Abstract
Cardiac troponin (cTn) is a sensitive and specific biomarker for assessing cardiac damage and should be utilized in drug safety assessment. Lactate dehydrogenase and creatine kinase isoenzyme analyses have historically been used in pre-clinical toxicity testing to assess cardiac injury, but since these assays are less sensitive and specific than cTn, isoenzyme analyses, as determined by the manual electrophoretic technique, are no longer warranted. Commercial cTn assays developed for humans do not have the same immunoreactivity and functional sensitivity in the common pre-clinical testing species, so it is important to show that the assay that is chosen is appropriate for the pre-clinical species being assessed. The kinetics of the cTn response depends on the dose and frequency of test article administration as well as the mechanism of the cardiac injury induced by the test article. Cardiac troponin should be used in the assessment of classes of compound that have previously been shown to induce cardiac necrosis or if cardiac necrosis is identified histologically with a novel compound. Next generation high sensitivity cTn assays are being developed and the low levels of cTn detected with these assays may be an early sign of possibly reversible damage to the heart.
The development of a novel biomarker is challenging, and acceptance by the scientific and regulatory communities is often a long and arduous process. Identifying biomarkers that have been used effectively in human medicine and applying these markers to preclinical safety testing is a more expedient way to implement them. Cardiac troponin (cTn), a biomarker of cardiac injury, is a good example of how this reverse translation has effectively been done and has led to the relatively rapid acceptance of this marker in preclinical toxicity testing. It is a sensitive and specific tool for assessing cardiac damage in those studies in which the test article has been shown to induce cardiac necrosis. It is not a direct marker of myocardial fibrosis or cardiomyocyte hypertrophy. There are recent excellent reviews of the use of cTn in preclinical testing, and this article is not meant to be another complete review of this subject (Berridge et al. 2009; O’Brien 2008; O’Brien et al. 2006; Walker 2006). The purpose of this commentary is to give a short historical perspective on how this marker has gained acceptance over the past several years, how it should be used, what is the current regulatory perspective on this use of this marker, and where we are headed with the next-generation assays.
Lactate dehydrogenase (LDH) and creatine kinase (CK) isoenzyme analyses have historically been used in humans to assess ischemia-induced cardiac injury. CK-MB isoenzyme analysis is still often used due to familiarity with this parameter, but cTn is the analyte recommended for assessing injury associated with myocardial ischemia and is replacing CK-MB. The American College of Cardiology and the European Society of Cardiology declared cTn the biomarker of choice for acute myocardial infarction in 2000 (Jaffe 2001). The use of cTn has been expanding to assess cardiac injury associated with many other disease processes as well. Historically, there were no good automated assays available to measure cardiac isoenzymes in animals, and the CK-MB assay that was developed for humans did not translate well to our preclinical species. Thus, manual electrophoretic separation and quantitation was needed, although there are some more recent non–activity-based assays that can be used to measure CK-MB in animals. In addition to the resource-intensive nature of the historical assays, they were not very sensitive for cardiac damage. It is my opinion that the advent of sensitive tests to detect cTn has eliminated the need for electrophoretic determination of CK or LDH isoenzymes as part of the assessment of cardiac injury in preclinical studies.
In 2008, Peter O’Brien reviewed the literature for articles in which cTn was used in animals for assessing cardiac injury (O’Brien 2008), and he found only a single publication in 1995. In contrast, there has been a marked increase since that time, with multiple publications in subsequent years on the use of cTn to assess cardiac injury in experimental studies, preclinical toxicity testing, and as a diagnostic tool in veterinary medicine. There has especially been an increase in our knowledge of use of this marker in dogs and rats, although there is a relative paucity of information on the use of cTn in nonhuman primates. This growing body of literature supporting the use of this marker, as well as much unpublished work within different companies, has led to the inclusion of this parameter in preclinical toxicity testing for both exploratory and regulatory studies.
