Hemoglobin variant study

Turbidimetry versus HPLC

Lenters-Westra, E., Siebelder, C., Slingerland, R.J., Weykamp, C.W.
IFCC/NGSP Reference Laboratory (European Reference Laboratory, NL)

Introduction Check every chromatogram

Ideally, a patient who has been diagnosed with diabetes

would also be screened for hemoglobinopathies and Chromatogram normal* Chromatogram abnormal
thalassemia. This information could then be used to choose
the right method to reliably measure hemoglobin A1c Chromatogram same as chromatogram
of a common Hb variant?
(HbA1c) and also for genetic counseling to avoid Hb major in (HbAS, HbAC, HbAD, HbAE)
newborns. Unfortunately, however, this is still not common
clinical practice. Yes No

The most common remark from technologists using cation

exchange high performance liquid chromatography (HPLC) Interference from Is the variant heterozygous
common Hb variant (HbA present)?
to measure HbA1c, as opposed to immunoassays, is that they with your HbA1c
method?
want to know if there is a variant present or not. In most See for information:
www.ngsp.org/interf.asp1
cases, a variant will result in an abnormal chromatogram, Yes No
whereas variant identification is not possible with
immunoassays. However, this means every chromatogram No Yes or ? Do not report HbA1c
should be checked manually for abnormalities prior to
reporting the HbA1c result. While this process may enable HbA1c can Determine HbA1c with another method,
the identification of potential interference from some Hb be reported preferably affinity chromatography
variants, more than 1,175* different variants exist.
Advice: refer patient to reference laboratory
Furthermore, some Hb variants co-elute with HbA generating to confirm Hb variant on DNA level and for
a normal chromatogram. In these cases, the HbA1c result genetic counseling
would be artificially low or high, depending on which Hb *A normal chromatogram does not guarantee that there is no Hb variant present. If a
variant is present. Figure 1 shows the complexity of HbA1c variant is present the HbA1c value will be falsely low or high, depending on the variant

measurements using cation exchange HPLC. If this flow chart

Fig. 1: Complexity of HbA1c measurement with cation exchange HPLC in the
is followed, cation exchange HPLC is a method for presence of Hb variants
determining HbA1c in the presence of Hb variants.
When using an immunoassay, technologists are often
concerned that the test gives no indication of whether or not
a variant is present. However, an immunoassay does not need
to identify individual Hb variants because all forms of Hb are
included for the HbA1c calculation. As long as there are no
mutations in the epitope for the antibody used in the assay
the HbA1c measurement is not affected.

* See Addendum II on page 13 for updated statistics

For the vast majority of the more than 1,175* Hb variants Furthermore, it is important to note that the prevalence of
identified to date,2 the mutation is not in the first four these 17* Hb variants is very low, and often they only occur
amino acids of the Hb protein and therefore does not in certain individuals (e.g. Hb Okayama in Japanese men)
interfere with the Tina-quant® HbA1c Gen.2 and Gen.3 or in a few families, e.g. (Hb Graz in Austria) globally. Figure
assays from Roche; the mutation is in this region for only a 2 shows the complexity of HbA1c measurement with an
very small number of variants (17/1,175*).2 immunoassay and Table 1 lists some of the rare Hb variants
that can interfere with the Tina-quant® HbA1c Gen.2 and
Gen.3 assays.

Can this HbA1c result be reported?

Yes Yes

If the hemoglobin concentration is too low the correct For HbA1c levels within the measuring range of the
result is that no HbA1c result is reported assay, the correct result is that a HbA1c result is
reported
HbA1c measurements cannot be reported for patients
with anemia who have low hemoglobin concentrations No post-analytic steps required (e.g., manual
(men < 13 g/dL; women < 12 g/dL) chromatogram checking)
In these cases, HbA1c is not the right marker. Other The Tina-quant® HbA1c assay uses a very specific
glycated serum proteins (e.g. glycated albumin) might antibody which can detect the important forms of
be more helpful hemoglobin and is not subject to interference from
HbS, HbC, HbD, HbE or rare hemoglobin variants
Out of 1,175* known hemoglobin variants, only 17* can
cause interference in the Tina-quant® HbA1c assay.
However, these variants are extremely rare and are
often present in only one family, so are very unlikely to
affect the results

Fig. 2: Complexity of HbA1c measurement with an immunoassay in the presence of hemoglobin variants

Hb variant Substitution of amino acid Study design

Hb Raleigh β1 Val Ala The study objective was to investigate the potential
Hb Niigata β1 Val Leu interference of different Hb variants with different analytical
Hb Deer Lodge β2 His Arg methods for measuring HbA1c.
Hb Okayama β2 His Gln Samples were analyzed at two sites using the following
HHb Graz β2 His Leu methods:
Hb Agrigente β2 His Pro
Site 1: IFCC/NGSP Reference Laboratory (European
Hb Fukuota β2 His Tyr Reference Laboratory, Isala, Zwolle, NL)
Table 1: Hemoglobin variants that can interfere with the Tina-quant® HbA1c • Tina-quant® HbA1c Gen.2 on the COBAS INTEGRA® 800
Gen.2 and Gen.3 assays and potentially also with cation exchange HPLC2 analyzer; IFCC and NGSP certified immunoassay (Roche)
• Trinity Ultra,2 IFCC and NGSP certified affinity
chromatography HPLC (Trinity Biotech)
• Tosoh G8; IFCC certified cation-exchange HPLC (Tosoh
Bioscience).

*See Addendum II on page 13 for updated statistics 2

Site 2: IFCC/NGSP Reference Laboratory (European Results and discussion
Reference Laboratory, Queen Beatrix Hospital,
Winterswijk, NL) Results using Tina-quant® HbA1c assay from Roche
The study shows that the Hb variants HbAS, HbAC, HbAD,
• Tina-quant® HbA1c Gen.2 on the cobas c 501 module; HbAE, HbA2, HbAJ, HbAG and the rare variants do not
immunoassay (Roche) interfere with the Tina-quant® HbA1c Gen.2 and Gen.3
• Tina-quant® HbA1c Gen.3 on the cobas c 501 module; assays run on the cobas c 501 module or the Tina-quant®
immunoassay (Roche) HbA1c Gen.2 assay run on the COBAS INTEGRA® 800
analyzer (Table 2, Figures 3–5, 8–10, 13–15, 17–19 and
• Menarini HA-8180V; IFCC and NGSP certified cation 21–23). Only HbF > 8 % and the Hb Okayama variant were
exchange HPLC (A. Menarini Diagnostics).** found to interfere with the Tina-quant® HbA1c Gen.2 assay
The IFCC reference system is currently the only valid (Figures 26–28).
method for standardizing HbA1c measurements.3 All The effect of Hb Okayama can be explained by the fact
instruments and methods were calibrated using the IFCC that the antibodies used in these assays target the four
secondary reference material (batch Berlin). The IFCC amino acids and the glucose at the N-terminal end of the
calibrators were run as samples and IFCC offline calibrated hemoglobin β-globin chain. In the case of Hb Okayama,
values calculated retrospectively. The final data evaluation there is a substitution of the second amino acid (β-2 His
to determine any potential interference of the variant was Gln) and therefore this variant will not be recognized by
based on the IFCC offline calibrated results for all methods. the antibodies in the assay. HbF does not glycate as fast as
As the majority of Hb variants are considered not to HbAA, but is included in the measurement of total Hb,
interfere with Trinity Ultra2 affinity chromatography, this which explains why the HbA1c result is falsely low. It is very
method served as a comparator and all samples were difficult to get samples from patients with different HbF
analyzed in duplicate using this method. and HbA1c values and for this reason the HbF samples
tested comprised mixtures of blood from adults and
For the other methods investigated, if 10 or more samples neonates. The literature confirms that most of the common
of one variant type were available, samples were measured HbA1c methods are affected when HbF is > 15 %.4,5 In
individually, otherwise samples were tested in duplicate. All these studies samples from adults were used instead of
samples were investigated over 5 runs on 5 different days mixtures of blood from adults and neonates and this could
of about equal length. account for the fact that the current study observed
Sample material consisted of whole blood from single interference at HbF > 8 % rather than HbF > 15 %.
donors. HbF samples comprised a mixture of blood from Results using Menarini HA-8180V
adults and neonates. The Menarini HA-8180V did not produce a HbA1c result
For variants Hb Volga and HbSS, no results could be for samples containing HbAD, HbAE, HbAJ, HbAG and
reported. In the case of HbF, the mean value of the most of the rare variants. However, reporting no result in
Menarini HA-8180V and the Tosoh G8 was taken as the the case of an abnormal chromatogram, where the Hb
reference method. variant is not completely separated from the A0 peak, is the
correct outcome. The Menarini HA-8180V did produce a
Significant differences between each method and the result for samples containing HbAS and HbAC and
comparator, and recoveries between normal HbAA samples elevated A2. The HbS and HbC peaks are completely
and samples containing variant Hb were calculated using a separated from the A0 peak in the chromatogram and
Deming regression at a target value of 42 mmol/mol and similarly the HbS1c and HbC1c peak are completely
75 mmol/mol, respectively. The maximum expected separated from the HbA1c peak. The software can
deviation was ~ 5 % and a deviation > 10 % was therefore accurately subtract the area of the variant peak
considered to be significant. from the total area and calculate the HbA1c value correctly
The results were also plotted in a graph and if the results of and these variants do not cause interference (Figures 6
samples with Hb variants fell within the dispersion of the and 11). All investigated variants except Hb Indonesia and
normal HbAA samples the variant was considered not to Hb Hopkins-2 gave abnormal chromatograms with the
interfere with the method. Menarini HA-8180V and, in most cases, gave no HbA1c

** Not available in the U.S.

*See Addendum II on page 13 for updated statistics 3

result (Table 2). Using this HPLC technique, the HbF peak HbAJ. However, interference due to HbAE and HbAJ was
is separated from the total area and therefore a true HbA1c observed, with false low values obtained in samples
result is obtained. containing these two variants (Figures 20 and 25).
Figure 7 shows lower results obtained with the Tosoh G8
Results using Tosoh G8
when samples contained the variant HbAS and the
Table 2 and Figures 7, 12, 16, 20 and 25 show that the
deviation from the HbAA samples was borderline (9.5 %
Tosoh G8 produced a result for samples containing
deviation at 42 mmol/mol and 8.5 % at 75 mmol/mol).
HbAS, HbAC, HbAD, HbAE, elevated A2, HbAG and

See Addendum III on page 14 for NGSP conversion rate

cobas c 501 module with Tina-quant® HbA1c Gen.2 assay
Linear Deming R2 42 mmol/mol Significant 75 mmol/mol Significant
Regression regression difference difference
AA y = 1.07x - 3.67 y = 1.09x - 4.42 0.978 41 77

AS y = 1.04x - 2.91 y = 1.05x - 3.26 0.987 41 No 75 No

AC y = 1.02x - 1.29 y = 1.03x - 1.64 0.985 42 No 76 No
AD y = 0.96x + 0.37 y = 0.97x - 0.05 0.983 41 No 72 No
AE y = 0.99x - 0.04 y = 1.00x - 0.44 0.958 42 No 74 No
HbF > HbF 8 %

A2 No

Rare Hb Okayama
variants

cobas c 501 module with Tina-quant® HbA1c Gen.3 assay

Linear Deming R2 42 mmol/mol Significant 75 mmol/mol Significant
Regression regression difference difference
AA y = 1.08x - 3.01 y = 1.09x - 3.70 0.980 42 78

AS y = 1.10x - 4.38 y = 1.10x - 4.68 0.990 42 No 78 No

AC y = 1.01x - 0.77 y = 1.03x - 1.59 0.965 42 No 76 No
AD y = 0.97x + 1.99 y = 0.98x + 1.41 0.976 43 No 75 No
AE y = 1.08x - 2.89 y = 1.10x - 3.78 0.971 42 No 78 No
HbF > HbF 8 %

A2 No

Rare Hb Okayama
variants

COBAS INTEGRA® 800 analyzer with Tina-quant® HbA1c Gen.2

Linear Deming R2 42 mmol/mol Significant 75 mmol/mol Significant
Regression regression difference difference
AA y = 1.08x - 4.28 y = 1.08x - 4.56 0.992 41 77

AS y = 1.07x - 4.40 y = 1.08x - 4.72 0.989 41 No 76 No

AC y = 1.00x - 0.72 y = 1.01x - 1.07 0.980 41 No 75 No
AD y = 0.98x – 0.52 y = 0.99x - 0.82 0.988 41 No 73 No
AE y = 0.99x - 0.47 y = 1.00x - 0.89 0.984 41 No 74 No
HbF > HbF 8 %

A2 No

Rare Hb Okayama
variants

4
However, it should be noted that the Deming regression Hopkins-2 gave abnormal chromatograms with the Tosoh
lines calculated for the common variants were based on 20 G8 and, in most cases, gave no HbA1c result (Table 2).
samples compared with 50 samples for the normal HbAA Using this HPLC technique, the HbF peak is separated from
and that the distribution of HbA1c over the clinically the total area and therefore a true HbA1c result is
important range for HbA1c samples containing the obtained.
common variants (HbAS, HbAC, HbAD and HbAE) was not
Conclusion
always optimal. It is debatable, therefore, whether the
The perfect method for measuring HbA1c still does not
method for calculating the deviation from normal samples
exist and which method is chosen will depend on the
is correct, because Figure 7 clearly shows a difference
laboratory’s priorities. Some will choose to use an
from normal HbAA samples. This phenomenon of lower
immunoassay to obtain a fast and reliable HbA1c result
results with HbAS variants using the Tosoh G7/G8 has
regardless of the presence or absence of a Hb variant.
been observed in previous studies (not published) and, as
Other laboratories, however, will want to know if a
yet, there is no explanation given that the chromatograms
variant is present and opt for cation exchange HPLC.
show that the S0 peak is completely separated from the A0
Hence, the advantages/ disadvantages of the different
peak, and hence a correct result would be expected. All
HbA1c methods will be viewed differently by different
investigated variants except Hb Indonesia and Hb
laboratories.

