December 2010

“Any day now” the FDA will be delivering its long-awaited draft guidance document on the development of companion diagnostics. The FDA’s history in this area may lend some skepticism about how much time may actually pass before draft guidance is actually released, or whether it will convert into an official guidance document or even full regulation. After all, there has been plenty of noise in this area over the past decade without much clarity. The Secretary's Advisory Committee on Genetic Testing (SACGT) suggested in 2001 that the FDA should regulate all genetic tests, and yet seven years later, the Secretary's Advisory Committee on Genetics, Health, and Society (SACGHS) was recommending genetic testing oversight including FDA regulation. Meanwhile, in 2005 the FDA released a white paper on the co-development of therapeutics and diagnostics, and here we are in a new decade with no additional clarity. The FDA’s decision to tackle laboratory developed tests (LDTs), including IVDMIAs, at the exact same time only adds to the complexity of providing clear guidance.

Alberto Gutierrez, director of the Office of In Vitro Diagnostics at the FDA, was recently quoted by Pharmacogenomics Reporter as saying, "We've been working on the companion diagnostic guidance for four or five years now, but we still can't get it done." So, we can take away that some day in the future – possibly early 2011 for initial guidance on the companion diagnostics regulatory pathway and later in the year for guidance on drug/diagnostic co-development – there will be some form of draft guidance, which may or may not translate into clear and actionable advice for industry.

These various FDA statements will of course be important for both diagnostic and drug companies to understand and use to guide development efforts. But, the FDA’s actions to date and the actions of developers in this area do provide insights into how drug companies can reduce their own commercial and regulatory exposure, without the risk of giving up significant upside: find an initial diagnostic partner early on in the development process.

The idea of “personalized medicine” is one that has been thrown around for years, but since Herceptin (and to a certain extent Gleevec), there have been few breakthrough triumphs in oncology, where much of the personalized medicine attention is focused. But, this is clearly changing as the science and commercial realities aim us towards a world of optimally selected patient/drug pairings.

Within this emerging reality, regardless of the uncertainty entered into the mix by the FDA, there are a number of reasons why companies developing targeted agents should look to involve a commercial diagnostic manufacturer at an early stage. From a scientific standpoint, engaging this process early on allows for the drug manufacturer to learn and adapt along the development road, rather than react at a later date. From a regulatory standpoint, pending additional FDA guidance, a package that connects a bona fide test with the drug should provide a clear picture for approval and labeling. And from a commercial perspective, a simultaneous offering of a test should allow for rapid uptake as there is reduced confusion on patient selection, and most importantly, access to a test that is essential for the prescription.

In other words, by pairing off at an early stage, a drug company can avoid the challenges that Amgen and BMS faced with their EGFR inhibitors in the face of emerging data on the importance of KRAS mutations. The somewhat backwards process of 1) discovery of the link between KRAS mutation and EGFR inhibitor non-response, then 2) initial adoption of testing and prescribing patterns based off of this finding in the real world, then 3) an FDA advisory panel blessing the link and recommending re-labeling, and finally 4) an updated label based on retrospective data, led to significant confusion in the market. This confusion was only exacerbated by the lack of an FDA approved test, and all told certainly impacted the revenues of Erbitux and Vectibix.

Industry seems to be taking the message from this example seriously, with diagnostic-drug tie-ups in early stages increasingly showing up on the ticker. There are challenges to these tie-ups, primarily due to the different business models and lack of goals alignment that can occur between a diagnostic and drug company. Most notable among these, the drug company’s goal of a 100% testing rate may not line up with the diagnostic company’s optimum price-volume point. But pharma companies appear to be weighing this challenge against the prospect of FDA roadblocks and market confusion – or worse yet, a complete lack of an available test upon launch – and coming down on the side of partnering early.

Considering the optimal partner is obviously an important and nuanced decision, but one interesting question is whether pharma companies could and should consider ensuring multiple testing modalities in their development process, whether this occurs through one partner developing 2+ approaches or simply partnering with more than one diagnostic company for a single drug. The discussion around LDT regulation has suggested a future where molecular tests that are linked with cancer treatment decisions will be considered higher risk, and that higher risk tests will require proof of clinical utility. It seems unlikely that each new test will need to “re-prove” clinical utility after it is initially established. But engaging more than one diagnostic modality during the clinical trial process could allow for clinical utility and validity establishment in multiple testing types, (and the availability of multiple testing types at drug launch). When considering the goal of a 100% testing rate, more modalities means more access / ability to test, suggesting that relationships that ensure multiple testing types may be an attractive option for pharma companies.

The FDA’s eventual guidance may help define these relationships and what they entail from a regulatory perspective, but it seems unlikely that any FDA guidance would disincentivize these tie-ups. If anything, if the FDA simultaneously restricts the use of laboratory developed molecular testing, pharma will be wise to pair off with a capable diagnostic partner at an early stage of development with their “personalized medicine” and enjoy the ride from studies to approval to commercial success on a bicycle built for two.








About the author: David Luckenbach is a Principal at the Frankel Group. To contact him regarding this post, email blog@frankelgroup.com.