Companion Diagnostics for Guiding Therapy Decisions: Bridging the Gap from Development to Commercialization

There has been significant controversy in the field of pharmacogenetics and its utility in the drug development process. For many years, its proponents have suggested the incorporation of pharmacogenetic approaches into the drug development process will yield safer and more efficacious therapies (Roses, 2004) or improve postmarketing safety (Lesko and Woodcock, 2004). Skeptics have countered that the role of genetic biomarkers is limited to salvaging therapies that otherwise would not be granted approval. While there has been increasing focus on biomarkers of drug action, there has traditionally been more impetus to adopt the use of assays that identify variations in drug metabolism. This focus is exemplified by the recent IVD clearance of the first diagnostics for identifying Cytochrome P450 2D6 and 2C19 variants (Roche Diagnostics). Other examples include the relabeling of Camptosar® (irinotecan) to include information on UGT1A1*28 genetic variants, and the identification of patients with an inherited deficiency of the enzyme thiopurine methyltransferase (TPMT) prior to treatment with Purinethol (6-mercaptopurine, 6MP). Nonetheless, the clinical utility of these tests has not been realized. The value of genostratification in drug development, absent clear safety benefits, has yet to be demonstrated.

Several initiatives from the US FDA, including the Drug-Diagnostic Co-Development Concept Paper and the Critical Path Initiative, seek to promote biomarker usage in an attempt to speed the development process and create safer compounds. The Co-Development Concept Paper provides a framework for combination product submission, but lacks a sustainable business model to account for the differences in timing between clinical development studies and diagnostic device trials. Indeed, the larger issue becomes the challenge of knowing which biomarkers will be needed for a particular compound early enough in development to obtain the clinical validation of the biomarker itself. Once those data have been compiled, prioritizing the biomarker(s) within an IVD company’s pipeline is another matter altogether.

One alternative approach that provides a model for immediate access to the test upon drug approval is the use of reference laboratories to launch an in-house developed method, or “home-brew” test. The regulation of companion diagnostics in general and pharmacogenetic tests in particular remains the most critical issue in the development of diagnostic business models to support new therapies. Drugs and diagnostics are regulated differently by the FDA (Melzer, 2003). While FDA regulates diagnostic devices sold as kits (Medical Device Amendments of 1976), CMS (Center for Medicare and Medicaid Services) regulates diagnostic tests that are developed and performed in clinical laboratories under CLIA.

Generally, FDA clearance of in vitro diagnostic devices (IVD) includes evaluation of the performance claims of the assay, while CLIA laboratory inspections focus on reference laboratory quality standards. Many state health departments require their own certifications for tests performed on patients from their state, and these inspections review assay validation reports for in-house developed tests in detail. The Clinical and Laboratory Standards Institute (CLSI) publishes standards on assay validation and performance. Furthermore, large reference laboratories generally have adopted assay validation and acceptance policies to comply with pharmaceutical industry guidelines and expectations, including ICH GCP and GLP, and ISO accreditation.

Access to a companion diagnostic assay is important, but only one of many factors influencing adoption of the test in clinical practice. Drug labels that currently mention pharmacogenomic associations vary in their assertions on the use of the test (Shah, 2004). For the test to be used, its intended use in clinical practice should be clear. Another factor that affects the adoption of pharmacogenetic biomarkers is the availability of an inexpensive monitoring test.

The applications of biomarkers to specific therapies can further impact the test itself in terms of availability, standardization, regulatory status, and drug label. Markers that assess the biological effect of a targeted therapy can be clearly linked to that therapy, both in terms of market education, test commercialization, and in the drug label. In these cases, the test itself may be branded on the drug association (e.g., HercepTest®). Contrast these specific markers with broadly applicable panels of markers having clinical value for many indications—drug metabolizing enzyme (DME) polymorphisms being a good example. Cytochrome P450 isoforms or other DME genetic tests are mentioned in many drug labels. While the test application itself may be targeted to particular medical specialties (e.g., Roche AmpliChip™ and psychiatry), DME genotyping has potential utility across many therapeutic areas. Because of this widespread value of such testing, panels of DME markers are unlikely to be associated with specific therapies. However, the emerging combination product scenario is not so dichotomous. There are several drugs in clinical development for which both specific markers of biological effect and DME markers have proven useful. These situations present a challenge to the industry in terms of analytical platforms, IVD submissions, reimbursement, and testing in reference laboratories. Reference laboratories can accommodate custom panels more easily when they utilize standard analytical platforms. Open applications that have flexible input for desired markers (e.g., bead-based flow applications, TWT, real-time PCR multiplexes) are well suited to these applications. Fixed array platforms or static kits are less desirable in terms of laboratory throughput, and take considerably more time and to update with additional markers for a given application. Regardless of the methodology, however, the combination of broadly applicable genetic biomarkers with specific markers of drug action presents a unique challenge to the diagnostics industry.

Pharmaceutical companies vary widely in their use of pharmacogenetics and novel biomarkers in clinical development studies. Concerns such as market segmentation, cost-benefit considerations, and patient enrollment numbers vs. allelic frequency affect adoption in clinical research studies. Biomarkers may have value in identifying patients most suitable for a given therapy, however, independent of their value in drug development. There is no doubt that the prudent application of clinically validated biomarkers can improve response to treatment and help reduce adverse drug reactions. The Drug-Diagnostic Co-Development framework and Critical Path Initiative comprise a good start to move the industry further towards precision healthcare. The reference laboratory will continue to play an important role in advancing this effort.