See Addendum III on page 14 for NGSP conversion rate

Menarini HA-8180V
Linear Deming R2 42 mmol/mol Significant 75 mmol/mol Significant
Regression regression difference difference
AA y = 1.03x -1.54 y = 1.03x - 1.90 0.989 41 76

AS y = 1.06x - 5.91 y = 1.07x - 6.31 0.989 39 No 74 No

AC y = 1.02x - 0.46 y = 1.03x - 1.08 0.973 42 No 76 No
AD No result

AE No result

HbF

A2 No

Rare All investigated rare variants except for

variants Hb Indonesia and Hb Hopkins-2

Tosoh G8
Linear Deming R2 42 mmol/mol Significant 75 mmol/mol Significant
Regression regression difference difference
AA y = 1.02x - 0.37 y = 1.02x - 0.67 0.990 42 76

AS y = 0.95x - 1.60 y = 0.96x - 1.97 0.984 38 No 70 No

AC y = 1.02x - 1.95 y = 1.04x - 2.91 0.960 41 No 75 No
AD y = 0.83x + 8.41 y = 0.85x + 7.19 0.935 43 No 71 No
AE y = 0.69x + 0.80 y = 0.70x + 0.29 0.957 30 Yes 53 Yes
HbF

A2 No

Rare All investigated rare variants except for

variants Hb Indonesia and Hb Hopkins-2

Table 2: Medical decision points calculated with Deming regression lines. A deviation > 10 % compared with HbAA samples was considered to be significant

5
 Lower results obtained with the Tosoh G8 when samples contain the variant HbAS
Hemoglobin
140
S (% variant: 31–42 % S) 140
HbAA samples HbAA samples
See Addendum III on page 13 fory = NGSP conversion
1.07x - 3.67 rate y = 1.08x - 3.01

Tina-quant® HbA1c Gen.3 assay

Tina-quant® HbA1c Gen.2 assay

140
120 R2 HbAA
= 0.978samples 140
120 R2 HbAA
= 0.980samples
n =y 50
= 1.07x - 3.67 n =y 50
= 1.08x - 3.01

assay
assay

cobas® c 501 Gen.3

501 module
501 module

HbAS
R2 =samples
0.978 HbAS
R2 =samples
0.980
120
100140 120
100140
y =nHbAA
1.04x samples
= 50 - 2.91 y =nHbAA
1.10x samples
= 50 - 4.38
2 y = 1.07x - 3.67 2 y = 1.08x - 3.01

c 501 Gen.3

assay
Gen.3cGen.3
c 501 module
Gen.2cGen.2
assay

R HbAS
= 20.987
c 501 module

samples R HbAS
= 20.990
samples
100
80120 n =yR20 = 0.978
= 1.04x - 2.91
100
80120 n =yR20 = 0.980
= 1.10x - 4.38
– x =Rn2y= 50 n = 50
– x =R2y = 0.990

c 501 Gen.3
= 0.987

HbA1c
c 501 module
HbA1c
cobas c 501 module

HbAS samples HbAS samples

HbA1ccobas
Tina-quant HbA1ccobas

80
60100 n = 20 80
60100 n = 20
y = 1.04x - 2.91 y = 1.10x - 4.38
– x =2 y – x =2 y
R = 0.987

®
R = 0.990

cobas cobas
cobas

cobas®cobas

®
®

4060
80 4060
80

Tina-quant
n = 20
Tina-quant

n = 20
–x=y –x=y

®
®

2040
60 2040
60

Tina-quant
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20
40 Ultra2 20
40 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20
Ultra2 20
Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Ultra2 Ultra2
4.0 P/B Regression
Fig. 3: Results from the Tina-quant® HbA1c Gen.2 assay on the Fig. 4: Results from the Tina-quant® HbA1c Gen.3 assay on
cobas
Y = 1.0 *
3.5 c 501 module in the presence of HbAS X -P/B 0.07
Regression
the cobas c 501 module in the presence of HbAS
4.0 30–100 mmol/mol
HbA1c: HbA1c: 30–100 mmol/mol
140 md(95)
Y =HbAA = 0.085
1.0 * samples 140 HbAA samples
g/L

3.0
3.5
module
assay

4.0 N =
X y
126
- =
0.07r1.08x - 4.28
= 0.9974
P/B 2Regression y = 1.03x - 1.54
analyzer
Immunoturbidimetry

2.5120 140 t =md(95)R HbAA

= 0.992 140
120 R2 HbAA
= 0.989samples
Y0.9641 =* 0.085
samples
501 g/L

3.0 = 1.0
NX =-n 0.07
=y 50 n =y 50
® module

3.5
AAGP2

assay

126 = r1.08x - 4.28

= 0.9974 = 1.03x - 1.54
cGen.2

2.0
800 analyzer
Immunoturbidimetry

2.5 120
140 HbASR2 =samples
0.992 120
140 HbAS
R2 =samples
0.989
100 tmd(95)
= 0.9641 = 0.085 100
y =nHbAA samples y =nHbAA samples
800
cobasg/L

3.0 1.07x
= 50 - 4.40 1.06x
= 50 - 5.91
HA-8180V
AAGP2

module

1.5
assay

=y 2= r1.08x - 4.28 2 y = 1.03x - 1.54

cGen.2

NR =2126 = 0.9974
HbA1c

0.989 R HbAS
= 20.989
501

2.0 HbAS samples samples

800 analyzer
Immunoturbidimetry

2.5
80100
120 R20= 0.992 100
80120 n =yR20 = 0.989
INTEGRA

t =n0.9641
=y = 1.07x - 4.40 = 1.06x - 5.91
1.0
HA-8180V

1.5 – x =Rn2y= 50 – x =Ry = 50

n =
AAGP2

cGen.2
cobasHbA1c

= 0.989 0.989
cobas

2
501

2.0
®

HbAS samples HbAS samples

®

80 80
INTEGRA

0.5 60100 n = 20 60100 n = 20

Tina-quant

1.0 y = 1.07x - 4.40 y = 1.06x - 5.91

HA-8180V

1.5 – x =2 y – x =2 y
® COBAS
HbA1c

0 R = 0.989 R = 0.989
®
®

04060
0.5 0.5
80 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4060
80
INTEGRA
Tina-quant

1.0 n = 20 n = 20
COBASCOBAS

AAG –x=y –x=y

0
040600.5 1.0 1.5 2.0 g/L
Immunonephelometry
0.5
20 2.5 BN 3.0 II3.5 4.0 2040
60
Tina-quant

