A main challenge of human papilloma (HPV)-based screening for cervical cancer is to adequately identify HPV-positive women at highest risk of cervical intraepithelial neoplasia grade 3 or worse, CIN3+. The prognostic value of currently used adjunct markers (HPV16/18 genotyping and reflex cytology) may change after multiple rounds of HPV-based screening because of a change in the proportion of HPV-positive women with incident infections. To this end, we re-analyzed results from the POBASCAM trial (Population Based Screening Study Amsterdam). Women were randomized to HPV/cytology cotesting (intervention group) or to cytology-only (HPV blinded; control group) at enrolment. Our analytical population consisted of women with an HPV-positive result at the second round, 5 years after enrolment (n = 381 intervention, n = 392 control). Nine-year CIN3+ risks were estimated by Kaplan-Meier. HPV-positive women were stratified by risk markers: HPV16/18 genotyping, reflex cytology and preceding HPV results. When comparing one to two rounds of HPV-based screening, the prognostic value of an abnormal cytology result did not change (40.0% vs. 42.3%, p = 0.5617), but diminished for an HPV16/18 positive result (25.4% vs. 38.0%, p = 0.0132). HPV16/18 genotyping was nondiscriminative in women with incident HPV infections (HPV16/18 positive 10.0% vs. negative 12.1%, p = 0.3193). Women from the intervention group were more likely to have incident infections compared to women from the control group (incident screen-positive results 75.6% vs. 64.6%, p = 0.001) Our results indicate that at a second round of HPV-based screening, risk differentiation by cytology remained strong, but was diminished for HPV 16/18 genotyping because of a larger proportion of incident infections.
Keywords: HPV; cervical cancer; human papillomavirus; risk stratification; screening.
© 2017 UICC.