**1. Introduction**

Successive generations of scientists and nephrologists have failed to prevent or cure Acute Kidney Injury (AKI) and thousands every year die because of this. Recent innovations in proteomics and genomics have brought hope and renewed interest in preventing this blight. Motivated by high incidence and lack of effective treatments, researchers have focussed on how to detect AKI early in the disease process so as to provide the maximum opportunity for early intervention and positive outcomes. AKI Incidence is greatest in the intensive care, at about 11-52% in larger studies (n>500) (Ahlstrom et al. (2006); Bagshaw et al. (2008); Cruz et al. (2007)). Cardiac surgery and procedures involving radiocontrast pose smaller, but significant, risk of AKI with an incidence of 3-15% depending on cohort (Harjai et al. (2008); Lassnigg et al. (2008)). From 13 studies the mortality with AKI was 31.2% and was associated with an increase in relative risk of death from 2.40 to 6.15 depending on AKI severity (Ricci et al. (2008)). Stimulating much recent research has been the discovery of new kidney injury biomarkers, some of which appear to have sufficient sensitivity and specificity to be clinically useful.

This chapter will outline the history of the development of the concepts of clearance, acute renal failure, and acute kidney injury. This history provides the context for the current clearance based AKI diagnostic paradigm. The discovery of novel kidney injury biomarkers is challenging that paradigm. We will discuss the nature of that challenge and the opportunity it provides for development of early intervention treatments. All epidemiology, biomarker studies and clinical trials rely on tools to quantify AKI and assess efficacy of diagnostic or treatment efficacy. In section 3 we will discuss those tools before moving on to considering how they may best be applied in practice (section 4).

### **2. From ARF to AKI**

#### **2.1 Clearance and the rise and fall of creatinine**

While suppression of urine flow, ischuria renalis, was recognised as a fundamental manifestation of renal disease from the 17th Century, clear metabolic manifestations of AKI

Classical measurement of clearance with timed urine collection is cumbersome, so logically, the surrogate of clearance, the serum creatinine (actually measured in plasma), is usually the sole measure used to define renal status including development of AKI. However, serum creatinine increases slowly in response to a single step alteration in GFR. Serum creatinine has a half-life of approximately 4 hours when GFR is normal and 77 hours when GFR is reduced to 5% (Chiou & Hsu (1975); see section 3.1). As 3 to 5 half-lives are required after any change in GFR to obtain a new steady-state estimate, a reliable GFR based on the serum creatinine will require at least 12 hrs after even a minimal change in GFR. The current consensus definition (RIFLE: Risk, Injury, Failure, Loss, End-stage) of AKI requires at least a 33% decline in GFR, but a 25% reduction is traditionally accepted for diagnosis of contrast-induced AKI) (Bagshaw et al. (2008); Endre et al. (1989); Pickering & Endre (2009a)). Thus a new steady-state creatinine-based estimate of GFR will take from 24 to 72 hours. Obviously, the extent of increase in creatinine, which is determined by the extent of initial decrease in GFR and creatinine production, may allow diagnosis of AKI prior to the time needed to reach steady-state. However, the limited precision of creatinine measurement and the extent of intra-patient variation mean that a minimum 10% change in creatinine has traditionally been required by clinicians to demonstrate measurably significant change. An increase in serum creatinine has also been used as a major trigger for intervention (renal replacement therapy, RRT) in both AKI and chronic kidney disease (Gibney et al. (2008)). The recognised imprecision in determining change in GFR led to removal of GFR from the new Acute Kidney Injury Network (AKIN) consensus classification of AKI so that an absolute or percentage increase in creatinine alone or in combination with oliguria has become the new consensus definition of AKI (Mehta et al. (2007)). Thus, despite many limitations, such as dependence on muscle mass, diet etc (Perrone et al. (1992)), serum creatinine became the accepted shorthand for estimating GFR and significant change in creatinine became the

The Metamorphosis of Acute Renal Failure to Acute Kidney Injury 127

Since there is an inverse relationship between serum creatinine and GFR, it is easy to forget that a small increase in serum creatinine represents a substantial decline in GFR. Consequently, prior to the now widespread reporting of eGFR estimated from creatinine, clinicians often ignored small increases in creatinine, and characterised these as "mild" or "moderate" increases. Combined with the uncertainty associated with the precision of measurement, years of ignoring such small increases in creatinine have impaired insight into how AKI is triggered. Even if an increase is observed, the delay required for diagnosis creates uncertainty regarding the timing of the renal insult leading to AKI. This represents a lost opportunity to investigate and identify the underlying pathophysiology of AKI in humans. Even mild grades of severity of AKI and transient (less than 24 hour increases in creatinine) are associated with increased hospital mortality (Chertow et al. (2005); Uchino et al. (2010;