20 40 60AAG80 100 120 140 20 40 60 80 100 120 140

Excellent slope
0 and correlation Ultra
is demonstrated
2 (r = 0.9974, Y = 1.0). Ultra2
020
40 0.5 1.0 1.5 2.0 2.5
Immunonephelometry g/L3.0BN 3.5
II 4.0 20
40
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Excellent slope and correlation AAG is demonstrated (r = 0.9974, Y = 1.0).
Ultra
Immunonephelometry g/L BN II
20
2
Ultra2
20
20 from
Fig. 5: Results 40 the 60 80 ® HbA1c
Tina-quant 100 120 Gen.2 140
assay on the 20 from40the Menarini
Fig. 6: Results 60 80HA-8180V
100 120 140
in the presence of HbAS
Excellent
COBAS slope and corr
INTEGRA ® elation is demonstrated
Ultra
800 analyzer 2 presence
in the (r of
= 0.9974,
HbAS Y = 1.0). HbA1c: 30–100 mmol/mol Ultra2
HbA1c: 30–100 mmol/mol

140 HbAA samples

y = 1.02x - 0.37
140
120 R2 HbAA
= 0.990samples
n =y 50
= 1.02x - 0.37
120
140 HbAS
R2 =samples
0.990
100 HbAA samples
y =n0.95x
= 50 - 1.60
Tosoh G8

2 y = 1.02x - 0.37
R HbAS
= 20.984
samples
100
80120 n =yR20 = 0.990
= 0.95x - 1.60
G8

– x =Rn2y= 50
= 0.984
80 HbAS samples
Tosoh Tosoh

60100 n = 20
y = 0.95x - 1.60
G8

– x =2 y
R = 0.984
4060
80 n = 20
–x=y
2040
60
20 40 60 80 100 120 140
20
40 Ultra 2

20 from
Fig. 7: Results 40 the60
Tosoh80
G8 in100 120 140
the presence of HbAS
HbA1c: 30–100 mmol/mol
20
Ultra2
20 40 60 80 100 120 140 6
Ultra2
 All methods free from interference from HbAC
Hemoglobin C (% variant: 36–42 % C)
See 140
Addendum III on page 13 for NGSP
HbAA samples conversion140
rate HbAA samples

assay assay
y = 1.07x - 3.67 y = 1.08x - 3.01
assay assay

140
120 R2 = 0.978
HbAA samples 140
120 R2 = 0.980
HbAA samples
yn==1.07x
50 - 3.67 yn==1.08x
50 - 3.01

Gen.3Gen.3
Gen.2Gen.2

c 501 module
c 501 module

120
100 RHbAC
2 samples
= 0.978 120
100 RHbAC
2 samples
= 0.980
ny==501.02x - 1.29 ny==501.01x - 0.77

c 501 module
c 501 module

HbA1c
HbA1c

R2 = 0.985
HbAC samples R2 = 0.965
HbAC samples
80
100 80
100
140 HbAA samplesyn==1.02x
20 - 1.29 140 HbAA samples yn==1.01x
20 - 0.77
y = 1.07x –-R3.67
x2 =
= y0.985 –Rx2 =
= y0.965

HbA1c
HbA1c

Tina-quant® HbA1c Gen.3 assay

y = 1.08x - 3.01
assay
cobascobas

cobascobas
®

®
12080
60 80
60
Tina-quant

Tina-quant
R2 = 0.978 n = 20 120 R2 = 0.980 n = 20
n = 50 – x = y n = 50 –x=y
HbA1c Gen.2

cobas c 501 module

module

HbAC samples HbAC samples

®

®
10060
40 100 60
40
Tina-quant

Tina-quant
y = 1.02x - 1.29 y = 1.01x - 0.77
R2 = 0.985 R2 = 0.965
cobas c 501

80 n = 20 80 n = 20
4020 40
20
–x=y –x=y
20 40 60 80 100 120 140 20 40 60 80 100 120 140
®

60 60
Tina-quant

20 Ultra2 20 Ultra2
40 20 40 60 80 100 120 140 40 20 40 60 80 100 120 140
Ultra 2
Ultra2
20 20
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Fig. 8: Results from theUltra
Tina-quant
2 ®
HbA1c Gen.2 assay on the Fig. 9: Results from
Ultra 2 the Tina-quant® HbA1c Gen.3 assay on the

cobas c 501 module in the presence of HbAC cobas c 501 module in the presence of HbAC
HbA1c: 26–104 mmol/mol HbA1c: 26–104 mmol/mol

140 HbAA samples 140 HbAA samples

y = 1.08x - 4.28 y = 1.03x - 1.54
assay assay
COBAS INTEGRA 800 analyzer

140
120 R2 = 0.992
HbAA samples 140
120 R2 = 0.989
HbAA samples
140 y ==1.08x
n 50 - 4.28 140 y ==1.03x
n 50 - 1.54
Gen.2

HbAA samples HbAA samples

analyzer

120 HbAC
y = 1.08x -R4.28
2 samples
= 0.992 120 y = 1.03x - 1.54 RHbAC
2 samples
= 0.989
100 100
assay

R2 = 0.992 ny==50 1.00x - 0.72 120 ny==501.02x - 0.46

analyzer

R2 = 0.989
HA-8180V

120
Gen.2

R2 = 0.980 R2 = 0.973
HbA1c

n = 50 HbAC samples n = 50 HbAC samples

®
Gen.2

80
100 HbAC samples n==1.00x
20 - 0.72 100 80
100 HbAC samples
yn==1.02x
20 - 0.46
800

100 y
HA-8180V
INTEGRA®®800

y = 1.00x –- 0.72 y = 1.02x - 0.46

Rx2 =
= y0.980 –Rx2 =
= y0.973
HA-8180V
Tina-quant® HbA1c

R = 0.980 R2 = 0.973
HbA1c

2
®

8080 80
INTEGRA

60 n = 20 80 60 n = 20
Tina-quant

n = 20 n = 20
–x=y –x=y –x=y –x=y
®

6060 60 60
40 40
Tina-quant
COBAS
COBAS

40 40
4020 40
20
20
20 40 60 80 100 120 140 20
20 40 60 80 100 120 140
20
20 40 60 80 Ultra
100 2120 140 20 40 6020 80 100 120 Ultra2
140
40 Ultra Ultra
2 2
20 60 80 100 120 140 20 40 60 80 100 120 140
Ultra HbA1c assay Gen.2 on the
Fig. 10: Results from the Tina-quant 2 ®
Ultra
Fig. 11: Results from the Menarini 2
HA-8180V in the presence of HbAC
COBAS INTEGRA® 800 analyzer in the presence of HbAC HbA1c: 26–104 mmol/mol HbAA
HbAA
HbA1c: 26–104 mmol/mol
140
140
HbAC HbAC
HbAA
120

