Biomarkers: Discovery, Qualification, and Application

1 Introduction

Biomarker discovery has rapidly progressed from a classical empirical approach to a high-priority, technology-supported research activity. According to the Food and Drug Administration's (FDA) Biomarkers, EndpointS and other Tools glossary ( https://www.ncbi.nlm.nih.gov/books/NBK326791/ ), a biomarker is a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions. Within this characterization, the types of measurements and observations that are included in the various categories of biomarkers are extremely broad in both technical attributes and application. Molecular, histologic, radiographic, or physiologic characteristics are types of biomarkers. In practice, many bioassay results that are now included in various biomarker categories were developed by clinical pathologists long before the terminology was used to describe them as such. In fact, clinical signs, hematology, clinical chemistry, and histopathology all fit within the definition of “biomarkers” in the hands of the toxicologic pathologist. One could argue that successful identification, interpretation, and application of many biomarkers are partially dependent upon the body of knowledge that has been obtained via histopathological evaluations. Histopathological interpretation is a critical standard that is often paired with development of safety biomarkers. Therefore, the toxicologic pathologist will continue to play a key role in providing the context, interpretation, and qualification of biomarkers.

1.1 Biomarker versus Surrogate

It is important to distinguish the term biomarker from that of surrogate endpoint or surrogate marker. Surrogate endpoints are considered a subset of biomarkers that serve as an indirect measure of clinical benefit. The benefit ascribed to the response of a surrogate endpoint can be used to support a traditional or accelerated approval of drug. Although all surrogate endpoints may be considered biomarkers, it is clear that only a few biomarkers will meet the rigorous scientific and regulatory requirements to achieve inclusion in this distinct subset. Since the term “surrogate” literally means “to substitute for,” the use of the term “surrogate marker” is discouraged. Readers are referred to the FDA's downloadable Table of Surrogate Endpoints for a complete listing that of clinical endpoints that is updated every 6 months ( https://www.fda.gov/drugs/development-resources/table-surrogate-endpoints-were-basis-drug-approval-or-licensure#top ).

1.2 Qualification versus Validation

Another key concept to introduce is the distinction between the term qualification and validation . The FDA describes biomarker qualification and validation as two distinct processes ( ). We define qualification as the measure of the evidence that will support the intended use or purpose for the biomarker. With this definition for qualification, a fit-for-purpose approach to biomarker qualification can be used. We use of the term “validation” to mean analytical validation, which is a measure of the performance characteristics of an assay that ensure that the assay is reliable, reproducible, and adequately sensitive for the intended use.

2 Categories of Biomarkers

Using the intended purpose of the biomarker as the defining characteristic, biomarkers may be broken into subsets as is provided in Table 14.1 . The narrative within this section will use specific examples (within text boxes) to illustrate the discovery process, qualification, and expectations for use and, if applicable, regulatory requirements that the biomarker would need to meet.

Table 14.1

Categories of Biomarkers According to Intended Purpose.

