Description and evaluation of serologic assays used in a multicenter trial of acellular pertussis vaccines.

OBJECTIVE

To describe and evaluate the assays used to measure the antibody responses in infants to 13 experimental acellular pertussis vaccines and 2 licensed whole-cell pertussis vaccines.

METHODS

During a 53-week period, preimmunization and postimmunization sera were assayed for immunoglobulin G antibodies to pertussis toxin, filamentous hemagglutinin, pertactin, and a mixture of type 2 and type 3 fimbriae by enzyme-linked immunosorbent assay (ELISA), for whole-cell agglutinins (AGG), and for pertussis toxin-neutralizing antibodies by the Chinese hamster ovary cell assay. All ELISA reagents were characterized to assure antigen and isotype specificity of the assays. Intralaboratory reproducibility and temporal stability were evaluated by analysis of results of control sera and by assessment of the response to the control whole-cell vaccine. Interlaboratory reproducibility was assessed by repeating the assays on preimmunization and postimmunization sera for 10% of the infants in a second laboratory.

RESULTS

For control sera having antibody concentrations at least four times the minimum level of detection, the coefficients of variation within and between the ELISAs consistently were less than 20%. Trend analysis indicated that none of the assays drifted by more than 20% during the study period, and no significant drift was seen in the response to the control whole-cell vaccine. Results from the two laboratories correlated well; correlation coefficients were .93 or greater for the four ELISAs, .79 for the Chinese hamster ovary cell assay, and .82 for the AGG assay. For four of the six assays, there was either no difference or a modest (< 15%) difference in the geometric mean values for sera tested in both laboratories. Larger quantitative differences were observed for the AGG (45% difference) and pertactin (61% difference) assays.

CONCLUSION

Assay reproducibility and stability indicate that the standardized methods can be transferred between laboratories, and that the results accrued during a 1-year period for the 15 vaccines can be compared.