Post-Marketing Assessments of Serious Adverse Drug Reactions Reported by the Manufacturer to the FDA Differ Markedly with Those from an Independent Pharmacovigilance Program: Empirical Findings Based on Bisphosphonate-Associated Osteonecrosis Descriptions by the Research on Adverse Events and Reports Project and the Product Manufacturer.

Background: The Research on Adverse Events and Reports (RADAR) project, an independent pharmacovigilance project, has reported that post-marketing dissemination of clinical information on systemic adverse drug reactions (sADRs) in the FDA approved package labels of cancer drugs is often incomplete, delayed, or absent.

Methods: We compared post-marketing clinical assessments describing one previously unreported sADR: zoledronate (Z) and pamidronate (P)-associated osteonecrosis (ON) of the jaw. Data sources included 1) information submitted to the FDA by the manufacturer in support of an FDA package insert revision, 2) RADAR investigators review of case reports of BPs associated ON of the jaw reported to the FDA MedWatch program and, 3) RADAR investigators review of medical and dental records of BP treated multiple myeloma (MM) patients with confirmed, clinically obvious ON of the jaw at a large NCI-designated comprehensive cancer center.

Results: The manufacturers summary material indicated that ON occurs at a rate of < 1 per 10,000 BPs treated individuals; occurs 4x more frequently in cancer versus non-cancer patients; has multiple causes including trauma, infection, radiation therapy, and long-term corticosteroids; and causal relationship with BPs is unknown. Radar investigators found that not a single case of BPs associated ON of the jaw has been reported to the FDA’s MedWatch program as of to date. RADAR investigators, however, identified 7 patients (on BPs) with confirmed ON of the jaw among 600 myeloma patients seen between March 2001 and June 2004 at our institution. Characteristics of these patients included presentation with localized jaw symptoms (n=7) requiring surgical and/or medical treatment.

Conclusions: Independent toxicity assessments from the RADAR program suggest that ON is a serious, previously unrecognized ADR which can occur at a rate > 1% in MM patients, appears to be causally associated with prolonged BP therapy, and is difficult to treat. In comparison, materials submitted by the manufacturer to the FDA in support of a package insert revision as well as the FDA’s MedWatch data (with 0 cases reported) indicate that this sADR is rare, has multiple etiologies, and its association with BP is unknown. The manufacturer has informed the FDA that it will conduct a retrospective single-site study to further evaluate this toxicity but this method is likely to be limited by underreporting. In contrast, our findings highlight the need for 1) a multi-site prospective study to identify the true extent of this problem in patients receiving BPs, 2) educating health care officials about the importance of reporting ADRs to the FDA MedWatch program and, 3) careful evaluation of the source of toxicity information when assessing discrepant reports of sADRs.