those seen in our
institution.
The use of serological phenotyping methods is limited in the presence of
autoantibodies and not recommended if the patient has been recently
transfused due to interference by transfused red cells (BCSH, 2013). In
these cases the phenotype may be predicted using genotyping methods
(BCSH, 2013). In this study red cell antigen status predicted by the
BLOODChip IDCOREXT blood group genotyping technique showed a
high correlation with the serological antigen status in the 47 patients
tested. In addition, the IDCOREXT included assays covering a wider
range of red cell antigens than that covered by the automated serological
assay. Although typing and matching of blood for all antigens detected by
the genotyping system is not currently possible, the additional information
provided is useful in the event of the patient developing alloantibodies
post transfusion.
The IDCOREXT technique was simple to use, with a low failure rate and
high concordance with phenotyping results. However, the DNA
preparation phases are manual and labour intensive, and, as such, do not
currently lend them to the hospital transfusion setting where staff may not
be specialised in genetic techniques and the risk of error makes manual
testing undesirable. The advent of platforms with integrated DNA
extraction and real-time PCR, such as those used for microbiological
300
assays (Riedlinger et al., 2010; Le Guern et al., 2012; Dalpke et al.,
2012), may mean that these types of specialised tests can be transferred
into the routine service setting of the hospital transfusion laboratory. The
cost of blood group genotyping is also currently prohibitive for routine
use, the cost per test of the IDCOREXT assay alone is approximately £47
with the additional costs for DNA extraction, amplification, staff time and
analyser costs pushing the total cost to around £73 per patient.
Consequently, this type of assay is currently reserved for chronically
transfused patients with multiple antibodies. As DNA testing in general
becomes more mainstream and automated, it is likely that the costs will
reduce and, in the future, this may succeed serological phenotyping as
the routine test for pre-transfusion typing and matching strategies and
confirmatory testing for patients with red cell antibodies.
In the future, the potential for next generation sequencing (NGS), or
massively parallel sequencing, is being explored as current methods
detecting SNPs will only detect known blood group genotypes whereas
NGS will detect unknown ones as well (Altayar et al., 2013). NGS also
allows for very high throughput testing making it ideal for the testing of
donated blood. The use of NGS has been suggested as a potential
platform to accurately characterise the blood group phenotypes in
patients with sickle cell disease, particularly those in the African American
population who have a high RH allelic diversity (Reid et al., 2014).
301
Treatment of patients with haematological conditions using allogeneic
bone marrow, or stem cell, transplants may introduce a further
complication if a strategy for red cell antigen typing and matching of blood
for transfusion is implemented. Allogeneic bone marrow transplant (BMT)
replaces the recipient haemopoetic cells with those taken from a healthy
donor. This technique has been in use for over 20 years and is mainly
used as a treatment for leukaemias, lymphoproliferative disorders, solid
tumours and some non-malignant disorders (EBMT, 2011). The
conditioning regimes available, designed to eliminate malignant disease
(Gyurkocza and Sandmaier, 2014), should also eradicate immune system
cells in the bone marrow of the recipient, leaving the recipient immune
suppressed and tolerant to any foreign red cell antigens introduced by the
transplant. However, there are reports of recipient antibodies being
detected many months, and even years after the transplant (Sparkes et
al., 1979; Witherspoon et al., 1978; Korver et al., 1987; Van Tol et al.,
1996; Ang et al., 1997) suggesting that some recipient plasma cells may
survive the conditioning regime and exist long term. Izumi and co-workers
(2003) report an interesting case of a patient with pre-existing
alloantibodies to Rh E and c antigens, the patient received a bone
marrow transplant from an Rh E+ c+ donor and then suffered from
chronic persistent haemolytic anaemia, with positive DAT and IAT
present, for approximately 20 months after the BMT (Izumi et al., 2003).
Implementation of red cell antigen typing and matching of blood for
transfusion for patients prior to, during and after allogeneic bone marrow
transplantation would require special attention to the red cell phenotype
302
of the recipient, donor and that of the blood selected for transfusion at the
various stages of the transplant.
The pilot study performed suggested that recruitment to a larger, multi-
centre RCT would not pose any real issues, however care would need to
be taken with the estimate of the number of recruits to the trial as the
death rate of patients in the pilot study is suggestive of a relatively high
mortality rate in this cohort. Also, the recruitment process would need to
include selection for patients considered most likely to require blood
transfusion in the course of their disease treatment as demonstrated in
this study, such as myeloma, MDS, MPD, acute and chronic leukaemias,
to ensure that sufficient blood transfusions were performed in patients
within each arm of the study. Adequate numbers of patients would need
to be recruited to ensure that the trial was powered to investigate
differences in the rate of development of allo/autoantibodies in patients
receiving standard care and those receiving antigen matched blood. An
estimate of the sample size required for a two arm, binary, superiority trial
designed with a 90% power, which would allow a sample size large
enough to be reasonably confident of demonstrating that one of the
treatment arms is superior to the other, can be achieved using a power
calculator (Sealed Envelope, 2012). In such a trial the binary outcome
would have two measures, chronically transfused patients develop Rh
and Kell related antibodies or they do not develop antibodies. An estimate
of the percentage “success” in the control group and the intervention
group need to be established. In this scenario, given that approximately
303
25% of chronically transfused haematology patients in the retrospective
review developed allo/autoantibodies and 67% of the alloantibodies were
Rh or K related antibodies, an estimate of 16.75% of patients in the
control arm (non matched blood) would make antibodies, compared to
0% in the intervention group (given RhCcEe and K matched blood). Using
the power calculator 106 patients (53 in each arm) would be required to
have a 90% chance of detecting, as significant at the 5% level, a
decrease in the primary outcome measure from 16.75% in the control
group to 0% in the intervention group. However, it is prudent to proceed
cautiously when determining sample size to reduce the risk of under
powering the trial and being unable to make conclusions from the data.
The pilot study revealed issues with availability of antigen matched blood
for the less common phenotypes, these issues could be overcome by
improving communication between the haematology/renal clinic and the
laboratory and early pre-transfusion sample taking, giving laboratory staff
sufficient time to order red cell units with appropriate antigen types. Re-
design of stock holding could also facilitate supply of matched units; in
this pilot study no changes were made to the way that blood units are
stored in the blood stock fridge. Current local practice is to store units
according to ABO and RhD group, a small stock of c, E and K negative
(group O and A) are held separately, but there is no further separation of
blood according to Rh and Kell antigen types. Further separation of the
blood units could enable easier identification of matching units, which is
currently performed by searching through the units, and provide the basis
304
for an optimum order level for antigen negative units to ensure there are
always matching units available, even for patients with the rarer Rh types.
In conclusion, implementation of a strategy for red cell antigen typing and
matching of blood for transfusion in patients with haematological
conditions and renal insufficiency would appear to be an achievable and
beneficial objective. This is supported by a recent study which showed a
significantly lower rate of alloimmunisation of haematology patients who
were phenotyped and matched for Rh/K antigens compared to a group
who only received matched red cells after developing an antibody (Baia
et al, 2015). An automated platform for rapid, cost effective high
throughput extended serological red cell antigen typing, using the
Immucor NEO® analyser system, has been demonstrated to be suitable
for use in hospital transfusion services, with the potential for development
of electronic transfer of results to the host Laboratory Information
Management System (LIMS), which would allow hospital transfusion
services to implement a type and match strategy encompassing large
numbers of patients. The feasibility of introducing a method for blood
group genotyping into a routine hospital transfusion service setting has
also been assessed. Although the manual aspects of blood group
genotyping would appear to preclude this assay from the majority of
hospital transfusion laboratories, it remains a valuable tool for reference
laboratories. Advances in automated genotyping techniques may see
these assays being used in hospital transfusion laboratories in the near
future. Cost analysis of the serological phenotyping assay used in this
305
research has demonstrated that this technique is cost effective if used
within a type and match program for Rh (CcEe) and K antigens. The
assay for extended serological phenotyping for other blood group
systems, assessed in this research was not found to be cost effective for
a type and match program. However, this assay is a valuable tool in the
hospital transfusion laboratory for serological phenotyping for non-
chronically transfused patients with alloantibodies, and has considerable
advantages over manual techniques. The small pilot study performed
demonstrated that a type and match program for Rh (CcEe) and K
antigens can be achieved using the automated serological phenotyping
assay for chronically transfused patients. Patients can be provided with
matched blood from routine blood stocks held in the laboratory, but
provision could be improved with changes to the stock ordering and
holding arrangements. In addition, assurance of correct matching of blood
could be improved by the addition of algorithms within the LIMS. The pilot
study has also demonstrated that a large scale, multi-centre, randomised
controlled trial could be conducted in the UK to investigate the potential of
a typing and matching approach in reducing the risk of allo and/or
autoantibody production and decreasing the cost of laboratory testing in
these patient cohorts. The retrospective review of chronically transfused
patients demonstrated that these cohorts of patients are at risk of
development of alloantibodies, particularly to the Rh and Kell blood group
systems. A prospective type and match program has the potential to
reduce the transfusion complications for these patients and also reduce
the cost of provision of blood for the health service.
306
The results of this research have been utilised to implement a type and
match program and drive the changes required within the hospital
transfusion service at the RD&E to ensure that the program is robust and
successful. Patients have been enrolled on the program from January
2017. The automated serological phenotyping system has been
introduced within the laboratory and the LIMS has been upgraded to
incorporate information on the patient antigen status. All new
haematology, renal, oncology and paediatric oncology patients are typed
for Rh (CcEe) and K antigens prior to first transfusion, patients are
identified by the laboratory staff based on the location and/or clinical
details provided on the test request form. Previously transfused patients,
with no red cell antibodies and with no evidence of transfusion within four
months are also included on the program. The LIM system has also been
upgraded to include an algorithm for cross checking the patient antigen
requirements (included in the computer records as “antigen negative” and
detailing the RH (CcEe) and K antigen for which the patient is negative)
against the “antigen negative” status of the red cell unit selected for
transfusion by the laboratory staff. The computer system allows an
override if there is a mis-match in the antigen negative status of the unit
and the patient, for cases of emergency transfusion when matched units
cannot be supplied from the routine stock. The antigen matching
algorithm is also included in the interface for red cell units dispensed from
the remote allocation fridges, a system which does not include interaction
from the laboratory staff. The blood stock storage arrangements have
307
been altered to allow red cell units to be stored in individual trays with the
Rh (CcEe) and K antigen status clearly marked to enable easy and swift
identification by laboratory staff. Changes have also been made to the
blood stock ordering process to ensure that sufficient stock of each
RhCcEe and K typed units are available.
A preliminary review of the data showed that, during the 12 month period,
273 patients were identified by the laboratory staff as being in a high risk
group for chronic transfusion. None of the patients had alloantibodies and
all were suitable for serological phenotype with no record of transfusion
within the preceding four months. Of the 273 patients included in the type
and match program, only 175 were transfused (64.1%), suggesting that
identification of patients could be improved. A study by Lin and co-
workers (2017) investigating the impact of prophylactic Rh and K antigen
matching in patients with MDS also noted challenges in identification of
patients. In their study the identification of patients was performed by the
clinical staff and communicated to the transfusion laboratory, they noted
that only 62% of patients that should have received matched blood were
phenotyped. In contrast, in our institution, the identification of patients is
performed by laboratory staff and this would appear to lead to excess of
phenotyping. Further work is planned within the RD&E to address this
issue. Over the study period the patients (n=175) were transfused a total
of 773 (range 1 – 50, average 4.4) units of red cells. No cases were noted
where matching could not be achieved. At the end of the study period
(December 2017) no patients had developed antibodies to Rh or K
308
antigens, however, 2 patients had developed anti-Jka. This is a significant
reduction compared to the alloimmunisation rate noted in the
retrospective review of chronically transfused haematology patients (1.1%
compared to 12.9%) (see chapter 3). The study by Lin and co-workers
(2017) also noted no alloimmunisation to Rh and K antigens for patients
receiving matched blood, but noted a 6% alloimmunisation rate for non-
Rh/K alloimmunisation rate. This difference could be explained by the fact
that the patients in their study were transfused over a longer time frame
(3 years compared to 1 year) and received more red cell units (median 39
units) compared to the RD&E patients (median 2 units). It is accepted that
a type and match strategy for Rh (CcEe) and K antigens cannot eliminate
alloimmunisation to non-Rh/K antigens. However, the fact that two
patients at the RD&E developed anti-Jka after exposure to a small
number of units (one patient received 1 unit and the other received 4
units) has prompted consideration for the inclusion of typing and
matching for Jka and Jkb antigens into the program.
The cost of serological phenotyping for the RD&E patients amounted to
£1629.81. As a crude measure of cost effectiveness, the costs for
antibody identification and serological crossmatching were compared for
the same time period before and after implementation of the program.
The reduction in additional testing post implementation of the program
resulted in savings of £1728.3, although it should be noted that this may
also be affected by other factors, such as changes in the patient
population and a general reduction in blood usage.
309
In conclusion, a type and match strategy would appear to be both
achievable, in terms of service provision, and beneficial, in terms of
laboratory workload, cost effectiveness and patient alloimmunisation rate
for patients likely to require long-term transfusion support.
310
9 REFERENCES
Abdelrazik, A. M., Elshafie, S. M., El Said, M. N., Ezzat Ahmed, G. M., Al-
Gamil, A.-K. A., El Nahhas, M. G. M. and Sady, A. A. B. (2016), Study of
red blood cell alloimmunization risk factors in multiply transfused
thalassemia patients: role in improving thalassemia transfusion practice in
Fayoum, Egypt. Transfusion, 56: 2303–2307.
Agenda for Change NHS pay scale. NHS employers
http://www.nhsemployers.org/case-studies-and-
resources/2017/03/agenda-for-change-pay-scales-and-points-april-2017
(website accessed 20.1.2018)
Ahrens, N., Pruss, A., Kahne, A., Kieswewetter, H. and Salama, A.
Coexistence of autoantibodies and alloantibodies to red blood cells due to
transfusion. Transfusion 2007; 47, 813-816.
Altayar, M., Halawani, A., Kiernan, M., Reynolds, A., Kaushik, N.,
Madgett, T. and Avent, N. Next Generation Sequencing based
genotyping of Duffy, Kidd and ABO blood group genes. Transfusion
Medicine 2013; 23, 62-63.
Alfinito, F., Sica, N., Luciano, L., Della Pepa R., Palladino, C., Ferrara, I.,
Giani, U., Ruggiero, G. and Terrazzano, G. Immune dysregulation and
311
dyserythropoesis in the myelodysplastic syndromes. British Journal of
Haematology 2010; 148, 90-98.
Altman, D. and Bland, J. How to randomize. British Medical Journal
1999a; 319, 703-704.
Altman, D. and Bland, J. Statistics notes. Treatment allocation in
controlled trials: why randomise? British Medical Journal 1999b; 318,
1209.
Alves, V., Martins, P., Soares, S., Araújo, G., Schmidt, L., Costa, S.,
Langhi, D. and Moraes-Souza, H. (2012). Alloimmunization screening
after transfusion of red blood cells in a prospective study. Revista
Brasileira de Hematologia E Hemoterapia, 34, 206–211.
Ang, H., Apperley, J. and Ward, K. Persistence of antibody to human
parvovirus B19 after allogeneic bone marrow transplantation: role of prior
recipient immunity. Blood 1997; 89, 4646-4651.
Arain, M., Campbell, M., Cooper, C. and Lancaster, G. What is a pilot or
feasibility study? A review of current practice and editorial policy. BioMed
Central Medical Research Methodology 2010; 10, 67.
Arber, D., Orazi, A., Hasserjian, R., Thiele, J., Borowitz, M., Le Beau, M.,
Bloomfield, C., Cazzola, M. and Vardiman, J. The 2016 revision to the
312
World Health Organization classification of myeloid neoplasms and acute
leukemia. Blood 2016; 127, 2391-2405.
Asfour, M., Narvios, A. and Lichtiger B. Transfusion of RhD-Incompatible
Blood Components in RhD-Negative Blood Marrow Transplant
Recipients. Medscape General Medicine. 2004; 6, 22.
Astor, B., Muntner, P., Levin, A., Eustace, J. and Coresh, J. Association
of kidney function with anaemia. The Third National Health and Nutrition
Examination Survey (1988-1994). Archives of Internal Medicine 2002,
162, 1401-1408.
Avent, N., Ridgwell, K., Tanner, M. and Anstee, D. cDNA cloning of a 30
kDa erythrocyte membrane protein associated with Rh (Rhesus) blood-
group-antigen expression. Biochemical Journal. 1990; 271, 821-815.
Avent, N., Martinez, A., Flegel, W., Olsson, M., Scott, M., Nogues, N.,
Pisacka, M., Daniels, G., van der Schoot, E., Muniz-Diaz, E., Madgett, T.,
Storrey, J., Beiboer, S., Maaskant-van Wijk, P., von Zabern, I., Jiminez,
E., Tejedor, D., Lopez, M., Camacho, E., Cheroutre, G., Hacker, A.,
Jinoch, P., Svobodova, I. and de Haas, M. The BloodGen project:
towards mass scale comprehensive genotyping of blood donors in the
European Union and beyond. Transfusion 2007; 47, 405-465.
313
Avent, N. Large-scale blood group genotyping – clinical implications.
British Journal of Haematology 2009; 144, 3-13.
Aygun, B., Padmanabhan, S., Paley, C. and Chandrasekaran, V. Clinical
significance of red cell alloantibodies in sickle cell patients who received
transfusions. Transfusion 2002; 42, 37-43.
Baia, F., Correia, F., Alves, B., Martinez, F., Koch, C., Carneiro, A. and
Araujo. Phenotyping Rh/Kell and risk of alloimmunization in
haematological patients. Transfusion Medicine 2015; 26, 34-38.
Barrett, J., Saunthararajah, Y. and Molldrem, J. Myelodysplastic
syndrome and Aplastic Anaemia: distinct entities or diseases linked by a
common pathophysiology? Seminars in Hematology 2000; 37, 15-29.
Bauer, M., Wiersum-Osselton, J., Schipperis, M., Vandenbroucke, J. and
Briet, E. Clinical predictors of alloimmunisation after red blood cell
transfusion. Transfusion 2007; 47, 2066-2071.
BCSH Working Party of the British Committee for Standards in
Haematology Blood Transfusion Task Force. Guidelines for pre-
transfusion compatibility procedures in blood transfusion laboratories.
Transfusion Medicine 2013; 23, 3-35.
314
BCSH Validation and Qualification, including change control, for hospital
transfusion laboratories. Transfusion Medicine 2012; 22, 5-43.
Beauregard, P and Blajchman, M. Haemolytic and pseudo-haemolytic
transfusion reactions and the clinical conditions that mimic them.
Transfusion Medicine Reviews 1994; 8, 184-199.
Belhacene, N., Maulon, L., Guerin, S., Ehrland, J., Colin, Y., Cartron, J.
and Auberger, P. Differential expression of the Kell blood group and
CD10 antigens: two related membrane metallopeptidases during
differentiation of K562 cells by phorbol ester and hemin. Federation of
American Societies for Experimental Biology 1998; 12, 531-539.
Bennet, J., Catovsky, D., Daniel, M., Flandrin, G., Galton, D., Gralnick, H.
and Sultan, C. Proposal for the classification of the acute leukaemias.
British Journal of Haematology, 1976; 33, 452-459.
Blumberg, N., Peck, K., Ross, K. and Avila, E. Immune response to
chronic red blood cell transfusion. Vox Sanguinis 1983; 44, 212-217.
Blumberg, N., Ross, K., Avila, E. and Peck, K. Should chronic transfusion
be matched for antigens other than ABO and Rh0 (D)? Vox Sanguinis
1984; 47, 205-208.
315
Brugnara, C. and Eckardt, K. Hematologic aspects of kidney disease. In
Taal MW, ed. Brenner and Rector’s The Kidney. 9 th ed. Philadelphia:
Saunders; 2011: 2081-2120.
Cartwright, R., Gilman, E. and Gurney, K. Time trends in incidence of
haematological malignancies and related conditions. British Journal of
Haematology 1999; 106, 281-295.
Castilho, L. and Pellegrino Júnior, J. Blood group genotyping. Revista
Brasileira de Hematologia e Hemoterapia 2004; 26, 135-140.
Champlin, R. and Gale, R. Acute myelogenous leukaemia: recent
advances in therapy. Blood 1987, 69, 1551-1562.
Chan, J., Banks, P., Cleary, M., Delsol, G., De Wolf-Peeters, C., Falini,
B., Gatter, K., Grogan, T., Harris, N., Isaacson, P., Jaffe, E., Knowles, D.,
Mason, D., Muller-Hermelink, H-K., Pileri, S., Prirs, M., Ralfkiaer, E.,
Stein, H. and Warnke, R. A proposal for classification of lymphoid
neoplasms (by the International Lymphoma Study Group). Histopathology
1994; 25, 517-536.
Chan, J., Banks, P., Cleary, M., Delsol, G., De Wolf-Peeters, C., Falini,
B., Gatter, K., Grogan, T., Harris, N., Isaacson, P., Jaffe, E., Knowles, D.,
Mason, D., Muller-Hermelink, H., Pileri, S., Piris, M., Ralfkiaer, E., Stein,
H. and Warnke, R. A revised European-American classification of
316
lymphoid neoplasms proposed by the International Lymphoma Study
Group. A summary version. American Journal of Clinical Pathology 1995;
103, 543-560.
Cherif-Zahar, B., Bloy, C., Le Van Kim, C., Blanchard, D., Bailly, P.,
