February 15, 2005 (Vol. 25, No. 4)
Scope and Depth of Agency’s Interest
Pharmacogenomics is said to have been created with the 1955 presentation of re-search relating excretory levels of isoniazid to therapeutic results (Frueh & Gurwitz; 2004). In the fifty years since then, the number of published articles referencing pharmacogenomics or pharmacogenetics has grown to the thousands.
But this expanding investigation of inter-individual variability of drug response has not produced any fundamental changes in product development or clinical practice. The sources of inhibitions within the pharmaceutical and biotech communities to embracing individualized drug therapy are probably legion.
However, one issue often referenced (and recognized by the FDA itself) has been uncertainty over the scope and depth of the Agency’s interest in placing pharmacogenomics within its regulatory decision-making activities.
FDA substantially clarified its expectations of pharmacogenomics through its “Draft Gui-dance for Industry: Pharmaco-genomics Data Submissions” (referred to in this article as the Draft Guidance) document on November 4, 2003. But the release of the Draft Guidance actually followed a series of public interactions intended to identify issues to be addressed in and to gauge support for the Draft’s key provisions.
The Agency sponsored a “Workshop On Pharmacogenetics/Pharmacogenomics In Drug Development and Regulatory Decision-Making” on May 16-17, 2002, in conjunction with companies represented on the Pharmaceutical Research and Manufacturers Of America (PhRMA), Preclinical Safety Committee (DruSafe), and the Pharmacogenetics Working Group (PWG).
This was followed by presentations to the FDA Science Advisory Board on April 9, 2003, which resulted in the Board’s endorsement of the Agency’s proposals to develop a draft guidance outlining when submission of pharmacogenomic data would be requested or required.
Following the release of the Draft Guidance, the FDA has expanded its solicitation of public comment on matters pertaining to this field. On July 29, 2004, the Agency conducted a workshop on pharmacogenomic combination product co-development.
In announcing this event, the FDA described its primary purpose to be to “provide an interactive forum for discussing industry and other perspectives and experience derived from the development of recently approved pharmacogenomic combination products…[and to] provide FDA with valuable information to consider during development of guidance for industry on the co-development of pharmacogenomic combination products for therapeutic and diagnostic use.”
The one day meeting, held in collaboration with the Drug Information Association (DIA), PhRMA, Biotechnology Industry Organization (BIO), Advanced Medical Technology Association (ADVAMed), Medical Device Manufacturers Association (MDMA), and the Pharmaco-genomics Working Group (PWG), is expected to lead to the eventual issue of a draft guidance on the co-development of pharmacogenomic combination products.
The FDA announced on August 11, 2004, that it had submitted the Draft Guidance and related materials to the Office of Management and Budget (OMB) for review and clearance. On December 27, 2004, the Agency reported in the Federal Register that OMB had approved the submission. Release of the final guidance is expected in early 2005.
The FDA makes clear at the outset of the Draft Guidance that it was written to facilitate “scientific progress in the field of pharmacogenomics” and “the use of pharmacogenomic data in informing regulatory decisions.” If finalized as drafted, the Draft Guidance will apply to sponsors holding pending investigational new drug applications, new drug applications, and biologics license applications.
It will also give recommendations to those sponsors on when to submit pharmacogenomic data during the product development and review process, the formats to use to submit that data and how the data will be used in regulatory decision making.
As used in the Draft, “pharmacogenomics” specifically does not include proteomics, metabolomics, or genetic or genomic techniques for biological product characterization.
The Agency concedes in the Draft Guidance that uncertainties over its use of pharmacogenomic data in the drug application review process has discouraged sponsors from initiating pharmacogenomic testing programs during product development.
While acknowledging the experimental nature of many such tests, the FDA clearly believes that fulfilling the potential of pharmacogenomics to individualize therapies that maximize effectiveness and minimize risk will require its active encouragement (and participation in the evaluation) of these tests in the context of product development.
Consequently, a key element of the Draft Guidance is the proposed program for voluntary submission of pharmacogenomic test data, which is referred to as a VGDS (voluntary genomic data submission). As proposed, VGDSs would be referred to a cross-center Interdisciplinary Pharmacogenomic Review Group (IPRG), which would act as a clearinghouse and resource for further policy development.
At the same time, the submission of pharmacogenomic test data, even if experimental, would be required to the extent the data are used by the sponsor in any preclinical or clinical trial to support decisions regarding such questions as dose selection, entry criteria, safety monitoring or subject selection, or otherwise to be used to support approval or in the drug label.
