An integrated biomarkers approach speeds the road to drug discovery and development.

Traveling a long road without sign posts to direct you makes getting to your destination all the more difficult. Drug discovery and development is certainly a long road, and for many years there were few sign posts to guide us. Merck Research Laboratories (MRL), Rahway, N.J., has established three highly integrated biomarker departments whose goal is to find the right sign posts to navigate the pathway to novel and differentiated drugs. In response to the enormous challenges facing pharmaceutical R&D today, Merck is using its integrated approach to help streamline R&D and make better decisions earlier in the drug development process. By embedding strategies involving the discovery, characterization, and measurement of biomarkers throughout discovery and development, Merck is increasing the probability of success and decreasing the potential for late-stage failures. The systematic integration of these three dedicated departments is enabling Merck to identify likely drug candidate failures sooner, advance promising candidates to proof of concept more rapidly, and, as a result, evaluate more novel mechanisms of action.

Three departments, one goal

Biomarkers—biological indicators of normal physiological processes, disease or drug response—provide critical information, or directions, along the drug discovery and development pathway. Like many companies, Merck has been identifying, characterizing, and applying biomarkers for use in R&D for many years. For example, we have long invested in imaging technologies such as positron emission tomography (PET) and novel imaging probes (e.g., radiolabeled tracers) to assess the distribution of drug candidates in the body as well as their ability to engage therapeutic targets. The use of imaging has, and continues to be, particularly important in neuroscience research, where non-invasive or minimally invasive means of evaluating candidate drug-receptor interactions and assessing other pharmacological effects are critical. Likewise, Merck has used experimental (translational) medicine for decades to provide an early evaluation of the likely efficacy and safety of specific drug candidates. Finally, our investment in powerful gene expression, genomics, and bioinformatics capabilities over the past decade has provided a strong foundation for the identification of molecular signatures that can serve as biomarkers of disease risk, progression and sub-types, as well as the efficacy and safety of investigational drugs.

In 2004, Merck recognized that the systematic integration of these approaches across our key areas of research would be critical to achieving the goal of bringing innovative products to patients faster, with a focus on scientific excellence, evidence-based medicine, and drug safety. We made the strategic decision to establish three dedicated biomarker departments–Molecular Profiling, Imaging, and Experimental Medicine–structured so that they work closely with each other and with other functions from the earliest stages of drug discovery and development.

Delving deeper into diseases

Developing innovative therapies begins with obtaining a better understanding of the underlying biology of common illnesses including cardiovascular disease, diabetes, obesity, and Alzheimer’s disease. MRL’s Molecular Profiling department applies leading-edge “genetics of gene expression” and systems biology methods to decipher the complex pathways and networks that drive disease, and to identify both novel therapeutic targets and biomarkers. Biomarkers can be used for a wide range of applications in drug development, from identifying disease sub-types, to stratifying patients for clinical studies, and evaluating drug efficacy. Disease-related data from molecular profiling experiments can also be used to develop novel imaging probes that serve as markers of disease risk and progression and drug response.

Our ongoing research collaboration with FoxHollow into the characteristics of high-risk atherosclerotic plaque provides an example of how the Molecular Profiling and Imaging departments are collaborating to better understand disease and identify and develop biomarkers for important applications. Taking a cue from methods used by oncology researchers who examine the diseased tissue in cancerous tumors, we are harvesting atherosclerotic plaque–the diseased tissue in coronary heart disease–using the FoxHollow SilverHawk intravascular cutting device. The harvested plaque is analyzed using several different techniques (e.g., histology, gene expression, and protein profiling) to identify novel targets associated with plaque sub-types and patient sub-populations, such as patients with diabetes. Molecular imaging agents for these novel targets hold the potential to improve drug discovery and serve as potential diagnostic agents. Combined with fluid biomarkers and cardiovascular imaging approaches, these agents could help localize, identify, characterize, and track lesions, thereby helping to choose the appropriate therapy and monitor its effects.

The ultimate objective of identifying and applying molecular biomarkers is to advance the goal of personalized medicine. For example, scientists from MRL’s Molecular Profiling department, working in collaboration with partners, conducted fundamental studies that led to the development of a breast cancer prognosis test (Mammaprint, Agendia). This diagnostic recently became the first multi-gene expression test to be approved by the U.S. Food and Drug Administration, and could help assess the risk of recurrence and determine the best individualized treatment course for patients with breast cancer.