Hermand, P., Salmon, C., Cartron. J. and Colin, Y. Molecular cloning of
protein structure of a human blood group Rh polypeptide. Proceedings of
the National Academy of Sciences 1990; 87, 6243-6247.
Chou, S. and Westhoff, C. Molecular biology of the Rh system: clinical
considerations for transfusion in sickle cell disease. American Society of
Hematology Education Book 2009; 1, 178-184.
Chou, S., Jackson, T., Vege, S., Smith-Whitley, K., Friedman, D. and
Westhoff, C. High prevalence of red cell alloimmunization in sickle cell
disease despite transfusion for Rh-matched minority donors. Blood 2013;
122, 1062-1071.
Churchill, D., Taylor, D., Cook, R., La Plante, P., Barre, P., Cartier, P.,
Fay, W., Goldstein, M., Jindal., K. and Mandin, H. Canadian
Hemodialysis Morbidity Study. American Journal of Kidney Disease 1992;
19, 214-234.
Coles, S., Klein, H. and Holland, P. Alloimmunisation in two
multitransfused patient populations. Transfusion 1981; 21, 462-466.
317
Cook, I. Primary rhesus immunization of male volunteers. British Journal
of Haematology 1971; 20, 369-75.
Coombs, R.R.A., Mourant, A.E. and Race, R.R., 1945. Detection of Weak
and “Incomplete” Rh Agglutinins: A New Test. In Rhesus haemolytic
disease (pp. 88-88). Springer Netherlands.
Coombs, R., Mourant, A. and Race, R. A new test for detection of weak
and “incomplete” Rh agglutinins. British Journal of Experimental
Pathology 1945(b); 26: 255-266.
Cornell, L., Smith, R. and Colvin, R. Kidney transplantation: mechanisms
of rejection and acceptance. Annual Review of Pathology 2008; 3, 189-
220.
Court-Brown, W. and Doll, R. Adult Leukaemia Trends in mortality in
relation to aetiology. British Journal of Medicine 1959; 1, 1063-1069.
Cvejic, A., Haer-Wigman, L., Stephens, J., Kostadima, M., Smethurst, P.,
Frontini, M., van den Akker, E., Bertone, P., Bielczyk-Maczyńska, E.,
Farrow, S., Fehrmann, R., Gray, A., de Haas, M., Haver, V., Jordan, G.,
Karjalainen, J., Kerstens, H., Kiddle, G., Lloyd-Jones, H., Needs, M.,
Poole, J., Ait Soussan, A., Rendon, A., Rieneck, K., Sambrook, J.,
Schepers, H., Silljé, H., Sipos, B., Swinkels, D., Tamuri, A., Verweij, N.,
318
Watkins, N., Westra, H-J., Stemple, D., Franke, L., Soranzo, N.,
Stunnenberg, H., Goldman, N., van der Harst, P., van der Schoot, C.,
Ouwehand, W. and Albers, C. SMIM1 underlies the Vel blood group and
influences red blood cell traits. Nature Genetics 2013; 45, 542-545.
Dalpke, A., Hofko, M. and Zimmerman, S. Comparison of the BD Max
Methicillin-Resistant Staphylococcus Aureus (MRSA) Assay and the BD
GeneOhm MRSA Achromopeptidase Assay with Direct- and Enriched-
Culture Techniques Using Clinical Specimens for Detection of MRSA.
Journal of Clinical Microbiology 2012; 50, 3365-3367.
Daniels, G., Fletcher, A., Garratty, G., Jørgensen, J., Judd, W., Levene,
C., Lomas-Francis, C., Moulds, J. J., Moulds, J.M., Overbeeke, M., Reid,
M., Rouger, P., Scott, M., Sistonen, P., Smart, E., Tani, Y., Wendel, S.
and Zelinski, T. Blood group terminology 2004: from the International
Society of Blood Transfusion committee on terminology for red cell
surface antigens. Vox Sanguinis 2004; 87, 304-316.
Daniels, G. Human Blood Groups. Third Edition 2013. Wiley-Blackwell.
Daniels, G. and Bromilow, I. Essential Guide to Blood Groups. Second
Edition 2010. Wiley-Blackwell.
Da Silva, S., Ferriera, G., da Silva, S., Alves, T., Ribeiro, I., Ribeiro, T.
and Cavalcante, M. Red blood cell and leukocyte alloimmunisation in
319
patients awaiting kidney transplantation. Revista Brasileira de
Hematologia e Hemoterapia 2013; 35, 185-188.
Davies, S., McWilliam, A., Hewitt, P., Devenish, A. and Brozovic, M. Red
cell alloimmunization in sickle cell disease. British Journal of
Haematology 1986; 63, 241-245.
Dean, L. Blood Groups and Red Cell Antigens. Bethesda (MD): National
Centre for Biotechnology Information (US); 2005.
(http://www.ncbi.nlm.nih.gov/books/NBK2261/ Accessed 26.02.13).
Denomme, G. and Van Oene, M. High-throughput multiplex single-
nucleotide polymorphism analysis for red cell and platelet antigen
genotypes. Transfusion 2005; 45, 660-666.
Deschamps-Latscha, B. and Chatenoud, L. T cells and B cells in Chronic
Renal Failure. Seminars in Nephrology 1996; 16, 183-191.
Dombret, H. and Gardin, C. An update of current treatments for adult
acute myeloid leukaemia. Blood 2016; 127, 53-61.
Domen, R. and Ramirez, G. Red Cell Alloimmunization in Chronic Renal
Failure Patients Undergoing Hemodialysis. Nephron 1988; 48, 284-285.
320
Drago, F., Karpasitou, K. and Poli, F. Microarray beads for Identifying
Blood Group Single Nucleotide Polymorphisms. Transfusion Medicine
and Hemotherapy 2009; 36, 157-160.
Drago, F., Karpasitou, K., Spinardi, L., Crespiatico, L., Scalamogna, M.
and Poli, F. A Microsphere-Based Suspension Array for Blood Group
Molecular Typing: An Update. Transfusion Medicine and Hemotherapy
2010; 37, 336-338.
Drüeke, T. Hyporesponsiveness to recombinant human erythropoietin.
Nephrology Dialysis Transplantation 2001, 16, 25-28.
Dutcher, J., Schiffer, C., Aisner, J. and Wiernik, P. Long term follow up
patients with leukaemia receiving platelet transfusions: identification of a
large group of patients who do not become alloimmunized. Blood 1981;
58, 1007-1011.
European Group for Blood and Marrow Transplantation (EBMT). Annual
report 2011:
(https://www.ebmt.org/Contents/Resources/Library/Annualreport/Docume
nts/Annual%20Report%2011%20EBMT%20vertical_Website.pdf
accessed 01/06/2016).
Eiberg, H. and Mohr, J. Dombrock blood group (DO): assignment to
chromosome 12p. Human Genetics 1996; 98, 518-521.
321
El Nemer, W., Rahuel, C., Colin, Y., Gane, P., Cartron, J. and Le Van
Kim, C. Organization of the Human LU Gene and Molecular Basis of the
Lua/Lub Blood Group Polymorphism. Blood 1997; 89, 4608-4616.
Endeward, V., Cartron, J-P. and Gros, G. RhAG protein of the Rhesus
complex is a CO2 channel in the human red cell membrane. FASEB
Journal (The official journal of the Federation of American Societies for
Experimental Biology) 2008; 22, 64-73.
Ehrlich, G., Viegas-Pequignot, E., Ginzberg, D., Sindel, L., Soreq, H. and
Zakut, H. Mapping the Human Acetylcholinesterase Gene to
Chromosome 7q22 by Fluorescent in situ Hybridization Coupled with
Selective PCR Amplification from a Somatic Hybrid Cell Panel and
Chromosome-Sorted DNA Libraries. Genomics 1992; 13, 1192-1197.
Erslev, A. Humoral regulation of red cell production. Blood 1953; 8, 349-
357.
Exeter Census Profile 2011. Devon County Council
http://www.devon.gov.uk/census_profile-exeter.pdf (accessed
29/01/2018).
322
Ferlay, J., Autier, P., Boniol, M., Haenue, M., Colombet, M. and Boyle, P.
Estimates of cancer incidence and mortality in Europe 2006. Annals of
Oncology 2007, 18, 581-592.
Finning, K., Bhandari, R., Sellers, F., Revelli, N., Villa, M. A., Muñiz-Díaz,
E., and Nogués, N. (2016). Evaluation of red blood cell and platelet
antigen genotyping platforms (ID CORE XT/ID HPA XT) in routine clinical
practice. Blood Transfusion, 14, 160–167.
Fleiss, J., Levin, B. and Park, M. A statistical method for rates and
proportion. How to randomize. Third edition. Hobohan NJ. John Wiley
and Sons, 2003.
Fluit, C., Kunst, V. and Drenthe-Schonk, A. Incidence of red cell
antibodies after multiple blood transfusion. Transfusion 1990; 30, 532-
535.
Frane, J. A method of biased coin randomisation, its implementation and
validation. Drug Information Journal 1998; 32, 423-432.
Freda, V, Gorman, J., Pollack, W., Robertson, J., Jennings, E. and
Sullivan, J. Prevention of Rh isoimmunization: progress report of the
clinical trial in mothers. Journal of the American Medical Association
1967; 199, 390-4.
323
Friedman, L., Furberg, C. and DeMets, D. Fundamentals of Clinical Trials
3. New York: Springer; 1998.
Frohn, C., Dumbgen, L., Brand, J., Gorg, S., Luhm, J. and Kirchner, H.
Probability of anti-D development in D- patients receiving D+ RBCs.
Transfusion 2003; 43, 893-8.
Gabardi, S., Martin, S., Roberts, K. and Grafais, M. Induction
immunosuppressive therapies in renal transplantation. American Journal
of Health System Pharmacy 2011; 68, 211-218.
Garratty, G. Novel mechanisms for immune destruction of circulating
autologous cells. In: Silberstein LE, editor. Autoimmune disorders of
blood. Bethesda: American Association of Blood Banks; 1996, 79-114.
Garratty, G. Autoantibodies induced by blood transfusion. Transfusion
2004; 44, 5-9.
Gassner, C., Meyer, S., Frey, B. and Vollmert, C. Matrix-assisted laser
desorption/ionisation, time-of-flight mass spectrometry-based blood group
genotyping – the alternative approach. Transfusion Medicine Reviews
2013; 27, 2-9.
324
Gaston, R. Maintenance immunosuppression in the renal transplant
recipient: an overview. American Journal of Kidney Diseases. 2001; 38,
S25-S35.
Gerrard-Marchant, R., Hamlin Lennart, K., Rilke, F. and Stansfield, A.
Classification of NHL. Lancet 1974; ii, 406-409.
Getman, D., Eubanks, J., Camp, S., Evans, G. and Taylor, P. The Human
Gene Encoding Acetylcholinesterase Is Located on the Long Arm of
Chromosome 7. American Journal of Human Genetics 1992; 52, 170-
177.
Goldman, M., Nogués, N. and Castilho, L. An overview of the Progenika
ID CORE XT: an automated genotyping platform based on a fluidic
microarray system. Immunohaematology 2015; 31, 62-68.
Grais, M. Mortality and Morbidity from Leukaemia and Aleukaemia in
Specific Countries. Bulletin of the World Health Organisation 1962; 25,
683- 688.
Guelsin, G., Sell, A., Castilho, L., Masaki, V., Melo, F., Hashimotot, N.,
Higa, T., Hirle, L. and Visentainer, J. Benefits of Blood Group Genotyping
in Multi-Transfused Patients from the South of Brazil. Journal of Clinical
Laboratory Analysis 2010; 24, 311-316.
325
Gunson, H., Stratton, F., Cooper, D. and Rawlinson, V. Primary
immunization of Rh-negative volunteers. British Medical Journal 1970; 1,
593-5.
Gyurkocza, B. and Sandmaier, B. Conditioning regimens for
hematopoietic cell transplantation: one size does not fit all. Blood 2014;
124, 344-353.
Hallorhan, P. Immunosuppressive drugs for kidney transplantation. New
England Journal of Medicine 2004; 351, 2715-2719.
Harris, N., Jaffe, E., Stein, H., Banks, P., Chan, J., Cleary, M., Delsol, G.,
De Woolf-Peeters, C., Falini, B., Gatter, K., Grogan, T., Isaacson, P.,
Knowles, D., Mason, D., Muller-Hermelink, H., Pileri, S., Piris, M.,
Ralfkiaer, E. and Warnke, R. A revised European-American classification
of lymphoid neoplasms: a proposal from the International Lymphoma
Study Group. Blood 1994; 84, 1361-1392.
Hashmi, G., Shariff, T., Seul, M., Vissavajjhala, P., Hue-Roye, K.,
Charles-Pierre, D., Lomas-Francis, C., Chaudhuri, A. and Reid, M. A
flexible array format for large scale rapid blood group DNA typing.
Transfusion 2005; 45, 680-688.
Hashmi, G., Shariff, T., Zhang, Y., Cristobal., J., Chau, C., Seul, M.,
Vissavajjhala, P., Baldwin, C., Hue-Roye, K., Charles-Pierre, D., Lomas-
326
Francis, C. and Reid, M. Determination of 24 minor red blood cell
antigens for more than 2000 blood donors by high-throughput DNA
analysis. Transfusion 2007; 47, 736-747.
Heddle, N., Soutar, R., O’Hoski, P., Singer, J., McBride, J., Ali, M. and
Kelton, J. A prospective study to determine the frequency and clinical
significance of alloimmunization post-transfusion. British Journal of
Haematology 1995; 91, 1000-5.
Hendrickson, J., Desmarets, M., Deshpande, S., Chadwick, T., Hillyer, C.,
Roback, J. and Zimring, J. Recipient inflammation affects the frequency
and magnitude of immunization to transfused red cells. Transfusion 2006;
46, 1526-1536.
Hinds, P. and Gattuso, J. From pilot work to a major study in cancer
nursing research. Cancer Nursing 1991; 14, 132-135.
Hirschfeld, H. A connection between correlation and contingency.
Proceedings of the Cambridge Philosophical Society 1935; 31, 520-524.
Hmida, S., Mojaat, N., Maamar, M., Bejaoui, M., Mediouni, M. and
Boukef, K. Red cell alloantibodies in patients with haemoglobinopathies.
Nouvelle Revue Francaise d’Hematologie 1994; 36, 363-366.
327
Hoeltge, G., Domen, R., Rybicki, L. and Schaffer, P. Multiple Red Cell
Transfusions and Alloimmunization: Experience With 6996 Antibodies
Detected in a Total of 159 262 patients from 1985 to 1993. Archives of
Pathology and Laboratory Medicine 1995; 119, 42-5.
Holohan, T., Terasaki, P. and Deisseroth, A. Suppression of Transfusion
Related Alloimmunization in Intensively Treated Cancer Patients. Blood
1981; 58, 122-128.
Huang, C-H., Johe, K., Moulds, J., Siebert, P., Fukuda, M. and
Blumenfeld, O. δ glycophorin (glycophorin B) gene deletion in two
individuals homozygous for the S-s-U- blood group phenotype. Blood
1987; 70, 1830-1835.
Huang, C-H., Reid, M. and Blumenfeld, O. Remodelling of the
Transmembrane Segment in Human Glycophorin by Aberrant RNA
Splicing. Journal of Biological Chemistry 1994; 269, 10804-18012.
Huang, C-H. and Blumenfeld, O. MNSs blood groups and major
glycophorins: Molecular basis for allelic variation. In: Cartron, J-P,
Rouger, P. eds. Molecular basis of human blood group antigens. New
York: Plenum Press 1995, 153-188.
Huang, C-H., Blumenfeld, O., Reid, M., Chen, Y., Daniels, G. and Smart,
E. Alternative Splicing of a Novel Glycophorin Allele GPHe(GL)
328
Generates Two Protein Isoforms in the Human Erythrocyte Membrane.
Blood 1997; 90, 391-397.
Hu, F., Hu, Y., Ma, Z. and Rosenberger, W. Adaptive randomization for
balancing over covariates. WIREs Computational Statistics 2014; 6, 288-
303.
Huh, Y. and Lichtiger, B. Transfusion Reactions in Patients with Cancer.
American Journal of Clinical Pathology 1987; 87, 253-257.
Human Genome Organisation: Gene Nomenclature Committee
http://www.genenames.org/cgi-bin/genefamilies/set/454 (accessed
04/12/2015).
Isacke, C. and Horton, M. The Adhesion Molecule Facts Book, Edition 2.
Academic Press. London. 2000.
Ishikura, H., Endo, J., Saito, Y., Kosaka, Y., Matsuo, S., Nagasue, N.,
Nakamura, T., Tsunematsu, T., Miyazaki, T. and Matsumoto, H. Graft
versus host antibody reaction causing a delayed hemolytic anaemia after
blood transfusion. Blood 1993; 82, 3222-3.
Issitt, P. and Anstee, D. Applied Blood Group Serology. Fourth Edition,
1998. Montgomery Scientific Publications. Durham, North Carolina, USA.
329
Izumi, N., Fuse, I., Furukawa, T., Uesugi, Y., Tsuchiyama, J., Toba, K.,
Togashi, K., Yamada, K., Ohtake, S., Saitoh, Y., Yanagisawa, N. and
Aizawa, Y. Long-term production of pre-existing alloantibodies to E and c
after allogeneic BMT in a patient with aplastic anemia resulting in delayed
hemolytic anemia. Transfusion 2003; 43, 241-245.
Jacobson, L., Goldwasser, E., Fried, W. and Plzak, L. Role of the kidney
in erythropoiesis. Nature 1957; 179, 633-634.
JPAC - Joint United Kingdom (UK) Blood Transfusion and Tissue
Transplantation Services Professional Advisory Committee.
transfusionguidelines.org (www.transfusionguidelines.org.uk/transfusion-
handbook accessed 03.04.2015). Transfusion Handbook 5th ed. Jan
2014.
JPAC - Joint United Kingdom (UK) Blood Transfusion and Tissue
Transplantation Services Professional Advisory Committee.
transfusionguidelines.org. Guidelines for the Blood Transfusion Services
in the United Kingdom 8th edition, 2013.
Juji, T, Kano, K. and Milgrom, F. Mixed agglutination with platelets.
International Archives of Allergy. 1972; 42, 474-484.
330
Kadar, J. Impact of phenotype-matched red cell transfusions on the long
term blood supply of patients under hematolooncologic outpatient care.
Vox Sanguinis 2010, 99, 465-465.
Kalish, L. and Begg, G. Treatment allocation methods in clinical trials; a
review. Statistics in Medicine 1985; 4, 129-144.
Karpasitou, K., Drago, F., Crespiatico, L., Paccapelo, C., Truglio, F.,
Frison, S., Scalamogna, M and Poli, F. Blood group genotyping for
Jka/Jkb, Fya/Fyb, S/s, K/k, Kpa/Kpb, Jsa/Jsb, Coa/Cob and Lua/Lub with
microarray beads. Transfusion 2008; 48, 505-512.
Kato, S., Chmielewski, M., Honda, H., Pecoits-Filho, R., Matsuo, S.,
Yazawa, Y., Tranaeus, A., Stenvinkel, P. and Lindholm, B. Aspects of
immune dysfunction in end stage renal disease. Clinical Journal of the
American Society of Nephrology 2008; 3, 1526-1533.
King, L., Kozono, D. and Agre, P. From structure to disease: the evolving
tale of aquaporin biology. Nature Reviews Molecular Cell Biology 2004
(a); 5, 687-698.
King, K., Shirey, R., Thoman, S., Benson-Kennedy, D., Tanz, W. and
Ness, P. Universal leukoreduction decreases the incidence of febrile
nonhemolytic reactions to RBCs. Transfusion 2004 (b); 44, 25-29.
331
Kipps, T. and Carson, D. Autoantibodies in Chronic Lymphocytic
Leukaemia and Related Systemic Autoimmune Diseases. Blood 1993;
81, 2475-2487.
Klapper, E., Zhang, Y., Figueroa, P., Ness, P., Stubbs, J., Abumuhor, I.,
Bailey, J., Epperson, L., Tauscher, C., Enriquez, E., Hashmi, G. and Seul,
M. Towards extended phenotype matching: a new operational paradigm
for the transfusion service. Transfusion 2010; 50, 536-546.
Korver, K., de Lange, G., van den Burgh, R., Schellekens, P. and van
Loghem, E. Lymphoid chimerism after allogeneic bone marrow
transplantation. Y-chromatin staining of peripheral T and B lymphocytes
and allotyping of serum immunoglobulins. Transplantation 1987; 44, 643-
650.
Kościelak, J. The Hypothesis on Function of Glycosphingolipids and ABO
Blood Groups Revisited. Neurochemical Research 2012; 37, 1170-1184.
Kraemer, H., Mintz, J., Noda, A., Tinklenberg, J. and Yesavge, J. Caution
regarding the use of pilot studies to guide power calculations for study
proposals. Archives of General Psychiatry 2006; 63, 484-489.
Kudo, S. and Fukuda, M. Identification of a Novel Human Glycophorin,
Glycophorin E, by Isolation of Genomic Clones and Complementary DNA
332
Clones Utilizing Polymerase Chain Reaction. Journal of Biological
Chemistry 1990; 265, 1102-1110.
Lancaster, G., Dodd, S. and Williamson, P. Design and analysis of pilot
studies: recommendations for good practice. Journal of Evaluation in
Clinical Practice 2004; 10, 307-312.
Lapierre, Y., Rigal, D., Adam, J., Josef, D., Meyer, F., Greber, S. and
Drot, C. The gel test: a new way to detect red cell antigen-antibody
reactions. Transfusion 1990; 30, 109-113.
Laundy, R., Karamatic Crew, V., Davies, H., Harris, C., Baglow, L.,
Walser, P. and Thornton, N. A novel high incidence antigen in the Yt
blood group system. Transfusion Medicine 2017; 27, 19-20.
Lee, E., Whitehead, A., Jacques, R. and Julious, S. The statistical
interpretation of pilot trials: should significance thresholds be
reconsidered? BioMed Central; Medical Research Methodology 2014; 14,
41.
Lee, R-A and Gabardi, S. Current trends in immunosuppressive therapies
for renal transplant recipients. American Journal of Health-System
Pharmacy 2012; 69, 1961-1975.
333
Lee, S., Zambas, E., Marsh, W. and Redman, C. The Human Kell Blood
Group Gene Maps to Chromosome 7q33 and its Expression Is Restricted
to Erythroid Cells. Blood 1993; 81, 2804- 2809.
Le Guern, R., Herwegh, S., Grandbastien, B., Courcol, R. and Wallet, F.
Evaluation of a new molecular test, the BD Max Cdiff, for detection of
toxigenic Clostridium difficile in fecal samples. Journal of Clinical
Microbiology 2012; 50, 3089-3090.
Leon, A., Davis, L. and Kraemer, H. The role and interpretation of pilot
studies in clinical research. Journal of Psychiatric Research 2011; 45,
626-629.
Levine, P., Stock, A., Kuhmichel, A. and Bronikovsky, N. A new human
blood factor of rare incidence in the general population. Proceedings of
the Society of Experimental Biology 1951; 77, 402-403.
Lin, Y., Saskin, A., Wells, R., Lenis, M., Mamedov, A., Callum, J. and
Buckstein, R. Prophylactic RhCE and Kell antigen matching: impact on
alloimmunization in transfusion-dependent patients with myelodysplastic
syndromes. Vox Sanguinis 2017; 112, 79-86.
Loscalzo, J. Pilot trials in clinical research: of what value are they?
Circulation 2009; 119, 1694-1696
334
Lucien, N., Sidoux-Walter, F., Olives, B., Moulds, J., Le Pennec, P-Y.,
Cartron, J-P. and Bailly, P. Characterization of the Gene Encoding the
Human Kidd Blood Group/Urea Transporter Protein. Evidence for splice
site mutations in Jknull individuals. Journal of Biological Chemistry 1998;
273, 12973-12980.
Lukes, R. and Butler, J. The pathology of the nomer data of HD. Cancer
Research, 1986; 26, 1077-1081.
Maciejewski, J., O’Keefe, C., Gondek, L. and Tiu, R. Immune mediated
bone marrow failure syndromes of progenitor and stem cells: molecular
analysis of cytotoxic T cell clones. Fiolia Histochemica et Cytochemica
2007; 45, 5-14.
Makins, G. Injuries to the Blood Vessels. In History of the Great War
Medical Services. Surgery of the War London, 1922, 2, 170-296. His
Majesty’s Stationery Office, London.
Mallinson, G., Soo, K., Schall, T., Pisacka, M. and Anstee, D. Mutations
in the erythrocyte chemokine receptor (Duffy) gene: the molecular basis
of the Fya/Fyb antigens and identification of a deletion in the Duffy gene of
an apparently healthy individual with the Fy(a-b-) phenotype. British
Journal of Haematology 1995; 90, 823-829.
335
Marini, A-M., Urrestarazu, A., Beauwens, R. and Andre, B. The Rh
(Rhesus) blood group polypeptides are related to NH4+ transporters.
Trends in Biochemical Sciences 1997; 22, 460-461.
Martin-Vega, C., Massuet, L., Julia, A. Autoimmune Haemolytic
anaemia (AIHA) in Myelodysplastic Syndromes (MS). In: Proceedings of
the XX congress of the International Society of Blood Transfusion in
association with the British Blood Transfusion Society. London 1988.
Manchester: British Blood Transfusion Society, 1988; 81-81.
Master, S., Ong, M. and Mansour, R. Blood Transfusion in Adult Patients
with Sickle Cell Disease: Incidence of Alloimmunization. Blood 2016;
128:2459.
Mathew, S., Chaudhuri, A., Murty, V. and Pogo, A. Confirmation of Duffy
blood group antigen locus (FY) at 1q22-q23 by fluorescence in situ
hybridization. Cytogenetic and Cell Genetics 1994; 67, 68-68.
Mauthe, J., Coghlan, G. and Zelinski, T. Confirmation of the Assignment
of the Dombrock Blood Group Locus (DO) to chromosome 12p:
Narrowing the Boundaries to 12p 12.3 – p13.2). Vox Sanguinis 2000; 79,
53-56.
336
McBean, R., Hyland, C. and Flower, R. Molecular genotyping platforms
for blood group antigen prediction. Pathology 2014a; 46, S87.
McBean, R., Hyland, C. and Flower, R. Approaches to Determination of a
Full Profile of Blood Group Genotypes: Single Nucleotide Variant
Mapping and Massively Parallel Sequencing. Computational and
Structural Biotechnology Journal 2014b; 11, 147-151.
McBean, R., Hyland, C., Roscioli, T. and Flower, R. The low frequency
SARAH blood group antigen: evidence for new MNS antigen. Vox
Sanguinis. 2014c; 107, 17-17.
McDonald, A., Knight, R., Campbell, M., Entwistle, V., Grant, A., Cook, J.,
Elbourne, D., Francis, D., Garcia, J., Roberts, I. and Snowdon, C. What
influences recruitment to randomised controlled trials? A review of trials
funded by two UK funding agencies. BioMed Central; Trials 2006; 7, 9.
Mental Capacity Act 2005 (legislation.gov.uk) accessed 16/10/2015.
Michail-Merianou, V., Phamphili-Panousophoulou, L., Piperi-Lowes, L.,
Pelegrinis, E. and Karaklis, A. Alloimmunization to red cell antigens in
Thalassaemia: comparative study of usual versus better-match
transfusion programmes. Vox Sanguinis 1987; 52, 95-98.
Mikhail, A., Shrivastava, R. and Richardson, D. Anaemia of CKD. UK
Renal Association, 5th edition, 2009-2012, Final Version 2010.
337
Milic, M., Webb, G. and Bowen, D. 344 Single centre experience of
transfusion-dependent MDS demonstrates a high frequency of allo-
antibody formation in heavily transfused patients. Leukemia Research
2011, 35, S138.
Mohanram, A., Zhang, Z., Shahinfar, S., Keane, W., Brenner, B. and
Toto, R. Anemia and end-stage renal disease in patients with type 2
diabetes and nephropathy. Kidney International 2004; 66, 1131-1138.
Möller, M., Jöud, M., Storry, J., and Olsson, M. L. Erythrogene: a
database for in-depth analysis of the extensive variation in 36 blood
group systems in the 1000 Genomes Project. Blood Advances, 2016;
1:240-249. https://doi.org/10.1182/bloodadvances.2016001867
(accessed 26/01/2018).
Moncharmont, P., Barday, G. and Meyer, F. Red blood cell
alloimmunisation in 18 to 50-year old transfused women: a 3-year study.
