Background: Infection with human papillomavirus (HPV) is a necessary step in the progression to cervical cancer. Many methods for HPV testing are currently available, most developed to detect pools of HPV types.
Objectives: To evaluate the HPV typing by molecular methods and to compare commercial kits with an established laboratory method.
Study design: Eighty-four cervical samples found to be positive for HPV DNA by GP5+/6+-polymerase chain reaction-enzyme immunoassay-reverse line blotting (PCR-EIA-RLB) were re-tested with two commercial methods, INNO-LiPA and Amplisense HPV typing, able to identify the HPV type predicted by PCR-EIA-RLB in 76 and 67 samples, respectively.
Results: The INNO-LiPA assay revealed HPV DNA in 75/76 samples (98.7%; 95% CI, 0.93-0.99) that would contain HPV types identifiable by this assay. The Amplisense HPV assay revealed HPV DNA in 58/67 samples (86.6%; 95% CI, 0.76-0.93) containing HPV types detectable by this assay. For samples with a single infection, the unweighted kappa for concordance of HPV typing was 0.87 (95% CI, 0.78-0.97) for PCR-EIA-RLB versus INNO-LiPA, 0.94 (95% CI, 0.87-0.99) for INNO-LiPA versus Amplisense HPV, and 0.82 (95% CI, 0.70-0.94) for PCR-EIA-RLB versus Amplisense HPV typing. PCR-EIA-RLB revealed 12 multiple infections, INNO-LiPA revealed 14, and Amplisense HPV revealed 5. The agreement among tests for samples with multiple infections was lower, giving kappa values of 0.44 (95% CI, 0.18-0.70) for PCR-EIA-RLB versus INNO-LiPA, 0.52 (95% CI, 0.19-0.85) for PCR-EIA-RLB versus Amplisense HPV and 0.43 (95% CI, 0.12-0.74) for INNO-LiPA versus Amplisense HPV.
Conclusions: In HPV-positive samples, the agreement among tests for HPV typing was high for single infections but markedly lower for infections with multiple HPV types.