Inevitably, much of our interpretation of the pathophysiology of AKI is based on an animal model; the classical model utilises temporary cessation of renal blood flow to induce injury, usually through bilateral renal artery clamping. While this model may parallel human AKI after aortic surgery or renal transplantation, it greatly exaggerates the degree of tubular injury compared with that in the limited number of available human renal biopsies and is probably less relevant to AKI that does not follow hypoperfusion (Heyman et al. (2010)). Even in this model there is limited understanding of how major pathophysiological events in AKI are integrated, for example, the mechanism and timing of the switching "off" and "on" of

definition of AKI and marker of AKI severity.

2006)).

were not documented until World War I in the German and World War II in the English literature (see Eknoyan (2002); McGrath (1852)). The term "Acute Renal Failure" first appeared in the literature in 1946 (Frank et al. (1946)), although it has been attributed to Homer Smith (Eknoyan (2002)). In keeping with a recent change in nomenclature we shall use the term "Acute Kidney Injury" (AKI) unless we are specifically referring to an historic use of ARF.

The historical development of clearance techniques and the relationship to glomerular filtration rate (GFR) are discussed by Berliner in his tribute to the great renal physiologist Homer Smith (Berliner (1995)). The term clearance was introduce in 1928 with reference to urea and to clearance of a defined volume of plasma in unit time (Moller et al. (1928)). The idea of creatinine clearance as a measure of glomerular filtration rate (GFR) was beautifully first demonstrated by Rehberg in experiments on himself (Rehberg (1926)). These also highlighted how variations in serum creatinine could be induced by diet: Rehberg ingested different quantities of creatinine to vary his serum creatinine. The utility of clearance as a technique was further established in the laboratory of Homer Smith, especially with para-amminohippuric acid clearance, a measure of secretion and renal blood flow (Smith et al. (1945)).

Creatinine is formed non-enzymatically from creatine in muscle, has a molecular weight of 113.12 Da, is freely filtered at the glomerulus and completely cleared by renal excretion when renal function is normal. The proximal tubules secrete creatinine, which accounts for 10 to 20% of the excreted load, and results in overestimation of GFR when measured by creatinine clearance (Perrone et al. (1992); Shemesh et al. (1985)). The contribution of tubular creatinine secretion to clearance, is increased when GFR is reduced and may reach 50%, but is highly variable amongst individuals (Perrone et al. (1992)). In contrast, the tubules reabsorb creatinine in some clinical settings such as decompensated heart failure and uncontrolled diabetes (Levinsky & Berliner (1959); Perrone et al. (1992)).

With creatinine clearance firmly established as a reasonable approximation to GFR, the next step was to estimate creatinine clearance based on the reciprocal relationship with plasma creatinine. This was popularized by the Cockroft-Gault equation which was derived by, firstly, a regression to estimate creatinine excretion/kg body weight according to age in hospitalised male patients; then clearance was calculated by multiplying by weight and dividing by the serum creatinine (Cockcroft & Gault (1976)). An untested 15% reduction for female gender was included, based on the observation that, on average, females had less fat and muscle mass than males (Cockcroft & Gault (1976)). This formula has been widely replaced by other estimates of GFR (eGFR), most notably by the Modification of Diet in Renal Disease (MDRD) equation originally developed in a population of CKD patients (Levey et al. (1999)).

Many alternative algorithms for creatinine-based eGFR have been developed, including those regularly in use for children; these equations are more accurate and precise than estimates from measurement of creatinine alone (*KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification* (2002)). The various iterations of the MDRD equation rely on plasma creatinine plus several other variables, including gender and race, and originally albumin and urea (as blood urea nitrogen) concentrations, but excluding mass although they do incorporate body surface area via the units; the latest version is more accurate than the MDRD equation for patients with GFR>60 ml/min (Levey et al. (2009)) but caveats remain (Rule (2010)) and drug dosing tends to be based on Cockroft-Gault (Ryzner (2010)).