Blood Transfusion 2015; 13, 345-346.
Montpetit, A., Phillips, M., Mongrain, I., Lemieux, R. and St-Louis, M.
High-throughput molecular profiling of blood donors for minor red blood
cell and platelet antigens. Transfusion 2006; 46, 841-848.
338
Moon, C., Preston, G., Griffin, C., Jabs, E. and Agre, P. The human
aquaporin-CHIP gene. Structure, organization, and chromosomal
localization. Journal of Biological Chemistry 1993; 268, 15772-15778.
Moore, C., Carter, R., Nietert, P. and Stewart, P. Recommendations for
Planning Pilot Studies in Clinical and Translational Research. Clinical and
Translational Science 2011; 4, 332-337.
Moreira, G., Bordin, J., Kuroda, A. and Kerbauy, J. Red blood cell
alloimmunization in sickle cell disease: The influence of racial and
antigenic pattern differences between donors and recipients in Brazil.
American Journal of Hematology 1996; 52, 197-200.
Mouro, I., Colin, Y., Sistonen, P., Le Pennec, P., Cartron, J. and Le Van
Kim, C. Molecular basis of the RhCW (Rh8) and RhCX (Rh9) blood group
specificities. Blood 1995; 86, 1196-1201.
Mufti, G., Figes, A., Hamblin, T. Oscier, D. and Copplestone, J.
Immunological abnormalities in the myelodysplastic syndromes. I. Serum
immunoglobulins and autoantibodies. British Journal of Haematology
1986; 63, 143-147.
Murakami, M. and Honjo, T. Anti-red blood cell autoantibody transgenic
mice: murine model of autoimmune hemolytic anemia. Seminars in
Immunology 1996; 8, 3-9.
339
Murphy, M., Morrison, N., Miles, J., Fraser, R., Spurr, N. and Boyd, E.
Regional chromosomal assignment of the Kell blood group locus (KEL) to
chromosome 7q33-q35 by fluorescence in situ hybridization: evidence for
the polypeptide nature of antigenic variation. Human Genetics 1993; 91,
585-588.
Murphy, M., Templeton, L., Fleming, J., Ferguson, M., Peterkin, M. and
Fraser, R. Comparison of the Fyb status as determined serologically and
genetically. Transfusion Medicine 1997; 7, 135-141.
Murphy, M., Pamphilon, D and Heddle, N. Practical Transfusion
Medicine. 4th edition 2013. Wiley-Blackwell.
Nashan, B. Antibody Induction Therapy in Renal Transplant Patients
Receiving Calcineurin-Inhibitor Immunosuppressive Regimens: A
Comparative Review. BioDrugs 2005; 19, 39-46.
National Comparative Audit of Blood Transfusion (NCABT). Audit of the
use of blood in adult medicine – Part 1 2011. www.hospital.blood.co.uk
(accessed 03.04.2015).
National Comparative Audit Reports (NCABT)– National Survey Red Cell
Use – 2014 (http://hospital.blood.co.uk/audits/national-comparative-
audit/national-comparative-audit-reports/) accessed 01/12/2017.
340
Neote, K., Mak, J., Kolakowski, L. and Schall, T. Functional and
Biochemical Analysis of the Cloned Duffy Antigen: Identity With the Red
Blood Cell Chemokine Receptor. Blood 1994; 84, 44-52.
Ness, P., Shirey, R., Thoman, S. and Buck, S. The differentiation of
delayed serologic and hemolytic transfusion reactions: incidence, long-
term serologic findings, and clinical significance. Transfusion 1990; 30,
688-693.
NICE 2003; Improving Outcomes in Haematological Cancers. The
Manual. National Institute for Clinical Excellence.
(https://www.nice.org.uk/guidance/ng47/evidence/improving-outcomes-in-
haematological-cancers-the-manual-2487893581) accessed 18/04/2018
NICE 2011; Anaemia management in people with chronic kidney disease.
www.publications.nice.org.uk/anaemia-management-in-people-with-
chronic-kidney-disease-cg114 (accessed 07.03.2014).
NIH National Institute of Diabetes and Digestive and Kidney Diseases
(http://www.niddk.nih.gov/health-information/health-topics/kidney-
disease/anemia-in-kidney-disease-and-dialysis/Pages/facts.aspx
accessed 16/04/2016).
341
National Institute for Health Research: Guidance Feasibility and Pilot
studies (www.nets.nihr.ac.uk/glossary) accessed 17/10/2015.
Norman, D. and Turka, L. Primer on transplantation. 2 nd ed. Mt Laurel NJ:
American Society of Transplantation 2001; 85-102.
Novaretti, M., Sopelete, C., Velloso, E., Rosa, M., Dorlhiac-Llacer, P. and
Chamone, D. Immunohematological findings in myelodysplastic
syndrome. Acta Haematologica 2001; 105, 1-6.
Office for National Statistics. Cancer Statistics Registrations, England:
2012;
(https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocial
care/conditionsanddiseases/bulletins/cancerregistrationstatisticsengland/
2014-06-19) Accessed 25/12/2017.
Olives, B., Martial, S., Mattei, M-G., Matassi, G., Rousselet, G., Ripoche,
P., Cartron, J. and Bailly, P. Molecular characterization of a new urea
transporter in the human kidney. FEBS Letters 1996; 386, 156-160.
Olsson, M., Smythe, J., Hansson, C., Poole, J., Mallinson, G., Jones, J.,
Avent, N. and Daniels, G. The Fy(x) phenotype is associated with a
missense mutation in the Fy(b) allele predicting Arg89Cys in the Duffy
glycoprotein. British Journal of Haematology 1998; 103, 1184-1191.
342
Olujohungbe, A., Hambleton, I., Stephens, L., Serjeant, B. and Serjeant,
G. Red cell antibodies in patients with homozygous sickle cell disease: a
comparison of patients in Jamaica and the United Kingdom. British
Journal of Haematology 2001; 113, 661-665.
Parasol, N., Reid, M., Rios, M., Castilho, L., Harari, I. and Kosower, N. A
Novel Mutation in the Coding Sequence of the FY*B Allele of the Duffy
Chemokine Receptor Gene Is Associated With an Altered Erythrocyte
Phenotype. Blood 1998; 92, 2237-2243.
Parsons, S., Mallinson, G., Daniels, G., Green, C., Smythe, J. and
Anstee, D. Use of Domain-Deletion Mutants to Locate Lutheran Blood
Group Antigens to Each of the Five Immunoglobulin Superfamily
Domains of the Lutheran Glycoprotein: Elucidation of the Molecular Basis
of the Lua/Lub and the Aua/Aub Polymorphisms. Blood 1997; 88, 4219-
4225.
Plapp, F., Sinor, L., Rachel, J., Beck, M., Coenen, W. and Bayer, W. A
Solid Phase Antibody Screen. American Journal of Clinical Pathology
1984; 82, 719-721.
Petz, L. and Garratty, G. Acquired Immune Hemolytic Anemias. New
York: Churchill Livingstone, 1980:331-337.
343
Petz, L. and Garratty, G. Immune Hemolytic Anemias. 2nd Edition.
Philadelphia: Churchill Livingstone, 2004.
Pocock, S. and Simon, R. Sequential treatment assignment with
balancing for prognostic factors in the controlled clinical trial. Biometrics,
1975; 31, 103-115.
Pruenster, M. and Rot, A. Throwing light on DARC. Biochemical Society
Transactions 2006; 34, 1005-1008.
Priyadarshi, A. and Shapiro, J. Erythropoietin resistance in the treatment
of anaemia of chronic renal failure. Seminars in Dialysis, 2006; 19, 273-
278.
Purohit, K., Weber, J., Ward, L. and Keats, B. The Kell blood group locus
is close to the cystic fibrosis locus on chromosome 7. Human Genetics
1992; 89, 457-458.
Race, R. An “Incomplete” Antibody in Human Serum. Nature 1944; 153,
771-772.
Rahuel, C., London, J., Vignal, A., Ballas, S. and Cartron, J-P.
Erythrocyte Glycophorin B Deficiency May Occur by Two Distinct Gene
Alterations. American Journal of Hematology 1991; 37, 57-58.
344
Ramsey, G. and Larson, P. Loss of red cell alloantibodies over time.
Transfusion 1988; 28, 162-165.
Ramsey, G., Hahn, L., Cornell, F., Boczkowski, D., Staschak, S., Clark,
R., Hardesty, R., Griffith, B. and Starzl, T. Low rate of Rhesus
immunization from Rh-incompatible blood transfusions during liver and
heart transplant surgery. Transplantation 1989; 47, 993-995
Redman, M., Regan, F. and Contreras, M. A Prospective Study of the
Incidence of Red Cell Allo-Immunisation following Transfusion. Vox
Sanguinis 1996; 71, 216-220.
Reid, M. Applications and Experience with PCR-Based Assays to Predict
Blood Group Antigens. Transfusion Medicine and Hemotherapy 2009; 36,
168-178.
Reid, M., Storry, J., Ralph, H., Blumenfeld, O. and Huang, C-H.
Expression and quantitative variation of the low-incidence blood group
antigen He on some S-s- red cells. Transfusion 1996; 36, 719-724.
Reid, M., Halter-Hipsky, C., Hue-Roye, K. and Hoppe, C. Genomic
analyses of RH alleles to improve transfusion therapy in patients with
sickle cell disease. Blood Cells Molecules and Diseases 2014; 54, 195-
202.
345
Renoud, K., Barracchini, K., Byrne, K., Adams, S., Pickett, A., Caruccio,
L. and Stroncek, D. KEL6 and KEL7 genotyping with sequence-specific
primers. Transfusion 2006; 46, 1510-1514.
Reverberi, R. The persistence of red cell alloantibodies. Blood
Transfusion 2008; 4, 225-234.
Ribeiro, S., Costa, E., Belo, L., Reis, F. and Santos-Silva, A. rh EPO for
the Treatment of Erythropoietin Resistant Anemia in Hemodialysis
Patients – Risks and Benefits. Hemodialysis 2013;
(http://www.intechopen.com/books/hemodialysis accessed 02/05/2016).
Riedlinger, J., Beqaj, S., Milish, M., Young, S., Smith, R., Dodd, M.,
Hankerd, R., Lebar, W. and Newton, D. Multicenter evaluation of the BD
Max GBS assay for detection of group B streptococci in prenatal vaginal
and rectal screening swab specimens from pregnant women. Journal of
Clinical Microbiology 2010; 48, 4239-4241.
Robertson, O. Transfusion with preserved red blood cells. British Medical
Journal 1918; 1, 691-695.
Rolih, S., Eisinger, R., Moheng, M., Dean, W. and Eatz, R. Solid Phase
Adherence Assays: Alternatives to Conventional Blood Bank Tests.
Laboratory Medicine 1985; 16, 766-770.
346
Rosenberger, W. and Lachin, J., 2016. Randomization in clinical trials:
theory and practice. John Wiley & Sons.
Rosse, W., Gallagher, D., Kinney, T. Castro, O., Dosik, H., Moohr, J.,
Wang, W. and Levy, P. Transfusion and alloimmunization in sickle cell
disease. Blood 1990; 76, 1431-1437.
Royal College of Obstetricians and Gynaecologists. Red Cell Antibodies
during Pregnancy, the Management of Women with (Green-top
Guidelines No 65) May 2014. https://www.rcog.org.uk/en/guidelines-
research-services/guidelines/gtg65/ (accessed 25/01/2018).
Rozman, P., Dovc, T. and Gassner, C. Differentiation of autologous ABO,
RHD, RHCE, KEL, JK and FY blood group genotypes by analysis of
peripheral blood samples of patients who have recently received multiple
transfusions. Transfusion 2000; 40, 936-942.
Sangole, V. and Chaudhury, D. Alloimmunisation of Red Blood Cells in
Multitransfused Patients. International Journal of Health Sciences and
Research 2013; 3, 39-41.
Sanquin Blood
Supplyhttp://www.sanquin.nl/en/research/departments/transfusion-
medicine/randomized-clinical-trials/ (accessed 11.2.13).
347
Sanz, C., Nomdedeu, M., Belkaid, M., Martinez, I., Nomdedeu, B. and
Pereira, A. Red blood cell alloimmunization in transfused patients with
myelodysplastic syndrome or chronic myelomonocytic leukaemia.
Transfusion 2013; 53, 710-715.
Sarnaik, S., Schornack, J. and Lusher, J. The incidence of development
of irregular red cell antibodies in patients with sickle cell anemia.
Transfusion 1986; 26, 249-252.
Schofield, A., Martin, P., Spillett, D. and Tanner, M. The structure of the
human red blood cell anion exchanger (EPB3, AE1, band 3) gene. Blood
1994; 84, 2000-2012.
Schonewille, H., Haak, H. and Van Zijl, A. Alloimmunization after blood
transfusion in patients with hematologic and oncologic diseases.
Transfusion 1999; 39, 763-771.
Schonewille, H., Rene, R., De Vries, P. and Brand, A. Alloimmune
response after additional red blood cell antigen challenge in immunized
hematooncology patients. Transfusion 2009; 49, 453-457.
Schonewille, H., Honohan, A., van der Watering, L., Hudig, F., te
Boekhorst, P., Koopman-van Gemert, A. and Brand, A. Incidence of
alloantibody formation after ABO-D or extended matched red blood cell
348
transfusions: a randomized trial (MATCH study). Transfusion 2016; 56,
311-320.
Schulz, K and Grimes, D. Allocation concealment in randomised trials:
defending against deciphering. Lancet 2002a; 359, 614-618.
Schulz, K and Grimes, D. Unequal group size in randomised trials:
guarding against guessing. Lancet 2002b; 359, 966-970.
Sealed Envelope Ltd. 2012. Power calculator for binary outcome
superiority trial. [Online] Available from:
https://www.sealedenvelope.com/power/binary-superiority/ [Accessed
Nov 14 2016].
Seitanides, B., Antonopoulou, A. and Karabelis, A. Immunological
abnormalities in myelodysplastic syndrome. Journal of Clinical Pathology
1988; 41, 922-923.
Seltsam, A. and Doescher, A. Sequence-Based Typing of Human Blood
Groups. Transfusion Medicine and Hemotherapy 2009; 36, 204-212.
Serious Hazards of Transfusion (SHOT) annual report 2014
www.shotuk.org (accessed 21/05/2016).
349
Serious Hazards of Transfusion (SHOT) annual report 2016.
https://www.shotuk.org/wp-content/uploads/SHOT-Report-
2016_web_11th-July.pdf (accessed 25/01/2018).
Seyfried, H. and Walewska, I. Analysis of immune response to red blood
cell antigens in multitransfused patients with different diseases. Materia
Medica Polona 1990; 22, 21-5.
Shapiro, M. Serology and genetics of a new blood factor: hr s. Journal of
Forensic Medicine 1960; 7, 96-105.
Shapiro, M., Le Roux, M. and Brink, S. Serology and genetics of a new
blood factor: hr B. Haematologia 1972; 6, 121-128.
Shirey, R., Boyd, J., Parwani, A., Tanz, W., Ness, P. and King, K.
Prophylactic antigen‐matched donor blood for patients with warm
autoantibodies: an algorithm for transfusion management. Transfusion,
2002; 42, 1435-1441.
Sidoux-Walter, F., Lucien, N. and Olives, B. At Physiological Expression
Levels the Kidd Blood Group/Urea Transporter Protein Is Not a Water
Channel. Journal of Biological Chemistry 1999; 274, 30228-30235.
Sidoux-Walter, F., Lucien, N., Nissinen, R., Sistonen, P., Henry, S.,
Moulds, J., Cartron, J-P. and Bailly, P. Molecular heterogeneity of the
350
Jknull phenotype: expression analysis of the Jk (S291P) mutation found in
Finns. Blood 2000; 96, 1566-1573.
Spanos, T., Karageorga, M., Ladis, V., Peristeri, J., Hatziliami, A. and
Kattamis, C. Red cell alloantibodies in patients with thalassaemia. Vox
Sanguinis 1990; 58, 50-55.
Sparkes, R., Sparkes, M. and Gale, R. Immunoglobulin synthesis
following allogeneic bone marrow transplantation in man. Conversion to
donor allotype. Transplantation 1979; 27, 212-213.
Springer, G., Horton, R. and Forbes, M. Origin of anti-human blood group
B agglutinins in White Leghorn chicks. Journal of Experimental Medicine
1959a; 110, 221-244.
Springer, G., Horton, R. and Forbes, M. Origin of antihuman blood group
B agglutinins in germfree chicks. Annals of the New York Academy of
Sciences 1959b; 78, 272-275.
Springer, G. and Horton, R. Blood group isoantibody stimulation in man
by feeding blood group-active bacteria. Journal of Clinical Investigation
1969; 48, 1280-1291.
351
Sokol, R., Hewitt, S. and Brooker, D. Erythrocyte autoantibodies,
autoimmune haemolysis, and myelodysplastic syndrome. Journal of
Clinical Pathology 1989; 42, 1088-1091.
Solal-Celigny, P., Vazeux, R., Vroclans, M., Amar, M., Herrera, A.,
Bernard, J. and Boivin, P. Positive Coombs tests in acute leukaemia.
British Journal of Haematology 1984; 57, 563-569.
Sourial, N., Wolfson, C., Zhu, B., Quail, J., Fletcher, J., Karunananthan,
S., Bandeen-Roches, K., Beland, F. and Bergman, H. Correspondence
analysis is a useful tool to uncover the relationships among categorical
variables. Journal of Clinical Epidemiology 2010; 63, 638-646.
Stainsby, D., Jones, H., Asher, D., Atterbury, C., Boncinelli, A., Brant, L.,
Chapman, C., Davison, K., Gerrard, R., Gray, A., Knowles, S., Love, E.,
Milkins, C., McClelland, D., Norfolk, D., Soldan, K., Taylor, C., Revill, J.,
Williamson, L. and Cohen, H. Serious Hazards of transfusion: a decade
of hemovigilance in the UK. Transfusion Medicine Reviews 2006; 20,
273-282.
Starr, D. Blood: An epic history of medicine and commerce. 1999; Little
Brown and Company, London.
Steegmann, J., Requena, M., Martin-Regueira, P., De La Cámara, R.,
Casado, F., Salvanés, F. and Fernández Rañada, J. High incidence of
352
autoimmune alterations in chronic myeloid leukemia patients treated with
interferon-alpha. American Journal of Hematology 2003; 72, 170-176.
Storry, J., Reid, M., Fetics, S. and Huang, C-H. Mutations in the GYPB
exon 5 drive the S-s-U+var phenotype in persons of African descent:
implications for transfusion. Transfusion 2003; 43, 1738-1747.
Strumia, M. and McGraw, J. The Development of Plasma Preparations
for Transfusions. Annals of Internal Medicine 1941; 15, 80-88.
Sully, B., Julious, S. and Nicholl, J. A reinvestigation of recruitment to
randomised, controlled, multicenter trials: a review of trials funded by two
UK funding agencies. Trials 2013; 14, 166.
Suresh, K. An overview of randomization techniques: An unbiased
assessment of outcome in clinical research. Journal of Human
Reproductive Sciences 2011; 4, 8-11.
Swerdlow, S., Campo, E., Harris, N., Jaffe, E., Pileri, S., Stein, H., Thiele,
J. and Vardiman, J., 2008. WHO Classification of Tumors of
Hematopoietic and Lymphoid Tissues 4th Ed. (2008). IARC Publications.
Swerdlow, S., Campo, E., Pileri, S., Harris, N., Stein, H., Siebert, R.,
Advani, R., Ghielmini, M., Salles, G., Zelenetz, A. and Jaffe, E. The 2016
353
revision of the World Health Organization classification of lymphoid
neoplasms. Blood 2016; 127, 2375-2390.
Tahhan, H., Holbrook, C., Braddy, L., Brewer, L. and Christie, J. Antigen-
matched donor blood in the transfusion management of patients with
sickle cell disease. Transfusion 1994; 34, 562-569.
Tanner, M. Molecular and cellular biology of the erythrocyte anion
exchanger (AE1). Seminars in Hematology 1993; 30, 34-57.
Teare, M., Dimairo, M., Shephard, N., Hayman, A., Whitehead, A. and
Walters, S. Sample size requirements to estimate key design parameters
from external pilot randomised controlled trials: a simulation study.
BioMed Central; Trials 2014; 15, 264.
Terasaki, P. and Mizutani, K. Antibody Mediated Rejection: Update 2006.
Clinical Journal of the American Society of Nephrology 2006; 1, 400-403.
The International Society of Blood Transfusion Red Cell Immunogenetics
and Blood Group Terminology Working Party.
http://www.isbtweb.org/working-parties/red-cell-immunogenetics-and-
terminology (accessed 03/12/2015).
354
Timura, G., Duman, A., Ongut, G. and Saka, O. The Significance of
Autoantibodies in Non-Hodgkin’s Lymphoma. Leukemia and Lymphoma.
2000; 40, 119-122.
Tournamille, C., Colin, Y., Cartron, J-P. and Le Van Kim, C. Disruption of
a GATA motif in the Duffy gene promoter abolishes erythroid expression
in Duffy-negative individuals. Nature Genetics 1995; 10, 224-228.
Tournamille, C., Le Van Kim, C., Gane, P., Le Pennec, P., Roubinet, F.,
Babinet, J., Cartron, J-P. and Colin, Y. Arg89Cys substitution results in
very low membrane expression of the Duffy antigen/receptor for
chemokines in Fyx individuals. Blood 1998; 92, 2147-2156.
Türkmen, T., Qiu, D., Cooper, N., Sachs, U. J., Wößmann, W., Schranz,
D., Zimmer, K.-P., Ehrhardt, H., Hackstein, H. and Bein, G. (2017), Red
blood cell alloimmunization in neonates and children up to 3 years of age.
Transfusion, 57: 2720–2726.
Van den Berghe, H. Morphologic, immunologic and cytogenetic (MIC)
working classification of the acute myeloid leukaemias. British Journal of
Haematology 1988; 68, 487-494.
Van der Putten, K., Braam, B., Jie, K. and Gaillard, C. Mechanisms of
Disease: erythropoietin resistance in patients with both heart and kidney
failure. Nature Clinical Practice Nephrology 2008; 4, 47-57.
355
van Tol, M., Gerritsen, E., de Lange, G., van Leeuwen, A., Jol-van der
Zijde, C., Oudeman-Gruber, N., de Vries, E., Radl, J. and Vossen, J. The
origin of IgG production and homogeneous IgG components after
allogeneic bone marrow transplantation. Blood 1996; 87, 818-826.
Viitala, J. and Järnefelt, J. The red cell surface revisited. Trends in
Biochemical Sciences 1985; 10, 392-395.
Wagner, F., Flegel, W., Bittner, R. and Döscher, A. Molecular typing for
blood group antigens within 40 min by direct polymerase chain reaction
from plasma or serum. British Journal of Haematology 2017; 176, 814-
821.
Walter, C. and Murphy, W. A closed gravity technique for the preservation
of whole blood in ACD solution utilizing plastic equipment. Surgery,
Gynecology & Obstetrics 1952; 94, 687-692.
Weigle, W. The induction of autoimmunity in rabbits following injection of
heterologous or altered homologous thyroglobulin. Journal of
Experimental Medicine 1965; 121, 289-308.
Weir, C and Lees, K. Comparison of stratification and adaptive methods
for treatment allocation in an acute stroke clinical trial. Statistics in
Medicine 2003; 22, 705-726.
356
Wiener, A. Genetic Theory of the Rh Blood Types. Experimental Biology
and Medicine 1943, 54, 316-319.
Wiener, A., Unger, L. and Cohen, L. Distribution and heredity of blood
factor U. Science 1954; 119, 734-735.
Westhoff, C. The structure and function of the Rh antigen complex.
Seminars in Hematology, 2007; 44, 42-50.
Westhoff, C. and Shaz, B. Kell and Kidd Blood Group Systems.
Transfusion Medicine and Haemostasis (second edition) 2013.
https://www.sciencedirect.com/topics/immunology-and-microbiology/kell-
antigen-system (accessed 11/02/2018).
Winters, J., Pineda, A., Gorden, L., Bryant, S., Melton, J., Vamvakas, E.
and Moore, S. RBC alloantibody specificity and antigen potency in
Olmstead County, Minnesota. Transfusion 2001; 41, 1413-1420.
Witherspoon, R., Schanfield, M., Storb, R., Thomas, E. and Giblett, E.
Immunoglobulin production of donor origin after bone marrow
transplantation for acute leukemia or aplastic anemia. Transplantation
1978; 26, 407-408.
357
Woldie, I., Swerdlow, P., Bluth, M., Mohammad, U., Landolfi, E.,
Chaudrhy, S., Dyson, G. and O’Malley, B. Lifetime risk and
characterization of red blood cell alloimmunization in chronically
transfused patients with sickle cell disease. International Journal of Blood
Transfusion and Immunohematology 2015; 5, 1-5.
Woodrow, J and Donohoe, W. Rh-immunization by Pregnancy: Results of
a Survey and Their Relevance to Prophylactic Therapy. British Medical
Journal 1968; 4, 139-144.
Yazer, M., Triulzi, D., Shaz, B., Kraus, T. and Zimring, J. Does a febrile
reaction to platelets predispose recipients to red blood cell
alloimmunization? Transfusion 2009; 49, 1070-1075.
Zhu, X., Rivera, A., Golub, M., Peng, J., Sha, Q., Wu, X., Song, X.,
Kumarathasan, P., Ho, M., Redman, C. and Lee, S. Changes in red cell
ion transport, reduced intratumoral neovascularization, and some mild
motor function abnormalities accompany targeted disruption of the Mouse
Kell gene (Kel). American Journal of Hematology 2009; 84, 492-498.
Young, P., Uzielbo, A., Trulock, E., Lublin, D. and Goodnough, L.
Autoantibody formation after alloimmunization: are blood transfusions a
risk factor for autoimmune hemolytic anemia? Transfusion 2004; 44, 67 –
72.
358
Yu, Y-B., Gau, J-P., You, J-Y., Chern, H-H., Chau, W-K., Tzeng, C-H.,
Ho, C-H. and Hsu, H-C. Cost-effectiveness of postremission intensive
therapy in patients with acute leukaemia. Annals of Oncology 2007; 18,
529-534.
Zalpuri, S., Zwaginga, J., Cessie, S. Elhuis, J., Schonewille, H. and Van
der Bom, J. Red-blood-cell alloimmunization and number of red-blood-
cell transfusions. Vox Sanguinis 2012; 102, 144-149.
Zalpuri, S., Evers, D., Zwaginga, J., Schonewille, H., de Vooght, K., le
Cessie, S. and van der Bom, J. Immunosuppressants and
alloimmunisation against red blood cell transfusions. Transfusion 2014;
54, 1981-1987.
Zelinski, T., Coghlan, G., Myal, Y., Shiu, R., Phillips, S., White, L. and
Lewis, M. Genetic linkage between the Kell blood group system and
prolactin –inducible protein loci: provisional assignment of KEL to
chromosome 7. Annals of Human Genetics 1991; 55, 137-140.
Zhang, R. Modern Immunosuppressive Therapy in Kidney
Transplantation. Open Journal of Organ Transplant Surgery 2013; 3, 22-
31.
359
Zheng, Y. and Maitta, R. Alloimmunisation rates of sickle cell disease
patients in the United States differ from those in other geographical
regions. Transfusion Medicine 2016; 26, 225-230.
Zimring, J., Spitalnik, S., Roback, J. and Hillyer, C. Transfusion induced
autoantibodies and differential immunogenicity of blood group antigens: a
novel hypothesis. Transfusion 2007; 47, 2189-2196.
360
10 APPENDICES
10.1 Appendix 1: Raw data for retrospective review of haematology cohort
Key: N = No M=Male
Y = Yes F=Female
Alloimmunised No. Age at first

