As published in CLI April/May 2009 Renal disease

Emerging biomarkers for Acute

D is ea s e Fo c us

Kidney Injury
Early diagnosis of AKI currently depends on detection of logical variables and capable of both
reduced kidney function by the rise in serum creatinine early detection and risk stratifica-
tion would represent a tremendous
concentration, an unreliable measure in the acute advance in clinical medicine [4].
setting. This article discusses the pros and cons of
emerging early biomarkers which may prove useful. The search for AKI biomarkers
is an area of intense contempo-
by Dr Prasad Devarajan rary research. Conventional uri-
nary biomarkers such as casts and
fractional excretion of sodium are
The incidence of acute kidney general, there are several non-renal insensitive and non-specific for the
injury (AKI), previously referred to factors influencing the serum cre- early recognition of AKI. Other tra-
as acute renal failure, has reached atinine concentration such as body ditional urinary biomarkers such
epidemic proportions world-wide, weight, muscle mass, race, age, gen- as filtered high molecular weight
affecting about 7% of hospitalised der, total body volume, drugs, mus- proteins and tubular proteins or
patients. In the critical care setting, cle metabolism and protein intake. enzymes have also suffered from
the prevalence of AKI requiring In the face of AKI, serum creatinine lack of specificity and dearth of
dialysis is about 6%, with a mortal- is an even poorer reflection of kid- standardised assays. Fortunately, the
ity rate exceeding 60% [1, 2]. A sig- ney function, because the subjects application of innovative technolo-
nificant increase in morbidity and are not in steady state, and serum gies such as functional genomics
mortality associated with AKI has creatinine therefore lags far behind and proteomics to human and ani-
been demonstrated in a wide vari- renal injury. Furthermore, signifi- mal models of AKI has uncovered
ety of clinical situations, including cant chronic kidney disease can several novel genes and gene prod-
exposure to radiocontrast dye, car- exist with minimal or no change in ucts that are emerging as biomark-
diopulmonary bypass, mechanical creatinine because of renal reserve ers. The most promising of these are
ventilation and sepsis. Once and enhanced tubular secretion of listed in Table 1, and detailed in this
established, the treatment is creatinine. Ironically, experimental article.
largely supportive, at an annual studies have identified interven-
cost surpassing $10 billion in the US tions that may prevent or treat AKI Neutrophil gelatinase-associ-
alone. if instituted early in the disease ated lipocalin (NGAL)
process, well before the serum cre- Preclinical studies in ani-
The early diagnosis of AKI cur- atinine rises [2, 3]. The lack of pre- mal models identified NGAL
rently depends on detection of dictive biomarkers has impaired our (also known as lipocalin 2) as
reduced kidney function by the rise ability to translate these promising one of the most upregulated genes
in serum creatinine concentration, findings to human AKI. A troponin- and proteins in the kidney very early
which is a delayed and unreliable like biomarker of AKI that is easily in the course of AKI. Following the
measure in the acute setting. In measured, unaffected by other bio- serendipitous finding that NGAL
protein in animals was eas-
Biomarker Sample Cardiopulmonary Contrast Critical care Kidney Commercial ily detected in the urine
name source bypass (CPB) administration setting transplant (tx) assay
soon after AKI, a number
NGAL Urine <2 h post-CPB 2 h post-contrast 48 h pre-AKI 12-24 h post-tx ELISA, of translational stud-
#
ARCHITECT ies have now established
IL-18 Urine 6 h post CPB Not increased 48 h pre-AKI 12-24 h post-tx ELISA NGAL as an early diagnos-
tic biomarker for AKI in
KIM-1 Urine 12 h post CPB Not tested Not tested Not tested ELISA common human clinical
L-FABP Urine 4 h post-CPB 24 h post- Not tested Not tested ELISA situations. In prospective
contrast
*
studies of children who
NL Plasma <2 h post-CPB 2 h post-contrast 48 h pre-AKI Not tested ELISA, Triage
underwent elective cardiac
Table 1. Novel biomarkers for the early prediction of AKI in humans. AKI: acute kidney injury, defined as a 50% surgery, AKI (defined as a
increase in serum creatinine from baseline; NGAL: neutrophil gelatinase-associated lipocalin; IL-18: interleukin 18; 50% increase in serum cre-
KIM-1: kidney injury molecule 1; L-FABP: liver-type fatty acid binding protein. The times indicated (in hours) refer to atinine) occurred 2-3 days
the earliest time points at which the biomarker is increased significantly from baseline. #The ARCHITECT Assay is after surgery [5]. In con-
manufactured by Abbott Diagnostics. * The Triage NGAL Test is manufactured by Biosite Inc. trast, NGAL levels in urine
 As published in CLI April/May 2009

