Biomarkers Utilitarian Classification

Biomarkers are the stepping-stones for modern drug discovery and development [1-4]. Biomarkers are defined as biological substances or biophysical parameters that can be monitored objectively and reproducibly and used to predict drug effect or outcome. This broad definition is, however, of little utility to the pharmaceutical process since it carries no qualification for the significance and use of the biomarker. The following classes and definitions of biomarkers are therefore offered:

1. Target validation: biomarkers that assess the relevance and the potential for a given target to become the subject of manipulation that will modify the disease to provide clear therapeutic benefits while securing a sufficient therapeutic index of safety and tolerability.

2. Compound-target interaction biomarkers: biomarkers that define the discrete parameters of the compound (or biological) interaction with the molecular target. Such parameters include binding of the compound to the target, its residency time on the target, the specific site of interaction with the target, and the physical or chemical consequences to the target induced by the compound (or biological).

3. Pharmacodynamic biomarkers: biomarkers that predict the conse-quence(s) of compound (biological) interaction with the target. The pharmacodynamic biomarkers include events that are desired thera-peutically and adverse events based on mechanism of action. Pharmacodynamic biomarkers can report on discrete molecular events that are proximal to the biochemical pathway that is modified by the manipulated target or remote consequences such as in vivo or clinical outcomes (morbidity or mortality). Pharmacodynamic biomark-ers are diverse and frequently nonobvious. Advanced and sophisticated bioinformatics tools are required for tracking the divergence and convergence of signaling pathways triggered by compound interaction with the target. A subset of the pharmacodynamic biomarkers are consequences induced by the compound outside its intended mechanism of action. Such pharmacodynamic effects are often termed "off-target" effects, as they are not the direct consequence of the compound interaction with the target. Usually, such pharmacodynamic events are due to unforeseen lack of selectivity or metabolic transformations that yielded metabolites not present (or detected) in the animals used for safety and metabolic studies prior to launch of the compound into human trials or into human use. These issues are not dealt with in this chapter.

4. Disease biomarkers: biomarkers that correlate statistically with the disease phenotype (syndrome) for which therapeutics are developed. Correlation of levels (in the circulation, other fluids or tissue) or expression patterns (gene, protein) in peripheral blood cells or tissues should signify disease initiation, progression, regression, remission, or relapse. In addition, duration of aberrantly expressed biomarkers could also be associated with risk for disease, even if the level of the biomarker does not change over time. Since disease biomarkers are defined by their statistical correlation to features of the disease, it is imperative that the clinical phenotyping is clearly defined. Stratification of all possible phenotypic variables is clearly a prerequisite for accurate assessment of the discrete relationships of the biomarker to the disease. Gender, age, lifestyle, medications, and physiological and biochemical similarities are often not sufficiently inclusive, resulting in a plethora of disease bio-marker claims that are often confusing and futile.

5. Patient selection: biomarkers that are used for selection of patients for clinical studies, specifically proof-of-concept studies or confirmation phase III clinical trials that are required for drug registration. These biomarkers are important in helping to select patients likely to respond (or conversely, not respond) to a particular treatment or a drug's specific mechanism of action, and potentially predict those patients who may experience adverse effects. Such biomarkers are frequently genetic (single-nucleotide polymorphism, haplotypes) or pharmacogenomic biomarkers (gene expression), but could be any of the primary pharma-codynamic biomarkers. Biomarkers for patient selection are now mainstream in exploratory clinical trials in oncology, where genotyping of tumors in view of establishing the key oncogenic "driver(s)" are critical for the prediction of potential therapeutic benefits of modern treatments with molecular targeting drugs. The success of the new era of molecular oncology (as compared to the cytotoxic era) will depend largely on the ability to define these oncogenic signaling pathways via biomarkers such as phosphorylated oncogenes, or the functional state due to mutations that cause gain or loss of function.

6. Adaptive trial design: The objectives of adaptive design trials are to establish an integrated process to plan, design, and implement clinical programs that leverage innovative designs and enable real- time learning. The method is based on simulation-guided clinical drug development. In a first step, the situation is being assessed, the path forward and decision criteria defined, and assumptions analyzed. Adaptive trials have become an enabler strategy, and they work to integrate competing positions and utilities into a single aligned approach and to force much clearer articulation and quantification on the path forward. Once this framework is established, a formal scenario analysis that compares the fingerprints of alternative designs through simulation is conducted. Designs that appear particularly attractive to the program are further subjected to more extensive simulation. Decision criteria steer away from doses that are either unsafe or nonefficacious and aim quickly to hone in onto the most attractive dose range. Response-adaptive doseranging studies deploy dynamic termination rules (i.e., as soon as a no-effective-dose scenario is established, the study is recommended for termination). Bayesian approaches are ideally suited to enable ongoing learning and dynamic decision making [5]. The integrator role of adaptive trials is particularly strong in establishing links between regulatory accepted "confirm"-type endpoints and translational medicine's efforts to develop biomarkers. Search for biomarkers that may enable early decision making need to be read out early to gain greater confidence in basing decisions on them. A biomarker can be of value even if it only allows a pruning decision. These considerations highlight the importance of borrowing strength from indirect observations and use mathematical modeling techniques to enhance learning about the research question. For example, in a dose-ranging study, it is assumed that there should be some relationship between the response of adjacent doses, and this assumption can be used to model an algorithm. Both safety and efficacy considerations can be built into this model: ideally, integration of all efforts, from disease modeling in discovery to PK/PD modeling in early clinical development to safety/risk and business case modeling in late development [4-7].

The utility of this system is represented in Figures 1 and 2, which suggest a semiquantitative scoring system that helps assess the strength of the program

Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment