The diagnosis of acute kidney injury (AKI) has relied on serum creatinine and urine output, two biomarkers that are insensitive and nonspecific, particularly early in the course of the syndrome. In addition, creatinine and urine output are functional markers and not markers of frank injury. The lack of sensitive and specific injury biomarkers has greatly impeded the early diagnosis of AKI and limited the ability to predict outcomes. Multiple potential therapies for AKI that have found success in animal models have failed to translate into humans. It is likely that the absence of early, injury-specific biomarkers has impaired the ability of investigators to design appropriate clinical trials to evaluate the potential therapeutic efficacy of such treatments. A large number of biomarkers of kidney injury have been proposed that may aid in the early detection, differential diagnosis, and prognosis of AKI. We review the rationale for biomarker development and the status of some of the more promising biomarkers, both in the acute setting and as predictors of long-term outcomes. Such biomarkers of kidney injury will enable the development of more efficient strategies to evaluate new therapeutic approaches to this common clinical condition, which continues to be associated with high morbidity and mortality.
Keywordsbiomarkers, inflammation, NGAL, serum creatinine
Biomarkers in Acute Kidney Injury, 713
Creatinine as a Biomarker, 714
U.S. Food and Drug Administration Critical Path Initiative, 715
Need for New Biomarkers, 715
Specific Biomarkers of Acute Kidney Injury, 715
Repurposed Biomarkers, 722
The Future of Biomarkers in Acute Kidney Injury, 723
The authors would like to thank Drs. Joseph Bonventre and Venkata Sabbisetti for their contributions from a previous version of this chapter.
Biomarkers in Acute Kidney Injury
Acute kidney injury (AKI), previously referred to as acute renal failure (ARF), represents a common clinical problem with high mortality. It is estimated that the overall incidence of AKI among intensive care unit (ICU) patients is as high as 58%. A multinational study of 29,269 critically ill patients admitted to the ICU revealed an overall occurrence of AKI requiring renal replacement therapy (RRT) of 5.7% with an associated mortality of 60.3%. It is estimated that AKI in high-income countries costs US $1 billion, claims 300,000 lives, results in 170,000 end-stage renal disease (ESRD) diagnoses, and contributes to the development of 300,000 advanced chronic kidney disease (CKD) cases on an annual basis. The poor outcome associated with AKI has not improved in the past few decades despite progress in the understanding of the pathophysiology of AKI and advances in therapeutics and supportive care. AKI has been increasing in frequency and continues to be associated with an unacceptably high in-hospital mortality of 40% to 80% in the intensive care setting.
Historically, the definition of AKI was not standardized, with some authors defining it by a change in urine output, others by myriad combinations of an absolute increase in creatinine from baseline, a relative change from baseline, an absolute threshold creatinine had to rise above or even the need for RRT. The barriers to research that such heterogeneity represents are obvious. In an attempt at standardization, several guidelines have been put forth defining AKI using either absolute or relative changes in serum creatinine (sCR) levels or acute changes in urine output. The Acute Dialysis Quality Initiative (ADQI) developed a set of criteria for defining AKI, based on sCR levels and urine output. These are called the RIFLE ( R isk, I njury, F ailure, L oss, and E nd-stage) criteria. RIFLE uses relative changes in sCR and glomerular filtration rate (GFR) as criteria for its first three categories of risk, injury, and failure. In a further effort to standardize the definition of AKI, the ADQI and the Acute Kidney Injury Network (AKIN) modified the RIFLE criteria. The AKIN group defined AKI as an abrupt (within 48 hours) reduction in kidney function currently defined as an absolute increase in sCR of ≥0.3 mg/dL (≥25 μmol/L), a percentage increase in sCR of ≥50% (1.5-fold from baseline), or a reduction in urine output (documented oliguria of <0.5 mL/kg/h for >6 hours). Finally, in 2011, the Kidney Disease: Improving Global Outcomes (KDIGO) Clinical Practice Guideline proposed something of a synthesis of these two systems and defined AKI as an increase in creatinine of ≥0.3 mg/dL within 48 hours or an increase to ≥1.5 times baseline within 7 days or urine output ≤0.5 mL/kg for 6 hours. The three classification systems are summarized in Table 47.1 .
|Classification System||Criteria by Changes in Creatinine||Criteria by Changes in Urine Output|
|150%–200% of baseline creatinine or GFR decreased by 25% (within 7 d) |
200%–300% of baseline sCR or GFR decreased by >50%
>300% of baseline sCR or GFR decreased by >75% or sCR of >4 mg/dL in the setting of an increase of ≥0.5 mg/dL
Complete loss of kidney function requiring dialysis for >4 wk
Complete loss of kidney function requiring dialysis for >3 mo
|<0.5 mL/kg/h for <6 h |
<0.5 mL/kg/h for >12 h
<0.3 mL/kg/h × 24 h or anuria × 12 h
|Stage 1 |
|150%–200% of baseline sCR or increase of ≥0.3 mg/dL (within 48 h) |
200%–300% of baseline sCR
>300% of baseline sCR or sCR of >4 mg/dL in the setting of an increase of ≥0.5 mg/dL
|<0.5 mL/kg/h for >6 h |
< 0.5 mL/kg/h for >12 h
<0.3 mL/kg/h × 24 h or anuria × 12 h
|Stage 1 |
|150%–200% of baseline creatinine (within 7 d) or increase of ≥0.3 mg/dL (within 48 h) |
200%–300% of baseline sCR
>300% of baseline sCR or increase in sCR to ≥4 mg/dL or initiation of renal replacement therapy
|<0.5 mL/kg/h for 6 h |
<0.5 mL/kg/h for 12 h
<0.3 mL/kg/h for 24 h or anuria for ≥12 h
A biomarker is defined as a characteristic that can be objectively measured and evaluated as an indicator of normal biological or pathogenic processes (a diagnostic biomarker), or pharmacological responses to a therapeutic intervention (a therapeutic biomarker). Any parameter that can be measured—for example, proteins, lipids, genomic or proteomic patterns, imaging methods, blood pressure, electrical signals, and cells present in urine—may serve as a biomarker. Biomarkers are of different types: disease, toxicity, mechanistic, efficacy, predictive, and drug-target interaction biomarkers. Some of these markers can serve as translational markers that can be used in both preclinical and clinical settings. A surrogate endpoint marker is a biomarker that can substitute for a clinical endpoint. A surrogate endpoint marker is expected to predict clinical benefit (or harm or lack of benefit) based on epidemiological, therapeutic, pathophysiological, or other scientific evidence. An ideal biomarker is easily measurable, noninvasive, reproducible, sensitive, cost-effective, easily interpretable, and would use readily available specimens (blood and urine). The widely accepted measure of biomarker sensitivity and specificity is the receiver operating characteristic (ROC) curve. An ROC curve is a graphical display of tradeoffs between the true-positive rate (sensitivity) and the false-positive rate (1-specificity), when the biomarker is a continuous variable. A curve is generated; the closer the curve to the left-hand and top borders of the graph, the better the accuracy of the biomarker. The area underneath the ROC curve can range from 0.5 ( useless test performing at the level of chance ) to 1 ( perfect test ). A perfect biomarker will have true-positive rate of 1 and false-positive rate of 0. As will be seen in the evaluation of biomarkers for the diagnosis of AKI, the validity of all such tests is contingent on the validity of the gold standard against which they are compared.