140
120
X=Y X=Y
HbAC
cobas® c 501 Gen.2
cobas® c 501 Gen.2

140
100

140
100
HbAA samples
120
80
HbAA samples
y = 1.02x - 0.37 X=Y
120 100 80
R2 = 0.990 y = 1.02x - 0.37
cobas® c 501 Gen.2

140
120 60
n = 50 R2 = 0.990
HbAA samples
y ==1.02x
HbAC samplesn 50 - 0.37
60
10040 80
y = 1.02x - 1.95
HbAC samples
Tosoh G8

R2 = 0.960 R = 0.990
2
120
10040
80 2060
20

ny == 1.02x
50 - 1.95
G8 G8

n = 20
40 60 80 100 120 140

Ultra2
20 R2 = 0.960
– x = y 140 HbAC samples
80
100
TosohTosoh

yn==1.02x
20 - 1.95
40 20 40 60 80 100 120
60 Ultra2
20
–Rx2 =
= y0.960
4080
6020 40 60 80 100 120 140
n = 20
Ultra2
–x=y
2060
40
20 40 60 80 100 120 140
Ultra2
40
20
20 40 60 80 100 120 140
Ultra
Fig. 12: Results from the Tosoh
20 G8 2in the presence of HbAC
HbA1c: 26–104
20 mmol/mol
40 60 80 100 120 140
Ultra2 7
 Presence of HbAD causes no results on Menarini HA-8180V
Hemoglobin
140 D (% variant: 37–42 % D) 140
HbAA samples HbAA samples
See140Addendum III on page 13y 2=for
1.07xNGSP conversion rate

Tina-quant® HbA1c Gen.3 assay

- 3.67 y = 1.08x - 3.01
Tina-quant® HbA1c Gen.2 assay

120 R HbAA
= 0.978samples 140 120 R2 HbAA
= 0.980samples
n =y 50 n =y 50

assay
= 1.07x - 3.67 = 1.08x - 3.01
assay

module
module

140
120 HbADRHbAA
2
= samples
0.978
samples 140
120 HbADRHbAA
2
= samples
0.980
samples
100 100
y =ny0.96x
= 50 + 0.37 y =ny0.97x
= 50 + 1.99

assay
1.07x - 3.67 1.08x - 3.01
assay

c Gen.3
c Gen.2

501 module
501 module

120 R HbAD
2
=20.983
R samples
= 0.978 120 R HbAD
2
=20.976
R samples
= 0.980

501
501

100
80 100
80
n =yn20==0.96x
50 + 0.37 n =yn20==0.97x
50 + 1.99

Gen.3
Gen.2

c 501 cmodule
c 501 cmodule

– x =RHbAD
2y – x =RHbAD
2y

HbA1c
HbA1c

= 0.983
samples = 0.976
samples
cobas

100 100

cobas
6080 ny = 20
0.96x + 0.37 6080 ny = 20
0.97x + 1.99
– xR= 2 y – xR= 2 y

HbA1c
HbA1c

= 0.983 = 0.976
cobascobas

cobascobas
80 80
®

®
4060 n = 20 4060 n = 20
Tina-quant

Tina-quant
–x=y –x=y
®

®
60 60
Tina-quant

Tina-quant
2040 2040
20 40 60 80 100 120 140 20 40 60 80 100 120 140
40
20 Ultra2 40
20 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20 Ultra2 20 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Ultra2 Ultra2

Fig. 13: Results from the Tina-quant® HbA1c Gen.2 assay on the Fig. 14: Results from the Tina-quant® HbA1c Gen.3 assay on the
cobas c 501 module in the presence of HbAD cobas c 501 module in the presence of HbAD
HbA1c: 30–96 mmol/mol HbA1c: 30–96 mmol/mol
140 HbAA samples 140
y = 1.08x - 4.28
Tina-quant® HbA1c Gen.2 assay
analyzer

140
120 R2 HbAA
= 0.992samples 140
120
n =y 50
= 1.08x - 4.28
assay
800 analyzer

140
120 HbAD
RHbAA
2
= samples
0.992
samples 140
120
100 100
800

y =ny0.98x
= 50 - 0.52
1.08x - 4.28
HA-8180V
assay

Menarini HA-8180V
Gen.2Gen.2
800 ®analyzer

120 R2 HbAD
=20.988
R samples
= 0.992 120
®

100
80 100
80 No result for HbA1c for
INTEGRA

n =yn20
==0.98x
50 - 0.52
HA-8180V

– x =RHbAD
2y Menarini HA-8180V
= 0.988 samples containing HbAD
HbA1c

samples
100 100
6080 6080 No result for HbA1c for
INTEGRA

ny = 20
0.98x - 0.52
HA-8180V

– xR=2 y Menarini HA-8180V

= 0.988 samples containing HbAD
HbA1c
COBAS
®

80 80 No result for HbA1c for

INTEGRA
®

4060 n = 20 4060
Tina-quant

–x=y samples containing HbAD

COBAS
®

60 60
2040 2040
Tina-quant
COBAS

20 40 60 80 100 120 140 20 40 60 80 100 120 140

40
20 Ultra2 40
20 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20 Ultra2 20 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Fig. 15: Results from the Tina-quant
Ultra2 HbA1c Gen.2 assay on the
®

COBAS INTEGRA® 800 analyzer in the presence of HbAD

Ultra2
HbA1c: 30–70 mmol/mol A2, 36-54 mmol/mol AJ, 34-43 mmol/mol AG

140 HbAA samples

y = 1.02x - 0.37
140
120 R2 HbAA
= 0.990samples
n =y 50
= 1.02x - 0.37
140
120 HbAD
RHbAA
2
= samples
0.990
samples
100
y =ny0.83x
= 50 + 8.41
1.02x - 0.37
Tosoh G8

120 R2 HbAD
=20.935
R samples
= 0.990
100
80 n =yn20
==0.83x
50 + 8.41
G8 G8

– x =RHbAD
2y
= 0.935
samples
100
6080
TosohTosoh

ny = 20
0.83x + 8.41
– xR=2 y
= 0.935
80 n = 20
4060
–x=y
60
2040
20 40 60 80 100 120 140
40
20 Ultra2
20 40 60 80 100 120 140
Fig. 16: Results from the Tosoh G8 in the presence of HbAD
HbA1c:20
30–96 mmol/mol Ultra2
20 40 60 80 100 120 140
Ultra2 8
 No results on Menarini HA-8180V and false low values on Tosoh G8 in samples containing HbAE
Hemoglobin
140 E (% variant: 27–33 %HbAA
E) samples 140 HbAA samples

