Gao Kun, Eurasian Menglan, Zhang Jieqing, Wei Yuluan, Zheng Qian, Ye Hongtao, Li Li
Department of Gynecologic Oncology, Cancer Institute and Hospital, Guangxi Medical University, Nanning, China.
Gynecol Obstet Invest. 2015;80(3):153-63. doi: 10.1159/000371760. Epub 2015 Apr 1.
To evaluate the effectiveness of five methods including the ThinPrep cytological test (TCT), liquid-based cytology, the human papillomavirus (HPV) test, detection of the TERC and C-MYC genes and visual inspection with acetic acid/Lugol's iodine (VIA/VILI) for opportunistic cervical cancer screening, and to explore whether genomic amplification of the human telomerase gene and C-MYC in liquid-based cytological specimens can be used as a method for opportunistic cervical cancer screening.
Data were collected prospectively from 1,010 consecutive patients who visited the gynecology clinic and agreed to participate in opportunistic cervical cancer screening at our institution from November 2010 to July 2011. The five methods mentioned above were used for the screening in all cases. The histopathological diagnosis served as the gold standard for the evaluation. A comparison between the five screening methods for the diagnosis of high-grade cervical intraepithelial neoplasia (CIN II and III) was performed for their sensitivity, specificity, false-positive rate, false-negative rate, accuracy rate, positive likelihood ratio and negative likelihood ratio. A comprehensive comparison of the different combination programs for screening was performed according to the analysis of the receiver operating characteristic (ROC) curve area. The accuracy of the five screening methods for the diagnosis of high-grade CIN (CIN II and III) was compared in the different age groups. A joint model for the diagnosis using different combinations of the five methods was developed according to the analysis by the SAS 8.0 software. The model was used to evaluate the accuracy of the different combination programs for the diagnosis of high-grade CIN, and the results were confirmed by the histopathological examination.
The sensitivity and specificity of the single screen method (TCT, HPV test, detection of the TERC and C-MYC genes, and VIA/VILI method) for CIN II was 80.9, 70.2, 72.3, 76.6, and 72.3%, as well as 98.0, 95.1, 96.3, 96.3, and 90.4%, respectively. The sensitivity of the single screening method in four different age groups (25-34, 35-44, 45-54 and 55-66 years) was as follows: TCT, 64.3, 90.9 76.5, and 85.7%; HPV test, 78.6, 72.7, 60.0, and 71.4%; the TERC gene, 50.0, 90.9, 80.0, and 71.4%; the C-MYC gene, 50.0, 90.9, 80.0, and 100%; VIA/VILI, 85.7, 81.8, 66.7, and 42.9%. The specificity was: TCT, 98.9, 98.1, 98.8, and 95.2%; HPV test, 96.7, 95.1, 92.2, and 100%; the TERC gene, 95.0, 98.9, 94.0, and 95.2%; the C-MYC gene, 97.2, 97.3, 93.4, and 97.6%; VIA/VILI, 91.2, 90.5, 89.8, and 88.1%, respectively. In the joint model for the diagnosis using different combinations, we found Logit (P) = 5.757 - 4.055 × TCT - 3.724 × HPV. The sensitivity and specificity in the combination program with TCT (primary screening) and HPV testing (adjunct screening) were 78.7 and 99.5%, while in the combination with HPV (primary) and TCT (adjunct), they were 53.2 and 99.7%, respectively. However, in the cytology-HPV parallel test, they were 97.9 and 93.4%. The ROC analysis revealed that the cytology-HPV parallel test is superior to the combinations of either TCT (primary) and HPV (adjunct) or HPV (primary) and TCT (adjunct; AUCTCT-HPV parallel test = 0.956; AUCTCT/primaryHPV/adjunct = 0.764).
Opportunistic cervical cancer screening is a practical approach to improve the efficiency of cervical cancer screening. Although the accuracy of TCT is the highest of the five screening methods for the diagnosis of high-grade CIN, it is still subject to sample acquisition and the practitioner's skill and experience. Since the efficacy of VIA/VILI may vary in all ages, it is not recommended for menopausal and perimenopausal women.
评估液基薄层细胞学检测（TCT）、液基细胞学、人乳头瘤病毒（HPV）检测、TERC和C-MYC基因检测以及醋酸/卢戈氏碘肉眼检查（VIA/VILI）这五种方法用于机会性宫颈癌筛查的有效性，并探讨液基细胞学标本中人端粒酶基因和C-MYC基因的基因组扩增是否可作为机会性宫颈癌筛查的一种方法。
前瞻性收集2010年11月至2011年7月期间在我院妇科门诊就诊并同意参与机会性宫颈癌筛查的1010例连续患者的数据。所有病例均采用上述五种方法进行筛查。组织病理学诊断作为评估的金标准。比较五种筛查方法诊断高级别宫颈上皮内瘤变（CIN II和III）的灵敏度、特异度、假阳性率、假阴性率、准确率、阳性似然比和阴性似然比。根据受试者工作特征（ROC）曲线面积分析，对不同组合的筛查方案进行综合比较。比较五种筛查方法在不同年龄组中诊断高级别CIN（CIN II和III）的准确性。根据SAS 8.0软件的分析，建立使用五种方法不同组合的联合诊断模型。该模型用于评估不同组合方案诊断高级别CIN的准确性，并通过组织病理学检查进行验证。
CIN II的单一筛查方法（TCT、HPV检测、TERC和C-MYC基因检测以及VIA/VILI方法）的灵敏度分别为80.9%、70.2%、72.3%、76.6%和72.3%，特异度分别为98.0%、95.1%、96.3%、96.3%和90.4%。单一筛查方法在四个不同年龄组（25 - 34岁组、35 - 44岁组、45 - 54岁组和55 - 66岁组）中的灵敏度如下：TCT分别为64.3%、90.9%、76.5%和85.7%；HPV检测分别为78.6%、72.7%、60.0%和71.4%；TERC基因分别为50.0%、90.9%、80.0%和71.4%；C-MYC基因分别为50.0%、90.9%、80.0%和100%；VIA/VILI分别为85.7%、81.8%、66.7%和42.9%。特异度分别为：TCT为98.9%、98.1%、98.8%和95.2%；HPV检测为96.7%、95.1%、92.2%和100%；TERC基因为95.0%、98.9%、94.0%和95.2%；C-MYC基因为97.2%、97.3%、93.4%和97.6%；VIA/VILI分别为91.2%、90.5%、89.8%和88.1%。在使用不同组合的联合诊断模型中，发现Logit（P）= 5.757 - 4.055×TCT - 3.724×HPV。TCT（初筛）和HPV检测（辅助筛查）组合方案的灵敏度和特异度分别为78.7%和99.5%，而HPV（初筛）和TCT（辅助筛查）组合方案的灵敏度和特异度分别为53.2%和99.7%。然而，在细胞学-HPV平行检测中，灵敏度和特异度分别为97.9%和93.4%。ROC分析显示，细胞学-HPV平行检测优于TCT（初筛）和HPV（辅助筛查）或HPV（初筛）和TCT（辅助筛查）的组合（细胞学-HPV平行检测的AUCTCT - HPV平行检测 = 0.956；TCT/初筛HPV/辅助筛查的AUCTCT/primaryHPV/adjunct = 0.764）。
机会性宫颈癌筛查是提高宫颈癌筛查效率的一种实用方法。虽然TCT在诊断高级别CIN的五种筛查方法中准确性最高，但仍受样本采集以及从业者技能和经验的影响。由于VIA/VILI的效果在各年龄段可能有所不同，不建议用于绝经和围绝经期女性。
