Pharmacogenomics Draft Guidance Needs Clarification, Industry Tells FDA

Industry is pressing FDA to enhance the utility of its "companion guidance" on pharmacogenomic data submissions as the agency refines the draft document.

The new ¹ guidance is "intended to be used as a companion" to the March 2005 Pharmacogenomic Data Submissions guidance, according to FDA, but industry says the draft leaves them wanting more.

The original guidance highlights benefits to sponsors of voluntarily sharing pharmacogenomics data with FDA, separate from regulatory submissions (² (Also see "Pharmacogenomics Guidance Cites Benefits Of Voluntary Submissions" - Pink Sheet, 28 Mar, 2005.), p. 12).

The companion guidance "reflects experience gained since the issuance of that guidance with voluntary genomic data submissions, as well as with review by the FDA of numerous protocols and data submitted under INDs, NDAs, and BLAs," the draft states.

Industry comments uniformly praise FDA for its work on the guidance, but request revision in the document's scope, implications and clarity.

Abbott commends the agency for the draft, but says "This companion guidance contains certain recommendations that are not, in our view, relevant to the collection and reporting of pharmacogenomic data."

Johnson & Johnson endorses "the underlying theme... that the efficient and effective evaluation of pharmacogenomic data generated within the context of the pharmaceutical industry will be dependent on the generation of high quality analytical data and submission of sufficient supporting clinical information." However, the companion and original guidance have a "significant disconnect" in content, the company adds.

"The original document is directed toward the submission of data, whereas the proposed companion document is directed almost entirely to considerations for sample and assay quality," J&J states.

The content of the draft guidance could instead be addressed in a white paper, perhaps "on best practices for conduct of pharmacogenomics studies," using current or emerging future technologies, the Pharmaceutical Research and Manufacturers of America says.

The emphasis on microarray data and handling of RNA and DNA is "well intended," Sanofi Aventis notes, but "detracts from the actual focus and proposed requirements...The overall format and presentation is unlike other FDA guidance documents and lacks conciseness."

"A guidance providing advice at the level of laboratory practice would require regular, frequent revision lest it become obsolete. In the rapidly evolving field of genomics, this iterative revision process would represent a significant ongoing burden for agency personnel," PhRMA adds.

Pfizer expresses concern that the "detailed technical recommendations inevitably entail the use of designated products and potential endorsement of a particular product or system."

The use of microarray data is a screening, not diagnostic, tool, and therefore the draft guidance is "devoted to a scenario that we feel will occur only infrequently," Pfizer and PhRMA point out.

The draft guidance is part of FDA's effort to embrace the growing potential of "personalized" medicine.

"We need a consensus on how to do [pharmacogenomic] analysis and submit data," Federico Goodsaid, senior staff scientist in the Genomics Group at FDA, said at a Nov. 14 Drug Information Association webinar on the draft companion guidance.

FDA hopes to "facilitate scientific progress in the field of pharmacogenomics and to facilitate the use of pharmacogenomic data in drug development," and believes the original and companion guidances will "benefit sponsors considering the submission of either VGDS or marketing submissions containing genomics data."

Industry is hoping FDA can strike the right balance in encouraging the emerging field. Richard Hockett, a medical fellow at Lilly, said during the webinar, "After an initial burst of activity, the field stagnates without standardization, so guidance by regulators is a good thing - but not too much regulation."

New Designs For Clinical Trials?

A template could be helpful in designing focused, flexible and affordable trials, as well as data mining, collection and submission, according to industry. FDA is becoming more open to flexible trial design with population subset analyses (³ , p. 9).

PhRMA's comments on the draft companion guidance point out that "Pharmacogenomic studies, be they genetic, gene expression, or other, are typically add-ons to clinical trials that are powered and recruited to meet clinical endpoints."

With regard to sample size in clinical trials, FDA's request for "adequate coverage for ethnic/racial groups" in genotyping reports implies prospective studies and could be a problem with retrospective analyses, Novartis says.

"Many clinical studies do not have enough non-Caucasians to make definitive conclusions about these groups and analysis frequently needs to be limited to include only Caucasians," the company adds.

A critical issue is whether a company would have to start from the beginning with new clinical trials to get new pharmacogenomic data on drug labels, Raymond Woosley, president of the Critical Path Institute, told "The Pink Sheet."

"You don't want to start from scratch," Woosley said. "Not only would that be costly, but it's not the best use of clinical research ... What's encouraging [is that] FDA has demonstrated a willingness to be flexible."

A Future Companion

Several additions to the content of the companion guidance were suggested.

The distinction between exploratory and confirmatory data needs clarification, Merck Serono International said, as does reporting of data on different biomarker classifications and guidance on the level of validation.

Abbott and others request that FDA clarify requirements for Clinical Laboratory Improvement Amendments, certification of labs, and statistical analyses in data submissions.

"Lessons learned from electronic submissions of microarray gene expression data will be used for developing guidance for e-submission of other pharmacogenomic data," such as alternative gene expression platforms, proteomics and metabolomics, Weida Tong, director of FDA's Center for Toxicoinformatics, told the webinar.

There have been about 40 voluntary genomic data submissions and 25 face-to-face meetings since the original guidance was issued, Goodsaid added.

The next big class of drugs to use genomic data to guide therapy, Woosley believes, is antidepressants. There are some gaps in knowledge, but "they will be filled and we'll begin to subcategorize depression to better understand the chemical imbalance going on," he said.

Making Deals Along The Critical Path

"Hopefully, soon after that will be advances in subcategorizing Alzheimer's to better understand subpopulations - we don't know what those are yet, but every disease had myriad etiologies."

Alzheimer's has been the greatest revelation from the genome, Woosley said, because of diagnostic tests predicting biomarkers and disease progression.

"Point-of-care diagnostics are going to really expedite the use of the tests that we now know about. If you look at all the meetings, on every agenda is personalized medicine."

A new trend is pharma merging with diagnostic companies that can develop tests to enhance profiling and use of drugs.

In one example, Swiss firm Roche entered a confidentiality agreement Nov. 13 with Arizona-based cancer diagnostics company Ventana Medical Systems after several months of negotiations. The deal would give Roche access to Ventana's Pathway HER-2/neu tissue-based test, used to assess breast cancer patients being considered for treatment with Genentech cancer drug Herceptin (trastuzumab), of which Roche is majority owner (⁴ (Also see "Roche Seeks Diagnostics To Pair With Cancer Drugs In $3 Billion Ventana Bid" - Pink Sheet, 2 Jul, 2007.), p. 21).

Ventana and Woosley's CPI are developing a prototype for an industry-wide validation model for companion diagnostics.

"We're on the exponential phase of implementation" of pharmacogenomics, Woosley said. "Technology is really moving to point-of-care, which has been a major stumbling block ... Companies need to recognize that if they encourage use [of technology] they'll get good outcomes and the field will advance. I'd characterize the field as a rapidly growing teenager."

- Becky Jungbauer ([email protected])