Biomarker purpose (category)	Expectation	Example
Patient/clinical trial subject selection	Predict a response to a molecularly targeted agent; has the potential to enrich clinical trials such that the patient pool includes patients more likely to benefit. For this purpose, the use of biomarkers intended to be prognostic or predictive biomarkers may be employed. This category is largely populated by genomic biomarkers	EGFR mutations in NSCLC are strongly associated with sensitivity to gefitinib and erlotinib and a higher rate of clinical benefit to treatment with EGFR tyrosine kinase inhibitors was demonstrated in patients with EGFR mutations than in those without such mutations
Pharmacogenomic/Genomic	A measurable DNA and/or RNA characteristic that indicates normal biologic processes, pathogenic processes, and/or response to therapeutic or other interventions	Dihydropyrimidine dehydrogenase (DPD) is used to predict responsiveness to capecitabine. High levels of DPD relative to TP associated with poor response or outcome in capecitabine-treated patients
Surrogate endpoint	Used as a substitute for a direct measure of a clinical efficacy or toxicologic endpoint.	Disease-free survival is considered a surrogate endpoint for overall survival in anticancer therapy trials
Prognostic	Used to predict the natural course of the disease/effect and to distinguish the outcome	Elevated CA125 antigen in adenosarcomas of the ovary; CA 125 half-life < or = 20 days versus > 20 days provides an independent prognostic factor for patient survival in stage III–IV patients early in the course of therapy; must be FDA approved
Disease state or pathological process	Indicates the existence of pathological changes in tissue or organs or occurrence of abnormal cells or tissue function. Biomarkers intended for tissue injury may also be used for this purpose	Aspartate aminotransferase to platelet ratio index (APRI) for liver fibrosis
Mechanistic	Linked directly to the modulation of a specific target or signaling pathway	Adipsin is used as a biomarker for disruption of NOTCH-1 signaling and induction of goblet cell proliferation in the small intestine. Pharmacological inhibition of γ -secretase is one means to disrupt NOTCH-1 processing
Exposure	Indicators of absorbed or target dose (including an absorbed pollutant, its metabolite(s), or products resulting from interaction with endogenous substances), which are measured in a body tissue, fluid, or excreta.	Urinary dialkylphosphate (and other metabolites) is used to indicate exposure to organophosphates; guidelines for the interpretation of these biomarkers of exposure have not been established
Dose optimization	Used to measure responses linked to variable exposure or metabolism; includes many pharmacogenomic biomarkers	Variability in the anticoagulation response (INR) has been linked to the genetic variants of CYP2C9 and VKORC1. Knowledge of the polymorphisms in these genes in addition to other clinical and patient considerations, such as age and BMI, helped in selecting the optimal initial dose of warfarin to be prescribed, achieving a target INR more efficiently and lowering the risk of bleeding adverse events. The US label of warfarin (Coumadin®) was updated with this information
Drug response or pharmacodynamic (PD)	Measure of a specific biological response that occurs after receiving a drug	Plasma vascular endothelial growth factor (VEGF)-C, and soluble VEGF receptor-3 as measures of response to tyrosine kinase inhibitors targeting vascular endothelial growth factor receptors
Toxicity	Measure of a specific toxicologic response (e.g., cellular response, organ/tissue response, or physiologic response that is considered a signal of an adverse response). Biomarkers intended for tissue injury may also be used for this purpose. May include subsets of pharmacogenomic markers	Transmembrane tubular protein kidney injury molecule-1 (Kim-1) is markedly induced in response to renal tubular injury; cystatin C to monitor generalized decreases in renal function; serum cardiac troponin I for myocardial necrosis
Predictive	Baseline characteristics that can indicate the likelihood a specific response will occur to a given treatment.	Use of mutation of the KRAS oncogene as a negative predictive biomarker to identify patients with metastatic colorectal cancer who do not benefit from EGFR-I therapy; PTEN loss in metastases may be predictive of resistance to cetuximab
Tissue injury or damage	A measure that is abundant preferentially (or exclusively) produced in the tissue of interest, be typically present at low concentrations in the blood and other body fluids when there is no tissue injury, increase upon tissue injury, and released into the systemic circulation or other body fluid, where it can be detected and measured	Aspartate transaminase (AST) and alanine transaminase (ALT) are among the first biomarkers of tissue injury; enzymes that are released into the circulation after injury to the liver or muscle
Organ function	Measure of a shift from normal to abnormal organ physiology or organ dysfunction	A change in glomerular filtration rate measured by a clearance method (e.g., radiolabeled-tracer plasma or urinary clearance)

2.1 Biomarkers of Tissue Injury/Damage

It was first recognized in the 1950s that cellular, tissue, and organ injury, as a result of toxic or pathological insults, resulted in the movement of cellular constituents, such as cytoplasmic enzymes, into blood or urine because of altered structural integrity of affected cells, tissues, and organs. Traditional markers of cell and tissue integrity and function such as serum transaminases (liver), creatine kinases (heart and muscle), lipases (pancreas), blood urea nitrogen (BUN, kidney), and urinary albumin (kidney) have been successfully exploited to monitor organ damage. These biomarkers have become an indispensable element of clinical pathology, toxicological assessment, and clinical practice. The utility of these markers has stood the test of time, and most of those currently used in regulatory practice have been in use for almost 50 years. However, changes in biofluids without histopathologic evidence of cellular damage may lead to inaccurate conclusions; therefore, histopathology remains the cornerstone and “gold standard” confirmatory test for organ damage.