Fail fast, fail cheap

Historically, most key decisions about drug candidates in clinical development have been made in Phase II or Phase III, stages at which considerable time and resources have already been invested. A primary objective of integrating biomarker data throughout drug development is to identify failures as early as possible so that resources can be invested in evaluating alternative candidates perhaps targeting other mechanisms of action with a higher overall probability of success. Thus, “fail fast and fail cheap” is a mantra that guides our three biomarker departments and our overall approach to R&D at Merck.

Our integrated biomarkers approach is providing us with the information we need to make better decisions earlier in the development pathway. For example: We use RNA interference technology and gene expression analysis to determine whether down-regulating a novel target is associated with a biological effect in the preclinical setting. This helps in early target validation.

Culture change

Culturally, it is often difficult for scientists to make critical decisions based on reasonable certainty–the realm of most biomarkers–rather than absolute proof. At Merck, we now provide incentives to scientists to discontinue work on candidates or hypotheses that fail to demonstrate a reasonable expectation of early proof of concept. “Rewarding” the early acknowledgment of failure may be the best way to effect the cultural changes needed to make pharmaceutical R&D a more pragmatic process.

Driving to earlier proof of concept

Another key objective of our integrated biomarker strategy is to push promising candidates to proof of concept earlier, which is part of a larger MRL initiative called “Target Through Phase IIb”. As mentioned previously, molecular profiling and imaging data can help stratify patient populations and enrich proof-of-concept studies, potentially reducing sample size and study duration and improving the probability of success. In addition, imaging probes that assess target engagement can also help to establish appropriate dosing earlier in clinical development.

Experimental Medicine integrates the tools created by the other biomarker departments and creates novel tools and innovative, small-scale studies that help to establish proof of concept earlier. For example, we use an ad-libitum food intake model to assess the pharmacologic activity of novel anti-obesity compounds before large-scale clinical trials are initiated. Investigational agents that reduce food intake in this model, as benchmarked against an established anorexic agent, may be considered viable candidates for full clinical evaluation.

The Experimental Medicine department is also working with Merck’s cardiovascular franchise to identify novel antihypertensive agents with positive cardiovascular outcomes beyond peripheral blood pressure. Based on this need, we optimized pulse-wave analysis for use in clinical decision-making. In one trial, we benchmarked the positive cardiovascular effects of marketed anti-hypertensive agents using an augmentation index, which is an indicator of vascular compliance, central blood pressure and theoretically overall cardiovascular health. Once qualified, this platform was used to assess two novel anti-hypertensive compounds. In a single-dose study in healthy subjects, one compound did not meet pre-specified criteria for effects on these endpoints and was discontinued. A second compound behaved beautifully in another single-dose study in healthy subjects and was therefore selected for more extensive development in patients with hypertension.

Recent breakthroughs in science and technology may further expedite the path to proof of concept. For example, RNAi technology potentially could allow us to proceed from target selection directly to an “optimized lead,” which could be used for both preclinical and clinical proof of concept. This could significantly shorten the drug development pathway.

Embracing strategic collaborations

There are technical and practical challenges to using an integrated biomarker approach to facilitate earlier decision-making in R&D. Innovative therapies typically involve novel targets, and novel targets rarely have validated biomarkers. The process of validating or clinically qualifying novel biomarkers is long and arduous. Even for biomarkers that are not fully validated (the vast majority of biomarkers) adequate characterization for use in R&D requires significant investment.

External collaborations may help address this challenge. For example, consortia approaches, which combine the efforts of pharmaceutical companies with academic, government, and advocacy groups, can help to define new surrogate biomarkers. The benefits of such collaborations include access to broad expertise, patient samples, novel technologies, and proprietary information, along with facilitation and exchange of ideas and new technologies.

Collaboration with leading external scientists and participation in public-private consortia are key elements of our R&D strategy. Merck is a sponsor of the High-Risk Plaque Initiative, a major public-private collaboration that aims to discover better ways to predict and prevent cardiovascular disease and monitor the effectiveness of treatments. Scientists at Merck are also participating in the Alzheimer’s Disease Neuroimaging Initiative (ADNI), a consortium involving pharmaceutical companies, academic institutions and the National Institute on Aging. ADNI is using imaging technologies to track the progression of Alzheimer’s disease in people currently on the best therapies available for this condition. The data will serve to create or inform surrogate biomarkers, which help to evaluate whether new medicines can modify the progression of the disease.

Implementation of an integrated biomarker-driven R&D approach has had a profound impact on strategic decision making at Merck and has helped advance specific candidates more expeditiously, while maintaining a focus on scientific rigor and safety. We expect to see the full fruits of this approach over the coming years, as we continue to eliminate likely failures early and focus our resources on candidates with the highest probability of success. Ultimately, this will help deliver the best therapies to patients.