No. units Alloimmunised post transfusion recorded Platelet Irradiated
Diagnosis transfused at admission transfusion Gender episodes transfusion requirement requirement
Anaemia of chronic disease 53 N N M 24 81 N N

AIHA 17 N Y F 7 27 N N
AIHA 49 N Y F 23 85 N N
AIHA 4 N N M 1 66 N N
AIHA 13 Y Y F 4 50 N N
AIHA 7 N Y F 3 70 N Y
AIHA 35 N Y F 14 60 N N
AIHA 9 Y Y F 4 86 N N
AITL 6 N Y M 3 81 Y Y
AITL 15 Y Y M 7 74 Y Y
ALL 3 N N F 1 44 Y N
ALL 8 N N F 3 66 Y N
ALL 11 N N F 6 72 Y N
1|A p p e n d i x
ALL 46 N N F 20 54 Y Y
ALL 38 N N M 12 38 Y Y
ALL 27 N N F 12 75 Y N
ALL 84 N N F 38 55 Y Y
ALL 3 N N M 2 72 Y N
ALL 92 N N M 43 41 Y Y
ALL 35 N N M 17 23 Y N
ALL 17 N N F 8 51 Y N
ALL 8 N N M 4 84 N N
ALL 76 N N M 31 66 Y N
ALL 52 N N M 25 66 Y N
AML 2 Y Y M 1 81 N Y
AML 2 N N F 1 67 Y N
AML 8 N N F 4 52 Y Y
AML 9 N N M 3 80 N N
AML 10 N N F 4 88 N N
AML 11 N N F 4 62 Y N
AML 11 N N M 5 51 Y Y
AML 12 N Y F 6 71 Y N
AML 14 N N M 7 60 Y N
AML 17 N N M 7 55 Y Y
AML 20 N N M 10 78 Y N
AML 22 N N F 8 91 Y N
2|A p p e n d i x
AML 23 N N M 10 76 Y N
AML 25 N N F 11 69 Y N
AML 28 Y Y M 13 79 Y Y
AML 28 N N M 12 51 Y Y
AML 30 N N M 16 37 Y Y
AML 30 N N M 14 36 Y N
AML 32 N N M 15 46 Y Y
AML 35 N N M 15 37 Y Y
AML 36 N N M 15 57 Y Y
AML 37 N Y F 18 63 Y Y
AML 37 N Y M 17 32 Y Y
AML 41 N N F 15 73 Y N
AML 43 N Y M 20 56 Y Y
AML 47 N Y F 22 39 Y Y
AML 48 N N M 23 69 Y N
AML 59 N N F 28 71 Y N
AML 61 N Y F 26 49 Y Y
AML 64 N Y M 29 54 Y Y
AML 64 N Y M 32 49 Y Y
AML 64 N Y M 22 68 Y Y
AML 66 N Y F 33 60 Y Y
AML 73 N Y F 35 82 N N
AML 75 N Y M 30 19 Y Y
AML 79 N N F 36 56 Y Y
AML 85 N N M 43 74 N N
AML 168 N N F 79 51 Y Y
3|A p p e n d i x
AML 53 N Y M 26 42 Y Y
AML 51 N Y F 24 52 Y Y
AML 144 N Y F 67 71 Y Y
AML 14 N N F 7 73 N N
AML 53 N N F 25 74 Y N
AML 75 N Y M 34 63 Y Y
AML 32 N N M 15 40 Y Y
AML 8 N N F 4 89 N N
AML 18 N N F 9 80 Y N
AML 22 N N F 9 75 Y N
AML 56 N N M 26 64 Y Y
AML 52 N N F 26 50 Y Y
AML 44 N N F 18 60 Y Y
AML 13 N N M 6 62 Y Y
AML 8 N N M 4 80 Y N
AML 14 N N F 6 86 N N
AML 80 N N F 35 64 Y Y
AML 33 N N M 14 24 Y Y
AML 31 N N F 15 57 Y N
AML 4 N N F 2 55 N N
AML 30 N N F 13 93 Y N
AML 12 N N F 6 58 Y Y
AML 15 N N M 7 82 N N
AML 6 N Y F 3 74 N N
AML 6 N N M 3 85 Y N
AML 70 N N M 34 67 Y N
4|A p p e n d i x
AML 79 N N M 34 67 Y N
AML 13 N N M 5 61 Y N
AML 14 N N M 6 76 N N
AML 25 N N F 13 49 Y Y
AML 80 N Y F 33 67 Y N
AML 94 N Y M 46 56 Y Y
AML 53 N Y F 25 56 Y Y
AML 94 N N M 39 81 Y N
AML 89 N N F 43 63 Y Y
AML 42 N Y F 18 59 Y N
AML 33 N N M 16 34 Y Y
AML 17 N N M 7 67 Y N
AML 39 N N M 18 78 N N
AML 44 N N M 19 77 Y N
AML 30 N Y F 13 65 Y N
AML 22 N N M 10 61 Y Y
AML 112 N N F 54 52 Y Y
AML 34 N N F 16 75 Y N
AML 33 N N M 16 70 Y N
AML 85 N N F 42 54 Y Y
AML 42 N N M 21 71 Y N
AML 39 N N F 18 55 Y Y
AML 69 N N F 29 30 N Y
AML 41 N N F 19 24 Y Y
AML 53 N Y F 25 51 Y Y
AML 28 N N M 10 76 Y Y
5|A p p e n d i x
AML 27 N N F 12 74 Y Y
AML 31 N N M 17 64 Y N
AML 18 N N F 9 83 Y N
AML 10 N N F 4 83 Y N
AML 33 N N M 15 45 Y Y
AML 44 N N M 21 89 Y N
AML 21 N N M 10 70 Y N
AML 68 N N M 30 73 N N
AML 10 N N F 5 76 Y Y
AML 29 N N F 14 75 Y N
AML 28 N N M 13 65 Y N
AML 54 N N M 25 69 Y N
AML 10 N N F 4 72 N N
AML 35 N N F 17 54 Y N
AML 52 N N M 25 59 Y Y
AML 22 N N M 11 81 Y N
AML 157 N Y M 72 58 Y N
AML 104 Y Y F 53 66 Y Y
AML 40 N Y M 18 66 Y N
AML 15 N N F 6 84 Y N
AML 28 N N M 11 74 Y N
AML 10 N N F 6 63 Y Y
AML 10 N N F 4 48 Y Y
AML 42 N N F 21 67 Y Y
AML 166 N N M 63 71 Y N
6|A p p e n d i x
AML 18 N Y F 8 60 N N
AML 51 N N M 21 75 Y N
AML 38 N N F 17 60 Y Y
AML 53 N N F 24 54 Y Y
AML 115 N N M 47 59 Y Y
AML 26 N N M 9 84 Y N
AML 23 N N F 12 58 Y N
AML 27 N Y F 14 72 Y N
AML 62 N N M 30 73 Y Y
AML 61 N Y F 26 73 Y Y
AML 88 N N M 39 67 Y Y
AML 2 N N M 1 78 N N
AML 18 N N F 9 65 Y Y
AML 43 N N F 21 18 Y N
AML 36 N N F 17 74 Y N
AML 50 N Y M 23 61 Y Y
AML 36 N Y M 17 81 Y N
AML 31 N N M 13 68 Y Y
AML 22 N N F 11 61 Y N
AML 20 N Y F 10 75 Y N
AML 28 N N F 14 76 Y Y
AML 36 N N F 16 66 Y N
AML 48 N Y F 23 84 Y N
AML 65 N N F 29 69 Y N
7|A p p e n d i x
AMML 8 N N M 4 85 Y N
AMYLOID 4 N N F 2 48 N N
AMYLOID 47 N N M 10 31 Y Y
AMYLOID 72 N Y F 36 55 N N
AMYLOID 4 N N F 2 49 N N
ANAEMIA 7 N N F 2 45 N N
ANAEMIA 80 N Y M 53 42 N N
ANAEMIA 60 N N M 21 71 N N
ANAEMIA 32 N Y F 13 74 N N
ANAEMIA 2 N N F 1 40 N Y
APLASTIC 15 N N F 10 22 Y Y
APLASTIC 19 N Y F 9 87 Y N
APLASTIC 36 N Y M 14 26 Y Y
8|A p p e n d i x
APLASTIC 107 N Y M 52 57 Y Y
APLASTIC 26 N N M 13 27 Y N
APLASTIC 10 N N M 3 80 N N
APLASTIC 68 N Y F 30 27 Y Y
APLASTIC 4 N N M 2 64 N Y
APLASTIC 93 N Y F 44 58 N N
APLASTIC 9 N N F 4 33 N N
APLASTIC 9 Y Y M 4 53 N N
APML 15 N N M 7 44 Y Y
Autoimmune Neutropenia 6 N N F 3 80 N N
B-Cell Lymphoma 26 N N M 12 67 Y N
Cancer (solid organ) 5 N N M 2 71 N N
Cancer (solid organ) 7 N Y F 3 60 N N
Cancer (solid organ) 7 N N F 2 80 N N
Cancer (solid organ) 7 N N F 4 49 Y Y
Cancer (solid organ) 8 Y Y M 3 53 N N
9|A p p e n d i x
Cancer (solid organ) 8 N N M 4 68 Y N
10 | A p p e n d i x
Cancer (solid organ) 32 Y Y F 12 65 N N
Cancer (solid organ) 2 N N F 1 61 Y N
Cancer (solid organ) 6 N N F 3 66 Y N
Cancer (solid organ) 19 N Y M 9 51 Y M
Cancer (solid organ) 54 N Y M 20 80 N N
Cancer (solid organ) 14 N Y M 7 69 Y N