In drug discovery as well as toxicology, the use of mechanistic models and nonhistopathologic markers is relevant to both safety and efficacy; however, the challenges for toxicology (safety) are significant because there are often multiple mechanisms of toxicity in play and multiple pathologic changes observed. Consequently, in toxicology studies, the “gold standard” remains histopathology, which emphasizes morphologic changes and response of cells and tissues to injury. In toxicology studies, histopathology is used to identify and characterize the nature of the morphologic changes seen within the phenotypically altered cells that are the target of toxicity. The histologic evaluation plays an important role in forming the basis for initiation and development of biomarker strategies for translation medicine and for the current industry and regulatory focus to predict, identify, and manage target organ toxicity in humans. Ideally, biological markers of toxicity should be cellular products of the phenotypically altered cells that are damaged within the specific tissue and/or organ. Because the nature and character of the histopathology observations can be quite diverse, minimal tissue damage/injury does not necessarily reflect major disturbances in organ function or specific mode of action responsible for the toxicity. Furthermore, a concerted effort to identify mechanistically linked biomarkers using histopathology observations is questionable because identification and use of biological markers that truly reflect mechanism(s) of toxicity is rare. Rather, the currently used biomarkers are “reporters” of cell and/or tissue injury and have gained acceptance through use over many years. This explains why most biomarkers used in toxicity testing are diagnostic and reflect damage already done, rather than being predictive or mechanistic. In recent years, there are only a few limited examples of newly introduced novel biomarkers that have been identified and qualified preclinically and later translated further for use in clinical evaluation. Given these constraints, in the process of identification and qualification of injury biomarkers, histopathology still plays a crucial role because it links structural morphological changes with alterations in biofluids organ physiology and function that reflect the status of cellular and tissue integrity.

Examples of Organ-specific Biomarkers of Tissue Injury

Biomarkers of Liver Injury

(See also Liver and Gallbladder , Vol 3, Chap 2; Clinical Pathology in Nonclinical Toxicity Testing, Vol 1, Chap 18 ; and Interpretation and Reporting of Clinical Pathology Results in Nonclinical Toxicity Testing, Vol 2, Chap 14.)

Despite the known drawbacks of tissue injury biomarkers, additional biomarkers of liver injury, including hepatocellular toxicity and biliary effects, are being explored using comparative biomarker discovery efforts and “omics” methods ( ; ; ). Novel promising biomarkers include glutamate dehydrogenase, keratin 18, sorbitol dehydrogenase, glutathione S-transferase, bile acids, cytochrome P450, osteopontin (OPN), high-mobility group box-1 protein, fatty acid–binding protein 1, cadherin 5, miR-122, genetic testing, and omics technologies, among others.

Drug-induced hepatocellular injury biomarkers considered most likely to be qualified in the near future for preclinical, clinical, and translational applications include the following enzymatic or immunoassay multiplexes: arginase I, glutamate dehydrogenase, glutathione S-transferase alpha , 4-hydroxyphenylpyruvate dioxygenase, malate dehydrogenase, paraoxonase, purine nucleoside phosphorylase, serum F protein, and sorbitol dehydrogenase. These biomarkers show utility based on receiver operating characteristic (ROC) curves performed to correlate biomarker measurements to histopathology findings. However, these biomarkers have not yet been adequately qualified for use in preclinical and early clinical development. Furthermore, large qualification studies and robust assay development are required to determine a biomarker combination that provides additional information relative to alanine aminotransferase (ALT). For example, the performance of new biomarkers in the qualification paradigm might define beneficial activities to aid in interpretation of subtle elevations in ALT. Added cost, increased statistical complexity and data analysis, use of multiple platforms, and greater complexity of qualification and validation efforts are some of the disadvantages of a multimarker strategy, but these potential downsides must be weighed against the cost associated with unanticipated toxicity and more effective safety monitoring for patients.

A Classic: Alanine Aminotransferase ( )

ALT is the classic example of a liver-specific biomarker and extensively qualified measure of hepatocellular injury. A colorimetric assay for ALT was first described by Wroblewski and Cabaud in 1957 when they compared values in healthy individuals to patients with pathologic processes, including infectious hepatitis. Over the next few years, the clinical significance of elevated ALT activity in hepatic disease was established. Although ALT represents a clinical chemistry gold standard for detection of liver injury for both preclinical and clinical utility, guidelines for interpretation are still evolving. Current guidelines for the interpretation of biomarkers of hepatotoxicity in preclinical studies recommend that increases in ALT activity that correlate with histopathologic lesions should be considered adverse. However, the guidelines do not provide guidance regarding the interpretation of elevated ALT in the absence of correlative histopathologic changes. Therefore, the combined use of several serum biomarkers for hepatotoxicity is expected to add interpretive value. A correlation between histopathology and biomarker response serves as the principle standard for biomarker qualification. Since liver histopathology is not as readily attainable in a human clinical trial setting, translatable and analytically validated biomarkers to be compared with a standard measure like ALT are required. Investigations are focused on identifying biomarkers for evaluation of liver injury and integration of multiple analyte measurements to enhance the utility of ALT as a standard. The most promising biomarkers will be able to bridge preclinical toxicity testing and monitor drug-induced liver injury in the clinical setting.

