The rising costs of drug development, mainly attributed to late-stage compound failures, is intensifying efforts within the industry to discover new methods and technologies that can predict the safety and efficacy of new compounds early in the development process. With only one in 11 compounds advancing from first-in-man studies to regulatory approval, it is becoming vital for pharma companies to make reliable go/no-go decisions on candidates as early as possible, ideally while the candidate is still in preclinical development.
Scientific innovation — particularly in the areas of genomics, proteomics, bioinformatics and systems biology — is transforming the drug discovery and development process and facilitating our understanding of the molecular basis of disease. Translational research involves expediting the translation of these scientific discoveries into clinical practice, including the identification of biomarkers that can be used to diagnose and measure the progress of disease and/or the effects of treatment. Predictive biomarkers can provide early evidence of safety and efficacy while a molecule is still in preclinical development. These biomarkers can provide assurance that lead candidates will have a high probability of success in subsequent milestones, potentially reducing the time and cost of drug development. For this reason, predictive biomarkers are becoming an increasingly important tool in preclinical drug discovery and development and remain an important focus of translational activities.
Predictive biomarkers that facilitate the selection of lead compounds for development can include markers of efficacy, safety and pharmacokinetics/pharmacodynamics (PK/PD). Understanding the exact mechanisms of drug action that lead to the breadth of patient responses, both on- and off-target effects, remain a significant challenge. However, certain classes of molecules lend themselves as reliable biomarkers due to their intrinsic role in the pathophysiology of disease (e.g., kinases) and/or their known role in molecular targeted drug activity (e.g., drug target(s), drug metabolizing enzymes). The use of these biomarkers may ultimately support proof of mechanism, proof of principle and proof of concept studies. By evaluating biomarkers in relevant preclinical disease models, companies can optimize safety and efficacy of their compounds and enable the design of successful translational strategies into the clinic, including establishing safe first-in-human (FIH) dosing strategies and stratifying patients in early clinical trials.
Although “-omics” and molecular imaging technologies are rapidly transforming the drug discovery and development process, much work is still needed to bridge animal and clinical studies and to validate these much needed predictive biomarkers.
Seeing is Believing
The biopharma industry is actively pursuing new technologies that can provide more reliable data on which to base go/no-go decisions. “-Omics” and imaging (PET, MRI) studies are largely contributing to the field of biomarker research. The differential expression of genes, proteins and metabolites under conditions of drug exposure reflects a “signature” of drug exposure and can yield great insight into the effect of a drug on the system. Some of the most successful examples of preclinical-to-clinical translation have involved the application of imaging technologies to measure treatment response (e.g., Glivec, Herceptin), as imaging is able to deliver noninvasive and quantitative cellular and molecular information. Many companies are increasingly relying on imaging in preclinical development to increase the accuracy, sensitivity and specificity of predictive biomarkers that can be translated between the preclinical and clinical phases of drug development.
In preclinical development, MRI, CT, ultrasound and optical imaging are being used to quantify tumor incidence and burden in small animal models. Imaging is especially useful in models for metastasis and deep-tissue tumors, which are difficult to assess using traditional methods. Imaging at the molecular level provides a direct measure of mechanism of action, offering a more predictive measure of drug activity through the use of targeted image-based predictive biomarkers. As a result of these efforts, imaging is now being used to drive go/no-go decisions in early discovery, and there is a significant focus on developing and validating image-based biomarkers that are intended to be applicable for use in clinical trials.
A critical component of the drug development process is to gain an understanding of pharmacokinetics (the absorption, distribution, metabolism and elimination properties of the drug/metabolite) and pharmacodynamics, which defines the physiological and biological response to the drug. Together, PK and PD represent the driving force behind pharmacological and toxicological effects, and PK/PD testing is performed at every stage of the drug development process.
PK/PD modeling is the integration of in silico, in vitro, and in vivo preclinical data with mechanism-based models to better predict drug action in humans and assist in the selection of drug candidates. These models aid in the design and/or selection of promising lead compounds, first-in-human dosing, early clinical trial design and proof of concept studies.
PK/PD modeling is also proving valuable in determining relationships among biomarker response, drug levels and dosing regimens (see Figure 1). Linking biomarkers with PK/PD models can provide a better understanding of the mechanism of action and provide a quantitative basis for translational research as well as more accurate interspecies scaling.
PD biomarkers can help determine whether a new candidate has altered its target/pathway as intended. Ideally, these biomarkers are easy to measure, are present in easily accessible tissue (if not used with imaging) and correlate with disease activity. Target-specific PD biomarkers may include measuring the target itself as well as mRNAs, miRNAs and/or proteins downstream of the target pathway. Abnormal expression of a target-specific PD biomarker can be used as inclusion criteria in early clinical trials to demonstrate biological activity and mechanism of action, to help determine and adjust dosing and to demonstrate early evidence of clinical activity.
Identifying safety issues as early as possible in the drug development process is becoming very common, as the benefits of dropping potentially harmful candidates sooner are realized. Multiple toxicities — immunotoxicity, cardiotoxicity, hepatotoxicity and nephrotoxicity — must be evaluated prior to clinical studies, and several safety biomarkers are already widely accepted, including alanine aminotransferase (ALT) and aspartate aminotransferase (AST) for kidney and liver toxicity, respectively. Traditional animal models cannot always predict toxicity in humans, however, and the all-too-frequent result is late-stage drug attrition. The use of predictive toxicity biomarkers can help increase our understanding of toxicity mechanisms, better inform clinical safety in humans, assist in earlier decision making and ultimately reduce late-stage failures.
The Critical Path Institute’s (C-Path) Predictive Safety Testing Consortium (PSTC) recently developed a drug-induced kidney injury biomarker panel, which includes seven urinary protein biomarkers for testing in preclinical rat safety studies. This biomarker panel is now considered qualified by the Food and Drug Administration (FDA), European Medicines Agency (EMA), and Japan’s Pharmaceuticals and Medical Devices Agency (PMDA) for voluntary use in nonclinical safety studies to support regulatory decision-making, and they are now being evaluated for use in humans. Further development of similar predictive biomarker panels for safety testing will provide greater confidence in candidate molecules that enter human clinical testing and reduce drug attrition.
There is an ongoing collaborative effort among industry, FDA and academia to assess how imaging can fill unmet needs in safety testing. A consortium launched by the Health and Environmental Sciences Institute (HESI, which you can find at www.hesiglobal.org) is building a multi-sector and multi-specialty team of scientists to identify and pursue opportunities to integrate imaging approaches into current safety assessment paradigms for drugs. This group is focused on translational imaging in preclinical development and will guide safety biomarker development in collaboration with FDA and other regulatory agencies.
Pharmaceutical companies are beginning to regularly incorporate biomarker platforms into the drug discovery and development process as far upstream as possible. The earlier candidate efficacy and safety profiles can be established, the earlier go/no-go decisions can be made, leading to a more efficient process. The use of predictive biomarkers in preclinical studies in integration with other data generated may reduce the high level of late attrition so common in drug development and optimize selection of the best agents for further clinical evaluation.
Laura Caberlotto is head of Biomarkers and Immunoassays Laboratory at Aptuit, Inc. She can be reached at firstname.lastname@example.org.