PURPOSE: Bacterial agglutination antibodies against Bordetella pertussis Yamaguchi and Tohama strains are frequently measured for the serodiagnosis of pertussis in Japan. To determine the serological criteria, we comparatively evaluated the titers of the bacterial agglutination antibody and anti-pertussis toxin (PT) antibody.
METHODS: We analyzed 36 definitive (fourfold increase in an agglutination antibody) and 137 presumptive (single-antibody high titer) cases with B. pertussis infection in adolescents and adults and 318 healthy volunteers as control.
RESULTS: When a single Yamaguchi agglutinin titer of ≥1:1,280 was taken as a diagnostic criteria (more than 3 standard deviations greater than the geometric mean of the control group), the sensitivity and specificity at 4-5 weeks after cough onset were 58% and 98%, respectively. Using this criterion, the clinical findings of the presumptive cases were almost identical to the definitive cases. When 318 control sera were compared between these two tests, no association was found between the Tohama agglutinin titer and that of the anti-PT antibody, whereas a weak association was observed between the Yamaguchi agglutinin titer and the anti-PT antibody. When a high antibody titer was compared between these two tests using pertussis cases, 60% of cases with a Yamaguchi agglutinin titer of ≥1:1,280 showed an anti-PT antibody titer of ≥100 EU/mL.
CONCLUSIONS: Bacterial agglutination test is a low sensitivity method for the diagnosis of B. pertussis infection. Approximately 40% cases were missed at 4-5 weeks after cough onset when a Yamaguchi agglutinin titer ≥1:1280 was used as the criterion. The present results also indicated that the association between bacterial agglutination and anti-PT antibody titers was weak, and bacterial agglutination antibody titers were not found to be useful for predicting anti-PT antibody titers.
CLINICAL IMPLICATIONS: Bacterial agglutination test is not reliable diagnostic test. Anti-PT antibody which is known to be highly sensitive as well as highly specific might be measured to yield an accurate diagnosis.
DISCLOSURE: The following authors have nothing to disclose: Naoyuki Miyashita, Kazunobu Ouchi, Niro Okimoto
No Product/Research Disclosure Information