Thus, pharmacogenomic data can be made relevant to the Agency’s regulatory decision making by the use a sponsor may make of it during the product development process. However, the Draft Guidance goes on to suggest that the Agency may expect sponsors to conduct certain pharmacogenomic testing where such tests have become generally accepted within the scientific community.
The results of certain tests— identified in the Draft Guidance as “valid” biomarkers—can become appropriate for regulatory decision making because they are generated through a well-characterized analytical system. In addition, there exists an established scientific framework or body of evidence that explains their significance in measuring safety or efficacy.
But tests belonging to the distinctive class of such valid biomarkers are further distinguished as either “known” or “probable.” As defined in the Draft Guidance, a known valid biomarker is one which has been accepted in the broad scientific community, while a probable valid biomarker appears to have predictive value but has not been widely accepted or has been independently replicated.
A probable valid biomarker is likely to be one for which a sponsor has substantial data (and, thus, could be one it elects to rely upon in its own decision making), but which is not as well known or understood within the general scientific community.
The Draft Guidance makes clear that test results for known valid biomarkers should be submitted, although the extent of the reporting will be determined by whether the sponsor itself intends to rely on the results.
The attention to biomarkers in the Draft Guidance reflects the importance attributed to them in the FDA’s Critical Path Initiative (see GEN, FDA News & Analysis, January 15, 2005).
In the context of that Initiative, the Agency anticipates that today’s biomarkers will become tomorrow’s diagnostics, and continued progress in biomarker development will have wide applicability from in vitro diagnostics to therapeutic monitoring and assessment.
The recent market clearance issued by the Center for Devices and Radiological Health (CDRH) for the first DNA microarray diagnostic test (the Affymetrix GCS 3000Dx system for analyzing the genes from the cytochrome p450 family) is a case in point.
A set of charts are appended to the Draft Guidance to help sponsors understand when or whether pharmacogenomic data should be submitted, with the variables being the nature of the filing, the sponsor’s intentions regarding the data, and the classification of the data as a known or probable valid biomarker.
While it may be that the present experimental nature of most pharmacogenomic data will make it unreliable for regulatory decision making, the authors of the Draft Guidance clearly hope that the framework they have articulated for characterizing that data in that context will encourage sponsors to submit such data where they would not have in the past.
Such voluntary submissions, when aggregated under the auspices of the IPRG, could help establish future known valid biomarkers that will move the field of pharmacogenomics closer to its promise of individualized therapy.
Several pharmaceutical and biotechnology companies have submitted written comments to the Draft Guidance docket. Many of these comments offer extensive evaluations of the document while a number of others offer specific changes to be made to the text.
Although the commentators are almost uniformly supportive of the Agency’s interest in expanding the use of pharmacogenomic data in its regulatory decision making, many raised a number of questions regarding the document’s apparent interposition of a sponsor’s “decision making” with its “regulatory decision making” for purposes of identifying required submissions as well as the process for and subsequent treatment of voluntary submissions.
Several comments noted that pharmacogenomic data may be used through the discovery process to prioritize drug candidates without subsequently serving to address questions of safety or efficacy during the development testing of specific candidates.
Issues of confidentiality and control of intellectual property were prominent among comments addressed to the outline for voluntary submissions set forth in the Draft Guidance. Many also urged the Agency to develop in greater detail the process by which a pharmacogenomic test would become recognized and accepted for regulatory decision making as a known valid biomarker.
Several comments noted that the submission of pharmacogenomic data could be in conflict with state and federal medical and genetic privacy legislation. In fact, the Draft Guidance had alluded to the sponsor’s duty to establish its authority to disclose potentially personally identifiable health information through an appropriately structured informed consent process.
But the comments made on this point actually introduce a larger issue: the interaction between the Agency’s proposed collection and use of pharmacogenomic data for its regulatory purposes and the potentially conflicting activities of other entities able to influence the relevance of this data in the development and introduction of new medical products.
For example, the grouping effects of efforts by governmental and private health insurers to develop formularies to control the number and type of products prescribed for certain conditions could discourage progress toward “individualized” therapies without concurrent refinement in the classification of disease states by the medical profession.
The Agency acknowledges in the Draft Guidance that it functions at the crossroads where scientific research, product development, and clinical testing combine and interact to produce new drugs.
As it works toward finalizing the Draft Guidance, especially as part of a larger task to explore the goals and means of integrating the field of pharmacogenomics into its regulatory mission, the FDA can play a key role in bringing other agencies and organizations together with its core constituencies.
The goal is to identify and ameliorate potential conflicts that could forestall the introduction of well-characterized products that enhance human health as much by the therapeutic benefits they deliver as the adverse reactions they avoid.