| ObjectivesThis ariticle aims at comparing the results of using cytology, VIA/VILI and hrHPV testing as primary tests for cervical cancer and precancerous lesions screening programs performed by county level health providers in real world, instead of strict quality control studies, evaluating the approper triage tests for hrHPV positive women, investigate the utility of directed biopsy and random biopsy under the guidance of colposcopy and the preference of histologically proven lesions distribution on the cervix, providing scientific evidence for cervical cancer screening in China.Contents1. To compare the results of using cytology, VIA/VILI and hrHPV testing as primary tests respectively for cervical cancer and precancerous lesions screening programs performed by local health providers in real world;2. To compare the results of using cytology or VIA/VILI as triage tests for hrHPV positive women;3. To investigate the utility of directed biopsy and random biopsy;4. To investigate if the preference of histologically proven lesions differs by cervical four-quadrant location or by 12 o’clock surface locations.MethodsAt first, we conducted a cervical cancer and precancerous screening study based on the nationwide government supported cervical cancer screening program. According to the economic status, two different screening strategies were designed for lower economic sites as rural sites and higher economic sites as urban sites. In rural sites, cytology, VIA/VILI and hrHPV testing were used for primary screening respectively among women who were randomized into three arms. Women who were abnormal in primary cytology or VIA/VILI arms were referred to colposcopy directly, women who where hrHPV positive were randomly triaged by cytology, VIA/VILI or colposcopy directly. In urban sites, cytology and hrHPV genotyping were used for primary screening respectively among women who were randomized into two arms. Women who have had abnormal cytology or HPV16/18 positive were referred to colposcopy directly; women with other hrHPV positive were triaged by cytology or colposcopy respectively. Data on test positivity, positive predictive value and detection rate of high-grade cervical neoplasia for both primary screening and triaging test were analyzed. Secondly, retrospective data analysis was performed to investigate the utility of directed biopsy and random biopsy among different population. Besides, a pooled analysis of 19 population based cervical cancer screening studies was conducted to evaluating the preference of histologically proven lesions distribution on the surface of the cervix among Chinese women.Results1. The positive rates of different primary screening arms in rural sites showed significant difference, hrHPV testing was much higher than cytology and VIA/VILI (16.3% vs.8.4%,8.9%, P<0.001). In urban sites, the primary positive rate of HPV genotyging was 8.1%, which 1.9% for HPV16/18 and 6.2% for other types. The difference of primary positive rate between HPV genotyping and cytology showed statistical significance (8.1% vs.5.8%, P<0.001). If all women with positive primary screening test were referred to colposcopy examination directly, the estimated detection rate for CIN2+lesions would be much higher among hrHPV than cytology or VIA/VILI, which the estimated rate among rural sites were 1.2%,0.7%,0.7% respectively, and for CIN3+ lesions were 1.2%,0.4%,0.5%. In urban sites, the estimated detection rate of CIN2+lesions for HPV genotyping and cytology were 1.3% and 0.6%, for CIN3+lesions were 0.8% and 0.4% respectively.2. In rural sites, the propotion of hrHPV positive women who should be referred to colposcopy were 2.9%,2.9% and 16.3% for triaging with cytology, VIA/VILI or directly colposcopy respectively. The detection rate of CIN3+cases were significantly different (1.5%,1.6%,7.5%, P=0.012). The average number of colposcopy needed to detect one CIN2+case was 13.1 for directly colposcopy, 5.3 for triaging with cytology and 7.7 for triaging with VIA/VILI, the average number of colposcopy needed to detect one CIN3+case was 13.1,8.0 and 11.5 respcetively. In urban sites, the detection rate of CIN2+and CIN3+cases for HPV 16/18 positive women were 24.1% and 13.8%. The propotion of other hrHPV positive women who should be referred to colposcopy were 2.6% and 6.2% for women triaging with cytology or directly colposcopy respectively. The detection rate of CIN2+and CIN3+cases among directly colposcopy women were significantly higher than triaging with cytology (13.0% vs.5.9%,P=0.003; 6.3% vs.0.9%, P=0.003). The average number of colposcopy needed to detect one CIN2+or CIN3+case were 4.1 and 7.1 for HPV 16/18 positive women with directly colposcopy,6.8 and 14.1 for other hrHPV positive women with directly colposcopy and 6.0 and 39.0 for triaging with cytology respectively.3. In hrHPV negative women, the prevalence of CIN2+was 0.1%, no CIN2+ lesion was found by random biopsy. In VIA abnormal but negative colposcopy, the