Although there are some studies that have assessed which of the assays originally developed for humans could be used effectively in the different preclinical species, there was not a published article that systematically evaluated how to measure cTn in preclinical species using most of the more common automated assays available in safety assessment laboratories. The Health and Environmental Sciences Institute (HESI) Cardiac Troponins Biomarker Working Group, of which I was a participating member, recognized this knowledge gap and completed a study that compared which automated instrument/assay(s) were the best ones for cTn analyses in the preclinical species: rats, dogs, and nonhuman primates (Apple et al. 2008). All except one of the assays evaluated were specifically developed for use in humans. We clearly found that the assays had different comparative immunoreactivity, and not all could be used in each preclinical species. We also clearly found that each assay has a different lower limit of detection with greater than 20-fold variation across the species and platforms tested. A precision profile was developed for each assay using samples at or near the limits of detection of the assays. The functional sensitivity (ie, the level at which you can reproducibly detect a change from background) of each assay was quite different and is a critical factor in reliably assessing minimal cardiac damage. Based on the differences in these assays, laboratories involved in safety testing should ensure that the assay chosen is suitable, sensitive, and precise in the species being evaluated to generate accurate assessment of potential cardiac damage.
There have also been several studies evaluating the kinetics of cTn in response to different cardiac injuries, with isoproterenol often being used as model acute toxicant and doxorubicin as a model of chronic toxicity (Brady et al. 2010; Engle et al. 2009; Herman et al. 1998; Koh et al. 2004; Tonomura et al. 2009; York et al. 2007; Zhang et al. 2008). The time of peak cTn response and duration of the response depends on the mechanism of cardiac injury, dose, and frequency of test article administration. In many situations, cTn release can occur within minutes after injury. Thus, if the process has not been previously characterized, it is important to take samples early in a toxicity study to evaluate for cardiac damage, especially if only a single dose of compound is administered. If it is a multidose study and there is ongoing cardiac necrosis, however, cTn likely will be elevated at the end of study. On a whole-study basis, elevations in cTn will typically correlate with cardiac myonecrosis/degeneration seen histologically with a slight shift in temporal relationship in that the peak cTn response will be seen prior to the peak response in severity identified by histological examination. There will not always be a 1:1 correlation on an animal basis between cTn and histopathology results, probably due to a combination of limited sampling of the cardiac muscle on routine histopathological assessment as well as high sensitivity of the cTn assays. In fact, in some studies, you may find that there is correlation between the histopathology and cTn results at the higher doses, but at lower doses, there may only be positive cTn results. It is my opinion that if there is a correlation at the higher dose between the histopathology and cTn results, and there are significant dose-related elevations at a lower dose without histological correlation, then these are biologically relevant increases in cTn. This is in contrast to most other clinical pathology parameters, where we often use histopathology as the gold standard to set the nonadverse dose. For a review of other factors to consider for using cTn effectively in preclinical studies, the reader is referred to a recent article by the HESI Cardiac Troponins Biomarker Working Group (Berridge et al. 2009).
Based on the literature and scientific experience within each company, it is believed that cTn can be effectively used as a biomarker of cardiac damage in preclinical regulatory studies, and the Food and Drug Administration (FDA), European Medicines Agency (EMA), and other regulatory agencies are accepting this biomarker as part of new drug applications. To formalize how cTn should be used in regulatory studies, a Request for Qualification by FDA of Cardiac Troponin as a Blood Biomarker for Non-clinical Toxicology Studies was submitted to the FDA at the end of 2008. Peter O’Brien (University College of Dublin, Ireland) is leading this effort with Malcolm York (GlaxoSmithKline), Matt Jacobsen (AstraZeneca), and me also participating. The FDA requested additional information for the submission, which was provided in 2009, and the document is currently being reviewed by the agency. A similar document is also being reviewed by the EMA. The basis of this qualification is the published scientific literature, and it does not entail new unpublished experimental data. The document promotes the use of cTn as a marker of cardiac damage that can be used to assess acute and ongoing injury. It recommends that cTn should be used when the test article being evaluated is from a class of compound that has previously been shown to induce injury that results in cardiac necrosis. Alternatively, if a novel compound is being evaluated and cardiac necrosis is identified histologically, cTn should be added to further monitor and assess the injury in subsequent studies. Based on the historical information available on this analyte, it can be effectively used in rats, mice, dogs, rabbits, and, to a lesser extent, nonhuman primates. A deficit of published studies using this marker in nonhuman primates was identified, as well relatively few publications that have evaluated this marker in chronic studies. We look forward to receiving final acceptance by the agencies of our proposal. Ongoing use of this biomarker in preclinical studies will increase our knowledge of how to effectively detect and monitor cardiac toxicity. In addition, the knowledge gained using this assay in preclinical testing can be used in subsequent studies in humans, when appropriate, to monitor for cardiac toxicity.
