| Abstract: | In order to confirm the data reported in the previous papers, variously prepared pertussis vaccines were employed in the present investigation. Pertussis organisms grown either on a solid or in a liquid semisynthetic medium were treated by: (1) heating at 56 C for 30 min, (2) storage in 0.1% formalin at 37 C for 5 days, (3) storage in 0.1% formalin at 25 C for 5 days, and (4) simple addition of sodium ethyl-mercuri thiosalicylate (merthiolate) as a preservative. Freeze-dried vaccines were made from these preparations four and a half months later. Mice were immunized intraperitoneally and challenged intracerebrally with a virulent strain of Bordetella pertussis 10 or 17 days later. The data were analyzed statistically assuming the probit corresponding to the percentage of survivors at any dose be a linear function of the logarithms of the dose. The 50% effective doses (ED50) and slopes of each vaccine were found to be uniform. More accurate estimates of ED50 were obtained by employing a pooled slope in each experiment. From these ED50 values, the relative potency was estimated by comparing the value of a vaccine to that of a dried merthiolate-vaccine. For vaccines derived from solid cultures, with an immunization period of 17 days, the relative potency of the vaccine heated at 56 C was 0.63 (95% fiducial limits=0.52 to 0.76); the value for the formalinized vaccine at 37 C was 0.40 (0.30 to 0.53) and one at 25 C was 0.51 (0.34 to 0.77). Vaccines derived from liquid cultures showed a relative potency of 20 to 50% less than that of corresponding vaccine derived from a solid culture. The potency obtained for the 17 day immunization period was usually higher than that for the 10 day period. Using the overall-pooled slope, an experimental design which will be appropriate for statistical analysis is discussed. |