Creatinine as a Biomarker
Creatinine levels in the serum were first reported to be a marker of chronic nephritis more than 100 years ago and proposed as a means of distinguishing between what has become known as AKI from CKD soon after. Despite such an impressive run of clinical utility, creatinine as a biomarker of AKI has several widely recognized shortcomings. First, creatinine production and its release into the circulation is highly variable with age, sex, meat intake, muscle mass, and diseases. For example, in certain disease states such as rhabdomyolysis, sCR levels may rise more rapidly, due to the release of preformed creatinine from the damaged muscle. Second, a static measure of creatinine does not depict the real-time changes in GFR resulting from acute changes in kidney function. Given the large amounts of functional renal reserve in a healthy individual and the variable amounts of renal reserve in patients with mild renal diseases, creatinine is not a sensitive marker and does not change until significant renal damage has occurred. When creatinine levels do increase, it often takes 24 to 48 hours after the AKI (as the degree to which serum levels can rise is limited by the daily production), and at this point in most cases the acute event is remote in time and the likelihood that an intervention will alter the patient’s course of kidney injury will be markedly diminished. Third, the delay in detection of AKI is exacerbated and the underestimation of its severity potentiated by a decrease in creatinine production in critically ill patients. Fourth, a drug-induced alteration in tubular secretion of creatinine might result in underestimation of renal function. Fifth, the creatinine assay is subject to interference because of intake of certain drugs or because of certain pathophysiological states including hyperbilirubinemia and diabetic ketoacidosis. Finally, and perhaps most importantly, sCR is a marker of renal filtration, not structural injury. An acute rise in creatinine signals only an acute change in GFR. It is a functional rather than structural marker and thus cannot distinguish functional causes of changes in GFR such as prerenal azotemia from those involving frank structural injury such as acute tubular necrosis (ATN). Given the vast differences in mechanistic pathways, appropriate therapies, and ultimate outcomes between such disparate entities, it is likely that the inability of creatinine to distinguish between functional and structural AKI is one of the most fundamental causes of the vexing lack of therapies for acute tubular kidney injury. Urine microscopy is a time-honored test for evaluation of AKI, and seasoned physicians will attest to its value in detecting evidence of structural injury. Nevertheless, the sensitivity of this test as an early indicator of tubular injury in the kidney remains controversial. Because of these multiple significant limitations of creatinine as a marker, there has been a great deal of interest in the identification of improved biomarkers for kidney injury.
U.S. Food and Drug Administration Critical Path Initiative
As a major initiative of the U.S. Food and Drug Administration (FDA) focusing on biomarkers, The Critical Path Initiative is an effort to stimulate and facilitate a national effort to modernize the scientific process through which a potential human drug, biological product, or medical device is transformed from a discovery or “proof of concept” into a medical product. The FDA has provided guidelines that a biomarker can be considered “valid” only if the following conditions are met: (1) It is measured in an analytical test system with well-established performance characteristics; and (2) there is an established scientific framework or body of evidence that elucidates the physiological, pharmacological, toxicological, or clinical significance of the test result.
Need for New Biomarkers
Clinicians and researchers are in need of new biomarkers of AKI for the following reasons: (1) Rather than being injury markers, the current blood and urine markers are functional consequences of the injury itself ; (2) creatinine, a central component of many of the definitions of AKI, is a poor biomarker because of its aforementioned characteristics; and (3) we have an urgent need for novel biomarkers to diagnose AKI at early stages, to predict outcomes in patients with AKI, to identify who will respond to a given intervention, and to ascertain whether the intervention is actually working. Critically, better biomarkers will permit better stratification of patients for clinical trials and potentially lead to identification of new therapeutics for AKI. The absence of sensitive and specific early biomarkers of AKI not only delays the diagnosis of AKI but also greatly impairs early intervention strategies and clinical trial design, thus delaying the initiation of potential therapies. As a marker in changing function, a rise in creatinine after an AKI is analogous to a change in ejection fraction (EF) after a myocardial infarction (MI). Any attempts to develop therapies for MI based solely on changes in EF would be bedeviled by poor sensitivity, poor specificity, and delayed diagnosis. By more accurately phenotyping patients with AKI due to frank structural injury, biomarkers can enhance statistical power at a given sample size by reducing misclassification, reduce trial costs by identifying those patients most likely to have progression of early AKI and prioritizing them for trial enrollment, and serve as potential outcomes measures to identify therapies that deserve further investigation in costly, large, multicenter trials. In summary, the need for new biomarkers can be seen through the following list of potential applications: (1) early diagnosis; (2) to differentiate AKI subtypes (prerenal, intrinsic, postrenal); (3) to identify AKI associated with various etiologies; (4) to determine the primary location of the injury in the kidney; (5) to predict outcomes; and (6) to predict response to various therapies. The same biomarker may not perform in all of these settings.
Blood and urine are two candidate fluids used to measure biomarkers of kidney injury. Urine has the advantage of being readily available noninvasively and amenable to straightforward testing by both healthcare professionals and patients themselves. In addition, the low protein content of the urine in most clinical states makes urine more favorable for proteomic approaches. On the other hand, changes in urine flow rate and concentration will have effects on the concentration of an analyte, and variations in physical and chemical properties of urine may affect the stability of the analyte and reliability of the test. Serum samples are also readily available and may be more stable than urine. The presence of abundant proteins such as albumin and immunoglobulins in the blood leads to high interference and makes proteomic approaches more challenging.