2 Will-be-set-by-IN-TECH

were not documented until World War I in the German and World War II in the English literature (see Eknoyan (2002); McGrath (1852)). The term "Acute Renal Failure" first appeared in the literature in 1946 (Frank et al. (1946)), although it has been attributed to Homer Smith (Eknoyan (2002)). In keeping with a recent change in nomenclature we shall use the term "Acute Kidney Injury" (AKI) unless we are specifically referring to an historic

The historical development of clearance techniques and the relationship to glomerular filtration rate (GFR) are discussed by Berliner in his tribute to the great renal physiologist Homer Smith (Berliner (1995)). The term clearance was introduce in 1928 with reference to urea and to clearance of a defined volume of plasma in unit time (Moller et al. (1928)). The idea of creatinine clearance as a measure of glomerular filtration rate (GFR) was beautifully first demonstrated by Rehberg in experiments on himself (Rehberg (1926)). These also highlighted how variations in serum creatinine could be induced by diet: Rehberg ingested different quantities of creatinine to vary his serum creatinine. The utility of clearance as a technique was further established in the laboratory of Homer Smith, especially with para-amminohippuric

Creatinine is formed non-enzymatically from creatine in muscle, has a molecular weight of 113.12 Da, is freely filtered at the glomerulus and completely cleared by renal excretion when renal function is normal. The proximal tubules secrete creatinine, which accounts for 10 to 20% of the excreted load, and results in overestimation of GFR when measured by creatinine clearance (Perrone et al. (1992); Shemesh et al. (1985)). The contribution of tubular creatinine secretion to clearance, is increased when GFR is reduced and may reach 50%, but is highly variable amongst individuals (Perrone et al. (1992)). In contrast, the tubules reabsorb creatinine in some clinical settings such as decompensated heart failure and uncontrolled

With creatinine clearance firmly established as a reasonable approximation to GFR, the next step was to estimate creatinine clearance based on the reciprocal relationship with plasma creatinine. This was popularized by the Cockroft-Gault equation which was derived by, firstly, a regression to estimate creatinine excretion/kg body weight according to age in hospitalised male patients; then clearance was calculated by multiplying by weight and dividing by the serum creatinine (Cockcroft & Gault (1976)). An untested 15% reduction for female gender was included, based on the observation that, on average, females had less fat and muscle mass than males (Cockcroft & Gault (1976)). This formula has been widely replaced by other estimates of GFR (eGFR), most notably by the Modification of Diet in Renal Disease (MDRD)

equation originally developed in a population of CKD patients (Levey et al. (1999)).

Many alternative algorithms for creatinine-based eGFR have been developed, including those regularly in use for children; these equations are more accurate and precise than estimates from measurement of creatinine alone (*KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification* (2002)). The various iterations of the MDRD equation rely on plasma creatinine plus several other variables, including gender and race, and originally albumin and urea (as blood urea nitrogen) concentrations, but excluding mass although they do incorporate body surface area via the units; the latest version is more accurate than the MDRD equation for patients with GFR>60 ml/min (Levey et al. (2009)) but caveats remain (Rule (2010)) and drug dosing tends to be based on Cockroft-Gault (Ryzner

acid clearance, a measure of secretion and renal blood flow (Smith et al. (1945)).

diabetes (Levinsky & Berliner (1959); Perrone et al. (1992)).

use of ARF.

(2010)).

Classical measurement of clearance with timed urine collection is cumbersome, so logically, the surrogate of clearance, the serum creatinine (actually measured in plasma), is usually the sole measure used to define renal status including development of AKI. However, serum creatinine increases slowly in response to a single step alteration in GFR. Serum creatinine has a half-life of approximately 4 hours when GFR is normal and 77 hours when GFR is reduced to 5% (Chiou & Hsu (1975); see section 3.1). As 3 to 5 half-lives are required after any change in GFR to obtain a new steady-state estimate, a reliable GFR based on the serum creatinine will require at least 12 hrs after even a minimal change in GFR. The current consensus definition (RIFLE: Risk, Injury, Failure, Loss, End-stage) of AKI requires at least a 33% decline in GFR, but a 25% reduction is traditionally accepted for diagnosis of contrast-induced AKI) (Bagshaw et al. (2008); Endre et al. (1989); Pickering & Endre (2009a)). Thus a new steady-state creatinine-based estimate of GFR will take from 24 to 72 hours. Obviously, the extent of increase in creatinine, which is determined by the extent of initial decrease in GFR and creatinine production, may allow diagnosis of AKI prior to the time needed to reach steady-state. However, the limited precision of creatinine measurement and the extent of intra-patient variation mean that a minimum 10% change in creatinine has traditionally been required by clinicians to demonstrate measurably significant change. An increase in serum creatinine has also been used as a major trigger for intervention (renal replacement therapy, RRT) in both AKI and chronic kidney disease (Gibney et al. (2008)).

