Background/Purpose: Prolia has marketing authorization in the EU, US, Canada, Japan, and over 40 countries or administrative districts worldwide for the treatment of postmenopausal women with osteoporosis at high/increased risk for fracture. As part of pharmacovigilance activities, Amgen Global Safety (AGS) continually assesses postmarketing adverse events reported by health care providers, patients, and other sources. Spontaneous adverse event reports, while often with insufficient information, are the cornerstone of safety surveillance programs and help detect rare and serious adverse drug reactions. Here we characterize the Prolia postmarketing experience for serious infections, including opportunistic infections.

Methods: A cumulative analysis of all non-clinical trial postmarketing serious infection reports (defined as those leading to ER visit or hospitalization) for Prolia in the AGS Database was conducted as of May 10, 2014 from solicited and spontaneous sources. Using postmarketing exposure estimates, we calculated overall reporting rates over time as well as cumulative rates for several serious infection subtypes, including opportunistic infections. Each infection subtype includes ≥ 1 related MedDRA Preferred Terms (PTs).

Results: Cumulatively through May 10, 2014, the total estimated postmarketing Prolia exposure was 1,963,794 patient-years (p-ys). There were 1,232 postmarketing reports of serious infection. The top 5 most frequently reported PTs were pneumonia (236), urinary tract infection (166), cellulitis (145), diverticulitis (59), and sepsis (59). Time to onset after the first dose of Prolia (reported for 305 cases) was highly variable, ranging from 1 to 916 days (mean 156, median 82.5). Among the 7 reported cases of endocarditis, only 2 were confirmed by echocardiography; both cases had confounding factors. The reporting rate of overall serious infection has decreased over time since product registration (first reporting period in 2010: 153 cases/100,000 p-ys; current reporting period in 2014: 57 cases/100,000 p-ys). Cumulative reporting rates of serious infection subtypes were low (Table) and below the background rates estimated from insurance claims data. Few cases of opportunistic infections were reported and included herpes zoster (32 cases; 1.6 cases/100,000 p-ys), unspecified fungal infections (10 cases; 0.5 cases/100,000 p-ys), and mycobacterium tuberculosis (5 cases; 0.3 cases/100,000 p-ys).

Conclusion: Recognizing the limitations of postmarketing safety data, these data suggest that reporting rates of serious infections decreased over time, and rates of events reported are lower than estimated background rates. The benefit/risk profile for Prolia remains favorable. Ongoing safety surveillance continues through clinical studies and pharmacovigilance activities.