Performance of HPV E6/E7 mRNA assay as primary screening test: Results from the NTCC2 trial

Abstract

As the primary screening test, E6/E7 mRNA has shown similar sensitivity for CIN3+ and lower positivity rate than the HPV DNA test. Nevertheless, the overall mRNA positivity is too high for immediate colposcopy, making a triage test necessary. The aim was to estimate the mRNA performance as a primary test with different triage strategies. All HPV DNA-positives were tested for mRNA, cytology and p16/ki67. A sample of HPV DNA-negatives was also tested for mRNA to estimate test specificity. We included all CIN3+ histologically diagnosed within 24 months since recruitment. Of the 41 127 participants, 7.7% were HPV DNA-positive, of which 66.4% were mRNA-positive. Among the HPV DNA-negatives, 10/1108 (0.9%) were mRNA-positive. Overall, 97 CIN3+ were found. If mRNA was used as the primary test, it would miss about 3% of all CIN3+ with a 22% reduction of positivity compared with HPV DNA. The weighted specificity estimate for <CIN2 was 94.5% (95% CI = 93.9%-94.9%) and sensitivity for CIN3+ was 96.9% (95% CI = 91.3%-99.1%). If all the weighted estimated 6.0% mRNA-positive women had been referred to colposcopy, PPV for CIN3+ would have been 4.2%. Cytology or p16/ki67 triage would decrease immediate referral to 1.7% and 2.0%, increasing PPV to 11.2% and 11.7%, respectively; total colposcopy referral would be 4.0% and 3.9%, respectively. As the primary screening test, the mRNA assay showed a positivity rate lower than that of HPV DNA, with a small number of CIN3+ missed. Triage with cytology or p16/ki67 would only marginally decrease overall colposcopy referral.

Keywords: E6/E7 mRNA; accuracy; cervical cancer; cervical intraepithelial neoplasia; human papillomavirus; mass screening.