detection rate of CIN2+by random biopsy was 0.7%. No significant difference was found between the detection rates of women with abnormal or normal VIA by directed biopsy (P=0.127). The detection rate of CIN2+lesions by four quadrants biopsy increase with the severity if cytology (P<0.001) which for HSIL+cytology is 25.0%.When co-testing was used for primary screening, the detection rate of CIN2+lesion by directed biopsy and four quadrants random biopsy showed no significant difference among women with LSIL+and hrHPV positive (P=0.283). No CIN2+lesion was found among HPV negative women with negative colposcopy irrespective of the cytology.4. CIN2+lesions were significantly more frequently found in the posterior cervix than in the anterior cervix (P<0.0001). When the cervix was divided in half on a sagittal plane, there was no significant difference in CIN2+occurrence between the left and the right sides (P=0.319). By directed biopsy, CIN2 and CIN3+ lesions were significantly more likely to be found in the second and third quadrants than in the first and fourth quadrants (P<0.0001). No significant differences in quadrant distribution were found for CIN2 and CIN3+lesions obtained by random biopsy (P= 0.1911). By directed biopsy, CIN2+lesions were most likely to occur at the 4-(OR=2.572,95% CI:1.900,3.481) and 7-(OR=1.689,95% CI:1.211,2.355) o’clock. The CIN3+lesions were most likely to occur at 4-and 7-o’clock positions as well, the ORs were 2.959 (95% CI: 2.026,4.323) and 1.678 (95% CI:1.095,2.572) respectively. By random biopsy, CIN2+lesions were more likely to occur at the 5-(OR=4.793,95% CI:2.462, 9.330),6-(OR=3.841,95% CI:1.530,9.644),7-(OR=4.185,95% CI:2.156, 8.121),9-(OR=3.657,95% CI:1.125,11.893), and 12-(OR=3.697,95% CI: 1.593,8.583) o’clock positions. CIN3+lesions were more likely to occur at the 3-(OR=6.033,95%CI:1.431,25.431),5-(OR=4.744,95% CI:1.695,13.277), 7-(OR=5.178,95%CI:2.046,13.106) and 12-(OR=4.575,95%CI:1.408, 14.861) o’clock positions.Conclusions1. In nationwide cervical cancer screening programs, hrHPV testing as primary screening, positive women referred to colposcopy directly could improve the detection rate of CIN2+cases compare with the strategy of using cytology or VIA/VILI as primary screening in rural sites, however, the propotion of women should be referred to colposcopy would be increased on a large margin. Similarly, in urban sites, compared with using cytology for primary screening, the referring rate for colposcopy and detection rate of CIN2+cases would be increased by using HPV genotyping.2. In rural sites with high hrHPV prevalence, the detection rate of CIN2+cases among hrHPV positive women triaged with cytology or VIA/VILI were nearly the same, which could largely reduce the referral rate of colposcopy. In large population screening programs, training programs to improve the quality of cytology and VIA/VILI for triaging hrHPV women should be employed, otherwise, the detection rate of high-grade cervical lesions would be unsatisfactory. In urban sites with lower hrHPV prevalence, referring all hrHPV positive women to colposcopy will not decrease the performance of colposcopy compare with the strategy that HPV 16/18 women undergo directly colposcopy and otherhrHPV positive women triaging with cytology.3. VIA or hrHPV used as primary screening alone provided limited information for biopsy under colposcopy. If VIA or hrHPV positive alone with negative colposcopy, four quadrants biopsy were not recommended. If cytology diagnosed as HSIL+, four quadrants biopsy should be performed. If cytology combined with hrHPV testing were used for primary screening, four quadrants biopsy were recommended for women with LSIL+cytology and HPV positive. If the hrHPV is negative, no random biopsy should be performed irrespective with the cytology.4. A predisposition of specific locations on the cervix to CIN occurrence. Quadrants 2 and 3, especially the 4-and 7-o’clock positions should be preferentially targeted during biopsy. The decision for random biopsy should be reconsidered in future studies.In conclusion, hrHPV testing as primary screening by local health providers in different economic regions could detect more CIN2+cases, but the number of referring to colposcopy would be increased on a large margin. Training programs for local health providers would be necessary if cytology or VIA/VILI were used or more objective method should be used to triaging hrHPV positive women. According to the primary screening and colposcopy, random biopsy should be performed conditionally to increase the detection rate. The preference of CIN2+ lesion distribution on the surface of the cervix of Chinese women may provide objective evidences to improve the accuracy of biopsy under colposcopy.by using objective screening and diagnosis techniques and proper screening strategies according to the real world performance, the morbidity of cervical cancer among Chinese women could be reduced effectively. |
PDF Full Text Request