Tina-quant® HbA1c Gen.3 assay

13yR2=for
1.07x - 3.67 y = 1.08x - 3.01
Tina-quant® HbA1c Gen.2 assay

See Addendum III on page

120140 =HbAA
NGSP conversion
0.978 samples 120140
rate R2 =HbAA
0.980 samples
n =y50= 1.07x - 3.67 n =y50= 1.08x - 3.01

assay
assay

501 module
501 module

120
100140 HbAE R2 =samples
0.978 120
100140 HbAE R2 =samples
0.980
y =HbAA
0.99x
n = 50
samples
- 0.04 y =HbAA
1.08x
n = 50
samples
- 2.89

assay
Gen.3cGen.3
R2 y=HbAE
=0.958
1.07x - 3.67 R2 y=HbAE
=0.971
1.08x - 3.01

c 501 module
assay
Gen.2cGen.2
c 501 module

80120
100 samples 80120
100 samples
n =Ry20== 0.99x
0.978 n =Ry20== 1.08x
0.980
2 2
- 0.04 - 2.89
n = 50
– x = Ry2 = 0.958 n = 50
– x = Ry2 = 0.971

c 501 module
HbA1c
HbA1c
c 501 module

® cobas
® cobas

80
60100 HbAE samples 80
60100 HbAE samples
n = 20 n = 20
y = 0.99x - 0.04 y = 1.08x - 2.89
– x2 = y – x2 = y

HbA1c
HbA1c

R = 0.958 R = 0.971

cobas cobas
cobas cobas

60
40 80 60
40 80

Tina-quant
Tina-quant

n = 20 n = 20
–x=y –x=y

®
®

40
20 60 40
20 60

Tina-quant
Tina-quant

20 40 60 80 100 120 140 20 40 60 80 100 120 140

20
40
Ultra2 20
40
Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20
Ultra2 20
Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Ultra2 Ultra2
Fig. 17: Results from the Tina-quant® HbA1c Gen.2 assay on the Fig. 18: Results from the Tina-quant® HbA1c Gen.3 assay on the
cobas c 501 module in the presence of HbAE cobas c 501 module in the presence of HbAE
HbA1c:
140 38–81 mmol/mol HbA1c: 38–81 mmol/mol
140
HbAA samples
y = 1.08x - 4.28
Tina-quant® HbA1c Gen.2 assay
analyzer

120140 R2 =HbAA
0.992 samples 120140
n =y50= 1.08x - 4.28
assay

HbAE R2 =samples
800 analyzer

120
100140 0.992 120
100140
y =HbAA samples
800

0.99x
= 50- 0.47
HA-8180V

n
2 y = 1.08x - 4.28 Menarini HA-8180V
assay
Gen.2

R =HbAE0.984samples
®

800 analyzer

80120
100 n =Ry20== 0.99x
0.992 100 No result for HbA1c for
80120
INTEGRA

2
- 0.47
HA-8180V

– x = Ry=2 =500.984
n Menarini
samples HA-8180V
containing HbAE
® Gen.2
HbA1c
®

80 HbAE samples
60100 80 No result for HbA1c for
INTEGRA

n = 20 60100
y = 0.99x - 0.47
HA-8180V

– x2 = y Menarini HA-8180V
COBAS

R = 0.984 samples containing HbAE

HbA1c
®

60
40 80 60 No result for HbA1c for
40 80
INTEGRA
Tina-quant

n = 20
COBASCOBAS

–x=y samples containing HbAE

®

40
20 60 40
20 60
Tina-quant

20 40 60 80 100 120 140 20 40 60 80 100 120 140

20
40
Ultra 2
Ultra2
20
40
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20
Ultra2 Ultra2
20
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Fig. 19: Results from the Tina-quant ®
HbA1c Gen.2 assay on the
COBAS INTEGRA® 800 analyzer Ultra 2
in the presence of HbAE Ultra2
HbA1c: 38–81 mmol/mol

140 HbAA samples

y = 1.02x - 0.37
120140 R2 =HbAA
0.990 samples
n =y50= 1.02x - 0.37
120 HbAE R2 =samples
0.990
100140
y =HbAA
0.69x
n
samples
= 50- 0.80
Tosoh G8

2 y = 1.02x - 0.37
R =HbAE0.957samples
80120
100 n =Ry20== 0.69x
2
0.990
- 0.80
– x =nRy=2 =500.957
G8

80 HbAE samples
Tosoh Tosoh

60100 n = 20
y = 0.69x - 0.80
G8

– x2 = y
R = 0.957
60
40 80 n = 20
–x=y
40
20 60
20 40 60 80 100 120 140
20
40
Ultra2
20 40 60 80 100 120 140
Fig. 20: Results from the Tosoh G82in the presence of HbAE
HbA1c: 20
38–81 mmol/mol Ultra
20 40 60 80 100 120 140
9
Ultra2
 Interference from HbAJ and HbAG is observed with Menarini HA-8180V and samples containing
HbAJ
140 gave false low results onHbAA
Tosoh G8
samples 140 HbAA samples
Hemoglobin A2, AJ and AG (% variant:
n = 504–7 % A2, 49–51 % AJ, approximately 18 % AG)

Tina-quant® HbA1c Gen.3 assay

n = 50
Tina-quant® HbA1c Gen.2 assay

120140 HbA2 samples 120140 HbA2 samples

See Addendum III on page 13n =for10 NGSP
HbAA samples conversion rate n =HbAA
10 samples

® cobas c 501 module

® cobas c 501 module

assay assay
n =samples
50 n =samples
50
assay assay

100140 HbAJ 100140 HbAJ

120 n =HbA2
HbAA
5 samples
samples 120 n =HbA2
HbAA
5 samples
samples
n
n =
= 10
50
HbAG samples n
n =
= 10
50
HbAG samples

Gen.3Gen.3
Gen.2Gen.2

c 501 module
c 501 module

80120
100 n =HbAJ
HbA2
2 samples
samples 80120
100 n =HbAJ
HbA2
2 samples
samples
– x =nny =
= 510 – x =nny =
= 10
5

c 501 module
c 501 module
HbA1c

HbA1c
60100 HbAG samples
HbAJ samples 60100 HbAGsamples
HbAJ samples
80 n 80 n= = 52
n= = 25 n
– HbAG
x = y samples – HbAG
x = y samples
HbA1c

HbA1c
cobascobas

cobascobas
40 80
60 40 80
60
Tina-quant

Tina-quant
n=2 n=2
–x=y –x=y
20 60 20 60
®

®
40 40
Tina-quant

Tina-quant
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Ultra2 Ultra2
40
20 40
20
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20 Ultra2 20 Ultra2
20 40 60 80 100 120 140 20 40 60 80 100 120 140
Ultra2 Ultra2
Fig. 21: Results from the Tina-quant® HbA1c Gen.2 assay on the Fig. 22: Results from the Tina-quant® HbA1c Gen.3 assay on the
cobas c 501 module in the presence of HbA2, HbAJ and HbAG cobas c 501 module in the presence of HbA2, HbAJ and HbAG
HbA1c: 30-70 mmol/mol A2, 36-54 mmol/mol AJ, 34-43 mmol/mol AG HbA1c: 30-70 mmol/mol A2, 36-54 mmol/mol AJ, 34-43 mmol/mol AG
140 HbAA samples 140 HbAA samples
n = 50 n = 50
Tina-quant® HbA1c Gen.2 assay
analyzer