Biomarkers have been proposed to reflect injury to various parts of the nephron or to reflect interstitial disease, although in many cases the specificity of particular biomarkers for specific nephron sites has not been sufficiently studied. The proximal tubule is the primary site of damage for ischemic injury or reperfusion injury as well as with most tubular toxins. Even in those cases where the primary site of injury is more distal in the nephron, the proximal tubule is often secondarily involved as well. Although there are some important exceptions to this generalization, such as lithium toxicity that predominantly occurs at distal nephron, in general a biomarker that is sensitive for proximal injury will be useful in many clinical scenarios.
An ideal AKI biomarker should be (1) easily and reliably measured in a noninvasive or minimally invasive manner; (2) stable; (3) rapidly and reliably measurable at the bedside; (4) inexpensive to measure; (5) able to detect AKI early in the course; and (6) predictive in its ability to forecast the course of AKI and potentially the future implications of AKI.
Specific Biomarkers of Acute Kidney Injury
There are many cellular proteins that are released into the urine and have been used in the past to monitor kidney injury in animals and humans. As the concept of kidney injury biomarkers has evolved, new biomarkers have continued to be proposed and some older ones have come to receive less attention. One of the earlier reviews on AKI biomarkers focused on β2-microglobulin (β2M), retinol-binding protein, N -acetyl-glucosaminidase (NAG), adenosine deaminase binding protein, and l -alanine aminopeptidase. Since then, a plethora of proteins have been proposed, including but certainly not limited to α1-microglobulin (α1M), α-glutathione S -transferase (αGST), kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL), interleukin (IL)-18, fatty acid binding protein (FABP), cystatin C (Cys-C), netrin-1, clusterin, trefoil factor-3, osteopontin (OPN), monocyte chemotactic protein-1 (MCP-1), uromodulin, YKL-40, IL-6, IL-10, tissue inhibitor of metalloproteinase 2 (TIMP2) and insulin-like growth factor binding protein 7 (IGFBP7). The location of production and physiological significance of multiple biomarkers are shown in Fig. 47.1 . Because of the limitations of the length of this review, we will focus on an overview of important features of some of these markers (particularly urinary rather than serum) without being able to do justice to the data available on the full scope of current research.
α1M is a glycoprotein of approximately 27 to 30 kDa primarily synthesized by liver, available both in free form and as a complex with immunoglobulin A. α1M is freely filtered at the glomerulus and completely reabsorbed and catabolized by the proximal tubule. Megalin mediates the uptake of this protein in the proximal tubule. α1M is more stable than β2M over a range of pH levels in the urine. α1M quantitation in the urine has been reported as a sensitive biomarker for proximal tubule dysfunction in both adults and children. In a small cohort of 73 patients, out of which 26 required RRT, comparing α1M, β2M, cys-C, retinol-binding protein, αGST, lactate dehydrogenase, and NAG early in the course of AKI, Herget-Rosenthal et al. found that urinary cys-C and α1M have the highest ability to predict the need for RRT. In addition, α1M also has been reported as a useful marker for proximal tubular damage and recovery in early infancy. In a study of 78 patients undergoing elective cardiac surgery, urinary levels of α1M were moderately predictive of the subsequent development of AKI when corrected for urinary creatinine, area under the curve (AUC) of 0.71. Limitations associated with α1M include the variation in serum levels with age; sex ; clinical conditions, including liver diseases, ulcerative colitis, HIV, and mood disorders ; and the lack of international standardization.
IL-18 is a proinflammatory cytokine that plays an important role in many human diseases and is produced in a number of tissues. Renally, IL-18 is produced in the proximal tubule and is converted from its proform to the active form by caspase-1. Unlike the majority of other novel proteins that serve solely as biomarkers for AKI, IL-18 has been shown to be a potent mediator of ischemic AKI in animal models and has been shown to contribute to tubular damage during ischemia-reperfusion injury. Interestingly, IL-18–deficient mice have been shown to be protected from ischemia/reperfusion-induced AKI. It has been reported that IL-18 levels are elevated in the urine of patients with AKI and delayed graft function (DGF) compared with healthy individuals and patients with prerenal azotemia, urinary tract infection (UTI), CKD, and nephritic syndrome. In the same study, the urinary concentration cutoff for IL-18 at 500 pg/mg creatinine gave an optimal sensitivity (0.85) and selectivity (0.88) for the diagnosis of ATN. Parikh et al. reported that IL-18 is an early predictive biomarker of AKI after cardiopulmonary bypass (CPB), and that NGAL and IL-18 are increased in tandem after CPB. Coca et al., in an analysis of published literature, reported that the urine levels of IL-18 are significantly greater in patients with established ATN (AUC = 0.95) compared with all other types of patients including CKD, UTI, and prerenal azotemia. Furthermore, IL-18 has been shown to be a selective marker for predicting severity of AKI and mortality in adult, and critically ill children. In four studies of urinary IL-18 as an early predictive biomarker of AKI in adults and children, authors carried out ROC analysis of biomarker performance. In the four studies, AUCs for IL-18 varied from 0.54 to 0.9, suggesting IL-18 as a biomarker with variable sensitivity but with higher specificity for the early diagnosis of AKI.
A metaanalysis of 11 studies investigating IL-18 was composed of 2796 patients ranging from newborns to elderly. The majority of studies were either set in the ICU or involved cardiac surgery. The overall AUC was 0.77, whereas the diagnostic odds ratio for urinary IL-18 to predict AKI was found to be 5.1 (95% confidence interval [CI], 3.3 to 8.1) with a sensitivity and specificity of 0.51 and 0.79, respectively. IL-18 performed similarly in cardiac surgery patients and those in the ICU. However, it was noted that performance was better in pediatric populations than in adults and, in subgroup analysis, the early (<12 hours) predictive time group had significantly greater odds ratio than the 24- or 48-hour groups.
In an exhaustive study, Arthur et al. investigated 32 potential biomarkers in 95 patients with stage 1 AKI after cardiac surgery to assess for the association with worsening of AKI or death (primary outcome) or stage 3 AKI or death (secondary outcome). IL-18 was the best predictor with AUC of 0.74 and 0.89, respectively. A combination of IL-18 and KIM-1 demonstrated an excellent AUC of 0.93 to predict stage 3 AKIN or death.