A major type of liver injury of concern is acute idiosyncratic hepatocellular injury (AIHI). The event is termed “idiosyncratic” because the vast majority of treated patients are able to take the drug safely at the recommended dose range, and it is not possible to identify or predict patient susceptibility to AIHI. If a drug treatment is associated with recurrent bouts of AIHI, progressive loss of hepatocytes can lead to irreversible liver dysfunction and, in some cases, mortality. Thus, discovery and development of biomarkers that measure the potential to cause idiosyncratic liver injury or identify individual susceptibility to a drug with established AIHI potential is an area of active research and investigation.

Candidate biomarkers for AIHI are being identified from many paths of investigation. Comparative pan-genomic approaches are being used in rats to identify biomarkers capable of distinguishing pairs of drugs that are structurally and pharmacologically similar, but where one compound is capable of AIHI and the other drug is not. Another path to identifying potential biomarkers for AIHI is studies focused on an assessment of patients who have experienced AIHI. The Severe Adverse Event Consortium began whole-genome single nucleotide polymorphism analysis on germ-line DNA obtained from patients who have experienced varying degrees of drug-induced liver injury, including AIHI. Several efforts are also underway in which various “omics” technologies are used to analyze blood and urine samples from large cohorts of subjects. Additionally, large postmarketing adverse events databases are being mined to elucidate drug/environment susceptibility factors. These efforts could lead to testable hypotheses or be used to provide supportive data for genetic associations observed in these networks. Studying differences in hepatotoxicity susceptibility across panels of inbred strains of mice and quantitative trait loci mapping is also a promising approach to generating hypotheses that would be testable in relatively small numbers of human subjects.

Biomarkers of Vascular Toxicity ( , ; )

(See also Cardi ov ascular System, Vol 4 , Chap 1.)

Clinical development of novel life-saving therapies is often hindered because the nonclinical safety profiles of candidate drugs are associated with occult pathological changes that are not detected by noninvasive, routine clinical monitoring. In nonclinical safety studies, drug-induced vascular injury is an example of such a profile, because there are no biochemical markers for monitoring this lesion clinically and our understanding of the pathophysiology is limited. This type of vascular toxicity is most often associated with candidate drugs that are pharmacologically active in the vascular bed. In addition, immune complex vasculitis can be induced in animal studies of biotherapeutics, and while the relevance of these to humans is uncertain, there is often a need to monitor for vascular injury in clinical trials for these programs. There are a number of drugs from different pharmacologic classes and chemical structures that have been approved for human use but known to cause occult mesenteric or coronary arterial vascular lesions in animal studies.

At high doses in dogs, structurally and pharmacologically diverse vasoactive agents induce a distinctive morphologic coronary arterial pathology. Generally, drug-induced cardiovascular toxicity is regional, with some predilection for the right side of the heart, particularly the right extramural coronary arteries. Macroscopically, hemorrhagic areas are seen in the right atrium with petechial hemorrhages or short, linear, hemorrhagic streaks overlying or adjacent to the right branches of extramural coronary arteries. Similar small areas of hemorrhage are also seen infrequently in the left atrium. The ventricles are usually not affected. Whereas this change is a consistent feature of potential vascular toxicity in dogs and might serve as a biomarker for vascular toxicity, the challenge remains to identify, qualify, and validate a preclinical marker in humans that will bridge the gap between the biomarker used in dogs and the marker(s) of toxicity in preclinical studies.