and plasma measured by ELISA Recent studies have also demon- multi-centre validation studies.
revealed a dramatic increase, within strated the utility of early NGAL
two hours of the surgery, in those measurements for predicting clini- Clearly NGAL represents a novel
who subsequently developed AKI. cal outcomes of AKI. In subjects predictive biomarker for AKI and
Both urine and plasma NGAL were undergoing cardiac surgery, the its outcomes. However, there are
excellent independent predictors of two hour post-operative plasma some limitations [2-4]. Plasma
AKI, with an area under the curve NGAL levels were strongly corre- NGAL measurements may be influ-
for the receiver-operating charac- lated with duration and severity of enced by coexisting variables such
teristic (AUC-ROC) of > 0.9 for the AKI and length of hospital stay. In as chronic kidney disease (CKD),
2-6 hour measurements [5]. These addition, the 12 hour plasma NGAL chronic hypertension, systemic
findings have now been confirmed was strongly correlated with mor- infections, inflammatory condi-
and published in several additional tality [6]. Similarly, the two hour tions and malignancies. Urine
prospective studies of adults and urine NGAL levels were strongly NGAL levels are also elevated in
children who developed AKI after correlated with duration and sever- subjects with CKD, lupus nephritis
defined clinical events such as car- ity of AKI, length of hospital stay, and urinary tract infections. How-
diac surgery (10 studies), contrast dialysis requirement, and death ever, the levels of both plasma and
administration (four studies) and [7]. A collective analysis of 10 pub- urine NGAL in these clinical situ-
kidney transplantation (two stud- lished studies of NGAL as a predic- ations are significantly lower than
ies). Urine and plasma NGAL have tive biomarker for dialysis require- those typically measured in AKI.
also emerged as robust biomarkers ment reveals a mean AUC-ROC These issues are therefore unlikely
that predict development of AKI in of 0.78 (95% confidence interval to impact on the clinical utility of
heterogeneous groups of patients 0.69-0.87). NGAL as an excellent biomarker
with unknown timing of kidney for AKI.
injury, such as subjects admitted to The NGAL results described thus far
emergency or critical care settings have been obtained using research- Interleukin 18 (IL-18)
(six published studies). A collective based ELISA assays, which are IL-18 is a pro-inflammatory
analysis of all the published per- impractical in the clinical setting. cytokine that is known to be
formances of NGAL as a predictive A major recent advance from the induced and cleaved in the proxi-
biomarker for AKI to date [reviewed clinical laboratory perspective has mal tubule, and subsequently can
in 2-4] reveals a mean AUC-ROC of been the development of a point- be easily detected in the urine
0.82 (95% confidence interval 0.7- of-care kit for the clinical meas- following ischaemic AKI in ani-
0.9). The mean derived sensitivity urement of plasma NGAL (Triage mal models. In an initial cross-
and specificity from these studies NGAL Device, Biosite Incorpo- sectional study, urine IL-18 levels
are in the 75-80% range. From the rated). The assay is user-friendly, measured by a commercially avail-
published literature, the derived with quantitative results avail- able ELISA were markedly ele-
optimal cut-off value of NGAL for able in 15 minutes, and requires vated in patients with established
the early prediction of AKI appears only microlitre quantities of whole AKI, but not in subjects with uri-
to be at a median value of approxi- blood or plasma. In subjects under- nary tract infection, chronic kid-
mately 150 ng/mL. going cardiac surgery, the two hour ney disease, nephrotic syndrome
plasma NGAL measurement meas- or prerenal azotaemia. In a sub-
NGAL is also emerging as an early ured by the Triage Device showed sequent study of patients who
biomarker in interventional trials. an AUC of 0.96, sensitivity of 0.84, developed AKI two to three days
For example, a reduction in urine and specificity of 0.94 for predic- after cardiac surgery, urinary IL-
NGAL has been employed as an tion of AKI using a cutoff value 18 levels peaked at 12 hours post
outcome variable in clinical trials of 150 ng/mL [6]. Furthermore, a surgery, and predicted AKI with
demonstrating the improved effi- urine NGAL immunoassay has also an AUC-ROC of 0.75 [8]. Urine
cacy of a modern hydroxyethyl- been developed for a standardised IL-18 is also considered an early
starch preparation compared with clinical platform (ARCHITECT biomarker of AKI in the intensive
albumin or gelatin for maintaining analyser, Abbott Diagnostics). care setting, being able to predict
renal function in cardiac surgery This assay is also easy to perform; this complication about two days
patients. Similarly, the response no manual pretreatment steps are prior to the rise in serum creati-
of urine NGAL was attenuated in necessary, a first result is available nine. Early elevated urine IL-18
cardiac surgery patients, who expe- within 35 minutes and only 150 levels correlated with the severity
rienced a lower incidence of AKI microlitres of urine are required. of AKI as well as mortality.
after sodium bicarbonate therapy as Following cardiac surgery, the two
compared to sodium chloride. Fur- hour urine NGAL measurement Thus, IL-18 may also represent a