Neutrophil Gelatinase-Associated Lipocalin
NGAL, also known as lipocalin-2 or Siderocalin, was first discovered as a 25-kDa protein in granules of human neutrophils. NGAL is currently the most widely studied AKI biomarker and has been investigated in numerous settings including the emergency room, generalized hospital patients, adult cardiac surgery, pediatric cardiac surgery, the ICU, transplant, and in cirrhosis. NGAL is normally expressed at low levels in a number of organs including kidney, breast, liver, small intestine, prostate, stomach, lymphoid cells, thymus, and lungs. NGAL expression is upregulated in several cancers, including pancreas, lung, colon, ovary, breast, and brain. NGAL is markedly elevated in many organs during inflammation and ischemia. Systemic NGAL is filtered and reabsorbed in the proximal tubules via a megalin-dependent pathway. Renal NGAL production occurs primarily in the thick ascending loop of Henle and the intercalated cells of the collecting ducts. Elevated urine levels with AKI therefore could be secondary to decreased reabsorption by damaged proximal tubule cells or increased production distally. NGAL exists in three distinct isoforms: 25-kDa monomer, 45-kDa homodimer, and 135-kDa heterodimer. Because NGAL can be produced by neutrophils and other organs in addition to the kidney, in clinical conditions such as sepsis or UTIs, the appearance of NGAL in the urine might not directly reflect the severity of kidney injury. However, the heterodimer, which is conjugated to gelatinase, is specific to neutrophils and thus fractionating urinary isoforms should allow for improved isolation of renally produced NGAL.
NGAL binds to iron-siderophore complexes, thereby exerting a bacteriostatic effect by limiting bacterial iron uptake ( Fig. 47.2 ). In addition, NGAL provides antiapoptotic effects and enhances proliferation of renal tubular cells, potentially making it protective in AKI. Intriguingly, recent data suggests that recombinant NGAL might be useful prophylactically for the prevention of cisplatin- induced AKI. An elevation in urinary NGAL is detectable as early as 3 hours after tubular injury, peaks between 6 to 12 hours, and can persist up to 5 days depending on the severity and resolution (or lack thereof) of the renal insult.
The early clinical studies with NGAL were done in children. Mishra and colleagues prospectively obtained serial urine and serum samples from 71 children undergoing CPB for surgical correction of congenital heart disease. Twenty children (28%) developed AKI, defined as a 50% increase in sCR. Both serum and urinary NGAL within 2 to 6 hours after CPB almost perfectly predicted which patients would subsequently develop AKI with an AUC of 0.91 and 0.99 for serum and urine, respectively. A larger follow-up study of 120 children by Dent et al. showed that 2-hour postoperative serum NGAL was predictive of AKI (AUC ROC = 0.96) and correlated with postoperative change in sCR, duration of AKI, and hospital length of stay. At 12 hours, NGAL levels strongly correlated with mortality. In a subsequent prospective study in children by Zappitelli et al., in which the AKI population was more heterogeneous with unknown timing of AKI, the AUCs of urinary NGAL for prediction of AKI were lower than the AUC reported by Mishra et al. In a study of adults, NGAL levels were measured in a series of patients who entered the hospital via the emergency department. At a cutoff value of 130 mcg/g creatinine, sensitivity and specificity of NGAL for detecting acute injury were 0.9 and 0.995, respectively. These values were better than those for NAG, α1M, α1-acid glycoprotein, fractional excretion of sodium, and sCR, although the AUC was equivalent for NGAL and creatinine.
The most frequently studied setting for NGAL is in cardiac surgery. A recent metaanalysis identified 28 studies investigating AKI biomarkers measured either intraoperatively or immediately postoperatively; 16 of the studies containing 2906 patients measured NGAL. AKI was defined via the AKIN criteria in the majority of studies with the remainder using RIFLE or KDIGO. The composite AUC for NGAL across studies was 0.72. In a large multicenter study of 1191 adults undergoing cardiac surgery, plasma NGAL (drawn both pre- and postoperatively) was assessed with long-term survival at 3 years. In univariate analysis both pre- and postoperative levels were associated with survival but after adjustment only the presurgery level was independently associated.
In addition to being studied for the early detection of AKI, NGAL also has been investigated for differential diagnosis and to predict progression of clinical disease. Singer and colleagues investigated 161 hospitalized patients with established AKI via RIFLE criteria. After 16 patients were excluded due to post-renal obstruction or insufficient information, the remainder were clinically adjudicated as having intrinsic AKI (75), prerenal AKI (32), or indeterminant (38). Urinary NGAL discriminated between intrinsic and prerenal with an AUC of 0.87 (NGAL values in the indeterminate group were very similar to prerenal). An NGAL level >104 μg/L resulted in a likelihood ratio (LR) for intrinsic of 5.97, whereas a level <47 μg/L rendered intrinsic unlikely, LR 0.2. NGAL values were higher in patients who would meet a composite endpoint of progression to a higher RIFLE stage, dialysis, or death than in those who did not. A sampling of the ability of NGAL and other urinary biomarkers to predict prognosis is presented in Table 47.2 .
