[Performance of vaginal self-sampling high-risk HPV genotyping as primary and combining cytology or viral load as secondary in cervical cancer screening]
C L Guo,H X Luo,C Wang,X F Qu,Bin Yang,J L Belinson,H Du,R F Wu
DOI: https://doi.org/10.3760/cma.j.cn112141-20200824-00357
2021-04-25
Abstract:Objective: To evaluate the efficacy of high-risk HPV (HR-HPV) genotyping with vaginal self-sampling in primary screening and combining cytology or viral load for HR-HPV positive as secondary screening strategies. Methods: The data referring to HR-HPV genotyping of self-collected sample with mass array matrix-assisted laser desorption-ionization time-of-flight mass spectrometry (MALDI-TOF-MS), HR-HPV viral load of physician-collected sample with hybrid capture Ⅱ (HC-Ⅱ), liquid-based cytology and histology of 8 556 women were from Shenzhen cervical cancer screening trial Ⅱ (SHENCCAST-Ⅱ) conducted between April 2009 and April 2010. The data were reanalyzed to determine the sensitivity and specificity to cervical intraepithelial neoplasia (CIN) of grade 2 or worse (CIN Ⅱ+), CIN of grade 3 or worse (CIN Ⅲ+) when HR-HPV genotyping combining with colposcopy as primary screening strategy based on varied HR-HPV subtype (strategy 1, including 5 sub-strategies: 1a: HPV 16/18 positive; 1b: HPV 16/18/58 positive; 1c: HPV 16/18/58/31/33 positive; 1d: HPV 16/18/58/31/33/52 positive; 1e: any HR-HPV positive). The data were also compared to determine the efficacy of cytology (strategy 2, including 5 sub-strategies: 2a, 2b, 2c, 2d, 2e) or HR-HPV viral load (strategy 3, including 4 sub-strategies: 3a, 3b, 3c, 3d) of physician-collected sample as a triage with HR-HPV genotyping for self-sampling HR-HPV positives. Results: (1) The HR-HPV positive rate was 13.77% (1 178/8 556) in the self-collected samples of 8 556 pregnant women. Of them,the prevalences of HPV 16/18, HPV 16/18/58, HPV 16/18/58/31/33 and HPV 16/18/58/31/33/52 were 3.16% (270/8 556), 5.14% (440/8 556), 6.66% (570/8 556) and 9.81% (839/8 556), respectively. The HR-HPV viral load ≥10 relative light units/control (RLU/CO) was 8.87%(759/ 8 556), while cytological results ≥atypical squamous cell of undetermined signification (ASCUS) were 12.05% (1 031/8 556). (2) The strategy 1e had the highest sensitivities for CIN Ⅱ+, CIN Ⅲ+ which were 92.70% and 94.33%,respectively,among 14 sub-strategies,while the lowest specificity and positive predictive value (PPV). Meanwhile,the required colposcopy referral rates were much higher than other 13 sub-strategies (13.77%). The other 4 sub-strategies of strategy 1 (1a, 1b, 1c, 1d), strategy 1a had the highest specificities for CIN Ⅱ+ and CIN Ⅲ+ (97.92%, 97.69%, respectively), while 1d had the highest sensitivities for CIN Ⅱ+ and CIN Ⅲ+ (88.41%, 92.20%, respectively). (3) Both strategies of referring self-sampling HPV 16/18 positives for immediate colposcopy followed by triage physician-collected sample cytology (≥ASCUS) or viral load (≥10 RLU/CO) for non-HPV 16/18 positives had significantly higher sensitivity and specificity for CIN Ⅱ, CIN Ⅲ+, as well as lower referral rates (strategy 2a and 3a). Additionally, based on these two secondary screening strategies, cumulatively using the other four HR-HPV (HPV 58, 31, 33 and 52) positives as triage for immediate colposcopy showed an enhanced sensitivity. Conclusions: Primary HR-HPV cervical cancer screening strategy based on self-sampling with triage of cytology (≥ASCUS) or viral load (≥10 RUL/CO) provides a good balance among sensitivity, specificity for CIN Ⅱ+ and CIN Ⅲ+ and the number of tests required, referral rates. The efficacy of HR-HPV genotyping combining cytology or viral load secondary screening strategies will have a spiral escalation when HPV 58, 31, 33, 52 are included.