Department of Pathology, University of Virginia Health System, Charlottesville, VA.
Roche Molecular Systems Inc., Pleasanton, CA.
Int J Cancer. 2020 May 1;146(9):2599-2607. doi: 10.1002/ijc.32669. Epub 2019 Oct 6.
The objective of our study was to assess the performance of different triage strategies for high-risk human papillomavirus (hrHPV)-positive results utilizing either extended genotyping or a p16/Ki-67 dual-stained cytology (DS) approach, with or without partial genotyping. A subset of women with hrHPV infections participating in the Addressing the Need for Advanced HPV Diagnostics (ATHENA) study were analyzed to determine the number of cervical intraepithelial neoplasia grade 3 or worse (≥CIN3) cases detected, and the absolute risk for ≥CIN3 of each genotype. A clinical utility table was constructed to compare the impact of different triage strategies. In all, 2,339 women with single-genotype hrHPV infections were identified. Among these were 171 ≥CIN3 cases. The U.S. Food and Drug Administration (FDA)-approved algorithm (HPV16/18 positive, or 12-other hrHPV positive and Pap positive, i.e., ≥ atypical squamous cells of undetermined significance) for primary HPV screening detected 132/171 (77.2%) ≥CIN3 cases and required 964 colposcopies (colposcopies per ≥CIN3 ratio: 7.3). An approach that uses DS instead of cytology in the FDA-approved algorithm detected 147/171 (86.0%) ≥CIN3 cases, requiring 1,012 colposcopies (ratio: 6.9). Utilizing DS for triage of all hrHPV-positive women identified 126/171 (73.7%) ≥CIN3 cases, requiring 640 colposcopies (ratio: 5.1). A strategy that detected HPV16/18/31/33/35+ captured 130/171 (76.0%) ≥CIN3 cases, requiring 1,025 colposcopies (ratio: 7.9). Inclusion of additional genotypes resulted in greater disease detection at the expense of higher colposcopy ratios. Substituting cytology with a DS triage approach improved disease detection and the colposcopy detection rate. Further reduction of colposcopy rates can be achieved by using DS without partial genotyping. Extended genotyping strategies can identify a comparable number of cases but requires an increased number of colposcopies.
我们的研究目的是评估不同的分诊策略对于高危型人乳头瘤病毒（hrHPV）阳性结果的性能，这些策略利用了扩展基因分型或 p16/Ki-67 双重染色细胞学（DS）方法，同时包括或不包括部分基因分型。对参与 Addressing the Need for Advanced HPV Diagnostics（ATHENA）研究的患有 hrHPV 感染的女性亚组进行分析，以确定检测到的宫颈上皮内瘤变 3 级或更高级别（≥CIN3）病例数量，以及每种基因型的绝对≥CIN3 风险。构建了一个临床效用表来比较不同分诊策略的影响。共有 2339 名患有单一基因型 hrHPV 感染的女性被确定。其中有 171 例≥CIN3 病例。美国食品和药物管理局（FDA）批准的算法（HPV16/18 阳性，或 12 种其他 hrHPV 阳性且 Pap 阳性，即不典型鳞状细胞意义不明确）用于原发性 HPV 筛查，检测到 132/171（77.2%）≥CIN3 病例，需要 964 次阴道镜检查（每例≥CIN3 的阴道镜检查率：7.3）。在 FDA 批准的算法中使用 DS 代替细胞学检测，可以检测到 147/171（86.0%）≥CIN3 病例，需要 1012 次阴道镜检查（比率：6.9）。对所有 hrHPV 阳性女性进行 DS 分诊，可以发现 126/171（73.7%）≥CIN3 病例，需要 640 次阴道镜检查（比率：5.1）。检测 HPV16/18/31/33/35+的策略可捕获 130/171（76.0%）≥CIN3 病例，需要 1025 次阴道镜检查（比率：7.9）。纳入其他基因型会增加疾病检测率，但会导致阴道镜检查率升高。用 DS 替代细胞学进行分诊可以提高疾病检测率和阴道镜检查检出率。通过不进行部分基因分型而仅使用 DS 可以进一步降低阴道镜检查率。扩展基因分型策略可以识别出相当数量的病例，但需要增加阴道镜检查的数量。
