London School of Hygiene & Tropical Medicine, London, UK.
Cytology Department, Manchester Royal Infirmary, Manchester University NHS Foundation Trust, Manchester, UK.
Br J Cancer. 2023 May;128(10):1933-1940. doi: 10.1038/s41416-023-02227-9. Epub 2023 Mar 23.
Long-term follow-up of large cohorts is needed to determine the effects of HPV and screening on CIN3 (grade 3 cervical intraepithelial neoplasia) and ICC (invasive cervical cancer).
Women were recruited when attending for routine cervical screening in Greater Manchester, UK: 1987-93 for the Manchester Cohort (MC: 47,625 women) and 2001-03 for the ARTISTIC Cohort (AC: 24,496 women). Both were followed through national registration for cancer incidence and mortality to 2020.
Risk patterns following HPV infection differed for CIN3 and ICC. Risk of ICC in the MC rises for 30 years following a single positive HPV test, reaching 2.5% (95% CI: 1.3-4.5%). A similar pattern was seen in the AC, but the risks of cancer were approximately halved. CIN3 was diagnosed much sooner in the AC due to more efficient cytology. More sensitive HPV testing was able to better predict future risk.
The sensitivity of HPV testing and cytology influences the CIN3 detection rate. Sensitive HPV testing enables effective risk stratification. Increased risk of ICC is observed 15-30 years after HPV infection. Women testing HPV + should be followed until their infection clears. Discharging women from screening programmes whilst they remain HPV + may not be safe, even if cytology and colposcopy tests are normal.
需要对大型队列进行长期随访,以确定 HPV 和筛查对 CIN3(宫颈上皮内瘤变 3 级)和 ICC(浸润性宫颈癌)的影响。
英国大曼彻斯特地区的女性在常规宫颈筛查时被招募:1987-93 年参加曼彻斯特队列(MC:47625 名女性),2001-03 年参加 ARTISTIC 队列(AC:24496 名女性)。两者均通过国家癌症发病率和死亡率登记进行随访,截至 2020 年。
HPV 感染后 CIN3 和 ICC 的风险模式不同。MC 中单次 HPV 检测阳性后,ICC 的风险在 30 年内上升,达到 2.5%(95%CI:1.3-4.5%)。AC 中也出现了类似的模式,但癌症风险降低了约一半。AC 中更早诊断出 CIN3,因为细胞学检查更有效。更敏感的 HPV 检测能够更好地预测未来的风险。
HPV 检测和细胞学的敏感性会影响 CIN3 的检出率。敏感的 HPV 检测能够进行有效的风险分层。HPV 感染后 15-30 年观察到 ICC 风险增加。HPV 检测呈阳性的女性应在感染清除前接受随访。即使细胞学和阴道镜检查正常,从筛查计划中释放仍 HPV 阳性的女性可能不安全。