|Urine IL-18 (77)||Pediatric cardiac surgery||AUC 0.72 for AKI progression|
|Urine IL-18 (78)||Cardiac surgery||AUC 0.91 for AKI progression|
|Urine IL-18 (79)||All hospitalized patients||AUC 0.68 for AKI progression, dialysis, or death|
|Urine IL-18 (50)||ICU||>100 rg/mL, HR, 1.6 (95% CI, 0.8–2.9) for death |
>200 rg/mL, HR, 2.3 (95% CI, 1.2–4.4) for death
>500 rg/mL, HR, 5.1 (95% CI, 2–13.1) for death
|Urine IL-18 (80)||Cardiac surgery||>185 rg/mL, OR, 3 (95% CI, 1.3–7.3) for AKI progression|
|Urine IL-18 (81)||ICU||AUC 0.64 for AKI progression, dialysis, or death|
|Urine IL-18 (82)||Cardiorenal syndrome||OR, 3.6 (95% CI, 1.4–9.5) for AKI progression|
|Urine IL-18 (83)||AKI in cirrhosis||AUC = 0.71 for AKI progression or death|
|Urine NGAL (82)||Cardiorenal syndrome||OR, 4.7 (95% CI, 1.7–13.4) for AKI progression|
|Urine NGAL (81)||ICU||AUC = 0.65 for AKI progression, dialysis, or death|
|Urine NGAL (79)||All hospitalized patients||AUC = 0.75 for AKI progression, dialysis, or death|
|Urine NGAL (83)||AKI in cirrhosis||AUC = 0.77 for AKI progression or death|
|Urine NGAL (77)||Pediatric cardiac surgery||AUC = 0.71 for AKI progression|
|Urine NGAL (84)||ICU||AUC = 0.70 for AKI progression or dialysis|
|Urine KIM-1 (79)||All hospitalized patients||AUC = 0.69 for AKI progression, dialysis, or death|
|Urine KIM-1 (83)||AKI in cirrhosis||AUC = 0.66 for AKI progression or death|
|Urine KIM-1 (78)||Cardiac surgery||AUC = 0.70 for AKI progression|
|Urine KIM-1 (79)||All hospitalized patients||AUC = 0.70 for AKI progression|
|Urine KIM-1 (81)||ICU||AUC = 0.62 for AKI progression, dialysis, or death|
|Urine L-FABP (83)||AKI in cirrhosis||AUC = 0.76 for AKI progression or death|
|Urine L-FABP (81)||ICU||AUC = 0.79 (95% CI, 0.70–0.86) for progression, dialysis, or death|
|Urine L-FABP (86)||Cardiac surgery||AUC = 0.78 for nonrecovery of kidney function|
|Urinary YKL-40 (87)||All hospitalized patients||OR, 3.4 (95% CI, 1.5–7.7) for AKI progression or death|
|Urine [TIMP-2]×[IGFBP7] (88)||ICU||>2, HR, 2.16 (95% CI, 1.3–3.5) for death or dialysis|
Multiple studies have evaluated urinary biomarkers, including NGAL, as a means of distinguishing between ATN and hepatorenal syndrome (HRS) in patients with cirrhosis. In one of the largest, Belcher and colleagues investigated 110 patients with cirrhosis and AKI, 55 adjudicated as having prerenal azotemia (PRA), 16 with HRS, and 39 with ATN. Median NGAL values for PRA (54 ng/mL) and HRS (115 ng/mL) were significantly less than for ATN (565 ng/mL). At a cutoff of 365 ng/mL, NGAL had an AUC of 0.78 (95% CI, 0.69 to 0.88) to identify ATN. A metaanalysis identified eight studies comprising 1129 patients with cirrhosis and AKI which assessed NGAL for differential diagnosis and prognosis. NGAL performed extremely well in identifying patients with ATN, AUC of 0.89 (95% CI, 0.84 to 0.94) as well as predicting 90-day mortality, AUC of 0.76 (95% CI, 0.71 to 0.82).
Several studies have evaluated urinary NGAL levels in the immediate postoperative period to predict DGF in kidney transplant recipients. Parikh and colleagues found NGAL obtained on day 0 in recipients of kidneys from living or deceased donors performed well in predicting DGF (AUC = 0.9). In a follow-up study, NGAL was able to distinguish between patients who would go on to have immediate graft function, slow graft function, and DGF and could predict at 6 hours postoperatively which patients would require dialysis, AUC of 0.81 (95% CI, 0.70 to 0.92). Overall, NGAL represents a promising candidate as a biomarker for the early diagnosis, differential diagnosis, and prognosis of AKI.
Liver Type Fatty Acid-Binding Protein
L-FABP, also known as fatty acid-binding protein 1 (FABP1) is a 14-kDa protein encoded by the FABP1 gene in humans. It is expressed not only in the liver but also in the lung, intestine, stomach, and kidney. L-FABP binds and transports fatty acids to mitochondria and peroxisomes and also functions to mitigate H 2 O 2 -associated oxidative stress. In the kidney, L-FABP is expressed in the proximal tubule. Urinary levels of L-FABP are shown to be significantly upregulated in transgenic rodent models of kidney injury, including ischemia-reperfusion, cisplatin, folic acid, adenine, and cephaloridine. The levels of L-FABP were markedly elevated as early as 2 hours after cisplatin administration, whereas a rise in creatinine was not detectable until after 72 hours of cisplatin treatment.
Portilla and colleagues reported that in humans, urinary L-FABP levels at 4 hours after surgery were an independent risk indicator for AKI (defined as a 50% increase in creatinine over baseline) with an AUC of 0.81, sensitivity of 0.71, and specificity of 0.68 for a 24-fold increase in urinary L-FABP. Increases in the urinary levels of L-FABP also have been found to predict renal outcomes in idiopathic membranous nephropathy patients with a calculated sensitivity and specificity of 81% and 83%. In a recent study of 12 living kidney transplant related patients, immediately after reperfusion of their transplanted organs, a significant direct correlation was found between urinary L-FABP levels and both peritubular capillary blood flow and the ischemic time of the transplanted kidney. A systematic review and metaanalysis by Susantitahphong and colleagues identified 23 studies evaluating L-FABP and AKI spanning cardiac surgery, cardiac catheterization, ICUs, emergency departments, general hospital wards, liver transplantation, and hematopoietic cell transplantation. Unfortunately, only 7 could be included in a metaanalysis. L-FABP was found to have sensitivity and specificity for AKI diagnosis of 75% and 78%, 69% and 43% for predicting the need for dialysis, and 93% and 79% for in-hospital mortality. More recently, Parr and associates investigated several biomarkers in 152 ICU patients to predict a composite outcome of progression of AKI, dialysis, or death. L-FABP was the only biomarker with a strong ability to predict the composite outcome (AUC = 0.79 compared with NGAL, IL-18, and KIM-1 which were 0.65, 0.64, and 0.62, respectively) and the only one to improve the discrimination when added to a clinical model. The levels of urinary L-FABP should be interpreted with caution, however, as they may be influenced by a number of preexisting renal diseases, such as early diabetic nephropathy, nondiabetic CKD, polycystic kidney disease, and idiopathic focal glomerulosclerosis. Because L-FABP is expressed in other organs such as the liver, urinary L-FABP may lose specificity for kidney disease when there is coexisting liver disease. Nevertheless, L-FABP remains an intriguing candidate biomarker.