The recognised imprecision in determining change in GFR led to removal of GFR from the new Acute Kidney Injury Network (AKIN) consensus classification of AKI so that an absolute or percentage increase in creatinine alone or in combination with oliguria has become the new consensus definition of AKI (Mehta et al. (2007)). Thus, despite many limitations, such as dependence on muscle mass, diet etc (Perrone et al. (1992)), serum creatinine became the accepted shorthand for estimating GFR and significant change in creatinine became the definition of AKI and marker of AKI severity.

Since there is an inverse relationship between serum creatinine and GFR, it is easy to forget that a small increase in serum creatinine represents a substantial decline in GFR. Consequently, prior to the now widespread reporting of eGFR estimated from creatinine, clinicians often ignored small increases in creatinine, and characterised these as "mild" or "moderate" increases. Combined with the uncertainty associated with the precision of measurement, years of ignoring such small increases in creatinine have impaired insight into how AKI is triggered. Even if an increase is observed, the delay required for diagnosis creates uncertainty regarding the timing of the renal insult leading to AKI. This represents a lost opportunity to investigate and identify the underlying pathophysiology of AKI in humans. Even mild grades of severity of AKI and transient (less than 24 hour increases in creatinine) are associated with increased hospital mortality (Chertow et al. (2005); Uchino et al. (2010; 2006)).

Inevitably, much of our interpretation of the pathophysiology of AKI is based on an animal model; the classical model utilises temporary cessation of renal blood flow to induce injury, usually through bilateral renal artery clamping. While this model may parallel human AKI after aortic surgery or renal transplantation, it greatly exaggerates the degree of tubular injury compared with that in the limited number of available human renal biopsies and is probably less relevant to AKI that does not follow hypoperfusion (Heyman et al. (2010)). Even in this model there is limited understanding of how major pathophysiological events in AKI are integrated, for example, the mechanism and timing of the switching "off" and "on" of

markers. The first, a trial of 71 children undergoing cardiopulmonary bypass of whom 20 developed AKI showed urinary and plasma NGAL to increase 10-fold in AKI 2-h post surgery (Mishra et al. (2005)). Similarly, interluekin-18, IL-18, first identified as being released into the urine following I/R injury in mice (Melnikov et al. (2001)) was found to appear in large quantities in the urine of patients with acute tubular necrosis (Parikh et al. (2004)). Proteomic and genomic approaches continue to be a rich source of new urinary proteins

The Metamorphosis of Acute Renal Failure to Acute Kidney Injury 129

AKI Biomarker discovery has followed the well trodden path of "early promise" with some highly sensitive and specific biomarkers in demographically homogeneous populations (eg paediatric cardiopulmonary bypass surgery), followed by a more tempered response as studies evaluated candidate biomarkers in demographically heterogeneous populations with multiple causes of AKI and co-morbidities. There is no one biomarker which will successfully diagnose or predict a decline in renal function in all situations. Three factors have emerged which are likely to become determinant factors in the choice of biomarker for a particular clinical context, namely: *(i)* likely aetiology of AKI, *(ii)* pre-existing renal function, *(iii)* time from renal insult. IL-18 is an example of a biomarker which has been shown to be elevated following ischaemic/repurfusion injury (Hall et al. (2010); Parikh et al. (2006)), but possibly not following radiocontrast induced nephrotoxic injury (Bulent Gul et al. (2008)). Mcilroy et al demonstrated that in a cohort of 426 adult cardiac surgery patients urinary NGAL concentrations post-operatively did not differ between those who developed and those who did not develop AKI when their estimated baseline GFR (eGFR) was less than 60 ml/min, yet for those with a normal baseline eGFR (90-120 ml/min) NGAL was significantly elevated in the AKI cohort (Mcilroy et al. (2010)). In our own head to head comparison of 6 urinary biomarkers (GGT, AP, NGAL, Cystatin C, IL-18, and KIM-1) in 529 adult patients on entry to an intensive care unit, we demonstrated that the performance of biomarkers is critically dependent on both baseline renal function and time from renal insult (Endre et al. (2011)). Peak diagnostic performance at a level that may be considered clinically useful was limited to patients with eGFR 90-120 ml/min within 12-h of insult for GGT, 6-h for NGAL and from 6 to 12-h for Cystatin C, IL-18 and KIM-1, and to patients with eGFR < 60 ml/min from 12 to 36-h

