Redefining non-inferiority in anamnestic antibody responses using the mean increase of log-transformed antibody titers after revaccination: secondary analysis of a randomized controlled rabies vaccination trial

Non-inferiority in the anamnestic antibody response is conventionally determined by comparing seroconversion rates after revaccination. However, this approach is inadequate in the case of high pre-booster antibody titers. Therefore, we propose an alternative method to determine non-inferiority of booster responses. We used anonymized data from a randomized controlled trial (NCT01388985; EudraCT 2011-001612-62) in 500 adults, comparing a two-visit primary vaccination schedule (two intradermal 0.1 mL rabies vaccine doses on day 0 and 7) with a three-visit schedule (single intradermal 0.1 mL dose on day 0, 7, and 28). Participants were revaccinated intradermally (single dose) 1 to 3 years later. Rabies virus neutralizing antibody titers were measured on day 0 and 7 after revaccination. After log(3)-transformation of antibody titers, the mean increase in titers after revaccination was compared between schedules. Non-inferiority was defined as the lower bound of the two-sided 95% confidence interval not exceeding -0.369. Four hundred and ten participants fulfilled the inclusion criteria. The mean increase in log(3) titer was 2.21 and 2.31 for the two-visit and three-visit schedule, respectively. The difference between these increases was -0.10 [-0.28, 0.08], meeting the non-inferiority criterion. In conclusion, comparing mean increases in log-transformed titers after revaccination appears to be a feasible and more informative method of studying non-inferiority regarding the anamnestic antibody response.