A Two-Stage Sequential Design to Minimize Risk in a Large Vaccine Efficacy Trial

• Activity Number: 534
• Type: Invited
• Date/Time: Wednesday, August 6, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Ehab Bassily*+
• Companies: Sanofi Pasteur
• Keywords: Two stage sequential design ; Vaccine efficacy

Abstract

To carry out an efficacy trial in clostridium difficile (CD) it is necessary to have 250 cases of CD infection. About 15000 high risk subjects will be followed for 2.5-3.0 years to observe the 250 cases. About 200 clinical sites are needed to recruit the subjects who need to be scheduled for a surgery within 2 months or have been admitted twice into hospital in addition to antibiotic use in the last year. This makes the trial very expensive and resource intensive. Since the vaccine is a first in class product little is known about the likely vaccine efficacy. Estimates of sample size and trial duration were based on theoretical assumptions and computer simulations. More than one design was evaluated; the solution is a two stage design in which only part of the patient population is recruited for the initial phase. In this phase the sample size is chosen to make a futility decision. Later after the occurrence of more cases a go/no go decision is made. Then a large expansion of clinical sites and subjects allows achieving the 250 cases, the designed power and precise estimates of vaccine efficacy. Calculations and decisions leading to final design will be described in detail.