Biochemical Markers of Drug-Induced Vascular Injury ( )

von Willebrand Factor (vWF) and its propeptide (vWFpp) have been studied as potential markers of endothelial cell (EC) perturbation and drug-induced vascular injury. Release of vWF and vWFpp into circulation is controlled by well-defined constitutive and regulated pathways as well as biochemical processes. In models of drug-induced vascular injury, plasma vWF was evaluated as a potential marker of vascular injury in the rat and dog. Minor increases in plasma vWF have been observed following administration of a potassium channel opener drug and also fenoldopam. This observation in conjunction with other reports raises the possibility that the transient 2- to 6-h increase in circulating plasma vWF could be a reporter of endothelial activation/perturbation prior to morphologic evidence of vascular damage because levels had returned to baseline when vascular injury was confirmed histologically. Studies in the dog have indicated that minimal lesions induced by a potassium channel opener drug do not result in increased vWF levels. Therefore, vWF is not a suitable marker for use in nonclinical safety studies to monitor progressive vascular damage. However, it has been suggested that measurement and analysis of the vWF:vWFpp ratio in humans, dogs, and baboons allows discrimination between chronic and acute phases of EC perturbation, activation, and injury. There is also experimental evidence which suggests that measurement of plasma vWFpp levels may be a useful biochemical marker of regional drug-induced vascular injury because vWFpp is found in low levels in plasma, it has a short terminal half-life, it does not bind to the vascular wall subendothelial collagen, and its major source is ECs. Based on these properties, transient physiological increases in vWFpp can be differentiated from pathological increases that would lead to a sustained elevation.

Seeking Site-specific and Selective Biomarkers of Drug-Induced Vascular Injury ( ; ; ; ; ; )

Microscopically, vasodilators and/or positive inotropic agents such as minoxidil, hydralazine, phosphodiesterase inhibitors, and endothelin receptor antagonist (ETRA) are associated with this unique coronary arterial lesion, characterized by acute segmental changes of medial necrosis and hemorrhage that can develop as early as 12–24 h following intravenous administration and between 3 and 7 days with oral dosing. Medial hemorrhage can be transmural and/or circumferential, with well-preserved extravasated red blood cells and perivascular edema. Vascular lesions are seen primarily in muscular branches of the coronary arteries, in coronary arterioles and possibly capillaries. Subepicardial branches of the right coronary arteries composed of five to eight cell-layer thickness were most frequently affected and mural branches of the right and left coronary arteries were affected less frequently. Terminal arterioles and capillaries, particularly those in the outer third of the right atrial wall, were quite prominent due to enlargement related to swelling and/or activation of ECs. Chronic lesions are intimal proliferation, smooth muscle cell (SMC) hyperplasia with deposition of mucinous ground substance, and adventitial fibroplasia. Also present are varying degrees of a mixed mononuclear inflammatory cell infiltrate.

In the dog, understanding drug-induced vascular injury pathophysiology and potential mode of action has been aided by data describing a relationship between predisposed toxicity site(s) and the distribution, density, type, and ratio of vasoactive endothelin receptors in the heart and coronary arteries. Additionally, the composite analysis of receptor subtype profiles, mRNA expression, and regional blood flow measurements clearly supported the hypothesis that a disproportionate receptor distribution was responsible in the initial studies for the marked functional differences in regional blood flow at the affected sites of injury, the right atrium and right coronary arteries. Collectively, these data strongly suggest that exaggerated pharmacology is the basis for selective, site-specific, mesenteric, or coronary arterial damage in rats and dogs caused by these drugs ( ; ). Concomitantly, there is loss of regulation and control of key biochemical pathways in ECs and SMC, breakdown in cell-to-cell communications, and, ultimately, arterial wall damage. Therefore, both physiological and biochemical markers of endothelial and SMCs should be evaluated to identify and develop translational markers of vascular toxicity.

Biomarkers of Renal Injury ( )

(See also Kidney , Vol 4, Chap 2; Clinical Pathology in Nonclinical Toxicity Testing, Vol 1 , Chap 10 ; Interpretation of Clinical Pathology Results in Nonclinical Toxicity Testing , Vol 2, Chap 14.)