CLL 2 N N M 1 78 N Y
CLL 2 N N F 1 74 N N
CLL 4 N N F 1 53 Y N
CLL 6 N Y F 3 88 N N
CLL 6 N N M 3 78 Y N
CLL 7 N N M 3 91 Y N
CLL 10 N N F 3 84 N N
CLL 11 N N M 4 85 N N
CLL 16 N Y M 6 72 N Y
CLL 17 N N M 9 90 Y N
CLL 21 N Y M 10 93 Y N
CLL 65 N N F 27 78 Y N
CLL 22 N N F 11 65 N Y
CLL 11 N N M 5 93 N N
CLL 10 N N M 5 82 N N
CLL 10 N N F 5 63 Y Y
CLL 12 N N M 6 71 N Y
CLL 11 Y Y M 4 75 Y Y
CLL 15 Y Y F 7 75 N N
CLL 36 N Y M 15 68 Y Y
CLL 44 N N M 20 63 Y Y
CLL 39 N N M 19 71 Y Y
CLL 212 N Y M 87 68 Y Y
CLL 16 N N M 8 77 N Y
CLL 18 N Y M 5 70 Y Y
CLL 30 N Y M 15 69 Y Y
CLL 6 N N M 3 86 N N
CLL 14 Y Y M 6 72 N Y
CLL 5 Y Y M 2 83 N N
CLL 11 N Y M 5 73 N Y
CLL 34 N Y M 17 76 Y N
CLL 4 Y Y F 2 83 N N
CLL 12 Y Y F 5 78 N N
CLL 3 Y Y F 1 87 N N
CLL 10 N Y M 4 79 N Y
CLL 4 Y Y M 2 66 N N
CLL 10 N N F 5 79 N Y
CLL 11 N N F 4 74 Y Y
CLL 18 N Y F 9 89 N N
CLL 19 N N M 7 69 Y Y
CLL 52 N N M 24 75 Y Y
CLL 46 Y Y F 18 62 N Y
CLL 33 N N M 17 64 Y Y
CLL 19 N N M 11 92 N N
CML 24 N N F 11 85 N N
CML 45 N N M 20 61 Y Y
CML 22 N Y F 10 73 N N
CML 7 N N F 3 73 N N
CML 9 N N M 3 74 Y N
CML 10 N Y F 4 86 N N
CML 44 N Y F 19 74 N N
CML 10 N N M 5 39 Y Y
CML 2 Y Y M 1 62 N N
CML 40 N N F 22 63 Y Y
CML 4 N N F 2 73 N N
CMML 2 N N F 1 66 Y N
CMML 5 N N F 2 90 N N
CMML 37 N N M 25 73 Y N
CMML 37 N N M 16 82 Y N
CMML 28 N Y M 12 70 N N
CMML 63 N N M 29 76 Y N
CMML 18 N N M 7 70 Y N
CMML 11 N Y M 36 77 N N
CMML 20 N N M 10 70 N N
Cold Haemagglutinin Diease 12 Y Y F 6 73 N N
Cold Haemagglutinin Diease 2 Y Y M 1 76 N N