There are many excellent recent reviews of biomarkers of AKI. The pathophysiology of AKI in relation to potential biomarkers has been reviewed and discussed specifically in relation to AKI following cardiopulmonary bypass (Haase et al. (2010)), AKI with varying aetiology (Vaidya et al. (2008)), and AKI involving biomarker mediators of inflammation (Akcay et al. (2009)). Reviews of biomarker performance specific to nephrotoxic injury (Bonventre et al. (2010); Ferguson et al. (2008)), septic AKI (Bagshaw et al. (2007)), ischemic injury following cardiopulmonary bypass (Haase et al. (2010)), acute allograft rejection and ischemic injury (Alachkar et al. (2010)) as well as broader reviews across aetiologies (Coca et al. (2008); Edelstein & Faubel (2010); Endre & Westhuyzen (2008); Malyszko (2010)) have been published within the last 4 years. We have published a more specialist review considering biomarkers in the early phase of injury (Pickering & Endre (2009c)) and there has been one meta-analysis

The relationship between clearance (the function paradigm) and injury has, to date, been studied in humans primarily in sample populations of about 20 to 600 in which a biomarker's

which may predict AKI (Bennett & Devarajan (2011); Devarajan (2008)).

for GGT, Cystatin C, NGAL and IL-18.

of the performance of NGAL (Haase et al. (2009)).

**2.3 Paradigm lost and paradigm found**

glomerular filtration. With few exceptions (Alejandro et al. (1995); Myers et al. (1984)), most of our experimental interventions have been validated only in this and nephrotoxic animal models (Vaidya et al. (2010)).

While there is a great deal of information about a large number of cellular (tubular and endothelial) events and the autonomic, inflammatory and renal vascular responses in experimental ischemia-reperfusion injury (Devarajan (2006); Heemskerk et al. (2009)), there is little corroborative and time-relevant clinical pathophysiological data. Usually, the clinical diagnosis of ischemic AKI is a diagnosis of exclusion. Thus, the potential delay imposed by reliance on creatinine is further delayed by investigations (eg exclusion of urinary outflow obstruction) or interventions (fluid loading to treat underlying "pre-renal" AKI) designed to exclude rather than confirm the diagnosis of ischemic AKI. Other investigations such as measurement of global or parenchymal renal blood flow, or renal biopsy, which might provide insight into human AKI, are delayed, difficult to interpret in the absence of baseline data, and usually not performed.

Since diagnosis is delayed it is not surprising, that there has been failure of pharmacological intervention in clinical trials, which are largely based on experimental interventions to prevent rather than treat AKI (Jo et al. (2007)). However, pharmacologic prevention of AKI has also largely failed, even when the apparent aetiology is known, eg parenteral administration of iodinated radiocontrast (Fishbane (2008); Nigwekar et al. (2009); Zoungas et al. (2009)). The failure to translate apparently effective pharmacologic preventive measures in animal models into clinical practice, suggests that the serum creatinine-inspired delay in diagnosis, merely complements a lack of fundamental understanding of pathophysiology in human AKI.

#### **2.2 The rise and rise of injury**

The change in nomenclature from acute renal failure to acute kidney injury recognised that an acute decline in renal function is usually secondary to injury (*American Society of Nephrology Renal Research Report* (2005); Mehta et al. (2007)). The need for a renal-specific biomarkers of injury akin to a troponin was presaged in the late 1990s (Star (1998)) and the discovery of such proteins was later accorded highest priority by The American Society of Nephrology (*American Society of Nephrology Renal Research Report* (2005)) with the expectation that such biomarkers would: "Diagnose AKI before the rise in serum creatinine; Stratify patients with respect to severity of injury and; Provide prognostic indicators." The term "secondary prevention" was introduced to highlight that biomarker detection would lead to early intervention ideally prior to loss of GFR (Pickering & Endre (2009c)).