Kidney Injury Molecule-1
KIM-1 is a type I cell membrane glycoprotein, which contains in its extracellular portion a novel six-cysteine immunoglobulin-like domain, two N -glycosylation sites, and a Thr/Ser/Pro-rich domain characteristic of mucin-like O -glycosylated proteins. KIM-1 was initially discovered in rodents when, after an acute ischemic kidney injury, Kim-1 (the rodent counterpart to human KIM-1) mRNA was found to be significantly upregulated in proximal tubules cells 24 to 48 hours postinjury. The ectodomain of KIM-1 sheds from cells both in vitro and in vivo into the urine in rodents and humans after proximal tubular kidney injury. Elevated levels of soluble KIM-1 ectodomain in the urine also were demonstrated in patients with renal cell carcinoma. KIM-1 appears to act as a scavenger receptor on renal epithelial cells, which converts the normal proximal tubule cell into a phagocyte, facilitates the clearance of dead cells in the lumen, and likely plays an important role in the innate immune response after injury.
KIM-1 is constitutively expressed at very low levels in the kidney as well as other organs, but its expression is significantly upregulated in the kidney after both ischemia-reperfusion and drug-induced AKI. Characteristics of KIM-1 commending it as a biomarker include much higher expression in proximal tubular cells of the kidney than in any other cell of the kidney or any other organ, stability of the soluble ectodomain in the urine over a broad range of pH, sustained expression in proximal tubular epithelial cells until complete recovery, and extremely low levels in the healthy kidney, providing a high signal-to-noise ratio. In one of the first studies linking urinary levels of KIM-1 and AKI, an increased expression of KIM-1 in kidney biopsy specimens was demonstrated from patients with a pathological diagnosis of ATN along with corresponding levels of KIM-1 ectodomain in the urine. Notably, KIM-1 appeared in the urine before the appearance of casts. A subsequent study revealed that KIM-1 expression in transplant was significantly correlated with levels of creatinine and inversely correlated with estimated GFR on the biopsy day. KIM-1 was expressed focally in affected tubules in 92% of kidney biopsies from patients with acute cellular rejection. Focal positive KIM-1 expression was found in 28% of protocol biopsies in the presence of no detectable tubular injury on histological examination. This observation demonstrates the superior sensitivity of KIM-1 expression in detecting proximal tubule injury compared with morphology alone. The clinical significance, however, of increased KIM-1 in this setting remains to be determined. Van Timmeren and coworkers also found that occurrence of renal allograft loss over time increased with rising levels of KIM-1 excretion measured at baseline. High KIM-1 levels were associated with low creatinine clearance, proteinuria, and high donor age. KIM-1 levels predicted graft loss independent of creatinine clearance, proteinuria, and donor age. The pathological basis for this association may have been demonstrated in another study by the same authors where it was found that, in 102 transplant biopsies, the amount of KIM-1 protein expression in proximal tubule cells correlated with tubulointerstitial fibrosis and inflammation.
A number of studies have demonstrated the potential of KIM-1 as a biomarker of AKI. In a cohort of 201 patients with clinically established AKI, Liangos and coworkers evaluated urinary KIM-1 and NAG in predicting adverse clinical outcomes and reported that elevated levels of urinary KIM-1 and NAG were significantly associated with the composite endpoint of death or dialysis, even after adjustment for disease severity or comorbidity. KIM-1 is also a sensitive marker of kidney injury in children undergoing cardiac surgery.
In a study of 1199 adults undergoing cardiac surgery, 407 (34%) developed AKI with 251 (62%) having an AKI duration of 1 to 2 days, 118 (29%) with duration 3 to 6 days, and 38 (9%) ≥7 days. Multiple urinary biomarkers collected 0 to 6 hours postoperatively were independently associated with duration of AKI including IL-18, KIM-1, NGAL, L-FABP, and albumin. Of these, KIM-1 had the highest odds ratio at 1.36 (95% CI, 2.21 to 2.52). Additional prognostic data was found in a study by Jungbauer and associates of 149 patients with chronic heart failure and CKD. Over a 5-year follow-up, 26 (19%) patients developed progression of their CKD. Urinary KIM-1, along with NAG, was associated with CKD progression, and KIM-1 was independently associated with a combined endpoint of CKD progression or death.
In a novel study of 22 marathon runners, Mansour and colleagues investigated multiple urinary injury biomarkers including KIM-1, tumor necrosis factor-α, IL-18, IL-6, IL-8, and NGAL along with putative repair biomarkers YKL-40 and MCP-1 in the periods before and after the race. SCR rose significantly from a median of 0.81 mg/dL (interquartile range [IQR] = 0.76 to 0.95) premarathon to 1.28 mg/dL (IQR = 1.09 to 1.54) immediately after the marathon. By the following day, however, this had returned to baseline, 0.90 mg/dL (IQR = 0.80 to 0.90). All urinary injury and repair biomarkers demonstrated a statistically significant rise after the race from before it but only KIM-1 remained significantly elevated the following day; premarathon 132.6 pg/mL (IQR = 67.6 to 220), after the marathon 723.3 pg/mL (IQR = 459.4 to 970.6), following day 702.4 pg/mL (IQR = 123.3 to 1009; P < 0.001).
KIM-1 has primarily been measured by enzyme-linked immunosorbent assay. Importantly, although not yet in widespread clinical practice, a point-of-care dipstick has been developed that can provide semiquantitative results in 15 minutes. This is critical for studies going forward as biomarker research has been consistently hamstrung by the requirement of specialized laboratories for measuring levels.
In recognition of its strong performance relative to traditional biomarkers of kidney injury in preclinical biomarker FDA and European Medicines Agency qualification studies, KIM-1 has been approved by the FDA as an AKI biomarker for preclinical drug development.
Tissue Inhibitor of Metalloproteinase 2 and Insulin-Like Growth Factor-Binding Protein 7
TIMP2 and IGFBP7 are two proteins involved in G1 cell cycle arrest during the very early phases of cellular injury. The cyclin-dependent kinase inhibitor p21 halts cell cycle succession from G1 to S phase. P21-deficient mice have been demonstrated to be more sensitive to cisplatin-induced AKI, develop a more severe injury and have increased mortality, implying that cell cycle arrest is critical both to preventing AKI and abrogating its sequelae. A discovery study identified TIMP2 and IGFBP7 out of 340 potential candidate markers. The authors then validated these biomarkers using the Sapphire cohort. In this study, TIMP2 and IGFBP7, along with [TIMP2]x[IGFBP7], outperformed numerous other biomarkers including NGAL, KIM-1, IL-18, L-FABP, and π-GST in predicting moderate to severe AKI within 12 hours of sample collection. Subsequent studies have confirmed the utility of [TIMP2]x[IGFBP7] to predict stage 2 and 3 AKI with AUC of 0.82 and 0.79, respectively. Gocze and colleagues measured [TIMP2]x[IGFBP7] in 107 surgical patients at high risk for AKI, of whom 45 (42%) developed AKI. The AUC for AKI was 0.85; for early use of RRT, it was 0.83; and for 28-day mortality, it was 0.77. In a multivariable model incorporating established perioperative risk factors, [TIMP2]x[IGFBP7] was the strongest predictor of AKI.