thermore, subjects who developed by ARCHITECT analyser showed promising biomarker for AKI. IL-
AKI after aprotinin use during car- an AUC of 0.95, sensitivity of 0.79, 18 is more specific to ischaemic
diac surgery displayed a dramatic and specificity of 0.92 for predic- AKI, and levels are largely unaf-
rise in urine NGAL in the immedi- tion of AKI using a cutoff value of fected by nephrotoxins, chronic
ate post-operative period, attesting 150 mg/mL [7]. Both clinical plat- kidney disease or urinary tract
to the potential use of NGAL for forms are CE-marked for the Euro- infections. However, IL-18 meas-
the prediction of nephrotoxic AKI. pean Union, and are undergoing urements may be influenced by a
 Renal Disease
number of coexisting variables, patients with established AKI, uri- who developed AKI post contrast
since renal IL-18 mRNA levels nary KIM-1 was associated with dye. In a recent prospective study
are known to be induced in other adverse clinical outcomes, includ- of children undergoing cardiac
disease states such as endotoxae- ing dialysis requirement and death. surgery, urine L-FABP increased
mia, immunological injury and at four hours post-bypass, with an
cisplatin toxicity. Furthermore, Thus KIM-1 may represent a prom- AUC of 0.810 for a cutoff value of
plasma IL-18 levels are known to ising AKI biomarker. An advan- 486 ng/mg creatinine [10].
be increased in various systemic tage of KIM-1 as a urinary biomar-
inflammatory states, and the rela- ker is the fact that its expression Thus, L-FABP also appears to rep-
tionships between plasma and seems to be limited to the injured resent a promising AKI biomarker.
urine IL-18 remain unexplored. or diseased kidney, and no sys- However, urinary L-FABP meas-
temic source of KIM-1 has been urements may also be influenced
Kidney injury molecule described. However, urinary KIM- by a number of confounding vari-
1 (KIM-1) 1 measurements may be influenced ables. Several studies have docu-
In animal models, kidney injury by a number of other confounding mented increased urinary L-FABP
molecule 1 is one of the most highly variables. KIM-1 is induced in the levels in patients with non-diabetic
induced proteins in the kidney after kidney and upregulated in the urine chronic kidney disease, early dia-
AKI, and a proteolytically proc- by a large number of nephrotoxins, betic nephropathy, idiopathic focal
essed ectodomain of KIM-1 is eas- including cyclosporine, cisplatin, glomerulosclerosis and polycystic
ily detected in the urine soon after cadmium, gentamicin, mercury kidney disease. Additionally, L-
AKI. Assays for KIM-1 (ELISA and and chromium. Similarly, KIM-1 FABP is also abundantly expressed
microbead-based) have been devel- in the kidney and urine is induced in the liver, and urinary L-FABP
oped in research laboratories, but in a variety of chronic proteinuric, may be influenced by serum L-
are not commercially available. In a inflammatory and fibrotic disease FABP levels.
small human cross-sectional study, states in humans.
KIM-1 was found to be markedly Summary
induced in proximal tubules from Liver fatty acid binding Of the AKI markers discussed in
kidney biopsies in patients with protein (L-FABP) this article, NGAL has entered
established AKI, and urinary KIM- Liver fatty acid binding protein (L- the final phases of the biomarker
1 measured by ELISA distinguished FABP) is a 14 kDa protein normally development process, facilitated
ischaemic AKI from prerenal azo- expressed in the proximal convo- by the development of commercial
taemia and chronic renal disease. luted and straight tubules of the tools for its measurement in large
In a case-control study of children kidney, and upregulated in animal populations across different labo-
undergoing cardiac surgery, uri- models of AKI. An ELISA for this ratories. But will any single biomar-
nary KIM-1 levels were enhanced analyte is commercially available; it ker such as NGAL suffice for AKI
in subjects who developed AKI, was first used to demonstrate that management? In addition to early
with an AUC of 0.83 at the 12 hour urinary L-FABP levels were signifi- diagnosis and prediction, it would
time point [9]. In a larger prospec- cantly increased prior to the increase be desirable to identify biomark-
tive cohorts study of hospitalised in serum creatinine in those patients ers capable of identifying aetiolo-
gies, predicting clinical
outcomes, allowing for risk
stratification and monitor-
ing the response to inter-
ventions. In order to obtain
all of this desired informa-
tion, a panel of validated
tandem biomarkers may be
needed, with temporal pro-
files as illustrated in Figure
1. However, one must also
take into account the tech-
nical and cost issues sur-
rounding the identification,
validation, commercial
development and accept-
ance of multi-marker pan-
els. From the recent cardi-
ology literature, we can see
that troponin as a stand-
Figure 1. Urinary biomarker profiles in subjects who develop AKI after surgery with cardiopulmonary bypass (CPB). alone biomarker provides
AKI: acute kidney injury, defined as a 50% increase in serum creatinine from baseline; NGAL: neutrophil gelatinase- excellent diagnostic and
associated lipocalin (data from reference 5); IL-18: interleukin 18 (from reference 8); KIM-1: kidney injury molecule 1 prognostic information in
(from reference 9); L-FABP: liver-type fatty acid binding protein (from reference 10). acute coronary syndromes
 As published in CLI April/May 2009