Some urinary proteins, notably *α*1-microglobulin, *β*1-microglobulin, and N-acetyl-*β*-D-glucosaminidase (NAG), were already known to be associated with acute tubular injury (Yu et al. (1983)), whilst others awaited discovery. In 1998 Kidney Injury Molecule-1 (KIM-1) was identified as being upregulated in the proximal tubule cells after ischemic/reperfusion (I/R) injury (Ichimura et al. (1998)) and soon discovered in the urine of patients with I/R injury (Han et al. (2002)). Around the same time a small study of ICU patients identified urinary tubular injury makers, *α* and *π*-glutathone S-transferase (*α* and *π*-GST), and the bursh border enzymes, *γ*-glutamyl-transpeptidase (GGT) and alkaline phosphatase (AP), as diagnostic of AKI (Westhuyzen et al. (2003)). A transcriptome wide interrogation study to identify genes induced early after I/R injury identified neutrophil gelatinase-associated lipocalin (NGAL) to be upregulated in a mouse model (Mishra et al. (2003)). Plasma and urinary NGAL quickly showed spectacular success as diagnostic 4 Will-be-set-by-IN-TECH

glomerular filtration. With few exceptions (Alejandro et al. (1995); Myers et al. (1984)), most of our experimental interventions have been validated only in this and nephrotoxic animal

While there is a great deal of information about a large number of cellular (tubular and endothelial) events and the autonomic, inflammatory and renal vascular responses in experimental ischemia-reperfusion injury (Devarajan (2006); Heemskerk et al. (2009)), there is little corroborative and time-relevant clinical pathophysiological data. Usually, the clinical diagnosis of ischemic AKI is a diagnosis of exclusion. Thus, the potential delay imposed by reliance on creatinine is further delayed by investigations (eg exclusion of urinary outflow obstruction) or interventions (fluid loading to treat underlying "pre-renal" AKI) designed to exclude rather than confirm the diagnosis of ischemic AKI. Other investigations such as measurement of global or parenchymal renal blood flow, or renal biopsy, which might provide insight into human AKI, are delayed, difficult to interpret in the absence of baseline data, and

Since diagnosis is delayed it is not surprising, that there has been failure of pharmacological intervention in clinical trials, which are largely based on experimental interventions to prevent rather than treat AKI (Jo et al. (2007)). However, pharmacologic prevention of AKI has also largely failed, even when the apparent aetiology is known, eg parenteral administration of iodinated radiocontrast (Fishbane (2008); Nigwekar et al. (2009); Zoungas et al. (2009)). The failure to translate apparently effective pharmacologic preventive measures in animal models into clinical practice, suggests that the serum creatinine-inspired delay in diagnosis, merely complements a lack of fundamental understanding of pathophysiology in human AKI.

The change in nomenclature from acute renal failure to acute kidney injury recognised that an acute decline in renal function is usually secondary to injury (*American Society of Nephrology Renal Research Report* (2005); Mehta et al. (2007)). The need for a renal-specific biomarkers of injury akin to a troponin was presaged in the late 1990s (Star (1998)) and the discovery of such proteins was later accorded highest priority by The American Society of Nephrology (*American Society of Nephrology Renal Research Report* (2005)) with the expectation that such biomarkers would: "Diagnose AKI before the rise in serum creatinine; Stratify patients with respect to severity of injury and; Provide prognostic indicators." The term "secondary prevention" was introduced to highlight that biomarker detection would lead to

Some urinary proteins, notably *α*1-microglobulin, *β*1-microglobulin, and N-acetyl-*β*-D-glucosaminidase (NAG), were already known to be associated with acute tubular injury (Yu et al. (1983)), whilst others awaited discovery. In 1998 Kidney Injury Molecule-1 (KIM-1) was identified as being upregulated in the proximal tubule cells after ischemic/reperfusion (I/R) injury (Ichimura et al. (1998)) and soon discovered in the urine of patients with I/R injury (Han et al. (2002)). Around the same time a small study of ICU patients identified urinary tubular injury makers, *α* and *π*-glutathone S-transferase (*α* and *π*-GST), and the bursh border enzymes, *γ*-glutamyl-transpeptidase (GGT) and alkaline phosphatase (AP), as diagnostic of AKI (Westhuyzen et al. (2003)). A transcriptome wide interrogation study to identify genes induced early after I/R injury identified neutrophil gelatinase-associated lipocalin (NGAL) to be upregulated in a mouse model (Mishra et al. (2003)). Plasma and urinary NGAL quickly showed spectacular success as diagnostic

early intervention ideally prior to loss of GFR (Pickering & Endre (2009c)).

models (Vaidya et al. (2010)).

usually not performed.