Koyner and colleagues again used the Sapphire cohort of ICU adults to evaluate the association between TIMP2 and IGFBP7 with long-term outcomes, assessing them with a composite outcome of death or need for RRT. Data was available for 692 patients, of whom 382 (55%) developed AKI and 217 (31%) met the composite endpoint. On univariate analysis, [TIMP2]x[IGFBP7] >2 was associated with an increased risk (hazard ratio [HR], 2.11; 95% CI, 1.37 to 2.23). In a significant landmark for biomarker clinical translation, the FDA has allowed marketing of the NephroCheck test, which measures [TIMP2]x[IGFBP7] to determine whether certain critically ill patients are at risk for developing moderate to severe AKI within 12 hours after testing.
OPN is a multifunctional 44 kDa phosphoprotein, widely distributed in a variety of tissue types ranging from bone to epithelial cells of the gastrointestinal (GI) tract, lungs, breast, salivary glands, inner ear, placenta, and kidneys. OPN has been associated with a variety of physiological and pathological functions, including bone modeling, immunity, cell adhesion, and migration. OPN is expressed in ureteric buds and few interstitial cells of fetal kidney, whereas in the normal adult kidney OPN is localized primarily to the distal tubule and thick ascending segments of the loop of Henle. During renal disease, however, OPN has been shown to be upregulated in all tubular segments, including proximal tubules. OPN is secreted in the urine and inhibits the formation of calcium oxalate kidney stones. OPN has been evaluated in various experimental models of kidney injury, and multiple studies have reported the detection of OPN mRNA in regenerating proximal tubules in renal ischemia models. The elevated levels of OPN are sustained for 7 days after the injury. In gentamicin-induced ATN in rodents, OPN is detected only in the cortical distal tubules during first few days of the injury, but is markedly elevated in regenerative proximal tubules after day 15.
Recent work has evaluated serum levels of OPN as predictive of survival and renal recovery in patients acutely requiring RRT. Lorsenzen and coworkers measured serum OPN (sOPN) in 109 ICU patients with AKI. Lower sOPN at RRT initiation was independently associated with 28-day survival (AUC = 0.82). Twenty-four of 69 surviving patients (35%) were still dependent on RRT 4 weeks after initiation. sOPN was significantly lower at initiation in those survivors who recovered renal function than in those who did not. In those with recovery, sOPN levels at 4 weeks were significantly lower than at initiation, whereas they were unchanged in those without recovery. In a subsequent study of 102 ICU patients with AKI requiring RRT, sOPN levels on RRT day 1 were evaluated for association with 60-day mortality. After adjusting for a robust panel of variables, including creatinine and urine output when coming off RRT, sOPN levels in the second and third tertiles had adjusted odds ratio for 60-day mortality of 2.3 and 5.3, respectively (AUC = 0.81). In this study, however, sOPN was not associated with renal recovery in survivors.
YKL-40, also known as chitinase-3-like-1 protein, is a 40kDA glycoprotein secreted by a variety of cells and is involved in the activation of the innate immune system, extracellular matrix remodeling, and angiogenesis. YKL-40 is intimately involved in endothelial dysfunction and atherosclerosis and it is independently associated with albuminuria among patients with type 1 and type 2 diabetes mellitus. YKL-40 modulates renal repair mechanisms after ischemic kidney injury and can act as an effective marker of renal injury in the setting of kidney transplantation. Puthumana and associates measured YKL-40 in 1301 deceased kidney donors and ascertained outcomes in the corresponding 2435 recipients, 756 of whom experienced DGF. Donors with existing AKI at the time of kidney recovery had higher urinary YKL-40 and more frequent ATN on procurement biopsy. Elevated donor urinary YKL-40 concentration was independently associated with reduced risk for DGF in both recipients of AKI donor kidneys (adjusted relative risk [aRR], 0.51; 95% CI, 0.32 to 0.80 for the highest versus lowest YKL-40 tertile) and recipients of non-AKI donor kidneys (aRR, 0.79; 95% CI, 0.65 to 0.97). In those patients who suffered GFR, elevated donor urinary YKL-40 concentration associated with higher 6-month estimated GFR (6.75 mL/min/1.73 m 2 ; 95% CI, 1.49 to 12.02) and lower risk for graft failure (aRR, 0.50; 95% CI, 0.27 to 0.94). These findings suggest that YKL-40 is produced in response to tubular injury and is independently associated with recovery from AKI and DGF.
Hall and associates measured YKL-40 249 hospitalized patients with AKI as defined by AKIN criteria. Seventy-two patients (29%) had progression of their AKI to a higher stage or death. Urinary YKL-40 collected on the first day of AKI was ≥5 ng/mL and had an adjusted odds ratio of 3.4 (95% CI, 1.5 to 7.7) for this combined outcome. In patients deemed at high risk for the combined outcome due to elevated NGAL levels, YKL-40 was able to partition these patients into moderate or very high-risk groups. When added to a clinical model, YKL-40 resulted in a continuous net reclassification improvement of 29% ( P = 0.04). In addition, it has been found that YKL-40 is associated with increased mortality among patients with coronary artery disease, type 2 diabetes mellitus, cardiac failure, and peripheral artery disease. Tatar and colleagues evaluated 100 patients with a median of 72 months posttransplant. Patients were divided into tertiles based on serum YKL-40 levels (<42, 42 to 69, >69 ng/mL). SCR increased across tertiles, 1.39, 2.03, 2.45 mg/dL, as did proteinuria, 0.34, 0.74, 1.22 g/g. YKL-40 levels were positively associated with C-reactive protein and negatively with serum albumin. In multivariate analysis, GFR, systolic blood pressure, and YKL-40 were independently associated with levels of proteinuria.