CRF 6 N N F 2 75 N N
CRF 12 N Y M 6 83 N N
CRF 32 N Y F 13 74 N N
CRF 28 N N F 9 86 N N
CRF 22 N N F 10 76 N N
CRF 6 N N M 2 80 N N
CRF 23 N Y F 9 81 N N
Diamond Blackfan Anaemia 382 N Y F 147 13 N N

Diffuse large B –cell
lymphoma 36 N N M 17 59 Y Y
Essential thrombocythaemia 4 N N M 2 73 N N
Essential thrombocythaemia 6 N N M 2 75 Y N
Essential thrombocythaemia 40 N N F 15 79 Y N
Essential thrombocythaemia 5 N N F 2 74 N N
F IX DEF 13 N N M 5 77 Y N
Haemolytic anaemia 19 N Y F 7 76 N N
Hairy cell leukaemia 2 N N M 1 60 N Y
Hairy cell leukaemia 6 N N M 3 79 Y Y
Hairy cell leukaemia 2 N Y M 1 76 N Y
Hodgkin's lymphoma 2 N N F 1 32 N Y
Hodgkin's lymphoma 4 N N F 3 36 Y Y
Hodgkin's lymphoma 6 N N M 3 21 Y Y
Hodgkin's lymphoma 13 N N M 6 74 N Y
Hodgkin's lymphoma 10 N N F 5 70 Y Y
Hodgkin's lymphoma 30 Y Y M 15 41 Y Y
Hodgkin's lymphoma 19 N N M 9 72 N N
Hodgkin's lymphoma 4 N N F 2 68 N N
Hodgkin's lymphoma 50 N Y F 24 56 Y Y
ITP 8 N N M 2 68 Y N
ITP 9 Y Y M 4 69 Y N
ITP 4 N Y F 2 82 N N
ITP 19 N N M 6 84 N N
ITP 4 N N F 2 72 Y N
ITP 6 N N M 2 64 Y N
ITP 4 N N M 2 82 N N
LEA 4 N N F 2 97 N N
Lymphoma 2 Y Y M 1 86 Y N
Lymphoma 2 N N M 1 81 N N
Lymphoma 2 N N F 1 87 N N
Lymphoma 5 N N M 3 45 Y Y
Lymphoma 10 N N M 5 75 Y N
Lymphoma 12 N Y M 6 77 N N
Lymphoma 21 Y Y M 8 46 Y Y
Lymphoma 30 N N F 13 62 Y Y
Lymphoma 25 N N M 12 61 N Y
Lymphoma 20 Y Y F 8 70 N Y
Lymphoma 12 N N F 6 57 Y N
Lymphoma 3 Y Y M 1 82 N N
Lymphoma 4 Y Y F 2 71 N N
Lymphoma 58 N Y M 24 43 Y Y
Lymphoma 51 N Y M 23 83 N Y
Lymphoma 4 N N F 2 87 N Y
Lymphoma 13 N Y M 6 82 N N
Lymphoma 21 N N F 10 61 N Y
Lymphoma 6 N Y F 3 70 N N
Lymphoma 4 N N F 2 87 Y N
Lymphoma 10 Y Y F 5 84 N N
Mantle cell lymphoma 2 N N M 1 62 Y Y
MDS 5 N Y F 2 94 N N
MDS 5 N N M 2 86 N N
MDS 8 N N F 3 80 N N
MDS 8 N N F 4 53 Y Y
MDS 13 N Y M 6 85 Y Y
MDS 13 N N F 6 85 N N
MDS 14 N N F 7 85 Y N
MDS 21 N N M 6 80 Y N
MDS 25 N N M 10 67 Y N
MDS 30 N Y M 14 84 Y N
MDS 35 N N F 46 90 N N
MDS 39 N Y M 18 95 N N
MDS 41 N N F 14 76 N N
MDS 43 N N F 19 77 N N
MDS 59 N N M 23 82 Y N
MDS 67 N N F 32 86 N N
MDS 70 N N M 27 73 Y N
MDS 74 N N F 33 85 Y N
MDS 80 N N F 42 89 Y N
MDS 114 N Y M 48 67 Y N
MDS 184 N Y F 61 88 N N
MDS 209 N N F 75 72 Y N
MDS 256 N Y F 94 25 N N
MDS 310 N Y F 101 75 Y N
MDS 338 N N M 120 69 Y N
MDS 183 N N F 77 73 N Y
MDS 8 N N M 4 97 Y N
MDS 15 N N F 6 86 N N
MDS 15 N Y F 6 73 Y N
MDS 53 N N M 23 68 Y Y
MDS 28 N N M 12 89 N N
MDS 16 N N F 8 71 N N
MDS 16 N N F 7 86 Y N
MDS 12 N N M 6 71 Y N
MDS 19 N N F 10 74 N N
MDS 92 N N M 37 82 N N
MDS 30 N N F 15 80 N Y
MDS 109 N N M 51 74 Y N
MDS 40 N Y M 16 76 N N
MDS 45 N N M 22 85 Y N
MDS 63 N N M 30 77 N N
MDS 50 N N M 23 68 N Y
MDS 32 N N F 10 59 N N
MDS 18 Y Y M 9 81 Y N
MDS 58 N N M 25 76 Y N
MDS 135 N N F 55 80 N N
MDS 2 N N F 1 71 N Y
MDS 111 N N F 54 83 N N
MDS 94 N Y F 36 80 N N
MDS 61 N N F 21 74 N N
MDS 38 N N M 19 65 Y N
MDS 37 N N M 18 50 Y Y
MDS 24 N N F 12 82 Y N
MDS 32 N N F 15 87 N N
MDS 224 N N M 87 62 Y N
MDS 66 N Y M 30 79 N N
MDS 143 N Y M 58 72 Y N
MDS 119 N N M 40 79 N N
MDS 10 N N M 5 73 N N
MDS 15 Y Y F 7 82 N Y
MDS 94 N N M 38 88 N N
MDS 20 N N M 8 68 N N
MDS 49 N N M 19 67 Y N
MDS 332 N Y M 155 69 Y Y
MDS 20 N N M 9 81 N N
MDS 67 N N M 28 68 N N
MDS 51 N N M 23 78 N Y
MDS 28 N N F 11 80 Y N
MDS 32 N N F 14 27 Y N
MDS 6 N N M 3 83 Y Y
MDS 60 N Y M 27 84 N N
MDS 287 N N M 124 76 Y N
MDS 43 N N M 15 88 N N
MDS 10 N N M 5 78 N N
MDS 45 N Y M 22 83 Y N
MDS 226 N Y M 81 60 Y Y
MDS 21 N N M 10 75 Y N
MDS 35 N N F 14 82 N N
MDS 33 N Y M 17 91 N N
MDS 10 N N M 4 82 N N
MDS 83 N N M 31 86 N N
MDS 14 N N M 7 65 N N
MDS 11 N N M 6 42 Y N
MDS 33 N Y M 16 79 N N
MDS 29 N Y M 13 76 Y N
MDS 119 N N F 60 75 N Y
MDS 69 N Y M 25 90 N N
MDS 19 N N M 6 57 Y N
MDS 16 N Y M 7 74 N N
MDS 53 N N F 24 78 Y N
MDS 52 N Y M 22 78 N N
MDS 70 N Y F 25 84 Y N
MDS 38 N N M 19 92 N N
MDS 47 N N M 21 72 N N
MDS 35 N Y M 16 79 Y N
MDS 18 N N M 9 81 Y N
MDS 32 N N F 14 88 Y N
MDS 14 N N F 6 80 N N
MDS 61 N Y M 27 67 Y N
MDS 21 N N F 7 77 Y N
MDS 34 N Y M 17 74 Y N
MDS 124 N N M 60 81 Y N
MDS 19 N N M 9 89 N N
MDS 52 N Y F 24 74 Y N
MDS 24 N Y M 14 76 N N
MDS 22 N N M 11 83 Y N
MDS 166 N Y F 59 85 N N
MDS 30 Y Y M 15 70 Y N
MDS 29 N N M 15 74 Y N
MDS 11 N N F 5 73 Y Y
MDS 14 N N M 7 85 N N
MDS 20 N Y F 7 80 N N
MDS 140 N Y F 63 87 Y N
MDS 12 N N M 6 85 N N
MDS 10 N N F 5 75 N N
MDS 154 Y Y F 62 84 N N
MDS 14 N Y F 5 75 Y N
MDS 119 N Y M 45 89 N N
MDS 21 N N M 9 84 Y N
MDS 69 N Y F 29 63 Y Y
MDS 11 N N M 5 88 N N
MDS 25 N Y F 11 65 N N
MDS 16 N Y M 6 81 N Y
MDS 32 N Y F 16 81 N N
MDS 5 N N F 2 63 Y N
MDS 179 N Y F 70 84 Y N
MDS 17 N N F 4 77 Y N
MDS 110 N N M 39 90 N N
MDS 10 N N F 4 80 N N
MDS 22 N Y M 10 69 Y Y
MDS 62 N N M 27 90 N N
MDS/AML 11 N Y M 5 77 Y N
MF 14 N N M 5 81 N N
MF 202 N Y M 83 69 Y Y
MF 17 N N F 8 67 Y N
MF 12 N N M 5 82 Y N
MF 80 N N M 37 68 N N
MF 33 Y Y F 11 80 N N
MF 10 N N F 5 82 N N
MF 17 N N F 11 72 Y N
MF 34 N N F 17 72 N N
MF 3 N N M 1 52 N N
MF 4 N N M 2 83 N N
MF 132 Y Y M 46 77 N N
MF 24 Y Y F 12 74 N N
MF 73 N N M 28 64 Y N
MF 134 N Y F 61 61 Y N
MF 26 N N M 10 66 Y N
MF 18 Y Y F 8 61 Y Y
MF 151 N N F 53 73 N N
MF 48 N N F 24 59 N N
MPD 55 Y Y M 22 52 Y Y
MPD 33 N N F 11 64 N N
MPD 11 N N M 4 84 N N
MPD 6 N N F 3 74 N N
Myeloma 2 N Y M 1 66 Y Y
Myeloma 2 N N M 1 58 Y Y
Myeloma 2 N N M 2 76 N N
Myeloma 3 N N F 1 80 N N
Myeloma 4 N N M 2 61 N Y
Myeloma 4 N N M 2 75 Y N
Myeloma 4 N N F 2 54 Y Y
Myeloma 6 N Y F 3 82 N N
Myeloma 9 Y Y F 4 76 N N
Myeloma 10 N Y F 5 60 Y Y
Myeloma 16 N Y F 8 68 Y N
Myeloma 19 N N F 8 62 N Y
Myeloma 20 Y Y M 11 56 Y Y
Myeloma 20 N N F 9 75 Y N
Myeloma 22 N Y M 10 71 Y N
Myeloma 35 N Y F 20 75 Y N
Myeloma 41 N Y M 16 71 N N
Myeloma 55 Y Y F 34 52 Y Y
Myeloma 127 N Y M 59 78 Y N
Myeloma 177 Y Y F 67 79 N N
NHL 2 N N M 1 69 N N
NHL 2 N N M 1 41 Y Y
NHL 2 N N F 1 56 N N
NHL 4 N Y M 2 90 N N
NHL 4 N N F 2 73 Y N
NHL 4 N N F 2 45 Y Y
NHL 6 N N M 2 79 Y N
NHL 6 N N F 3 56 N Y
NHL 8 N Y F 4 71 Y N
NHL 8 N Y F 2 75 Y Y
NHL 10 N Y F 5 79 Y Y
NHL 10 N N M 5 59 N N
NHL 10 N N M 5 80 N N
NHL 10 N N F 5 81 N N
NHL 10 N N M 4 64 Y Y
NHL 10 N N M 5 54 Y Y
NHL 11 N N F 6 52 Y Y
NHL 12 N N M 6 77 Y Y
NHL 13 N N F 7 75 Y Y
NHL 13 N N M 6 74 N N
NHL 13 N N M 5 61 N N
NHL 15 N N M 7 78 Y Y
NHL 16 N N M 8 68 Y N
NHL 16 N N F 7 54 Y N
NHL 16 N N F 8 21 Y Y
NHL 19 N N F 7 59 Y Y
NHL 19 N N F 11 62 Y N
NHL 20 N Y M 11 67 Y Y
NHL 20 N N M 9 53 Y Y
NHL 20 N N M 9 73 Y Y
NHL 21 N Y M 10 80 N Y
NHL 23 N N M 12 23 Y N
NHL 24 N Y F 13 60 Y Y
NHL 25 N N M 11 25 Y Y
NHL 27 N N M 11 38 Y Y
NHL 29 N N F 12 46 Y Y
NHL 32 N N M 13 64 Y N
NHL 35 N N F 17 67 N Y
NHL 35 N N M 16 81 Y Y
NHL 36 N N M 16 74 Y Y
NHL 37 N N M 19 54 Y Y
NHL 38 N N M 15 45 Y Y
NHL 39 N Y F 19 67 Y Y
NHL 39 N N M 18 63 Y Y
NHL 40 N N M 19 57 Y Y
NHL 46 N N M 8 47 Y N
NHL 59 N Y F 31 61 Y Y
NHL 91 N N M 37 84 N N
NHL 117 N N F 53 58 Y Y
NHL 31 N N M 16 73 Y Y
NHL 16 N N F 8 70 N Y
NHL 10 N N F 5 74 Y N
NHL 6 N N M 3 89 N Y
NHL 12 N N F 6 78 N Y
NHL 12 N N M 6 73 N Y
NHL 38 Y Y F 19 64 N Y
NHL 61 N Y F 27 70 N N
NHL 14 N N M 6 61 N N
NHL 15 N N M 7 73 Y Y
NHL 13 N N F 5 73 N N
NHL 12 N N F 5 79 N N
NHL 7 Y Y F 4 80 Y Y
NHL 25 N N M 11 88 Y N
NHL 15 N N F 7 61 Y Y
NHL 29 N N M 15 73 Y N
NHL 23 N N F 10 56 Y Y
NHL 10 N N M 5 84 N N
NHL 34 N Y M 16 70 Y N
NHL 63 N Y M 26 68 Y Y
NHL 37 N N M 14 57 Y Y
NHL 16 N N F 8 67 Y N
NHL 17 N N M 8 60 Y Y
NHL 15 N N M 7 71 Y Y
NHL 10 N N M 5 58 N N
NHL 59 N N M 27 56 Y Y
NHL 26 N N F 13 78 N N
NHL 24 N N M 11 59 Y Y
NHL 30 N Y M 15 78 N N
NHL 17 Y Y F 7 76 N N
NHL 16 N N M 8 65 N Y
NHL 5 Y Y F 3 74 Y Y
NHL 10 N N M 5 29 Y Y
NHL 24 N N F 11 89 N N
NHL 13 N N M 5 65 Y Y
NHL 25 N Y M 12 60 Y N
NHL 17 N N M 9 50 Y N
NHL 13 N N F 7 77 N Y
NHL 14 N N M 6 48 Y Y
NHL 10 N Y F 4 43 N Y
NHL 49 N N F 21 63 Y Y
NHL 17 N N M 6 87 N N
NHL 18 N Y F 7 79 N N
NHL 20 N N F 10 65 Y Y
NHL 10 N N M 5 73 N Y
NHL 13 N N F 6 63 N N
NHL 11 N N M 5 66 Y Y
NHL 26 N N M 11 59 Y Y
NHL 20 N N M 10 63 Y Y
NHL 24 N N F 9 32 Y N
NHL 15 N Y M 7 72 N N
NHL 14 N N M 6 65 Y Y
NHL 63 N N F 27 54 Y Y
NHL 10 N N M 5 96 N N
NHL 16 N N M 8 63 Y Y
NHL 27 Y Y F 10 71 Y Y
NHL 17 N Y F 6 83 N N
NHL 96 N Y F 47 65 N N
NHL 16 N N M 8 52 Y N
NHL 63 N Y F 22 63 Y Y
NHL 12 N N F 6 61 N N
NHL 10 N N M 5 65 Y Y
NHL 13 N N F 6 80 N N
NHL 15 Y Y M 6 76 Y N
NHL 41 Y Y F 8 53 Y N
NHL 46 N Y F 23 88 N N
NHL 12 Y Y M 5 87 N N
NHL 6 Y Y F 3 77 N N
NHL 10 N N F 5 49 Y Y
NHL 55 Y Y M 24 60 N N
NHL 8 Y Y M 4 65 N N
NHL 16 N N F 8 76 N N
NHL 18 N Y F 9 74 N N
NHL 18 N N F 7 75 Y N
NHL 70 N N F 27 55 Y Y
NHL 10 Y Y M 5 79 N N
NHL 54 N N M 24 68 Y Y
NHL 85 N N M 39 61 Y Y
NHL 33 N N M 12 59 Y Y
NHL 16 N Y M 7 63 N N
NHL 14 N N M 7 77 Y Y
NHL 47 N N F 22 76 N N
NHL 26 N N M 4 67 Y Y
NHL 14 N N F 7 77 Y Y
NHL 45 N N M 22 40 Y Y
NHL 11 N N M 5 81 N N
NHL 42 N N M 18 40 Y Y
NHL 13 N N M 6 67 Y Y
NHL 20 N N F 7 72 N Y
NHL 13 N N M 6 54 Y Y
NHL 20 N Y M 10 62 Y Y
NHL 11 Y Y M 4 69 N N
NHL 10 N N M 4 74 N N
NHL 25 N N M 12 67 Y N
NHL 14 N Y M 6 80 N Y
NHL 67 N N M 33 82 Y N
NHL 27 N Y M 13 81 Y Y
NHL 14 N N F 7 76 Y Y
NHL 16 N Y M 5 62 Y Y
NHL 31 N N F 16 85 Y N
NHL 44 N N M 18 77 N Y
NHL 12 N N M 6 63 Y N
NHL 33 N N M 15 52 Y N
NHL 11 N N F 5 64 Y N
NHL 129 N N F 57 27 Y Y
NHL 12 Y Y F 6 78 Y Y
NHL 89 N N M 37 66 N Y
NHL + AIHA 93 N Y F 34 56 Y Y
NHL 4 Y Y M 2 81 N N
Pancytopaenia 9 N N F 4 54 Y N
Pancytopaenia 30 N N F 10 72 Y N
Pancytopaenia 4 N N F 2 76 N N
Pancytopaenia 5 N N M 2 36 Y N
Pernicious anaemia 102 N N F 28 71 N N
Post transplant
lymphomprofilerative disease 6 N N F 3 24 N N
PRV 17 N N F 8 74 N N
PVR 9 N N F 4 70 N N
Rheumatoid arthritis 14 Y Y M 4 69 N N
Sideroblastic Anaemia 5 N N M 2 72 N N
Sideroblastic Anaemia 244 N Y M 104 66 N N
Sideroblastic Anaemia 76 N N F 28 87 N N
Sideroblastic Anaemia 124 N Y F 41 77 N N
Sideroblastic Anaemia 54 N Y M 26 88 N N
Sideroblastic Anaemia 98 N N F 39 79 N N
T cell prolymphocytic
leukaemia 68 N N M 28 69 Y Y
T-cell leukaemia 45 N Y M 22 71 N N
Thalassaemia 381 Y Y M 127 25 N N

TTP 11 N Y F 3 30 Y N
TTP 3 Y Y F 1 20 N N
VWD 2 N N M 1 52 N N
WM 32 N Y M 16 71 N Y
WM 2 N N F 1 78 Y N
WM 42 Y Y F 19 61 Y Y
WM 20 N N F 12 70 Y N
WM 23 Y Y F 11 75 N Y
WM 7 N N F 3 65 N N
WM 39 N Y F 16 77 N N
WM 21 N Y F 9 73 Y Y
WM 15 N N F 8 91 N Y
WM 4 N N M 2 80 Y Y
WM 16 N N F 8 75 Y Y
WM 89 N N M 41 71 Y Y
10.2 Appendix 2: Raw data for retrospective review of renal cohort
Key: AKI = Acute kidney injury

