Are We Closer to Finding Biomarkers for Identifying Acute Drug-Induced Liver Injury?

KEYWORS::

• acetaminophen
• HMGB1
• miR-122
• qualification

“Drug-induced liver injury is a complex multicellular and multimechanism disease; therefore, it is logical that a battery of complimentary biomarkers that reflect specific cellular processes is required in the first instancerather than simply displacing current markers.”

Over recent years, there has been a considerable effort to identify and develop new biomarkers of acute drug-induced liver injury (DILI). It is clear that the development and qualification of sensitive and specific hepatic biomarkers that permit quantitative assessment of the translation between preclinical and clinical studies is urgently required. The potential for novel biomarkers to provide enhanced understanding of the fundamental mechanisms that result in clinical DILI is becoming increasingly recognized. Furthermore, there is a great need for mechanistic biomarkers that can bridge between preclinical models and the clinical situation of DILI. These will therefore enable us to understand the human relevance of predictive models of acute DILI and the interpretation of data generated from such models. Such biomarkers would not only accelerate the pace of drug development but also provide a point-of-care test to enable patient and/or DILI-specific treatment stratification once a new drug is licensed. Currently, the number of putative biomarkers that hold some clinical utility is low, mainly because less attention has been placed on the science of drug safety compared with drug efficacy [1].

The current assessment of acute DILI & a critical evaluation

Presently, only a small number of blood-based tests are used to assess liver injury or dysfunction in humans, and assessment within drug development is heavily dependent upon hepatic histological interpretation. The ‘classical‘ blood-based tests include the serum concentration of total bilirubin (TBL) and the activity assessment of the enzymes ALP, AST and ALT. Elevations in the activity of these enzymes may indicate injury to biliary cells or hepatocytes, while changes in TBL represent changes in hepatic function [2]. Although the assay of ALT has become the primary screening tool to detect acute DILI, it is not without its problems. Changes in enzyme activities are not specific for DILI and can occur in a number of disease processes, including viral hepatitis, fatty liver disease and liver cancer [3]. In addition, elevations in ALP may also be attributed to hyperthyroidism or bone disease and increases in ALT and AST can also be the result of myocardial damage or muscle damage from extreme exercise. In addition, the methods used to quantify ALT activity have not been standardized and a robust definition of normal reference ranges has not been agreed upon; these inevitably depend upon the population group defined as normal. Assay measurements also vary between laboratories. Furthermore, although ALT activity is regarded as sensitive for detecting liver injury when it occurs, it is not sensitive with respect to time/kinetics. ALT activity alone therefore offers little predictive value given the rarity of idiosyncratic DILI and that an ALT elevation represents probable injury to the liver after injury has occurred. Therefore, ALT activity is often combined with the liver-specific assessment of TBL as part of Hy’s law. This is currently the only accepted regulatory model to assess significant acute DILI. However, there are several reported examples of when a subject qualifies as a Hy’s law case and this has been previously reviewed [1,4]. This diagnosis of a Hy’s law case is often delayed and often occurs when treatment is discontinued. This highlights the space for improvement and modification to Hy’s law criteria that novel biomarkers can fill. Therefore, taking these deficiencies into account, there has been a considerable effort to identify and develop new biomarkers that can inform the mechanistic basis of DILI and provide potential measures for patient management and treatment stratification.

Which biomarkers have been identified & developed & what is their clinical utility?