The HESI Cardiac Troponins Biomarker Working Group also recently addressed how this analyte can effectively be translated from preclinical testing to human clinical trials. The paper titled “A Translational Approach to Detecting Drug-Induced Cardiac Injury With Cardiac Troponins: Consensus and Recommendations From the Cardiac Troponins Biomarker Working Group of HESI” was a result of a conference that brought together preclinical and clinical scientists to assess the points of consensus, identify the gaps, and make recommendations on the use of cTn in preclinical species and man during drug development (Berridge et al. 2009). The entire paper will not be reviewed, but the main points of how to effectively use cTn in clinical trials are summarized.
Although the use of cTn in clinical trials can be challenging, the major recommendations were that cTn should be assessed when there is a risk for cardiomyocyte injury. Preclinical findings (including the type of the cardiac lesion, dose-response curve, and assessment of reversibility), drug class effects, and understanding the target patient populations with a preexisting disease that might exacerbate cardiac injury will all be useful to effectively implement cTn monitoring. Troponin should be measured before enrollment in studies to help find patients with existing cardiac disease who may be at a higher risk for cardiac injury. Troponin data should be interpreted, with additional information including results of preclinical studies, clinical reference ranges, baseline values, clinical signs, and history. Long-term clinical outcomes for healthy patients with transient drug-induced cardiac injury (as assessed by concentrations of cTn above the normal cutoff level) are lacking, particularly if drug treatment is stopped quickly.
The future of cTn analysis is now; that is, we are in the process of implementing new assays that have greater sensitivity and precision than the first-generation assays. At the same time, assays that are considered high-sensitivity assays are being developed and tested. Nanosphere, Singulex, Beckman, and Roche are all developing new assays that allow the detection of very low levels of cTn (Apple 2009). Whereas the current assays we are using detect cTn in the ng/mL range, the newer assays detect in the low pg/mL range. Previously, with the older assays, a normal animal or patient was typically below the limit of detection. With these newer high-sensitivity assays, cTn is being detected in normal animals and people (Mikaelian et al. 2009; Wu et al. 2009). Could these values reflect normal physiological release of cTn from the minimal turnover of cardiomyocytes that has recently been proposed to occur in humans (Bergmann et al. 2009)? Recent articles published by Schultze et al. (2009, 2008) nicely showed how the Singulex assays can effectively be used in the preclinical species (rats, dogs, and nonhuman primates). He further showed what is the normal variation in cTn over time in resting rats, animals receiving oral gavage, and simulated transport (Schultze et al. 2009, 2008). More needs to be done to characterize these new assays and define what the biological impact of minor cTn elevations is. But there is some clinical evidence that these assays may be markers of very early signs of damage and thus may be markers of early myocardial ischemia (Sabatine et al. 2009). If this early phase can be detected by transient elevations in cTn, and an intervention could occur at this early time point, would this prevent significant cardiac damage from occurring? Are we at the level of detection that the elevations that are seen are reflective of a reversible cardiac injury? A recent article supports this by reviewing the literature, which shows that the heart can undergo release of membranous blebs containing cTn during ischemia but without significant necrosis (Hickman et al. 2010). These blebs then release their content into the circulation and can be picked up as transient elevations in cTn. Much more research needs to be done on these assays, both preclinically and clinically, to understand how these high-sensitivity assays can be effectively used in regulatory studies.
In summary, the growing body of literature, as well as our collective experiences, concludes that standard cTn assays can and should be used today as a marker of cardiac damage in both preclinical and clinical drug development studies. We should also continue to share our collective experience on the use of this biomarker through consortium activities and publications to continue to gain additional knowledge to increase our effective use of this biomarker in drug development. The new high-sensitivity assays may provide us with tools to detect potentially reversible cardiac injury.