ARF = Acute renal failure
CKD = Chronic kidney disease
CRF = Chronic renal failure
ESRF = End stage renal failure
Y =Yes
N = No
M =Male
F = Female
No Alloimmunised
units Alloimmunised post Autoimmunised Autoimmunised No. Age at first
Diagnosis given at admission transfusion at admission post transfusion Gender Episodes transfusion Transplant
AKI 94 N N N N M 43 87 N
AKI 7 N N N N M 3 40 N
AKI 3 N N N N F 1 65 N
ARF 30 N N N N F 14 36 Y
ARF 2 N N N N F 1 54 N
ARF 29 N N N N M 13 80 N
No Alloimmunised
ARF 6 N N N N M 3 83 N
ARF 10 N N N N M 5 77 N
ARF 20 N N N N F 9 53 N
ARF 14 N N N N F 6 72 N
ARF 14 N N N N M 6 61 N
ARF 15 N N N N M 6 82 N
No Alloimmunised
ARF 13 N N N N M 5 93 N
ARF 17 N N N N M 7 67 N
ARF 44 N N N N F 13 44 N
ARF 10 N N N N F 4 82 N
ARF 18 N N N N M 8 76 N
CKD 2 N N N N M 1 63 N
CKD 47 N N N N M 13 72 N
CKD 2 N N N N F 1 73 N
No Alloimmunised
CKD 74 N N N N M 35 70 N
CKD 10 N N N N M 4 66 N
CKD 14 N N N N M 7 91 N
CKD 16 N N N N M 8 72 N
CKD 16 N N N N F 7 76 N
CKD 36 N N N N F 14 75 N
CKD 25 N N N N M 13 69 N
No Alloimmunised
CKD 14 N N N N F 7 79 N
CKD 12 N N N N M 6 52 N
CKD 10 N N N N M 5 87 N
CKD 44 N N N N M 16 95 N
CKD 13 N N N N F 6 55 N
CKD 16 N N N N M 5 65 N
CKD 39 N N N N F 20 59 N
CKD 12 N N N N M 5 71 N
CKD 16 N N N N M 8 83 N
No Alloimmunised
CKD 15 N N N N M 8 71 N
CKD 9 N Y N Y M 4 69 N
CKD 2 Y Y Y Y F 1 43 N
CKD 14 N Y N Y F 7 65 N
CKD 98 N Y N Y F 40 55 N
CKD 25 Y Y Y Y M 13 74 N
CRF 29 N N N N F 13 59 N
CRF 3 N N N N M 1 37 Y
CRF 3 N N N N F 1 44 Y
CRF 48 N N N N M 22 55 Y
CRF 26 N N N N M 13 15 Y
CRF 21 N N N N M 9 57 Y
CRF 12 N N N N M 5 57 Y
CRF 12 N N N N F 5 32 Y
CRF 22 N N N N F 11 67 Y
No Alloimmunised
CRF 11 N N N N M 4 45 Y
CRF 31 N N N N M 19 37 Y
CRF 12 N N N N F 6 54 Y
CRF 15 N N N N M 7 45 Y
CRF 22 N N N N F 8 32 Y
CRF 22 N Y N N M 7 42 Y
CRF 143 N Y N N M 54 45 Y
CRF 4 N Y N N M 2 38 Y
No Alloimmunised
CRF 9 N Y N Y M 4 44 Y
CRF 3 N N N N M 2 84 N
CRF 2 N N N N F 1 74 N
CRF 13 N N N Y M 6 53 N
CRF 10 N N N N F 4 63 N
CRF 12 N N N N M 5 65 N
CRF 20 N N N N F 9 77 N
CRF 11 N N N N M 4 80 N
No Alloimmunised
CRF 21 N N N N M 9 78 N
CRF 32 N N N N F 14 58 N
CRF 11 N N N N M 5 79 N
CRF 14 N N N N M 4 65 N
CRF 12 N N N N M 5 68 N
CRF 10 N N N N F 4 80 N
CRF 11 N N N N M 5 79 N
No Alloimmunised
CRF 17 N N N N M 6 83 N
CRF 34 N N N N F 13 46 N
CRF 246 N N N N M 67 75 N
CRF 27 N N N N M 14 42 N
CRF 22 N N N N M 8 73 N
No Alloimmunised
CRF 10 N N N N F 4 76 N
CRF 10 N N N N M 4 64 N
CRF 66 N N N N F 26 47 N
CRF 17 N N Y Y F 8 76 N
CRF 31 N N N N F 15 68 N
CRF 12 N N N N F 6 71 N
CRF 21 N N N N F 9 67 N
CRF 13 N N N N F 6 35 N
CRF 75 N N N Y M 36 65 N
CRF 15 N N Y Y M 6 76 N
No Alloimmunised
CRF 13 N N N N F 5 70 N
CRF 10 N N N N M 5 78 N
CRF 10 N N N N F 4 54 N
CRF 16 N N N N F 8 63 N
CRF 16 N N N N M 7 88 N
CRF 24 N N Y Y M 9 81 N
CRF 46 N N N N M 18 83 N
CRF 12 N N N N F 6 55 N
CRF 56 N N N N M 24 54 N
CRF 11 N N N N M 4 82 N
CRF 62 N N N N F 28 53 N
CRF 11 N N N N M 5 75 N
CRF 24 N N N N M 9 77 N
CRF 19 N N N N M 9 67 N
CRF 50 N N N N M 21 57 N
CRF 25 N N N N M 11 65 N
CRF 21 N N N N F 10 74 N
CRF 18 N N N N M 9 66 N
No Alloimmunised
CRF 25 N N N N M 9 80 N
CRF 6 N N Y Y M 3 85 N
CRF 18 N N N N F 9 77 N
CRF 10 N N N N M 4 78 N
CRF 14 N N N N M 6 60 N
CRF 38 N N N N M 18 61 N
CRF 162 N N N N M 34 76 N
CRF 12 N N N N M 6 78 N
CRF 10 N N N N M 4 84 N
CRF 12 N N N N F 4 56 N
CRF 21 N N N N F 4 79 N
CRF 10 N N N N M 4 72 N
No Alloimmunised
CRF 16 N N N N F 7 71 N
CRF 2 N N Y Y M 1 66 N
CRF 13 N N N N M 6 61 N
CRF 13 N N N N M 6 48 N
CRF 31 N N N N F 13 75 N
CRF 17 N N N N F 7 65 N
No Alloimmunised
CRF 41 N Y N Y M 16 71 N
CRF 5 Y Y N N F 2 62 N
CRF 15 N Y N N F 7 74 N
CRF 21 N Y N N F 11 24 N
CRF 9 N Y N N F 5 73 N
CRF 12 Y Y N Y M 6 79 N
CRF 2 Y Y N N M 1 82 N
CRF 4 N Y N N M 2 58 N
CRF 78 Y Y N Y F 37 67 N
CRF 36 N Y N Y M 11 77 N
CRF 12 N Y N N F 5 80 N
CRF 3 Y Y N N F 1 81 N
CRF 6 Y Y Y Y F 3 76 N
CRF 16 N Y N N F 7 68 N
CRF 22 N Y N N F 9 43 N
CRF 40 N Y N N M 18 49 N
CRF 14 N Y N Y M 6 43 N
No Alloimmunised
CRF 12 N Y N Y F 6 44 N
CRF 23 N Y N N F 9 81 N
CRF 5 Y Y N Y F 2 76 N
CRF 10 N Y N Y M 4 77 N
CRF 8 Y Y N Y F 3 80 N
CRF 7 N Y N Y M 3 86 N
CRF 12 N Y N N M 6 76 N
CRF 103 N Y N N F 51 65 N
CRF 13 Y Y N Y M 4 73 N
CRF 107 N Y N N F 58 69 N
CRF 25 N Y N Y F 11 68 N
CRF 42 N Y N Y M 18 63 N
CRF 10 N Y N Y F 5 63 N
CRF 4 Y Y Y Y F 2 62 N
CRF 10 N N N N F 5 79 N
CRF 3 N N N N M 1 49
ESRF 37 N N N Y F 17 34 N
ESRF 18 N N N Y M 8 68 N
ESRF 37 N N N N M 18 53 N
ESRF 5 N N N N M 2 38 Y
No Alloimmunised
ESRF 27 N N N N F 12 29 Y
ESRF 5 N N N N F 2 78 N
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
ESRF 4 N N Y Y M 2 77 N
No Alloimmunised
ESRF 132 N N N N F 58 70 N
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
ESRF 100 N N N N M 44 36 N
No Alloimmunised
No Alloimmunised
No Alloimmunised
No Alloimmunised
ESRF 63 N Y N Y M 29 67 N
ESRF 5 N Y N Y F 2 62 N
ESRF 73 Y Y Y Y F 36 52 N
ESRF 13 N Y N N F 6 33 N
ESRF 4 Y Y N N F 2 83 N
No Alloimmunised
ESRF 2 Y Y N N M 1 84 N
ESRF 36 N Y N N M 15 71 N
ESRF 16 Y Y N N M 7 79 N
ESRF 119 N Y N Y M 53 78 N
No Alloimmunised
ESRF 6 Y Y N Y F 2 85 N
ESRF 101 N Y N Y M 45 77 N
ESRF 8 N Y Y Y F 4 77 N
No Alloimmunised
ESRF 22 Y Y N Y M 11 65 N
ESRF 11 Y Y N Y M 5 61 N
ESRF 6 Y Y Y Y M 3 80 N
HUS 6 N N N N M 3 26 N
NHL 5 N Y N Y M 3 82 N
Pyelonephritis 5 N N N N M 2 37 N
Pyelonephritis 19 N N N N M 9 75 N
Wegener
granulomatosis 5 N N N N M 3 65 N
No Alloimmunised
Wegener
granulomatosis 5 N N N N F 2 57 N
Wegener
Wegener
Wegener
Wegener
Wegener
Wegener
Wegener
Wegener
Wegener
10.3 Appendix 3: Programmed cut off values for the serological phenotyping profiles on the NEO analyser
Antiserum Cut off low Cut off high Antiserum Cut off low Cut off high
Negative 0.000 23.998 Negative 0.000 23.998
Not determined 23.999 27.998
Not determined 23.999 75.998 Positive (1+) 27.999 35.998
Anti-C Anti-K
Positive (2+) 35.999 50.998
Positive (3+) 75.999 80.998 Positive (3+) 50.999 80.998
Postive (4+) 80.999 99.999 Postive (4+) 80.999 99.999
Not determined 23.999 75.998 Not determined 23.999 50.998

Anti-c Anti-Cw
Postive (4+) 80.999 99.999 Postive (4+) 80.999 99.999

Anti-E Anti-M

Postive (4+) 80.999 99.999 Postive (4+) 80.999 99.999
Anti-e Anti-N
Positive (2+) 35.999 50.998
Postive (4+) 80.999 99.999 Postive (4+) 80.999 99.999
Antiserum Cut off low Cut off high Antiserum Cut off low Cut off high
Anti-Fya Anti-S
Postive (4+) 90.999 99.999 Postive (4+) 90.999 99.999
Anti-Fyb Anti-s
Postive (4+) 90.999 99.999 Postive (4+) 90.999 99.999
Anti-Jka Anti-k
Postive (4+) 90.999 99.999 Postive (4+) 90.999 99.999
Anti-Jkb Control Negative
Postive (4+) 90.999 99.999 Postive (4+) 90.999 99.999
10.4 Appendix 4: Raw data for genotyping results
Interrogated
Blood Antigens Predicted
Group (ISBT Phenotype
Symbol Sample CSV System Alleles Assayed Genotype Result Phenotype) Result
New Batch
Green Unknown1 4_20140514_125813.csv Rh RHCE*ce RHCE*cE C (RH:2) 0
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*Ce E (RH:3) +
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*cE c (RH:4) +
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CE e (RH:5) 0
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CeCW CW (RH:8) 0
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*ceCW V (RH:10) 0
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CECW hrS (RH:19) +
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*ceAR VS (RH:20) 0
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CeFV hrB (RH:31) 0
Symbol Sample CSV Blood Alleles Assayed Genotype Result Interrogated Predicted
Group Antigens Phenotype
System (ISBT Result
Phenotype)
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CeVG
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*cEFM
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*ce[712G]
New Batch
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*ce[733G,1006T]
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*CE-D[2, 5, 7]-CE
New Batch
Green Unknown1 4_20140514_125813.csv RHCE*cE[697G,712G,733G]
New Batch RHD*r’s-
New Batch
Green Unknown1 4_20140514_125813.csv Kell KEL*K_KPB_JSB KEL*k_KPB_JSB K (KEL:1) 0
New Batch
Green Unknown1 4_20140514_125813.csv KEL*k_KPB_JSB k (KEL:2) +
New Batch
Green Unknown1 4_20140514_125813.csv KEL*k_KPA_JSB Kpa (KEL:3) 0
New Batch
Green Unknown1 4_20140514_125813.csv KEL*k_KPB_JSA Kpb (KEL:4) +
System (ISBT Result
Phenotype)
New Batch
Green Unknown1 4_20140514_125813.csv Jsa (KEL:6) 0
New Batch
Green Unknown1 4_20140514_125813.csv Jsb (KEL:7) +
New Batch
Green Unknown1 4_20140514_125813.csv Kidd JK*A JK*A, JK*B Jka (JK:1) +
New Batch
Green Unknown1 4_20140514_125813.csv JK*B Jkb (JK:2) +
New Batch
Green Unknown1 4_20140514_125813.csv JK*B_null(IVS5-1a)
New Batch
Green Unknown1 4_20140514_125813.csv JK*B_null(871C)
New Batch
Green Unknown1 4_20140514_125813.csv Duffy FY*A FY*A, FY*B Fya (FY:1) +
New Batch
Green Unknown1 4_20140514_125813.csv FY*B Fyb (FY:2) +
New Batch
Green Unknown1 4_20140514_125813.csv FY*B_GATA
New Batch
Green Unknown1 4_20140514_125813.csv FY*B[265T]_FY*X
New Batch
Green Unknown1 4_20140514_125813.csv MNS GYPA*M GYPA*N M (MNS:1) 0
New Batch
Green Unknown1 4_20140514_125813.csv GYPA*N N (MNS:2) +
System (ISBT Result
Phenotype)
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*S GYPB*s S (MNS:3) 0
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*s s (MNS:4) +
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*S_null(IVS5+5t) U (MNS:5) +
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*S_null(230T) Mia (MNS:7) 0
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*deletion
New Batch
Green Unknown1 4_20140514_125813.csv GYPB*Mur
New Batch
Green Unknown1 4_20140514_125813.csv Diego DI*A DI*B Dia (DI:1) 0
New Batch
Green Unknown1 4_20140514_125813.csv DI*B Dib (DI:2) +
New Batch
Green Unknown1 4_20140514_125813.csv Dombrock DO*A DO*A, DO*B Doa (DO:1) +
New Batch
Green Unknown1 4_20140514_125813.csv DO*B Dob (DO:2) +
New Batch
Green Unknown1 4_20140514_125813.csv DO*B_HY- Hy (DO:4) +
New Batch
Green Unknown1 4_20140514_125813.csv DO*A_JOA- Joa (DO:5) +
System (ISBT Result
Phenotype)
New Batch
Green Unknown1 4_20140514_125813.csv Colton CO*A CO*A Coa (CO:1) +
New Batch
Green Unknown1 4_20140514_125813.csv CO*B Cob (CO:2) 0
New Batch
Green Unknown1 4_20140514_125813.csv Cartwright YT*A YT*A Yta (YT:1) +
New Batch
Green Unknown1 4_20140514_125813.csv YT*B Ytb (YT:2) 0
New Batch
Green Unknown1 4_20140514_125813.csv Lutheran LU*A LU*B Lua (LU:1) 0
New Batch
Green Unknown1 4_20140514_125813.csv LU*B Lub (LU:2) +
New Batch
Green Unknown2 4_20140514_125813.csv Rh RHCE*ce RHCE*ce, RHCE*cE C (RH:2) 0
New Batch
New Batch
New Batch
Green Unknown2 4_20140514_125813.csv RHCE*CE e (RH:5) +
New Batch
New Batch
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
Green Unknown2 4_20140514_125813.csv RHCE*CeFV hrB (RH:31) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
System (ISBT Result
Phenotype)
New Batch
New Batch
Green Unknown2 4_20140514_125813.csv MNS GYPA*M GYPA*M M (MNS:1) +
New Batch
Green Unknown2 4_20140514_125813.csv GYPA*N N (MNS:2) 0
New Batch
Green Unknown2 4_20140514_125813.csv GYPB*S GYPB*S, GYPB*s S (MNS:3) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown2 4_20140514_125813.csv Dombrock DO*A DO*A Doa (DO:1) +
System (ISBT Result
Phenotype)
New Batch
Green Unknown2 4_20140514_125813.csv DO*B Dob (DO:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown3 4_20140514_125813.csv Kidd JK*A JK*B Jka (JK:1) 0
New Batch
New Batch
New Batch
100 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown3 4_20140514_125813.csv MNS GYPA*M GYPA*M, GYPA*N M (MNS:1) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
101 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
Green Unknown3 4_20140514_125813.csv Dombrock DO*A DO*B Doa (DO:1) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
102 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green Unknown4 4_20140514_125813.csv Rh RHCE*ce RHCE*ce, RHCE*Ce C (RH:2) +
New Batch
Green Unknown4 4_20140514_125813.csv RHCE*Ce E (RH:3) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
103 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch KEL*K_KPB_JSB,
Green Unknown4 4_20140514_125813.csv Kell KEL*K_KPB_JSB KEL*k_KPB_JSB K (KEL:1) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown4 4_20140514_125813.csv Kidd JK*A JK*A Jka (JK:1) +
104 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green Unknown4 4_20140514_125813.csv JK*B Jkb (JK:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
105 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
106 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
107 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
108 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
Green Unknown5 4_20140514_125813.csv Kidd JK*A JK*A Jka (JK:1) +
New Batch
Green Unknown5 4_20140514_125813.csv JK*B Jkb (JK:2) 0
New Batch
New Batch
New Batch
Green Unknown5 4_20140514_125813.csv Duffy FY*A FY*A Fya (FY:1) +
New Batch
Green Unknown5 4_20140514_125813.csv FY*B Fyb (FY:2) 0
New Batch
New Batch
New Batch
Green Unknown5 4_20140514_125813.csv MNS GYPA*M GYPA*N M (MNS:1) 0
New Batch
109 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown5 4_20140514_125813.csv Dombrock DO*A DO*A Doa (DO:1) +
New Batch
Green Unknown5 4_20140514_125813.csv DO*B Dob (DO:2) 0
New Batch
New Batch
110 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
111 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
112 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green Unknown6 4_20140514_125813.csv Kidd JK*A JK*B Jka (JK:1) 0
New Batch
New Batch
New Batch
New Batch
Green Unknown6 4_20140514_125813.csv Duffy FY*A FY*B Fya (FY:1) 0
New Batch
New Batch
113 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
114 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
115 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
116 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
117 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
118 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
119 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green Unknown8 4_20140514_125813.csv Rh RHCE*ce RHCE*Ce C (RH:2) +
New Batch
Green Unknown8 4_20140514_125813.csv RHCE*Ce E (RH:3) 0
New Batch
Green Unknown8 4_20140514_125813.csv RHCE*cE c (RH:4) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
120 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
121 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
122 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
123 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
Green EX1307335 5_20140514_171740.csv Rh RHCE*ce RHCE*Ce C (RH:2) +
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*Ce E (RH:3) 0
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*cE c (RH:4) 0
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CE e (RH:5) +
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CeCW CW (RH:8) 0
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*ceCW V (RH:10) 0
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CECW hrS (RH:19) +
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*ceAR VS (RH:20) 0
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CeFV hrB (RH:31) +
124 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CeVG
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*cEFM
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*ce[712G]
New Batch
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*ce[733G,1006T]
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*CE-D[2, 5, 7]-CE
New Batch
Green EX1307335 5_20140514_171740.csv RHCE*cE[697G,712G,733G]
New Batch
Green EX1307335 5_20140514_171740.csv Kell KEL*K_KPB_JSB KEL*k_KPB_JSB K (KEL:1) 0
New Batch
Green EX1307335 5_20140514_171740.csv KEL*k_KPB_JSB k (KEL:2) +
New Batch
Green EX1307335 5_20140514_171740.csv KEL*k_KPA_JSB Kpa (KEL:3) 0
New Batch
Green EX1307335 5_20140514_171740.csv KEL*k_KPB_JSA Kpb (KEL:4) +
125 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307335 5_20140514_171740.csv Jsa (KEL:6) 0
New Batch
Green EX1307335 5_20140514_171740.csv Jsb (KEL:7) +
New Batch
Green EX1307335 5_20140514_171740.csv Kidd JK*A JK*A, JK*B Jka (JK:1) +
New Batch
Green EX1307335 5_20140514_171740.csv JK*B Jkb (JK:2) +
New Batch
Green EX1307335 5_20140514_171740.csv JK*B_null(IVS5-1a)
New Batch
Green EX1307335 5_20140514_171740.csv JK*B_null(871C)
New Batch
Green EX1307335 5_20140514_171740.csv Duffy FY*A FY*B Fya (FY:1) 0
New Batch
Green EX1307335 5_20140514_171740.csv FY*B Fyb (FY:2) +
New Batch
Green EX1307335 5_20140514_171740.csv FY*B_GATA
New Batch
Green EX1307335 5_20140514_171740.csv FY*B[265T]_FY*X
New Batch
Green EX1307335 5_20140514_171740.csv MNS GYPA*M GYPA*M, GYPA*N M (MNS:1) +
New Batch
Green EX1307335 5_20140514_171740.csv GYPA*N N (MNS:2) +
126 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*S GYPB*S, GYPB*s S (MNS:3) +
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*s s (MNS:4) +
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*S_null(IVS5+5t) U (MNS:5) +
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*S_null(230T) Mia (MNS:7) 0
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*deletion
New Batch
Green EX1307335 5_20140514_171740.csv GYPB*Mur
New Batch
Green EX1307335 5_20140514_171740.csv Diego DI*A DI*B Dia (DI:1) 0
New Batch
Green EX1307335 5_20140514_171740.csv DI*B Dib (DI:2) +
New Batch
Green EX1307335 5_20140514_171740.csv Dombrock DO*A DO*A, DO*B Doa (DO:1) +
New Batch
Green EX1307335 5_20140514_171740.csv DO*B Dob (DO:2) +
New Batch
Green EX1307335 5_20140514_171740.csv DO*B_HY- Hy (DO:4) +
New Batch
Green EX1307335 5_20140514_171740.csv DO*A_JOA- Joa (DO:5) +
127 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307335 5_20140514_171740.csv Colton CO*A CO*A Coa (CO:1) +
New Batch
Green EX1307335 5_20140514_171740.csv CO*B Cob (CO:2) 0
New Batch
Green EX1307335 5_20140514_171740.csv Cartwright YT*A YT*A Yta (YT:1) +
New Batch
Green EX1307335 5_20140514_171740.csv YT*B Ytb (YT:2) 0
New Batch
Green EX1307335 5_20140514_171740.csv Lutheran LU*A LU*B Lua (LU:1) 0
New Batch
Green EX1307335 5_20140514_171740.csv LU*B Lub (LU:2) +
New Batch
Green EX1307336 5_20140514_171740.csv Rh RHCE*ce RHCE*Ce, RHCE*cE(10) C (RH:2) +
New Batch
Green EX1307336 5_20140514_171740.csv RHCE*Ce E (RH:3) +
New Batch
Green EX1307336 5_20140514_171740.csv RHCE*cE c (RH:4) +
New Batch
New Batch
New Batch
128 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
129 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1307336 5_20140514_171740.csv Kidd JK*A JK*A Jka (JK:1) +
New Batch
Green EX1307336 5_20140514_171740.csv JK*B Jkb (JK:2) 0
New Batch
New Batch
New Batch
Green EX1307336 5_20140514_171740.csv Duffy FY*A FY*B, FY*B[265T]_FY*X Fya (FY:1) 0
New Batch
New Batch
130 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
Green EX1307336 5_20140514_171740.csv GYPB*S GYPB*S S (MNS:3) +
New Batch
Green EX1307336 5_20140514_171740.csv GYPB*s s (MNS:4) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1307336 5_20140514_171740.csv Dombrock DO*A DO*B Doa (DO:1) 0
131 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
132 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
133 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
134 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307657 5_20140514_171740.csv Duffy FY*A FY*A, FY*B Fya (FY:1) +
New Batch
New Batch
New Batch
New Batch
Green EX1307657 5_20140514_171740.csv MNS GYPA*M GYPA*M M (MNS:1) +
New Batch
Green EX1307657 5_20140514_171740.csv GYPA*N N (MNS:2) 0
New Batch
Green EX1307657 5_20140514_171740.csv GYPB*S GYPB*s S (MNS:3) 0
New Batch
New Batch
New Batch
New Batch
New Batch
135 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1307657 5_20140514_171740.csv Colton CO*A CO*A, CO*B Coa (CO:1) +
New Batch
Green EX1307657 5_20140514_171740.csv CO*B Cob (CO:2) +
New Batch
New Batch
New Batch
New Batch
136 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1307898 5_20140514_171740.csv Rh RHCE*ce RHCE*ce C (RH:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
137 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
138 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
139 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
140 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
Green EX1307899 5_20140514_171740.csv Rh RHCE*ce RHCE*ce, RHCE*cE C (RH:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
141 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1307899 5_20140514_171740.csv Kell KEL*K_KPB_JSB KEL*k_KPB_JSB K (KEL:1) +
New Batch
New Batch
New Batch
142 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
143 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
144 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
145 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
146 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
147 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
Green EX1308240 5_20140514_171740.csv MNS GYPA*M GYPA*N M (MNS:1) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
148 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1308368 5_20140514_171740.csv Rh RHCE*ce RHCE*ce, RHCE*Ce C (RH:2) +
New Batch
New Batch
149 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
150 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
151 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1308368 5_20140514_171740.csv Duffy FY*A FY*A Fya (FY:1) +
New Batch
Green EX1308368 5_20140514_171740.csv FY*B Fyb (FY:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
152 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
153 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1308578 5_20140514_171740.csv Rh RHCE*ce RHCE*ce, RHCE*cE C (RH:2)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
154 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
155 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
156 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1308578 5_20140514_171740.csv Cartwright YT*A YT*A, YT*B Yta (YT:1) +
157 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1308578 5_20140514_171740.csv YT*B Ytb (YT:2) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
158 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
159 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
160 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
161 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
162 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
163 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
164 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
165 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
166 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
167 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
168 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
169 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
170 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
171 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1308959 5_20140514_171740.csv Kidd JK*A JK*B Jka (JK:1) 0
172 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
173 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
174 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
Green EX1311437 5_20140514_171740.csv Rh RHCE*ce RHCE*Ce, RHCE*CeCW C (RH:2) +
New Batch
New Batch
New Batch
New Batch
Green EX1311437 5_20140514_171740.csv RHCE*CeCW CW (RH:8) +
New Batch
New Batch
New Batch
New Batch
175 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
176 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
177 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
178 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
179 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
180 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
181 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
182 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
183 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
184 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
185 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
186 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
187 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
188 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
189 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
190 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
191 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
192 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
193 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
194 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
195 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
196 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
197 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
198 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1311989 6_20140515_111617.csv Dombrock DO*A DO*A Doa (DO:1) +
199 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1311989 6_20140515_111617.csv DO*B Dob (DO:2) 0
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1311989 6_20140515_111617.csv Lutheran LU*A LU*A, LU*B Lua (LU:1) +
New Batch
New Batch
New Batch
New Batch
200 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
201 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
202 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
203 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1312294 6_20140515_111617.csv Lutheran LU*A LU*A, LU*B Lua (LU:1) +
New Batch
204 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
205 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
206 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
207 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
208 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
209 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
210 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
211 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
212 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
213 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
214 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
215 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
216 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
217 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
218 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
219 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
220 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
221 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
222 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
223 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
224 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
225 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
226 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
227 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
228 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
229 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
230 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
231 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
232 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
233 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
234 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
235 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
236 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
237 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
238 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1312675 6_20140515_111617.csv Rh RHCE*ce RHCE*Ce, RHCE*CeCW C (RH:2) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
239 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
240 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
241 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
242 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
243 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
244 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
245 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
246 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
247 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
248 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
249 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
250 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1313154 6_20140515_111617.csv Cartwright YT*A YT*A, YT*B Yta (YT:1) +
New Batch
Green EX1313154 6_20140515_111617.csv YT*B Ytb (YT:2) +
New Batch
New Batch
New Batch
New Batch
New Batch
251 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
252 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
253 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
254 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
255 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
Green EX1313515 6_20140515_111617.csv Rh RHCE*ce RHCE*cE C (RH:2) 0
New Batch
New Batch
New Batch
Green EX1313515 6_20140515_111617.csv RHCE*CE e (RH:5) 0
New Batch
New Batch
New Batch
New Batch
New Batch
Green EX1313515 6_20140515_111617.csv RHCE*CeFV hrB (RH:31) 0
New Batch
New Batch
New Batch
256 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
257 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
258 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
259 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
Green EX1313516 6_20140515_111617.csv Rh RHCE*ce RHCE*ce, RHCE*CeCW(12) C (RH:2) +
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
260 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
261 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
262 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
263 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
264 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
265 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
266 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
267 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
268 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
269 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
270 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
271 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
272 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
273 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
274 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
275 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
276 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
277 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
278 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
279 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
280 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
281 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
282 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
283 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
284 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
285 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
286 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
287 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
288 | A p p e n d i x
System (ISBT Result
Phenotype)
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
New Batch
289 | A p p e n d i x
10.5 Appendix 5: Patient information leaflet given to the
participants in the pilot study
PATIENT INFORMATION SHEET

