October 9, 2014
Division of Docket Management (HFA-305)
Food and Drug Administration
5630 Fishers Lane, Rm. 1061
Rockville, MD 20852
Comments of the Cancer Prevention and Treatment Fund on
“Methodological Considerations in Evaluation of Cancer as an Adverse Outcome Associated With Use of Non-Oncological Drugs and Biological Products in the Postapproval Setting”
Docket No. FDA-2014-N-0731
The Cancer Prevention and Treatment Fund appreciates the opportunity to comment on Methodological Considerations in Evaluation of Cancer as an Adverse Outcome Associated With Use of Non-Oncological Drugs and Biological Products in the Postapproval Setting.
Challenges
Due to its biological complexity and often long latency, cancer represents a challenge for monitoring in the pre-approval or postapproval setting. Pre-market studies are usually too short-term or too small to adequately evaluate cancer as an adverse event. Unfortunately, many postapproval studies are also too short-term, too small, or have too many patients lost to follow-up to accurately evaluate cancer as an adverse event.
Existing voluntary databases for adverse event reporting in the post-market setting are known to greatly under-report and often do not contain sufficient information. Postapproval studies with pre-specified endpoints often have poor patient retention or are not completed. While cancer registries can provide another source of information they often do not contain detailed drug history and are not nationwide. Therefore, comprehensive systems for monitoring cancer as an adverse event for drugs and biologics do not exist at this time.
Long-term surveillance is critical
Cancer development is a multi-step biological process that can occur over several years, making long-term surveillance critical to identifying a cancer safety signal. Short pre-market studies that are not specifically designed to assess cancer outcomes are unlikely to identify this potential hazard. In addition, patient subgroups that may be particularly susceptible to cancer risks, such as children, adolescents and elderly adults taking multiple medications, are typically under-represented in pre-market studies, further decreasing the likelihood of detecting a cancer risk.
Post-market surveillance should consider all cancer types
While concerns over specific cancer types may arise from pre-market studies, post-market surveillance should include consideration of all cancer types. As any cancer signal can be difficult to detect pre-market, post-market surveillance needs to account for all possibilities. As mechanisms of carcinogenesis are often shared across different tissues of origin, the possibility of cancer at multiple sites cannot be ruled out.
In 2006, the Government Accountability Office issued a report stating that the FDA needs to address weaknesses in its post-market surveillance system, including the need for larger and more diverse datasets using uniform systems such as electronic health record information. That same year, the Institute of Medicine recommended that the FDA should take action to partner with other federal agencies and use all available resources to bolster its post-market surveillance efforts. We strongly agree. We recommend that the Sentinel System be used to mine specific data on the possible increased cancer risk due to the effects of non-oncological prescription medications and biological products, but Sentinel should not replace postapproval studies designed to evaluate those risks.
Concerns about cancer risks in pre-market development
If concern over a potential cancer risk arises during pre-market development, well-designed preclinical and clinical studies to directly examine this possibility need to definitively address this issue. Specifically, sponsors need to have a clear understanding of the biological mechanism of action of their product, and perform comprehensive carcinogenicity testing. This should include sensitized animal model systems as appropriate, and should also address non-genotoxic mechanisms of carcinogenicity, such as hormonal effects, which may display non-traditional pharmacological behavior, i.e. non-monotonic dose responses. Any cancer concerns identified in pre-market studies should also be adequately addressed in labeling information, special warnings, and post-market studies as needed, if the product is approved.
Conclusions
To better evaluate the challenges in designing postapproval studies to determine whether a non-oncological drug causes or influences cancer, the FDA should develop long-term studies that include susceptible subgroups such as children, adolescents and the elderly. The post-market surveillance should consider all types of cancer, and include large and diverse data sets utilizing electronic health records and the Sentinel System. However, Sentinel should not be used as a replacement for postapproval studies. Also, if pre-market studies flag a cancer concern, it should be addressed in the label of the drug with a special warning.
Cancer Prevention and Treatment Fund
The Cancer Prevention and Treatment Fund can be reached through Paul Brown at (202) 223-4000 or at pb@center4research.org