In 2010, the FDA issued the first formal biomarker qualification decision about novel safety biomarkers of drug-induced kidney injury for preclinical drug development. The ILSI Health and Environmental Sciences Institute (HESI) evaluated four urinary biomarkers of nephrotoxicity (α-glutathione S-transferase, μ-glutathione S-transferase, renal papillary antigen [RPA-1], and clusterin) and compared their performance against traditional measurements indicative of renal injury in male Sprague–Dawley and Wistar rats. In this biomarker discovery effort, compounds that cause injury to specific anatomical regions of the nephron and corresponding biomarker values to defined histopathologic lesions were used to generate a histopathological “gold standard” to assess biomarker performance. Immunohistochemistry (IHC) was applied to confirm the location of regional lesions. The discriminatory accuracy of each biomarker was assessed using ROC curve methods. Immunolocalization of RPA-1 in gentamycin-treated rats demonstrated that RPA-1 is increased in the cytoplasm of intact injured collecting duct epithelial cells and in necrotic cells within the proximal tubular epithelium. It is also detected in the urine using immuno-based assays. Based on the review of the data submitted by an ad hoc appointed Biomarkers Qualification Team, RPA-1 and clusterin were qualified for use in Good Laboratory Practice preclinical toxicology studies, but not for routine monitoring of drug-induced nephrotoxicity in the clinical setting. RPA-1 represented a novel biomarker not previously qualified for this purpose.

Over the course of several months in 2007, the C-Path Predictive Safety Testing Consortium (PSTC) submitted data to the FDA/EMEA supporting claims for additional renal toxicity biomarkers. The newly proposed biomarkers included urinary albumin, β2-microglobulin, cystatin C, kidney injury molecule (Kim-1), total protein, and urinary trefoil factor 3. Performance of each biomarker was compared to accepted standards of BUN and serum creatinine by comparing the area under the curve (AUC) of the ROC analysis for each biomarker. Similar to the HESI studies discussed in the previous paragraph, histopathology was the standard used to qualify biomarker measurements. Although an ad hoc panel of reviewers concluded that the panel of proposed biomarkers is acceptable for nonclinical drug development, they noted several limitations of the data set, including lack of a clear correlation between the biomarkers and the temporal pathogenesis of the renal lesions (as determined by histopathology) and a lack of correlation between reversibility and recovery of kidney function. The panel concluded that use of these biomarkers to predict or monitor renal toxicity was insufficiently demonstrated. The panel did, however, consider the group biomarkers acceptable for the detection of acute drug-induced nephrotoxicity in preclinical drug development. Hence, the biomarker review panel did recognize the potential of these biomarkers to be qualified for use as clinical markers of acute drug-induced renal injury. Albumin, β2-microglobulin, cystatin C, Kim-1, total protein, urinary clusterin, and urinary trefoil factor 3 thus provide additional and complementary information to BUN and serum creatinine to correlate the histopathology, which is considered the gold standard.

Region-Specific Kidney Damage

The use of renal injury biomarkers to define the specific region of the kidney subject to injury is promising. A marker that is strongly induced in proximal tubular injury is urinary Kim-1. Kim-1 has been cloned by representational difference analysis, a polymerase chain reaction–based technique, which was conducted to compare gene expression in normal versus postischemic rat kidney, to identify genes that were upregulated with renal ischemia. Researchers demonstrated that the ectodomain of Kim-1 is shed from cells in vitro and in vivo into the urine in rodents as well as humans after proximal tubular kidney injury. Thus, urinary Kim-1 serves as a translational biomarker in preclinical and clinical studies and is a diagnostic indicator of kidney injury that is detected earlier and with greater specificity compared to conventional biomarkers (e.g., serum creatinine, BUN, glycosuria, increased proteinuria or increased urinary N-acetyl-β- d -glucosaminidase, γ-glutamyltransferase , alkaline phosphatase levels). Initially, an enzyme-linked immunosorbent assay (ELISA) was developed to measure Kim-1 in rodent and human urine samples. More recent technological advances facilitated the development of a high-throughput microbead-based assay to quantify Kim-1 in rat urine that is more sensitive, has a greater dynamic range, and requires less urine volume and reagents than the conventional ELISA. Kim-1 has been shown to clearly outperform serum creatinine as a marker for tubular injury in rats in terms of sensitivity and specificity.

Biomarkers for Reversibility of Kidney Damage

Based on recommendations resulting from the FDA's biomarker qualification process, a study was recently reported in which reversibility and comparative injury data for several candidate urinary biomarkers of kidney injury were evaluated. The nephrotoxin gentamicin was given to rats once on each of 3 days and the animals were humanely euthanized at various time points. Between days 1 and 3, all biomarkers except albumin were elevated, peaked at day 7, and returned to control levels by day 10 (μ- and α-glutathione S-transferases, and RPA-1) or by day 15 in the case of Kim-1, lipocalin-2, OPN, and clusterin. All biomarkers performed better during injury than during recovery except OPN, which performed equally well at both time periods. During the evolution of injury, Kim-1, RPA-1, and clusterin correlated best with histopathologic lesions that included necrosis, apoptosis, and regeneration. During the recovery period, Kim-1, OPN, and BUN were the best markers of injury resolution based on correlation with histopathology. RPA-1 was the best indicator of tubular and/or collecting duct regeneration, especially during recovery.