120140 HbA2 samples 120140 HbA2 samples

n =HbAA
10 samples n =HbAA
10 samples
n =samples
50 – x =ny = 50
assay

HbAJ
800 analyzer

100140 100120
140
120 n =HbA2 samples HbA2
HbAA samples
assay800

HbAA
5 samples samples
HA-8180V

HbAGn
n= 10
= samples
50 n
n== 10
50
Gen.2
®

800 analyzer

80120 HbAJ samples 80100 – HbA2

x = y samples
INTEGRA

100 n =HbA2
2 samples 120
– x =nny =
= 510
HA-8180V

n = 10
Gen.2

HbAG samples
HbA1c

HbAJ samples –x=y

®

60100
80 60100
80
INTEGRA

n=
n = 52
HA-8180V
COBAS

– HbAG
x = y samples
HbA1c
®

40 80 40 60
®

80
INTEGRA

60 n=2
Tina-quant

–x=y
COBASCOBAS

20 60 20 40
®

40 60
Tina-quant

20 40 60 80 100 120 140 20 40 60 80 100 120 140

Ultra2 Ultra2
40
20 40
20
20 40 60 80 100 120 140 20 40 60 80 100 120 140
20 Ultra 2
20 Ultra 2

20 from
Fig. 23: Results 40 the60Tina-quant
80 ®100 120
HbA1c 140
Gen.2 assay on the 20
Fig. 24: Results 40 60 Menarini
from the 80 HA-8180V
100 120in the
140presence of HbA2
Ultra
COBAS INTEGRA® 800 analyzer 2 presence of HbA2, HbAJ and HbAG
in the HbA1c: 30-70 mmol/mol Ultra2
HbA1c: 30-70 mmol/mol A2, 36-54 mmol/mol AJ, 34-43 mmol/mol AG

140 HbAA samples

n = 50
120140 HbA2 samples
n =HbAA
10 samples
n =samples
HbAJ 50
100120
140 n =HbA2
HbAA
5 samples
samples
Tosoh G8

n
n=
HbAG 10
= samples
50
80120
100 n =HbAJ
HbA2
2 samples
samples
–x=yn
n =
= 5
10
G8 G8

HbAG samples
HbAJ samples
60100
80
TosohTosoh

n
n= = 25
– HbAG
x = y samples
40 80
60 n=2
–x=y
20 40
60
20 40 60 80 100 120 140
Ultra2
40
20
20 from
Fig. 25: Results 40 the60Tosoh80G8 in100 120 140
the presence of HbA2, HbAJ and HbAG
HbA1c: 30-70 mmol/mol A2,Ultra
36-54 2mmol/mol AJ, 34-43 mmol/mol AG
20
20 40 60 80 100 120 140 10
Ultra2
 All investigated rare variants except for Hb Indonesia and Hb Hopkins-2 gave an abnormal
chromatogram and no HbA1c result with Menarini HA-8180V and Tosoh G8
Rare
110 variants HbAA samples 110 HbAA samples
110 HbHbAA
Setif samples 110 HbHbAA
Setif samples

Tina-quant® HbA1c Gen.3 assay

Tina-quant® HbA1c Gen.2 assay

See100Addendum III on page 13Hbfor

100
Hb NGSP
(selected) conversion
Setif
Setif
100 rate HbHb Setif
Setif (selected)

assay
assay

100
90 HbHb Setif (selected)
Okayama 90 HbHb Setif (selected)
Okayama
110
90 HbAA samples 110
90 HbAA samples
module

module
80 HbHb Okayama
Philadelphia 80 HbHb Okayama
Philadelphia
Hb Setif Hb Setif
Gen.2

Gen.3
c 501 module

c 501 module
c 501assay
c 501assay

100
80 Hb
(HbH) Philadelphia 100
80 Hb
(HbH) Philadelphia
70 Hb Setif (selected) 70 Hb Setif (selected)
90 Hb(HbH)
Zurich 90 Hb(HbH)
Zurich
70 Hb Okayama
Hb Zurich 70 Hb
Hb Okayama
Zurich
60 HbEE 60 HbEE
HbA1c

HbA1c
Gen.2

Gen.3
501 module

501 module
80
60 Hb
HbEEPhiladelphia 80
60 Hb
HbEEPhiladelphia
HbSE HbSE
50 (HbH) 50 (HbH)
HbSE HbSE
cobas

cobas
70
50 HbO Indonesia 70
50 HbO Indonesia
Hb Zurich Hb Zurich
HbA1c
cobas ccobas

HbA1c
cobas
40 HbO
HbSD Indonesia 40 HbO
HbSD Indonesia
®

®
Tina-quant

Tina-quant
60
40 HbEE 60
40 HbEE
30 HbHbSD
Coushatta 30 HbHbSD
Coushatta
HbSE HbSE

cobas c
50
30 (C4Hb Coushatta
+ C9) 50
30 (C4Hb Coushatta
+ C9)
HbO Indonesia HbO+Indonesia
®

20 20

®
(C4 + C9)
Hb Queens (Q1) Hb(C4
Queens C9)(Q1)
Tina-quant

Tina-quant
40
20 HbSD 40
20 HbSD
10 HbHb Queens (Q1)
Hopkins-2 10 HbHb Queens (Q1)
Hopkins-2
1030 Hb
Hb Coushatta Hb
Hb Coushatta
1020 30 40 50 60 70 80 90 100 110 HbCC Hopkins-2 1030
1020 30 40 50 60 70 80 90 100 110 HbCC Hopkins-2
10 20 30 40 50 60
Ultra 2 70 80 90 100 110 (C4
HbCC+ C9) 10 20 30 40 50 60
Ultra 2 70 80 90 100 110 (C4
HbCC+ C9)
20 HbSC 20 HbSC
Ultra2 Hb Queens (Q1) Ultra 2 Hb Queens (Q1)
– x =HbSC
y – x =HbSC
y
10 – Hb
x = Hopkins-2
y 10 – Hb
x = Hopkins-2
y
10 20 30 40 50 60 70 80 90 100 110 HbCC 10 20 30 40 50 60 70 80 90 100 110 HbCC
Ultra2 HbSC Ultra2 HbSC
–x=y –x=y
Fig. 26: Results from the Tina-quant® HbA1c Gen.2 assay on the Fig. 27: Results from the Tina-quant® HbA1c Gen.3
cobas c 501 module in the presence of rare hemoglobin variants assay on the cobas c 501 module in the presence of rare hemoglobin variants

110 HbAA samples 140

110 HbHbAA
Setif samples 140
Tina-quant® HbA1c Gen.2 assay

100
800 analyzer

HbHb Setif
assay

100 Setif (selected) 120

800 analyzer

90 Hb Hb Setif (selected)
Okayama
110
90 120Menarini HA-8180V
140
HbAA samples
80 HbHb Okayama
Philadelphia Menarini HA-8180V
100 All investigated rare variants
Gen.2

