Risk stratification of HPV-positive results using extended genotyping and cytology: Data from the baseline phase of the Onclarity trial

Background. Optimizing the balance between colposcopy referrals and the detection of high-grade cervical in-traepithelial neoplasia (CIN) during cervical cancer screening requires robust triage strategies. We evaluated the performance of extended HPV genotyping (xGT), in combination with cytology triage, and compared it to previ-ously published performance data for high-grade CIN detection by HPV16/18 primary screening in combination with p16/Ki-67 dual staining (DS).Methods and materials. The baseline phase of the Onclarity trial enrolled 33,858 individuals, yielding 2978 HPV-positive participants. Risk values for >= CIN3 were determined for Onclarity result groupings corresponding to HPV16, not HPV16 but HPV18 or 31, not HPV16/18/31 but HPV33/58 or 52, not HPV16/18/31/33/58/52 but HPV35/39/68 or 45 or 51 or 56/59/66 across all cytology categories. Published data from the IMPACT trial for HPV16/18 plus DS was utilized as a comparator during ROC analyses.Results. There were 163 >= CIN3 cases detected. The >= CIN3 risk stratum hierarchy (% risk of >= CIN3) that re-sulted from this analysis included: >LSIL (39.4%); HPV16, <= LSIL (13.3%); HPV18/31, <= LSIL (5.9%); HPV33/58/ 52/45, ASC-US/LSIL (2.4%); HPV33/58/52, NILM (2.1%); HPV35/39/68/51/56/59/66, ASC-US/LSIL (0.9%); and HPV45/35/39/68/51/56/59/66, NILM (0.6%). For >= CIN3 ROC analysis, the optimal cutoff for sensitivity versus specificity was approximated between not HPV16 but HPV18 or 31, any cytology (>= CIN3 sensitivity = 85.9% and colposcopy-to->= CIN3 = 7.4) and not HPV16/18/31 but HPV33/58/52, NILM (>= CIN3 sensitivity = 94.5% and colposcopy-to->= CIN3 = 10.8). HPV16/18 with DS triage showed a sensitivity of 94.3%, with a colposcopy- to->= CIN3 ratio of 11.4.Conclusions. xGT performed similarly compared to HPV primary screening plus DS for detection of high-grade CIN. xGT provides results that stratify risk in a flexible and reliable manner for colposcopy risk thresholds set by different guidelines or organizations.(c) 2023 Becton, Dickinson and Company. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).