Cervical cancer screening strategies for evidence-based health care reform in China: a multicentre, open-label randomised trial

Abstract

BackgroundAs part of China's 2009 health-care system reform, a national cervical cancer screening programme was launched, offering free screening by cytology or visuaI inspection with acetic acid and LugoI's iodine (VIA/VILI). We evaluated high-risk HPV (hrHPV) testing as a new screening modality for the national programme. MethodsThis multicentre, open-label randomised trial was undertaken at nine urban and 11 rural primary-care centres in mainland China. Women aged 35–64 years without history of cervical cancer or hysterectomy, no clinical symptoms of pregnancy, who could understand the study procedures, and voluntarily participated were included. At baseline, women in urban areas were randomly assigned (1:2) to primary screening by cytology or hrHPV, and women in rural areas were randomly assigned (1:1:1) to cytology, hrHPV, or VIA/VILI. Women with baseline hrHPV positive tests were randomly assigned to triage strategies using cytology, VIA/VILI (rural sites), or directly refer to colposcopy. At 24-months, women in urban regions were rescreened with cytology and hrHPV co-testing; rural women received combined hrHPV, cytology, and VIA/VILI screening. Women who showed positive in cytology or VIA/VILI were referred to exams for further colposcopy, and biopsy was then required for participants with abnormal colposcopy results. The primary outcome was cervical intraepithelial neoplasia 2 or worse (CIN2+) and cervical intraepithelial neoplasia 3 or worse (CIN3+). Data analyses were done in the intention-to-screen population (ie, all women assigned to screening). This trial is registered with Chinese Clinical Trial Registry, number ChiCTR1900022530. FindingsBetween May 18, 2015, and Sept 30, 2016, 61 484 women were assessed as eligible and 60 732 were randomly assigned to screening. Among women in urban regions, 8 955 were assigned to cytology and 18 176 to hrHPV-genotyping; among women in rural regions, 11 136 were assigned to VIA/VILI, 7080 to cytology, and 15 385 to hrHPV-testing. Compared with cytology and VIA/VILI-based primary screening, hrHPV without triaging had 2·0–2·7 times higher baseline CIN2+ or CIN3+ detection (risk ratios: urban hrHPV vs cytology CIN2+ 2·2 [95% CI 1·5–3·1], p<0·0001, CIN3+ 2·0 [1·2–3·3], p=0·0084; rural hrHPV vs cytology CIN2+ 2·6 [1·7–4·0], p<0·0001, CIN3+ 2·7 [1·5–4·6], p<0·0001; rural hrHPV vs VIA/VILI CIN2+ 2·0 [1·5–2·8], p<0·0001, CIN3+ 2·3 [1·5–3·6], p=0·00012). At the 24-month follow-up, hrHPV without triage had lower new CIN2+ or CIN3+ than cytology in urban sites (risk ratio: CIN2+ 0·5 [0·3–0·97], p=0·038; CIN3+ 0·2 [0·03–0·8], p=0·026), and lower baseline false negative results in rural sites (risk ratio: cytology CIN2+ 0·3 [0·2–0·6], p=0·00019 and CIN3+ 0·3 [ 0·1–0·6] p=0·0025; VIA/VILI CIN2+ 0·3 [0·2–0·6], p=0·00012 and CIN3+ 0·4 [0·2–0·8], p=0·013). InterpretationHrHPV-testing offered better performance than current cytology and VIA/VILI-based primary screening and could be incorporated into China's national programme, but with the longer screening intervals recommended by WHO. FundingNational Health Commission of China.