The costs and effectiveness of large Phase III pre-licensure vaccine clinical trials.

Prior to the 1980s, most vaccines were licensed based upon safety and effectiveness studies in several hundred individuals. Beginning with the evaluation of Haemophilus influenzae type b conjugate vaccines, much larger pre-licensure trials became common. The pre-licensure trial for Haemophilus influenzae oligosaccharide conjugate vaccine had more than 60,000 children and that of the seven-valent pneumococcal conjugate vaccine included almost 38,000 children. Although trial sizes for both of these studies were driven by the sample size required to demonstrate efficacy, the sample size requirements for safety evaluations of other vaccines have subsequently increased. With the demonstration of an increased risk of intussusception following the Rotashield brand rotavirus vaccine, this trend has continued. However, routinely requiring safety studies of 20,000-50,000 or more participants has two major downsides. First, the cost of performing large safety trials routinely prior to licensure of a vaccine is very large, with some estimates as high at US$200 million euros for one vaccine. This high financial cost engenders an opportunity cost whereby the number of vaccines that a company is willing or able to develop to meet public health needs becomes limited by this financial barrier. The second downside is that in the pre-licensure setting, such studies are very time consuming and delay the availability of a beneficial vaccine substantially. One might argue that in some situations, this financial commitment is warranted such as for evaluations of the risk of intussusception following newer rotavirus vaccines. However, it must be noted that while an increased risk of intussusception was not identified in large pre-licensure studies, in post marketing evaluations an increased risk of this outcome has been identified. Thus, even the extensive pre-licensure evaluations conducted did not identify an associated risk. The limitations of large Phase III trials have also been demonstrated in efficacy trials. Notably, pre-licensure trials of pneumococcal conjugate severely underestimated their true effect and cost-effectiveness. In fact, in discussions prior to vaccine introduction in the USA for PCV7, the vaccine was said to be not cost-effective and some counseled against its introduction. In reality, following introduction, PCV7 has been shown to be highly cost-effective. In the last decade, new methods have been identified using large linked databases such as the Vaccine Safety Datalink in the USA that allow identification of an increased risk of an event within a few months of vaccine introduction and that can screen for unanticipated very rare events as well. In addition, the availability of electronic medical records and hospital discharge data in many settings allows for accurate assessment of vaccine effectiveness. Given the high financial and opportunity cost of requiring large pre-licensure safety studies, consideration could be given to 'conditional licensure' of vaccines whose delivery system is well characterized in a setting where sophisticated pharmacovigilance systems exist on the condition that such licensure would incorporate a requirement for rapid cycle and other real-time evaluations of safety and effectiveness following introduction. This would actually allow for a more complete and timely evaluation of vaccines, lower the financial barrier to development of new vaccines and thus allow a broader portfolio of vaccines to be developed and successfully introduced.