Characterization of serological responses to pertussis

Abstract

We have compared the use of five nonvaccine antigens to the use of conventional vaccine antigens, pertussis toxin (PT), and filamentous hemagglutinin (FHA) for the serological diagnosis of pertussis by enzyme-linked immunosorbent assay (ELISA). The nonvaccine antigens included the catalytic region of adenylate cyclase toxin (CatACT), the C-terminal region of FHA (C-FHA), lipooligosaccharide (LOS), the peptidoglycan-associated lipoprotein (PAL), and the BrkA protein. The serological responses of individuals with culture-confirmed pertussis were compared to those of adults with no recent history of a coughing disease. An immunoglobulin G (IgG) ELISA for PT was the most sensitive (92.2%) test for the serodiagnosis of pertussis. Of the nonvaccine antigens, ELISA for IgG responses to CatACT (sensitivity, 62.8%), C-FHA (sensitivity, 39.2%), and LOS IgA (sensitivity, 29.4%) were less sensitive but could also distinguish culture-positive individuals from control individuals. The use of a combination of multiple ELISA targets improved the sensitivity of the assay for serological diagnosis. Elevated IgG and IgA antibody titers persisted for more than a year in the individuals with culture-confirmed pertussis.

FIG. 1.

Distribution of responses to vaccine antigens in the culture-positive individual group and the control group. (A) PT IgG; (B) PT IgA; (C) FHA IgG; (D) FHA IgA. Visit 1, antibody values for serum obtained at the time of diagnosis; visit 2, antibody values for serum obtained approximately 1 year after diagnosis; visit 3, antibody values for serum obtained approximately 2 years after diagnosis. Box-whisker plots represent boxes for medians with 25th and 75th percentiles and whiskers for 10th and 90th percentiles. Closed circles represent outliers. The numbers in each box chart represent the probability values obtained by Student's t test.

FIG. 2.

Distribution of responses to nonvaccine antigens in the culture-positive individual group and the control group. (A) CatACT IgG; (B) CatACT IgA; (C) C-FHA IgG; (D) C-FHA IgA; (E) LOS IgG; (F) LOS IgA; (G) PAL IgG; (H) PAL IgA. Visit 1, antibody values for serum obtained at the time of diagnosis; visit 2, antibody values for serum obtained approximately 1 year after diagnosis; visit 3, antibody values for serum obtained approximately 2 years after diagnosis. Box-whisker plots represent boxes for medians with 25th and 75th percentiles and whiskers for 10th and 90th percentiles. Closed circles represent outliers. The numbers in each box chart represent the probability values obtained by Student's t test.

FIG. 3.

ROC analysis of tests. (A) ROC curves of seven assays with predictive value (PT IgG, CatACT IgG, FHA IgA, C-FHA IgG, FHA IgG, LOS IgA, and PT IgA); (B) ROC curves of assays that lack predictive value (CatACT IgA, PAL IgG, LOS IgG, PAL IgA, and C-FHA IgA).

FIG. 4.

Distribution of responses to LPS of B. parapertussis in the culture-positive individual group and the control group. (A) LPS IgG; (B) LPS IgA. Visit 1, antibody values for serum obtained at the time of diagnosis; visit 2, antibody values for serum obtained approximately 1 year after diagnosis; visit 3, antibody values for serum obtained approximately 2 years after diagnosis. Box-whisker plots represent boxes for medians with 25th and 75th percentiles and whiskers for 10th and 90th percentiles. Closed circles represent outliers. The numbers in each box chart represent the probability values obtained by Student's t test.