Limitations of the complement-fixation test for distinguishing naturally acquired from vaccine-induced yellow fever infection in flavivirus-hyperendemic areas.

On the basis of previous studies, it has long been stated that 17D yellow fever (YF) vaccine generally does not induce complement-fixing (CF) antibodies, and that the presence of CF antibodies could be used in epidemiological studies to distinguish individuals infected with wild YF virus from vaccinated persons. In January 1979, seroepidemiological investigations were conducted during a YF epidemic in The Gambia, West Africa. Since a mass vaccination campaign was also in progress, it was important to confirm that the CF test could be used for serodiagnosis and determination of the incidence of natural YF infections. The serological responses of 58 individuals who received 17D YF vaccine were studied. The vaccinees fell into three gorups: 1) those with prevaccination YF neutralizing (N) antibodies; 2) immunological virgins without prevaccination YF-N antibody or hemagglutination-inhibiting (HI) antibodies to heterologous flaviviruses (Zika, West Nile, dengue 1, Uganda S, Spondweni, or Ntaya; and 3) those without prevaccination YF-N antibodies but with heterologous flaviviral HI antibodies. Vaccination of persons without prior flaviviral immunological experience resulted in monotypic YF HI and/or N antibody seroconversions, but no CF antibody response. The presence of prevaccination YF N antibodies blocked serological response to the vaccine in a high proportion of the cases; however, 24% of vaccinees in this group had a marked rise in log2 YF CF antibody titer (mean increase of 3.9). Thirteen (46%) of 28 persons without prevaccination YF N, but with heterologous flaviviral HI antibodies demonstrated YF CF antibody seroconversion or increase in titer following vaccination; in this group the mean increase in log2/ YF CF antibody titer was 2.1. The CF antibody response was generally broadly cross-reactive; but in a few individuals, the YF CF antibody response was homotypic. Nine different patterns of HI and CF homologous and heterologous antibody responses were defined and are discussed. The practical significance of these studies is that they demonstrate that in a high percentage of persons with prior flavivirus exposure, anamnestic serological responses to YF vaccine result in CF antibodies similar to those induced by natural YF virus infection. In Africa and tropical America, where the background of flaviviral immunity is high, it is imperative that seroepidemiologic investigations during or after YF outbreaks be conducted prior to vaccination.