Uromodulin, also known as Tamm-Horsfall protein, is a glycoprotein expressed exclusively by renal tubular cells lining the thick ascending limb (TAL) of the loop of Henle. It is the only biomarker discussed in this chapter that is primarily produced in this location. Uromodulin has been implicated in the pathophysiology of multiple kidney diseases including polycystic kidney disease, UTIs, and nephrolithiasis. Uromodulin has been shown to confer protection to proximal tubules against AKI via a possible crosstalk between the two functionally distinct tubular segments. In a murine model of ischemia-reperfusion AKI, El-Achkar et al. demonstrated that although uromodulin expression in TAL is downregulated at the peak of injury, it is significantly upregulated 48 hours after. Uromodulin expression is redirected from the apical domain to the basolateral and interstitium. This corresponds with increased uromodulin in the serum but not in the urine. Compared with the wild-type, uromodulin (-/-) mice show significantly delayed renal recovery after ischemia-reperfusion, possibly due to reduced suppression by uromodulin of proinflammatory cytokines and chemokines such as MCP-1 during recovery. The redistribution of uromodulin in the TAL after AKI and its increased interstitial presence therefore may downregulate inflammatory signaling in neighboring proximal tubules, thereby augmenting kidney recovery. In a post hoc analysis of 218 patients undergoing on-pump cardiac surgery, Garimella and associates investigated the association between preoperative urinary uromodulin levels and postoperative AKI. Lower urinary uromodulin-to-creatinine ratio was associated with higher odds for AKI (adjusted odds ratio [aOR], 1.43; 95% CI, 0.99 to 2.07). The lowest uromodulin-to-creatinine ratio quartile also was associated with higher odds for AKI relative to the highest quartile (aOR, 2.43; 95% CI, 0.91 to 6.48). Each lower SD in uromodulin-to-creatinine ratio was associated with a higher adjusted mean peak sCR (0.07 mg/dL per SD; 95% CI, 0.02 to 0.13).
IL-6 and IL-10 are cytokines believed to play critical roles in the inflammatory process mediating the pathophysiology of AKI. IL-6 is one of the main proinflammatory cytokines, whereas IL-10 has anti-inflammatory and immunomodulatory functions related to the regulation of lymphocyte activity. Multiple studies both in children and adults have demonstrated that these cytokines may be potential biomarkers of AKI. In animal models, AKI itself, not just the proinflammatory state of sepsis or surgery, induces higher IL-6 and IL-10 levels. IL-6 and IL-10 also are known to play a key role in the systemic inflammatory response to cardiac surgery and CPB. Greenberg and colleagues evaluated 106 children 1 month to 18 years of age undergoing CPB as part of the TRIBE-AKI (Translational Research Investigating Biomarker Endpoints in AKI) cohort, assessing for an association between IL-6 and IL-10 and postoperative AKI. Stage 2/3 AKI, defined by at least a doubling of the baseline sCR concentration or dialysis, was diagnosed in 24 patients (23%). Preoperative serum IL-6 concentrations were significantly higher in patients with stage 2/3 AKI (median 2.6 pg/mL; IQR = 0.6 to 4.9 pg/mL) than in those without stage 2/3 AKI (median 0.6 pg/mL; IQR = 0.6 to 2.2 pg/mL; P =0.03). Adjusted for clinical and demographic variables, the highest preoperative IL-6 tertile was associated with a sixfold increased risk for stage 2/3 AKI compared with the lowest tertile (OR, 6.41; 95% CI, 1.16 to 35.35). IL-6 and IL-10 levels increased significantly after surgery, peaking postoperatively on day 1 but did not significantly differ between patients with and without stage 2/3 AKI. In a study of 960 adults undergoing cardiac surgery, preoperative concentrations of IL-6 and IL-10 were not significantly associated with AKI or mortality. However, elevated first postoperative IL-6 concentrations were significantly associated with higher risk for AKI, and the risk increased in a dose-dependent manner (second tertile aOR, 1.61; 95% CI, 1.10 to 2.36; third tertile aOR, 2.13; 95% CI, 1.45 to 3.13). Elevated first postoperative IL-10 concentration was significantly associated with higher risk for AKI (aOR, 1.57; 95% CI, 1.04 to 2.38) and lower risk for mortality (aHR, 0.72; 95% CI, 0.56 to 0.93).
With cytokines such as IL-6 and IL-10 perhaps not only serving as markers of AKI but also as mediators of outcomes, several studies have therefore looked at the effect on outcomes of lowering their levels with dialysis. Park et al. studied 212 patients with sepsis and AKI randomized to receive continuous venovenous hemodiafiltration (CVVHDF) at conventional (40 mL/kg/h) or high (80 mL/kg/h) doses. High-dose, but not conventional, CVVHDF reduced serum IL-6 and IL-10 levels. However, there was no difference in either overall survival or dialysis-free survival at 28 days. A smaller study by Kade and colleagues examined 28 patients receiving continuous venovenous hemodialysis using a high cutoff hemofilter. Cytokine levels were measured before the start of dialysis and after 24 hours of treatment. After 24 hours, IL-10 levels in the serum were significantly lower. Clearance of IL-6 was approximately four times higher than IL-10 yet IL-6 levels rose, significantly correlating with mortality. It remains to be seen whether using this method to remove IL-6 will alter patient outcomes.
Fractional Excretion of Sodium
The fractional excretion of sodium (FENa), defined as [(urine sodium × plasma creatinine)/plasma sodium ×x urine creatinine) × 100], has historically been used to distinguish between prerenal azotemia (functional AKI) and ATN (structural AKI). A FENa <1% indicates both increased sodium avidity and tubular integrity sufficient to meet sodium reabsorptive demands, whereas a value >2% to 3% suggests frank tubular injury. However, in some settings of extreme sodium avidity, such as advanced cirrhosis, virtually all patients, whether they have AKI or not, have a FENa <1% and so the test has historically not been employed. Several recent studies have investigated whether it still might have utility if a different cutoff is used. In a study by Fagundes and associates of 74 patients with cirrhosis and prerenal azotemia (16), CKD (14) HRS (33), and ATN (11), FENa varied significantly across groups and was lowest in HRS (0.2%) and highest with ATN (2.2%). Verna and colleagues studied 118 patients with cirrhosis, 17 of whom had prerenal azotemia, 20 HRS syndrome, and 15 with ATN. FENa was once again significantly lower in the HRS patients (0.17%) than in ATN (0.7%). Similarly, Belcher and associates investigated 110 patients with cirrhosis and AKI who were diagnosed with prerenal azotemia (55), HRS (16), or ATN (39). Once again FENa was significantly lower for patients with HRS (0.1%) than for those with ATN (0.31%). Table 47.3 presents FENa values for different diagnoses across studies for patients with cirrhosis.