TruScreen异常结果在非宫颈癌筛查人群中的影响因素分析

Analysis of clinical influencing factors of abnormal results of TruScreen detection for cervical cancer

目的分析非宫颈病变(癌)筛查人群中TruScreen(TS)人工智能筛查技术呈现异常结果的影响因素。方法选取2020年7月至2021年3月在湘潭市中心医院体检中心接受体检的1800例妇女为研究对象,分析TS检测结果的影响因素。采用卡方检验分析妇科检查、阴道微生态检查和人乳头瘤病毒(HPV)检测结果与TS阳性率之间的差异,并采用多因素Logistic回归分析探讨TS异常结果的可能影响因素。结果1800例体检妇女(经阴道镜或活检检查为宫颈正常)中TS异常者180例,占总研究对象的10.0%。TS正常组和异常组妇女的年龄、妇科检查和阴道微生态检查异常情况比较,差异均有统计学意义(P<0.05),而TS检测结果与HPV检测结果比较,差异无统计学意义(P=0.189)。正常宫颈、轻度、中/重度柱状上皮异位者的TS异常率比较,差异具有统计学意义(P<0.001),且TS异常率与宫颈柱状上皮异位严重程度的关联呈上升趋势。Logistic回归分析结果显示,相对于宫颈正常者,宫颈柱状上皮异位者、宫颈肥大/纳氏囊肿/息肉者TS检查结果异常的风险增加(OR=4.52,P<0.001;OR=2.64,P<0.001);相对于正常阴道微生态,霉菌性与混合性阴道炎的TS检查结果为异常的风险增加(OR=3.25,P=0.019;OR=3.359,P=0.007)。结论宫颈正常妇女中人工智能筛查技术检测异常率为10.0%,宫颈柱状上皮异位、肥大/纳囊/息肉、霉菌性和混合性阴道炎是TS异常的危险因素。如何降低TS异常的风险以及HPV感染对上皮细胞的损害程度,有待进一步研究。

Objective To analyze the clinical influencing factors of TruScreen artificial intelligence screening technology(TS) in the detection of cervical cancer. Methods 1 800 women who received physical examination in the physical Examination Center of Xiangtan Central Hospital between July 2020 and March 2021 were selected as the subjects. The TS detection results and clinical influencing factors were collected and analyzed. Chi square test was used to analyze the differences between the results of gynecological examination, vaginal microecology examination, human papilloma virus(HPV) DNA testing and TS detection. The multivariate Logistic regression analysis was used to explore the possible influencing factors of abnormal results of TS. Results 180 out of 1 800 women(with normal cervix by colposcopy or biopsy) were detected with abnormal results by TS, accounting for 10.0% of the total study population. There were significant differences in age, gynecological examination results, and vaginal microecological examination results between TS normal group and TS abnormal group(P<0.05), but there was no significant difference in the results between TS and HPV testing(P=0.189). There was significant difference in the rate of TS abnormal result among the groups with normal cervix, or mild, moderate/severe columnar epithelium ectopy(P<0.001). The TS abnormalities showed an upward trend in its relation with the severity of cervical columnar epithelium ectopy. Logistic regression analysis showed that the groups with cervical columnar epithelium ectopy, or cervical hypertrophy/cyst/polyp had an increased risk of abnormal results of TS, as compared with those with normal cervix(OR=4.52, P<0.001;OR=2.64, P<0.001);and the groups with fungal or mixed vaginitis had an increased risk of TS abnormality, as compared with the normal vaginal microecology group(OR=3.25, P=0.019;OR=3.359, P=0.007). Conclusion In women with normal cervix, there were 10% of them with abnormal results of TS examination. The cervical columnar epithelium ectopy, hypertrophy/cyst/polyp, fungal vaginitis and mixed vaginitis were the influencing factors of TS abnormalities. Further research is expected on how to reduce the risk of TS abnormalities and the degree of epithelial cell lesions by HPV infection.