and acute decompensated heart fail- The author

ure. If the current prospective mul- Prasad Devarajan, M.D.
ticentre studies measuring NGAL Professor of Pediatrics and Developmental
levels using standardised labora- Biology,
tory platforms provide promising Cincinnati Childrens Hospital, Univer-
results, we may already have the sity of Cincinnati, 3333 Burnet Avenue,
AKI troponin. Cincinnati, OH 45229-3039, USA.
e-mail: prasad.devarajan@cchmc.org

References
1. Parikh CR, Devarajan P. New
biomarkers of acute kidney injury.
Crit Care Med 2008; 36(4 Suppl):
S159-65.
2. Devarajan P. Neutrophil gelati- :_W]deij_Yi
nase-associated lipocalin (NGAL):
a new marker of kidney disease.
Scand J Clin Lab Invest Suppl 2008;
241:89-94.
3. Devarajan P. NGAL in acute kidney
injury: from serendipity to utility.
Am J Kidney Dis 2008; 52:395-9
4. Devarajan P. Neutrophil gelatinase-
associated lipocalin--an emerg- Early diagnosis
ing troponin for kidney injury.
Nephrol Dial Transplant 2008;
of acute
23:3737-43.
5. Mishra J, Dent C, Tarabishi R, et al.
kidney injury
Neutrophil gelatinase-associated
lipocalin (NGAL) as a biomarker for
t Critical illness
acute renal injury following cardiac
surgery. Lancet 2005; 365:1231-38. t Cardiac surgery
6. Dent CL, Ma Q, Dastrala S, et al.
t Nephrotoxicity
Plasma NGAL predicts acute kid-
ney injury, morbidity and mortal- t Kidney transplantation

NGAL
ity after pediatric cardiac surgery:
a prospective uncontrolled cohort
study. Crit Care 2007; 11:R127.
7. Bennett M, Dent CL, Ma Q, et al. NGAL Rapid ELISA Kit IVD

Urine NGAL predicts severity of

acute kidney injury after cardiac
surgery: A prospective study. Clin J
Am Soc Nephrol 2008; 3:665-673.
8. Parikh CR, Mishra J, Thiessen-
Philbrook H, et al. Urinary IL-18
is an early predictive biomarker
of acute kidney injury after car-
diac surgery. Kidney Int 2006;
70:199-203.
9. Han WK, Waikar SS, Johnson A, et
al. Urinary biomarkers in the early
diagnosis of acute kidney injury.
Kidney Int 2008; 73:863-869.
10. Portilla D, Dent C, Sugaya T, et al.
Liver fatty acid-binding protein as
a biomarker of acute kidney injury
BioPorto Diagnostics A/S Phone (+45) 4529 0000
after cardiac surgery. Kidney Int
Grusbakken 8 Fax (+45) 4529 0001
2008; 73:465-72.
DK-2820 Gentofte E-mail info@bioporto.com
Denmark Web www.bioporto.com
www.ngal.com