**2.2 The rise and rise of injury**

markers. The first, a trial of 71 children undergoing cardiopulmonary bypass of whom 20 developed AKI showed urinary and plasma NGAL to increase 10-fold in AKI 2-h post surgery (Mishra et al. (2005)). Similarly, interluekin-18, IL-18, first identified as being released into the urine following I/R injury in mice (Melnikov et al. (2001)) was found to appear in large quantities in the urine of patients with acute tubular necrosis (Parikh et al. (2004)). Proteomic and genomic approaches continue to be a rich source of new urinary proteins which may predict AKI (Bennett & Devarajan (2011); Devarajan (2008)).

AKI Biomarker discovery has followed the well trodden path of "early promise" with some highly sensitive and specific biomarkers in demographically homogeneous populations (eg paediatric cardiopulmonary bypass surgery), followed by a more tempered response as studies evaluated candidate biomarkers in demographically heterogeneous populations with multiple causes of AKI and co-morbidities. There is no one biomarker which will successfully diagnose or predict a decline in renal function in all situations. Three factors have emerged which are likely to become determinant factors in the choice of biomarker for a particular clinical context, namely: *(i)* likely aetiology of AKI, *(ii)* pre-existing renal function, *(iii)* time from renal insult. IL-18 is an example of a biomarker which has been shown to be elevated following ischaemic/repurfusion injury (Hall et al. (2010); Parikh et al. (2006)), but possibly not following radiocontrast induced nephrotoxic injury (Bulent Gul et al. (2008)). Mcilroy et al demonstrated that in a cohort of 426 adult cardiac surgery patients urinary NGAL concentrations post-operatively did not differ between those who developed and those who did not develop AKI when their estimated baseline GFR (eGFR) was less than 60 ml/min, yet for those with a normal baseline eGFR (90-120 ml/min) NGAL was significantly elevated in the AKI cohort (Mcilroy et al. (2010)). In our own head to head comparison of 6 urinary biomarkers (GGT, AP, NGAL, Cystatin C, IL-18, and KIM-1) in 529 adult patients on entry to an intensive care unit, we demonstrated that the performance of biomarkers is critically dependent on both baseline renal function and time from renal insult (Endre et al. (2011)). Peak diagnostic performance at a level that may be considered clinically useful was limited to patients with eGFR 90-120 ml/min within 12-h of insult for GGT, 6-h for NGAL and from 6 to 12-h for Cystatin C, IL-18 and KIM-1, and to patients with eGFR < 60 ml/min from 12 to 36-h for GGT, Cystatin C, NGAL and IL-18.

There are many excellent recent reviews of biomarkers of AKI. The pathophysiology of AKI in relation to potential biomarkers has been reviewed and discussed specifically in relation to AKI following cardiopulmonary bypass (Haase et al. (2010)), AKI with varying aetiology (Vaidya et al. (2008)), and AKI involving biomarker mediators of inflammation (Akcay et al. (2009)). Reviews of biomarker performance specific to nephrotoxic injury (Bonventre et al. (2010); Ferguson et al. (2008)), septic AKI (Bagshaw et al. (2007)), ischemic injury following cardiopulmonary bypass (Haase et al. (2010)), acute allograft rejection and ischemic injury (Alachkar et al. (2010)) as well as broader reviews across aetiologies (Coca et al. (2008); Edelstein & Faubel (2010); Endre & Westhuyzen (2008); Malyszko (2010)) have been published within the last 4 years. We have published a more specialist review considering biomarkers in the early phase of injury (Pickering & Endre (2009c)) and there has been one meta-analysis of the performance of NGAL (Haase et al. (2009)).

#### **2.3 Paradigm lost and paradigm found**

The relationship between clearance (the function paradigm) and injury has, to date, been studied in humans primarily in sample populations of about 20 to 600 in which a biomarker's

**3. Defining AKIs**

**3.1 Creatinine and cystatin C kinetics**

volume of distribution, *V* (Figure 2).

