Performance of Aptima and Cobas HPV testing platforms in detecting high-grade cervical dysplasia and cancer.

ABSTRACT

BACKGROUND: Human papillomavirus (HPV) tests and genotyping have been used in clinical risk assessment. The purpose of this study was to analyze the performance of 2 common HPV testing platforms in detecting high-grade cervical lesions (high-grade squamous intraepithelial lesion [HSIL] or worse [≥HSIL]). METHODS: Between January 1 and December 31, 2015, 2041 Papanicolaou (Pap) tests with biopsy confirmation were analyzed along with HPV tests performed on Cobas or Aptima platforms. A biopsy diagnosis of grade 2 cervical intraepithelial neoplasia was confirmed with p16/Ki-67 immunohistochemistry. RESULTS: In total, 1866 and 175 Pap cases were tested on Cobas and Aptima platforms, respectively. Both platforms were highly sensitive (97% for both) for biopsy-confirmed ≥HSIL. Cobas HPV testing had higher positive rates for the diagnosis of benign lesions (84% vs 51%) and low-grade squamous intraepithelial lesions (89% vs 63%) on biopsy compared with Aptima. Aptima testing had significantly higher specificity for ≥HSIL than Cobas (41% vs 13%; P < .0001). Overall, performance of the Aptima platform was superior to that of the Cobas platform in detecting biopsy-confirmed ≥HSIL, resulting from its significantly higher positive predictive value (25% vs 16%; P < .03) and overall accuracy (50% vs 26%; P < .0001). CONCLUSIONS: Although both the Cobas and Aptima platforms offer highly sensitive tests for high-grade cervical lesions, Aptima HPV testing demonstrated significantly higher specificity and positive predictive value than Cobas testing for biopsy-confirmed ≥HSIL. The considerable difference may be related to the significant increase in E6/E7 expression after HPV DNA integration. The significantly higher specificity and overall accuracy of Aptima testing for ≥HSIL, resulting in the identification of high-risk populations that require immediate treatment and close follow-up, may prove useful in clinical risk stratification. Cancer Cytopathol 2017;125652-7. © 2017 American Cancer Society.