De Strooper Lise M A, Hesselink Albertus T, Berkhof Johannes, Meijer Chris J L M, Snijders Peter J F, Steenbergen Renske D M, Heideman Daniëlle A M
Department of Pathology, VU University Medical Center, Amsterdam, the Netherlands.
Department of Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, the Netherlands.
Cancer Epidemiol Biomarkers Prev. 2014 Sep;23(9):1933-7. doi: 10.1158/1055-9965.EPI-14-0347. Epub 2014 Jun 24.
Primary screening for high-risk human papillomavirus (hrHPV) requires a triage protocol. Repeat cytology testing at baseline and after 6 to 12 months has emerged as a reasonable triage approach, but carries the risk of loss to follow-up. Repeat cytology testing may be omitted if cytology is supplemented with another, complementary triage test at baseline. In this study, the performance of combined triage by cytology and DNA methylation analysis was assessed. In hrHPV-positive cervical scrapes (n = 250), cytology [threshold: atypical squamous cells of undetermined significance (ASCUS)], bi-marker CADM1/MAL methylation testing (at different assay thresholds), and combinations of both were evaluated for endpoints cervical intraepithelial neoplasia grade 2 or worse (CIN2(+)) and grade 3 or worse (CIN3(+)). At a predefined methylation threshold of 70% specificity for CIN3(+), combined triage revealed a CIN3(+) sensitivity of 86.8% [95% confidence interval (CI), 76.1-97.6] compared with 65.8% (95% CI, 50.7-80.9) for sole cytology triage testing. Corresponding CIN3(+) specificity was 64.8% (95% CI, 58.1-71.5) for combined triage and 78.6% (95% CI, 72.8-84.3) for sole cytology triage testing. For CIN2(+), the sensitivity of combined triage testing was 84.5% (95% CI, 75.2-93.8) compared with 65.5% (95% CI, 53.3-77.7) for sole cytology triage, with corresponding specificities of 69.9% (95% CI, 63.1-76.6) and 83.5% (95% CI, 78.0-89.0), respectively. In conclusion, combined triage reached substantially higher CIN2(+)/3(+) sensitivities compared with sole cytology at a slight drop in specificity. Therefore, it is an attractive triage strategy for colposcopy of hrHPV-positive women with a high reassurance for cervical cancer and advanced CIN lesions.
高危型人乳头瘤病毒（hrHPV）的初筛需要一种分流方案。在基线时以及6至12个月后重复进行细胞学检测已成为一种合理的分流方法，但存在失访风险。如果在基线时细胞学检测辅以另一种互补的分流检测，则可省略重复细胞学检测。在本研究中，评估了细胞学和DNA甲基化分析联合分流的性能。在hrHPV阳性宫颈刮片（n = 250）中，针对宫颈上皮内瘤变2级或更严重（CIN2(+)）和3级或更严重（CIN3(+)）的终点，评估了细胞学[阈值：意义不明确的非典型鳞状细胞（ASCUS）]、双标记CADM1/MAL甲基化检测（在不同检测阈值下）以及两者的组合。在针对CIN3(+)预定义的70%特异性甲基化阈值下，联合分流显示CIN3(+)敏感性为86.8% [95%置信区间（CI），76.1 - 97.6]，而单纯细胞学分流检测为65.8%（95% CI，50.7 - 80.9）。联合分流对应的CIN3(+)特异性为64.8%（95% CI，58.1 - 71.5），单纯细胞学分流检测为78.6%（95% CI，72.8 - 84.3）。对于CIN2(+)，联合分流检测的敏感性为84.5%（95% CI，75.2 - 93.8），而单纯细胞学分流为65.5%（95% CI，53.3 - 77.7），相应的特异性分别为69.9%（95% CI，63.1 - 76.6）和83.5%（95% CI，78.0 - 89.0）。总之，联合分流在特异性略有下降的情况下，与单纯细胞学相比，CIN2(+)/3(+)敏感性显著更高。因此，对于hrHPV阳性女性的阴道镜检查，这是一种有吸引力的分流策略，对宫颈癌和高级别CIN病变有较高的排除把握。
