Prospective evaluation of an Australian pertussis toxin IgG and IgA enzyme immunoassay.

ABSTRACT

Serological diagnosis of recent pertussis infection is an important part of both clinical assessment and epidemiological documentation of this disease. Standardization of serological testing and interpretation remains challenging despite international efforts to improve it. Currently, determining the anti-pertussis toxin (PT) IgG titer is recommended as the most accurate serological test in Europe and the United States, while Australia relies predominantly on measurement of Bordetella pertussis IgA antibody responses. Using B. pertussis PCR and the WHO clinical case definition as reference standards, the diagnostic utility of in-house anti-PT IgG and anti-PT IgA assays was evaluated prospectively in an Australian community-based cohort (n = 327). Patients provided up to four consecutive serum samples to document the kinetics of antibody response and decay. Previously validated cutoffs for positivity were converted to international units by using WHO-approved reference sera. At currently used cutoffs, both anti-PT IgG (>94 IU/ml) and anti-PT IgA (>20 IU/ml) assays had good specificity (80% [95% confidence interval {95% CI}, 68 to 88%] and 87% [95% CI, 77 to 94%]), but anti-PT IgG assay was consistently more sensitive than anti-PT IgA assay across a range of cutoffs (60 to 79% [95% CI, 53 to 84%] versus 41 to 62% [95% CI, 34 to 69%]). The combination of anti-PT IgG and anti-PT IgA assays performed no better than anti-PT IgG assay alone. The anti-PT IgA response in children under 12 years of age was poor. The accuracy of serology was optimal between 2 and 8 weeks after symptom onset. Cutoffs of >94 IU/ml for anti-PT IgG and >20 IU/ml for anti-PT IgA correlated well with recent pertussis infection and were consistent with recent recommendations from the EU Pertstrain group. Anti-PT IgG assay was superior to anti-PT IgA assay as the test of choice for the diagnosis of pertussis from a single sample.