Interventions given to healthy persons to prevent disease are required to have a low risk-to-benefit ratio when compared to therapeutic interventions. Many childhood vaccines that are recommended for universal use by ACIP are required by states for attendance in licensed day care facilities and public schools, and thus administered to large segments of the population (e.g., nearly the entire annual birth cohort of more than 4 million children). Some adult vaccines are also universally recommended, others are recommended for specific occupations (e.g., health care workers) and, in some cases, required by employers. A substantial proportion of the adult population receives influenza vaccine each year (for example, between two-thirds and three-quarters of adults 65 years of age or older received influenza vaccination in 2008; CDC, 2006).
The process of anticipating, detecting, and quantifying the risks of rare adverse events following immunization presents an enormous challenge. Vaccine studies submitted as part of Biological License Applications to FDA’s Center for Biologics Evaluation and Research (CBER) have historically included several thousand individuals. Rare but serious adverse events may follow vaccination, sometimes at rates in the range of one in a million vaccine recipients. Even vaccine trials including 100,000 or more participants may not have adequate statistical power to detect such rare adverse events. Delaying licensure after efficacy has been shown in order to amass additional evidence related to rare adverse events associated with a candidate vaccine would result in continuing cases and deaths due to the preventable disease.
After FDA licensure, as knowledge about a vaccine’s safety increases when large numbers of individuals are immunized, additional safety assessment becomes possible, complementing pre-licensure data. Vaccine safety researchers both at FDA and outside government have emphasized the equal importance of adequate pre-licensure study and post-licensure surveillance for “signals” of adverse events. It is challenging to detect a true “signal” of a vaccine safety problem amidst the considerable “noise” of coincidental, only temporally related events.
Licensure of second generation rotavirus vaccines offers a clear example in which pre-licensure studies put a special emphasis on vaccine safety because of knowledge about the risk of intussusception acquired after introduction of the first licensed rotavirus vaccine. The large studies undertaken for the second-generation rotavirus vaccines—70,000 subjects for RV5 (bovine-based, RotatTeq) and nearly 75,000 for RV1 (human-based, Rotarix)—reflect a specific safety concern related to the first rotavirus vaccine (Ellenberg et al., 2005; GSK, 2008; Heyse et al., 2008).
The FDA Amendments Act of 2007 has strengthened CBER’s authority to require post-licensing studies. FDA may require the manufacturer to conduct post-licensure studies of vaccine safety that meet certain specifica-