Considering the need to develop and qualify new biomarkers for acute DILI (and injury to other organs), a significant level of resources have recently been directed towards this and a number of public–private consortia have been developed; namely the Predictive Safety Testing Consortium and Safer And Faster Evidence-based Translation consortium [5]. Important and noteworthy progresses have been made in the development of biomarkers for renal drug safety evaluation, which have been qualified for use by various regulatory authorities. The lessons learnt from these can be applied to the liver [6,7]. In particular, the concept that new biomarkers should be used to complement existing ones and not to replace them has arisen. With respect to DILI, a number of targeted biomarkers have been identified as having preclinical sensitivity and specificity for their utility to assess acute liver injury, such as α-glutathione S-transferase, arginase 1 and 4-hydroxyphenylpyruvate dioxygenase [8]. In addition, sensitive mechanistic biomarkers have also been reported. These range from highly liver-specific miRNAs such as miR-122 [9], to markers of epithelial cell apoptosis and necrosis such as KRT18 [10], and to markers of mitochondrial dysfunction such as GLDH [11] and markers of immune modulation and hepatic necrosis such as HMGB1 [10]. Clinical utility of miR-122, HMGB1, KRT18 and GLDH has been recognized and they have recently been demonstrated to report on liver injury associated with acetaminophen overdose in a more sensitive manner than currently used indicators of liver injury [11,12]. Moreover, clinical findings also reveal that a number of these markers hold enhanced prognostic capability over current methods following acetaminophen overdose [13,14]. However, while these markers offer a sensitive indication of liver injury, information about the mechanism of toxicity, enhanced prognostic utility and a clear route for patient stratification, a major limitation is that these conclusions are mainly drawn on clinical evidence following relatively common acetaminophen overdose and have not been assessed in rare cases of idiosyncratic DILI. A better understanding of the kinetics of these current biomarkers in human or animal biofluids is required so that we can have a stronger basis on how to interpret the data. Furthermore, the majority of biomarkers under evaluation reflect hepatic injury when it occurs and their assessment of hepatic regeneration/drug adaptation has not been evaluated. For example, considering ALT activity from a regulatory point of view, elevations in ALT activity are worrisome with respect to liver safety. Frequent and relatively large elevations in ALT activity are associated with treatments that do not pose a clinical liver-safety issue, such as heparins [15] and tacrine [16]. Indeed, the mechanism by which ALT is released from the hepatocyte has never been studied in detail and the quantitative relationship between cell injury per se and the number of hepatocytes damaged has not been established. The challenge now is to identify or assess currently proposed biomarkers (on their own or in combination) that can distinguish between benign elevations in ALT activity and those that represent serious DILI outcome. The refinement of a recently described clinically relevant porcine model of predictable acute liver failure following acetaminophen overdose may be appropriate to discover mechanisms and biomarkers of sublethal injury, recovery or adaptation [17].

Are we too quick to write off currently used markers of acute DILI?

The assay of serum ALT activity has become the primary method to assess acute liver injury. Although not formally qualified against human histology for DILI, ALT is widely ‘validated‘ and ‘qualified‘ through its widespread use and ease of assay. Although ALT activity has substantial limitations, it can be a very useful biomarker if understood and used properly. Often, elevations in serum ALT activity are understood and interpreted wrongly. The origins and reasons for why the misinterpretation of elevations in ALT activity have come into wide acceptance have been recently and elegantly reviewed [4].

Within this review, Senior proposes new interpretations relating to ALT activity and suggests how they might lend themselves to better general use, and that suitable revisions in guidance, teaching approaches and clinical practice will need to reflect this. We now build upon this to initiate discussion and debate as to the best course of action to develop new biomarkers of DILI:

• ▪ Serum ALT activity is not a test of liver function but of liver cell injury, and the quantitative relationship between ALT activity and the actual number of dying hepatocytes requires further characterization;

• ▪ Whole-liver dysfunction determines severity of injury and serum TBL is one measure of liver function;

• ▪ The combined biomarker of ALT and TBL has high sensitivity for when liver injury occurs and has great hepatic specificity that represents the current standard to refine, build on or surpass any putative acute liver injury biomarker;

• ▪ ALT has been shown to have different functional isoforms. Therefore, refinement of the current assay should reflect this and may improve data quality;

• ▪ There is a need for defined and standardized assays for ALT activity (or another chosen novel biomarker) and what should be adapted as a true normal reference value;

• ▪ If elevations in ALT activity (or the chosen novel biomarker) over the normal range are seen during assessment, the test should be repeated as a matter of urgency, so that evidence-based decisions can be made as to the direction of injury (progression or resolution);

• ▪ New mechanism-based biomarkers, which reflect particular cellular processes involved in DILI and can be used alongside current methods, should offer a clear rationale for patient stratification and novel pathways for therapeutic intervention.

Conclusion

It is clear that there is a need for novel biomarkers to assess liver injury associated with drug use and clear progress has been made and clinical utility shown regarding ‘mechanism-based‘ biomarkers. It is also clear that no single biomarker will be the answer and that a panel approach of novel biomarkers alongside a more intelligent use of currently used biomarkers represents the way forward. Studies should therefore be designed to understand what purpose the biomarker is fit for and to define its context of use in a prospective manner. With this guidance in mind, a note of caution should also be adopted as to the interpretation of data from the assessment of novel markers that are qualified or evaluated based on data in high-incidence populations (such as acetaminophen overdose) and applying them to rare idiosyncratic acute DILI when high negative predictive values can confound judgement of sensitivity and specificity. In summary, there are now clearly a number of DILI biomarkers that hold great potential. When using these markers, lessons from the qualification of recent renal safety biomarkers can be applied to the liver. DILI is a complex multicellular and multimechanism disease; therefore, it is logical that a battery of complimentary biomarkers that reflect specific cellular processes is required in the first instance rather than simply displacing current markers.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.