advertisement
Editors Selection IGR 9-1
Medical treatment: Postmarketing surveillance
Comment by Gary Novak on:
17211 Scientific challenges in postmarketing surveillance of ocular adverse drug reactions, Fraunfelder FW; Fraunfelder FT, American Journal of Ophthalmology, 2007; 143: 145-149
As a way to discuss the challenges of postmarketing surveillance for adverse events in the practice of ophthalmology, Fraunfelder and Fraunfelder (318) select three medications: systemic sildenafil-associated or amiodarone-associated optic neuropathy, and ocular chloramphenicol-associated blood dyscrasias. Based upon retrospective review of case reports in the medical literature and adverse event reports, they recommend that patients who experience optic neuropathy with sildenafil or amiodarone stop treatment and seek ophthalmological consult. With respect to chloramphenicol, they recommend that previous blood dyscrasias or family history be a contraindication for use, and that physicians should limit the dose and treatment duration.
The authors note that the voluntary systems in place in the U.S. has many limitations.
This perspective provided by the Drs. Fraunfelder gives the reader many issues to consider in a world where approved drugs are being withdrawn due to safety issues found in the post-approval period and changing benefit-risk ratios (e.g., Vioxx® (rofecoxib), Permax® (pergolide) and Zelnorm® (Tegaserod). Not mentioned in the authors' brief Perspective are existing systems, such as the Japanese requirement for a mandatory 6-year post-market surveillance study for all approved products, and the benefit of captive health care systems for evaluating drug use and safety (e.g., Canada and the Netherlands). These systems have the advantage not only of reporting the adverse event ('the numerator'), but also quantifying the exposure ('denominator'). Pre-approval studies have three major limitations in predicting safety after approval - they are relatively small, of relatively short-term duration, and enroll relatively homogenous patients who may not be reflective of who will get the medication once it is approved. A simple rule of thumb might also be helpful - sample size needs to be about three times the incidence rate for detection. So, in order to detect an adverse event with a frequency of 1%, you need to evaluate 300 patients.
Public policy shifts from demands for early approval to allow for the potential benefits of an effective therapy for affected individuals, to a demand for later approval to better assess safety - especially rarer events and those occurring only with long-term therapy. Not appreciated in the pages of our newspapers is that these concerns are really two ends of a continuum - the risk/benefit ratio which regulators must assess in their decision to approve a pharmaceutical for marketing, and thus wider use. Recent product withdrawals have stimulated ideas on improving our drug regulatory process, and we look forward to a collaborative, scientifically based effort to reach this goal.
Comments
The comment section on the IGR website is restricted to WGA#One members only. Please log-in through your WGA#One account to continue.Log-in through WGA#One
