A cohort study of cervical screening using partial HPV typing and cytology triage. Issue 11 (26th August 2016)

• Record Type: Journal Article
• Authors: Schiffman, Mark; Hyun, Noorie; Raine‐Bennett, Tina R.; Katki, Hormuzd; Fetterman, Barbara; Gage, Julia C.; Cheung, Li C.; Befano, Brian; Poitras, Nancy; Lorey, Thomas; Castle, Philip E.; Wentzensen, Nicolas
• Is Part Of: International journal of cancer. Volume 139:Issue 11(2016:Dec. 01)
• Page Start: 2606
• Page End: 2615
• Publication Date: 2016-08-26
• Subjects: cervix -- HPV -- screening -- cancer -- cytology; Cancer -- Periodicals; Cancer -- Prevention -- Periodicals; 616.994
• DOI: 10.1002/ijc.30375
• Languages: English
• ISSNs: 0020-7136

Abstract

Abstract : HPV testing is more sensitive than cytology for cervical screening. However, to incorporate HPV tests into screening, risk‐stratification ("triage") of HPV‐positive women is needed to avoid excessive colposcopy and overtreatment. We prospectively evaluated combinations of partial HPV typing (Onclarity, BD) and cytology triage, and explored whether management could be simplified, based on grouping combinations yielding similar 3‐year or 18‐month CIN3+ risks. We typed ∼9, 000 archived specimens, taken at enrollment (2007–2011) into the NCI‐Kaiser Permanente Northern California (KPNC) HPV Persistence and Progression (PaP) cohort. Stratified sampling, with reweighting in the statistical analysis, permitted risk estimation of HPV/cytology combinations for the 700, 000+‐woman KPNC screening population. Based on 3‐year CIN3+ risks, Onclarity results could be combined into five groups (HPV16, else HPV18/45, else HPV31/33/58/52, else HPV51/35/39/68/56/66/68, else HPV negative); cytology results fell into three risk groups ("high‐grade, " ASC‐US/LSIL, NILM). For the resultant 15 HPV group‐cytology combinations, 3‐year CIN3+ risks ranged 1, 000‐fold from 60.6% to 0.06%. To guide management, we compared the risks to established "benchmark" risk/management thresholds in this same population ( e.g ., LSIL predicted 3‐year CIN3+ risk of 5.8% in the screening population, providing the benchmark for colposcopic referral). By benchmarking to 3‐year risk thresholds (supplemented by 18‐month estimates), the widely varying risk strata could be condensed into four action bands (very high risk of CIN3+ mandating consideration of cone biopsy if colposcopy did not find precancer; moderate risk justifying colposcopy; low risk managed by intensified follow‐up to permit HPV "clearance"; and very low risk permitting routine screening.) Overall, the results support primary HPV testing, with management of HPV‐positive women using partial HPV typing and cytology. Abstract : What's new? Cervical cancer screening includes two elements: cytology and HPV testing. But many women who test positive for HPV won't need further examination; most HPV infections don't lead to cancer. How best to identify those at risk? These authors created a system to simplify risk assessment. By conducting a cohort study to document the likelihood of patients developing CIN3 within three years, they evaluated different patterns of screening results. They were able to stratify risk profiles generated by HPV typing and cytology into four "action bands, " or further screening recommendations, thus balancing the sensitivity of HPV typing with simplicity of implementation.