BLOOD GROUP TYPING AND MATCHING OF BLOOD FOR
TRANSFUSION IN CHRONICALLY TRANSFUSED PATIENTS
Part 1 of the Information Sheet

1. Invitation
You are being invited to take part in a research study. Before you
decide whether or not to take part, it is important for you to understand
why the research is being done and what it will involve. Please take time
to read the following information carefully, and discuss it with others if you
wish. One of our team will go through the information with you and
answer any questions you have.
Ask us if there is anything that is not clear, or if you would like more
information. Take time to decide whether or not you wish to take part.
Normally blood for transfusion is matched for major blood groups only.
The purpose of this study is to see if it is feasible to provide blood that is
more fully matched with your blood type.
2. What is the purpose of the study?

The study will give us information on whether it is possible to provide
blood that is more fully matched with your blood type. This information
can then be used for a larger study that will look at whether giving blood
that is a closer match to your own can reduce the amount of transfusions
needed for your disease type.
3. Why have I been invited?

You have been chosen because your condition means that you will need
regular blood transfusions.
4. Do I have to take part?

No. It is up to you to decide whether or not to take part. If you decide to
take part you will be given this information sheet to keep and be asked to
sign a consent form to confirm that you understand what is involved when
290 | A p p e n d i x
taking part in this study. If you decide to take part you are free to leave
the study at any time and without giving a reason. If you withdraw, unless
you object, we will still keep records relating to the treatment given to you,
as this is valuable to the study. A decision to withdraw at any time, or a
decision not to take part, will not affect the quality of care you receive
5. What will happen to me if I take part?

 You will have the normal blood samples taken as part of the pre-
transfusion testing protocol.
 You will only need to attend as usual for your normal blood tests
and blood transfusions as required by your doctor.
 Your blood will be tested for blood type using your blood cells and
your DNA.
 This study will last for up to 3 years.
 This study is a Randomised Trial. Sometimes we don‘t know which
way of treating patients is best. To find out, we need to compare
different treatments. We put people into groups and give each
group a different treatment. The results are compared to see if one
is better. To try to make sure the groups are the same to start with,
each patient is put into a group by chance (randomly). This study
involves two groups; one group will be given blood matched to
current level and the other group will be given the blood that is a
closer match. You have a 50% chance of being in the group being
given blood of a closer match to your own type.
6. Expenses and Payment

There are no expenses or payments associated with this trial.
7. What will I need to do?

You will only need to attend as usual for your normal blood tests and
blood transfusions as required by your doctor.
8. What are the alternatives for diagnosis or treatment?

The alternative is the current practice, providing blood that is matched
only for the major blood types.
9. What are the possible disadvantages and risks when taking

part?
No side effects are expected as a result of participation in this trial.
However, if you do decide to take part in the study, you must report any
problems you have to your study nurse or doctor. There is also a contact
number given at the end of this information sheet for you to phone if you
become worried at any time. In the unlikely event of an emergency
occurring during the conduct of the study, we may contact your
nominated next of kin.
10. What are the possible benefits of taking part?
291 | A p p e n d i x
We cannot promise the study will help you but the information we get
might help improve the future treatment of people with chronic transfusion
requirements.
11. What happens when the research study stops?

You will continue to receive your blood transfusion as required and
discussed with your doctor.
12. What if there is a problem?

If you have a concern about any aspect of this study, you should ask to
speak with the researchers who will do their best to answer your
question. The researchers contact details can be found at the end of this
information sheet. If you remain unhappy and wish to complain formally,
you can do this through the NHS Complaints Procedure. Details can be
obtained from the hospital.
In the event that something does go wrong and you are harmed during
the research study there are no special compensation arrangements. If
you are harmed and this is due to someone’s negligence then you may
have grounds for a legal action for compensation but you may have to
pay your legal costs. The normal National Health Service complaints
mechanisms will still be available to you.
13. Will my part in this study be kept confidential?

Yes. We will follow ethical and legal practice and all information
about you will be handled in confidence. Any information about you which
leaves the hospital will have your name and address removed so that you
cannot be recognised.
This completes part 1 of the information sheet.
If the information in part 1 has interested you and you are considering
participation, please read the information in part 2 before making any
decisions.
Part 2 of the information sheet
14. What if new information becomes available?
Sometimes during the course of a clinical trial, new information becomes

available on the procedures that are being studied. If this happens, we
will tell you about it and discuss with you whether you want to or should
continue in the study. If you decide to withdraw, we will make
arrangements for your care to continue. If you decide to continue in the
study you will be asked to sign an updated consent form.
292 | A p p e n d i x
On receiving new information, we might consider it to be in your best
interests to withdraw you from the study. If so, we will explain the reasons
and arrange for your care to continue.
If the study is stopped for any other reason, you will be told why and your
continuing care will be arranged.
15. What will happen if I don’t want to carry on with the study?
You can withdraw from the study at any time, this will not affect your care.
16. Informing your General Practitioner (GP)

Your GP will be made aware that you are involved in this trial.
17. What will happen to any samples I give?

Your samples will be tested as described in this information sheet. They
will not be used for any other research purposes and they will be
disposed of in accordance with Trust guidelines.
18. Will any Genetic testing be done?

Your DNA will be used to do blood group typing but no other genetic
analysis will be performed on your DNA and your DNA will not be used
for any other research.
19. What will happen to the results of this clinical trial?
The results of the study will be available after it finishes and will usually
be published in a medical journal or be presented at a scientific
conference. The data will be anonymous and none of the patients
involved in the trial will be identified in any report or publication.
Should you wish to see the results, or the publication, please ask your
study doctor.
20. Who is organising and funding this clinical trial?

This study is being organised by a member of staff in the Haematology
Department at the RD&E who is undertaking a Professional Doctorate at
the University of the West of England. It is being funded by the RD&E.
21. Who has reviewed the study?

All research in the NHS is looked at by independent group of people,
called a Research Ethics Committee, to protect your interests. This study
has been reviewed and given favourable opinion by
______________Research Ethics Committee.
22. Contact for further information
You are encouraged to ask any questions you wish, before, during or
after your treatment.
293 | A p p e n d i x
General information on research can be found on the Health Research
Authority website the web address is given below:
http://www.hra.nhs.uk/hra/patients-and-the-public/
If you have any questions about this study, please speak to your research
specialist/coordinator or your haematology doctor, who will be able to
provide you with up to date information about the treatment involved. If
you wish to read the research on which this study is based, please ask
your research specialist/coordinator.
If you would like advice as to whether you should participate please

contact your research specialist/coordinator or your haematology doctor.
If you have any concerns during the study then you can contact the
research specialist/coordinator on the details given in this leaflet, or you
can contact the Haematology ward on 01392 402882.
If you decide you would like to take part then please read and sign the
consent form. You will be given a copy of this information sheet and the
consent form to keep. A copy of the consent form will be filed in your
patient notes, one will be filed with the study records and one may be
sent to the Research Sponsor.
You can have more time to think this over if you are at all unsure.
Thank you for taking the time to read this information sheet and to
consider this study.
Contact Details
Your Research/Specialist Coordinator
Name Jennifer Davies Tel. Number: 01392 402959
If you have any further questions or require clarification please feel

free to contact:
Jennifer Davies
Haematology Department
Royal Devon and Exeter NHS Foundation Trust
Barrack Road
Exeter
EX2 5DW
01392 402959
Jeni.davies@rdeft.nhs.uk
Chairman: James Brent Chief Executive: Angela Pedder
294 | A p p e n d i x
10.2 Appendix 6: Consent form signed by participants in the pilot
study
295 | A p p e n d i x

Developing A Strategy For Red Cell Antigen Typing AND Matching of Blood For Chronic Transfusion

Uploaded by

Copyright:

Available Formats

Developing A Strategy For Red Cell Antigen Typing AND Matching of Blood For Chronic Transfusion

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Developing A Strategy For Red Cell Antigen Typing AND Matching of Blood For Chronic Transfusion

Uploaded by

Copyright:

Available Formats

DEVELOPING A STRATEGY

FOR RED CELL ANTIGEN TYPING

FOR CHRONIC TRANSFUSION

Jennifer Mary Davies

A thesis submitted in partial fulfilment of the requirements of

This research program was carried out in collaboration with the

Submission date: July 2018

transfusion reactions and can also create challenges with subsequent

compatibility testing, sourcing compatible blood, delays in the provision of

blood and cost implications. This is particularly relevant in patients who

are dependent on chronic transfusion support; such as those with sickle

cell disease, thalassaemia, haematological malignancies and renal

disease. A retrospective review of chronically transfused patients with

haematological malignancies and renal disease, treated at the Royal

Devon and Exeter NHS Foundation Trust (RD&E), revealed considerably

allo/autoantibodies than for non-immunised patients, providing new

evidence to support the cost effectiveness of extended antigen matching

for these groups of patients.

The risk of development of red cell allo/autoantibodies could be reduced

by a type and match strategy for, at least, Rh (CcEe) and K antigens as

novel technique was validated for high throughput extended automated

A large scale randomised controlled trial to investigate the benefit and

randomly assigned to a standard care or intervention (type and match)

and provided new evidence supporting the potential to implement a type

Implementation of an extended type and match strategy, utilising the new

malignancies and renal disease at the RD&E. Patients identified as

potential recipients of chronic transfusion support are serologically

phenotyped for Rh (CcEe) and K antigens prior to transfusion. Changes

information management system to support provision of matched blood

for transfusion. The program is being monitored for its effectiveness in

reducing both allo/autoimmunisation rate and the cost of provision of

Figure 1.1 Structure of the ABO antigens ............................................... 17

Figure 1.2 Diagrammatic representations of the RhD and RhCE antigen

Figure 1.3 The Kell and Xk heterodimer ................................................. 24

Figure 1.4 The Duffy glycoprotein ........................................................... 28

Figure 1.5 The Kidd glycoprotein ............................................................ 30

Figure 1.6 Diego antigens located on band 3 glycoprotein ..................... 32

Figure 1.7 Dombrock antigens on the ART4 glycoprotein ....................... 34

Figure 1.8 The Colton glycoprotein structure .......................................... 36

Figure 1.9 Structure of the Lutheran glycoprotein ................................... 39

Figure 1.10: Principles and reactions of the direct agglutination assay

Figure 1.11: Principles of the column agglutination assay ...................... 53

Figure 1.12: Reaction patterns of the column agglutination assay .......... 55

Figure 1.13: Principles of the solid phase phenotyping assay ................ 60

Figure 1.14: Comparison of reaction patterns in phenotyping assays ..... 62

Figure 1.15: The principles of the IDCOREXT Assay .............................. 67

Figure 1.16 Principles of the Luminex xMAP analyser ............................ 72

Figure 2.1: Rh and K antigen phenotyping by column agglutination ....... 89

Figure 2.2: Manual Serological phenotyping for Fy (a and b), Jk (a and b)

and s antigens by column agglutination .................................................. 91

Figure 2.3: Reaction patterns generated by automated red cell antigen

phenotyping for Rh(CCwcEe), K, M and N antigens ................................ 97

Figure 2.5: NHSBT ten cell ID panel profile .......................................... 104

Figure 2.7: NHSBT titration cell set profile ........................................... 106

Figure 3.1: Correspondence analysis of the association between disease

types and antibody types in the haematology cohort ............................ 131

Figure 3.2: Incidence of antibody specificities within the haematology

cohort .................................................................................................... 134

Figure 3.3: Immunisation rates in males and females within the