Biomarkers of Cardiac Injury ( )

(See also Cardi ovascular System, Vol 4 , Chap 1; Clinical Pathology in Nonclinical Toxicity Testing, Vol 1 , Chap 10 ; and Interpretation of Clinical Pathology Results in Nonclinical Toxicicity Testing, Vol 2, Chap 14 .)

Historically, lactate dehydrogenase, myoglobin, and creatinine kinase isoenzyme analyses have been used as a measure of ischemia-induced cardiac injury. These serum-based myocardial markers have limited cardiac specificity and relatively short serum half-lives. Cardiac troponins (cTns) T and I are more recent additions to the list of biomarkers of cardiac injury recommended for assessing injury associated with myocardial ischemia and drug-induced cardiac toxicity.

Troponins are a complex of three proteins that regulate the calcium-mediated interaction of actin and myosin. Separate genes encode each cTn isoform. The isoform for cTnC is common to all types of muscle, whereas those for cTnI and cTnT are considered cTns. To date, the cardiac isoform of cTnI has not been found outside of cardiac tissue, whereas isoforms of cTnT exist in skeletal muscle as well as cardiac tissue.

cTnI and cTnT are released from the cytoskeleton during myocardial infarction, cardiac myocyte injury, or other cardiovascular conditions, with increased levels detected in blood. The American College of Cardiology and the European Society of Cardiology declared cTns the biomarker of choice for acute myocardial infarction in 2000. According to the “gold standard,” analytical assays for cTns should include free cTnI, the I–C binary complex, the T–I–C ternary complex, and oxidized, reduced, and phosphorylated isoforms of the three cTnI forms; however, epitopes located on the stable part of the cTnI molecule should be a priority. Currently, cTns are the markers of choice for detection of clinical cardiotoxicity and cardiac ischemic necrosis based on their cardiac specificity and their longer serum half-life. Furthermore, cTnI and T have been promoted as the gold standard biomarkers of drug-induced cardiac toxicity in the preclinical setting as well. In fact, cTns are now routinely included in nonclinical safety studies ( ; ).

2.2 Biomarkers of Altered Organ Function

Functional biomarkers are expected to monitor specific changes in function associated with pathological events at the cellular, tissue, or systemic level (e.g., changes in biological function, diagnostic/disease state). A molecular biomarker is an entity whose release, abundance, and/or modification state is altered as a result of injury or disease and can be used to aid diagnosis.

The Discovery and Qualification of Cystatin C: A Measure of Glomerular Function ( ; ; ; )

The use of serum cystatin C as a measure of glomerular filtration rate was first proposed in 1985. Since then, numerous studies have been conducted to qualify cystatin C as a biomarker of glomerular function. Most of the studies have been comparative biomarker analyses comparing cystatin C to other standard references in healthy subjects compared to patients with impaired glomerular function. Cystatin C is an endogenous inhibitor of cysteine proteinases that is freely filtered by the glomeruli and reabsorbed from ultrafiltrate by the tubules, where it is almost completely catabolized, with the remainder eliminated in urine (similar to β2-microglobulin). Increased urinary cystatin C concentrations allow the detection of a direct impairment of the tubular protein reabsorption complex or an indirect impairment caused by a tubular protein overload due to glomerular alterations and injury in various nephropathies. Mean urinary cystatin C concentrations in patients with kidney tubular disease are elevated compared to healthy patients or patients with glomerular disease. Thus, increased urinary cystatin C reflects a direct or indirect functional renal tubular impairment independent of glomerular filtration rate. However, the concentration of serum cystatin C is mainly determined by the glomerular filtration rate. Serum cystatin C as an endogenous marker of glomerular filtration rate and kidney function is superior to serum creatinine because cystatin C is less dependent on extrarenal factors than creatinine and is completely cleared by the glomerulus, in contrast to creatinine, which is also cleared by tubular secretion. Metaanalysis of combined clinical studies done to qualify the performance of cystatin C as a marker of glomerular function confirms the superiority of cystatin C over creatinine with respect to glomerular filtration rate measurements, by both superior correlation coefficients and greater ROC plot AUC values.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here