Hb Setif
Gen.2 ®assay

100
80 Hb Philadelphia
(HbH)
800 analyzer

All investigated rare variants

HA-8180V

70 Hb Setif (selected) 100except

120
90 Hb(HbH)
Zurich for Hb Indonesia
HA-8180V

70 Hb
60 Hb Okayama Menarini
except HA-8180V
for Hb Indonesia
INTEGRA

Zurich 80 and
HbA1c

HbEE Hb Hopkins-2 gave an

®

80
60 Hb Philadelphia
INTEGRA

HbEE
HbSE 80 All
100 andinvestigated
Hb raregave
Hopkins-2 variants
an
50 (HbH) abnormal chromatogram and
HA-8180V

70
50 HbOHbSE
Indonesia
Hb Zurich 60 except for chromatogram
abnormal Hb Indonesia and
HbA1c

40
60no HbA1c result
®

HbO Indonesia
®

HbSD
Tina-quant

60
INTEGRA

80
COBAS

40 HbEE and Hb Hopkins-2

no HbA1c result gave an
30 HbHbSD
Coushatta
COBASCOBAS

50 HbSE 40
30 (C4Hb Coushatta
+ C9) abnormal chromatogram and
20
®

HbO +Indonesia 40
60
Tina-quant

40 Hb (C4
Queens C9)(Q1)
20 HbSD no HbA1c result
10 HbHb Queens (Q1)
Hopkins-2 20
30
101020 30 40 50 60 70 80 90 100 110 Hb Coushatta
Hb Hopkins-2 20
HbCC 20 40
40 60 80 100 120 140
(C4 + C9)
2010 20 30 40 50 602 70 80 90 100 110
Ultra HbCC
HbSC 20 40 60 802 100 120 140
Ultra
Ultra2 Hb Queens (Q1)
10 – x =HbSC
y Ultra2
Hb
x = Hopkins-2 20
10 20 30 40 50 60 70 80 90 100 110 – HbCC y
20 40 60 80 100 120 140
Ultra 2
Fig. 28: Results from the Tina-quant HbA1c Gen.2
®
HbSC Variant Type Ultra2 HbA1c (mmol/mol) n
assay on the COBAS INTEGRA® 800 analyzer in the
–x=y HbAA 0-100 50
presence of rare hemoglobin variants
Hb Setif 38-58 2
Hb Okayama 74 1
140 Hb Philadelphia (HbH) 41 1
140
Hb Zurich 20 1
120
120 Tosoh G8 HbEE 60 1
140
Tosoh G8 HbSE 36 1
100 All investigated rare variants
All investigated
100 except rare variants
Tosoh G8

120 for Hb Indonesia HbO Indonesia 39 1

G8

Tosoh
except G8
for Hb Indonesia
80 and Hb Hopkins-2 gave an HbSD 21 1
80 All
Tosoh Tosoh

100 andinvestigated
abnormal Hopkins-2raregave
Hb chromatogram variants
an
Hb Coushatta (C4 + C9) * 2
G8

60 except
abnormalfor chromatogram
Hb Indonesia
60and no HbA1c result
80 Hb Queens * 1
and Hb Hopkins-2
no HbA1c gave an
result
40 abnormal chromatogram Hb Hopkins-2 43 1
40
60
and no HbA1c result HbCC 51 1
20
20
20 40 60 80 100 120 140 HbSC 21 1
40
20 40 60 802 100 120 140
Ultra * No indications possible

Ultra2 Table 3: Rare hemoglobin variants used on all systems

20
20 40 60 80 100 120 140 11
Ultra2
Addendum I

HbA1c Multi Center Evaluation Study – cobas c 513

The cobas c 513 analyzer shows excellent correlation to different HPLC
systems in a method comparison utilizing whole blood samples

Lab 03, whole blood Lab 02, whole blood cobas c 513 vs Tosoh
application cobas c 513 application cobas c 513 vs G8 (NGSP values)
vs Menarini HA-8180V Tosoh G8 with IFCC SRM calib

Tina-quant HbA1c delivers accurate results, displays no interference

by common hemoglobin variants

12
Addendum II

Updated Statistics

As of February 2018, there are:

• 1,771 known hemoglobin mutations (hemoglobin variants and thalassemias).7
• 1,315 known human hemoglobin variants.7
• 19 variants with a mutation in the first four amino acids of the glycated
hemoglobin chain.7

13
Addendum III

Conversion Factor - IFCC mmol/mol to NGSP %8

Conversion Factor Equation: NGSP = (0.09148*IFCC) + 2.152

IFCC
31 42 53 64 75 86 97 108
mnol/mol

NGSP % 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0

SOURCE: NGSP – Convert between NGSP, IFCC, and eAG units

14
Acknowledgement
This study was conducted by the European Reference Laboratory (IFCC/NGSP Reference Laboratory) Zwolle and
Winterswijk, NL, supported by Roche.

1. NGSP. HbA1c methods: Effects of hemoglobin variants (HbC, HbS, HbE and HbD traits) and elevated fetal hemoglobin (HbF). Updated July 2013.
Available at: http://www.ngsp.org/interf.asp (accessed October 2013).
2. HbVar. A database of human hemoglobin variants and thalassemias. Available at: http://globin.cse.psu.edu/globin/hbvar/ (accessed October 2013).
3. Consensus Committee. (2007). Consensus statement on the worldwide standardization of the hemoglobin A1C measurement:
the American Diabetes Association, European Association for the Study of Diabetes, International Federation of Clinical Chemistry and Laboratory Medicine, and the
International Diabetes Federation. Diabetes Care 30, 2399–2400.
4. Rohlfing, C.L., Connolly, S.M., England, J.D., Hanson, S.E., Moellering, C.M., Bachelder, J.R., Little, R.R. (2008). The effect of elevated fetal hemoglobin on
hemoglobin A1c results: five common hemoglobin A1c methods compared with the IFCC reference method. Am J Clin Pathol 129, 811–814.
5. Little, R.R., Rohlfing, C.L., Hanson, S.E., Schmidt, R.L., Lin, C.N., Madsen, R.W., Roberts, W.L. (2012). The effect of increased fetal hemoglobin on 7 common
Hb A1c assay methods. Clin Chem 58, 945–947.
6. HbVar. A database of human hemoglobin variants and thalassemias. Available at: http://globin.cse.psu.edu/globin/hbvar/ (accessed February 2018).
7. NGSP.Org. IFCC Standardization of HbA1c. 2010. Available at: http://www.ngsp.org/ADA.asp (accessed February 2018).

COBAS, COBAS C and INTEGRA are trademarks of Roche.

©2018 Roche. MC-US-02505-1218

Roche Diagnostics
9115 Hague Road
Indianapolis, IN 46256
usdiagnostics.roche.com 15