*dq*

equation 1 becomes:

then equation 2 becomes:

constant may be determined from equation 3:

Fig. 2. One compartment pharmacokinetics model

and the rate at which it is leaving the compartment:

kinetics.

In its broadest sense renal failure is simply the rapid loss of renal filtration (ie decrease in GFR). As discussed, current clinical diagnosis of AKI and evaluation of novel biomarkers of kidney injury are largely dependent on observation of changes in creatinine as a surrogate for a change in GFR. An understanding of the kinetics of the relationship between creatinine and GFR provides a basis for understanding the definitions of AKI and their limitations. Plasma cystatin C, an alternative to creatinine (Nejat et al. (2010); Westhuyzen (2006)), follows similar

The Metamorphosis of Acute Renal Failure to Acute Kidney Injury 131

Creatinine and cystatin C are generated in tissue outside of the plasma compartment, diffuse into the plasma compartment from where they are lost by renal and non-renal excretion. Because both creatinine and cystatin C are not bound to plasma proteins the exchange between the extravascular and plasma compartments is rapid (compared with rates of production and elimination) allowing us to conflate the two compartments into one compartment with

The change in total mass (*q*) of creatinine/cystatin C depends on the rate at which it is entering

Under normal circumstances non-renal losses are much less than renal losses and may be ignored. The renal loss is the product of the renal elimination rate constant, *kr*, and the total mass, *q*. As the total mass is the product of the concentration *C* and volume of distribution *V*,

where *G* is the generation rate of creatinine/cystatin C (see Box 3.1). The volume of distribution of creatinine is equal to the total body water (*TBW*) and that of cystatin C to the extracellular fluid (about one third of *TBW*)(see Box 3.2). If this is assumed not to change

At equilibrium (*dC*/*dt* = 0 at *t* = 0), prior to any change in GFR the renal elimination rate

*kr*<sup>0</sup> <sup>=</sup> *<sup>G</sup>*

*dC*

*dC dt* <sup>=</sup> *<sup>G</sup>*

*<sup>C</sup> dV dt* <sup>+</sup>

*dt* <sup>=</sup> gain from generation <sup>−</sup> (renal loss <sup>+</sup> non-renal loss) (1)

*dt <sup>V</sup>* <sup>=</sup> *<sup>G</sup>* <sup>−</sup> *krCV* (2)

*<sup>V</sup>* <sup>−</sup> *krC* (3)

*CbV* (4)

ability to diagnose a subsequent rise in a surrogate marker of function (normally creatinine) is assessed. However, we need to explore the early stages of, and prior to, loss of renal function in order to understand the relevant pathophysiology of clinical AKI. We can conceptualise this as an evaluation of the evolution of injury phase which leads to loss of GFR and the early loss of function phase immediately following a GFR decrease (Figure 1). Detection and characterisation of this very early phase in man appears essential for progress. This highlights the need for quantifying the time course of injury biomarkers in relation to change in GFR and for real-time assessment of renal function. Two promising techniques are under development. The ambulatory renal monitor (ARM) is a shielded detector which monitors extracellular excretion of <sup>99</sup>*m*TcDTPA over up to 24 hours following injection and relates this to GFR (Rabito et al. (2010)). The ratiometric fluorescence approach monitors the plasma disappearance of both a rapidly filtered and a poorly filtered fluorescent marker introduced by a single bolus infusion, changes in the ratio of which enable a calculation of GFR (Wang et al. (2010)). Both techniques allow a rapid (5 to 15 min) measure of GFR. Ideally these or similar techniques would allow monitoring of kidney function from prior to the time of insult through the evolution time until change of GFR and beyond.

Fig. 1. The paradigm of function compared with the paradigm of injury. Following injury the early injury biomarker is elevated prior to change in function (here, a 65% loss of GFR, approximately 6 hours post-injury). The later biomarker takes takes longer to be elevated. Creatinine is only elevated following loss of GFR (following equation 6)

Much of the literature describes a biomarker as "predicting" AKI when it precedes an increase in plasma creatinine. However, a biomarker is only truly predictive if it precedes a decrease in GFR. If it is elevated shortly after such a decrease, then it should be described as *diagnostic* rather than predictive of AKI even though it may predict the latter increase in plasma creatinine.
