Mortality impact, risks, and benefits of general population screening for ovarian cancer: the UKCTOCS randomised controlled trial
Usha Menon,Aleksandra Gentry-Maharaj,Matthew Burnell,Andy Ryan,Jatinderpal K Kalsi,Naveena Singh,Anne Dawnay,Lesley Fallowfield,Alistair J McGuire,Stuart Campbell,Steven J Skates,Mahesh Parmar,Ian J Jacobs
DOI: https://doi.org/10.3310/BHBR5832
2023-05-11
Abstract:Background: Ovarian and tubal cancers are lethal gynaecological cancers, with over 50% of the patients diagnosed at advanced stage. Trial design: Randomised controlled trial involving 27 primary care trusts adjacent to 13 trial centres based at NHS Trusts in England, Wales and Northern Ireland. Methods: Postmenopausal average-risk women, aged 50-74, with intact ovaries and no previous ovarian or current non-ovarian cancer. Interventions: One of two annual screening strategies: (1) multimodal screening (MMS) using a longitudinal CA125 algorithm with repeat CA125 testing and transvaginal scan (TVS) as second line test (2) ultrasound screening (USS) using TVS alone with repeat scan to confirm any abnormality. The control (C) group had no screening. Follow-up was through linkage to national registries, postal follow-up questionnaires and direct communication with trial centres and participants. Objective: To assess comprehensively risks and benefits of ovarian cancer screening in the general population. Outcome: Primary outcome was death due to ovarian or tubal cancer as assigned by an independent outcomes review committee. Secondary outcomes included incidence and stage at diagnosis of ovarian and tubal cancer, compliance, performance characteristics, harms and cost-effectiveness of the two screening strategies and a bioresource for future research. Randomisation: The trial management system confirmed eligibility and randomly allocated participants using computer-generated random numbers to MMS, USS and C groups in a 1:1:2 ratio. Blinding: Investigators and participants were unblinded and outcomes review committee was masked to randomisation group. Analyses: Primary analyses were by intention to screen, comparing separately MMS and USS with C using the Versatile test. Results: 1,243,282 women were invited and 205,090 attended for recruitment between April 2001 and September 2005. Randomised: 202,638 women: 50,640 MMS, 50,639 USS and 101,359 C group. Numbers analysed for primary outcome: 202,562 (>99.9%): 50,625 (>99.9%) MMS, 50,623 (>99.9%) USS, and 101,314 (>99.9%) C group. Outcome: Women in MMS and USS groups underwent 345,570 and 327,775 annual screens between randomisation and 31 December 2011. At median follow-up of 16.3 (IQR 15.1-17.3) years, 2055 women developed ovarian or tubal cancer: 522 (1.0% of 50,625) MMS, 517 (1.0% of 50,623) USS, and 1016 (1.0% of 101314) in C group. Compared to the C group, in the MMS group, the incidence of Stage I/II disease was 39.2% (95% CI 16.1 to 66.9) higher and stage III/IV 10.2% (95% CI -21.3 to 2.4) lower. There was no difference in stage in the USS group. 1206 women died of the disease: 296 (0.6%) MMS, 291 (0.6%) USS, and 619 (0.6%) C group. There was no significant reduction in ovarian and tubal cancer deaths in either MMS (p = 0.580) or USS (p = 0.360) groups compared to the C group. Overall compliance with annual screening episode was 80.8% (345,570/420,047) in the MMS and 78.0% (327,775/420,047) in the USS group. For ovarian and tubal cancers diagnosed within one year of the last test in a screening episode, in the MMS group, the sensitivity, specificity and positive predictive values were 83.8% (95% CI 78.7 to 88.1), 99.8% (95% CI 99.8 to 99.9), and 28.8% (95% CI 25.5 to 32.2) and in the USS group, 72.2% (95% CI 65.9 to 78.0), 99.5% (95% CI 99.5 to 99.5), and 9.1% (95% CI 7.8 to 10.5) respectively. The final within-trial cost-effectiveness analysis was not undertaken as there was no mortality reduction. A bioresource (UKCTOCS Longitudinal Women's Cohort) of longitudinal outcome data and over 0.5 million serum samples including serial annual samples in women in the MMS group was established and to date has been used in many new studies, mainly focused on early detection of cancer. Harms: Both screening tests (venepuncture and TVS) were associated with minor complications with low (8.6/100,000 screens MMS; 18.6/100,000 screens USS) complication rates. Screening itself did not cause anxiety unless more intense repeat testing was required following abnormal screens. In the MMS group, for each screen-detected ovarian or tubal cancer, an additional 2.3 (489 false positives; 212 cancers) women in the MMS group had unnecessary false-positive (benign adnexal pathology or normal adnexa) surgery. Overall, 14 (489/345,572 annual screens) underwent unnecessary surgery per 10,000 screens. In the USS group, for each screen-detected ovarian or tubal cancer, an additional 10 (1630 false positives; 164 cancers) underwent unnecessary false-positive surgery. Overall, 50 (1630/327,775 annual screens) women underwent unnecessary surgery per 10,000 screens. Conclusions: Population screening for ovarian and tubal cancer for average-risk women using these strategies should not be undertaken. Decreased incidence of Stage III/IV cancers during multimodal screening did not translate to mortality reduction. Researchers should be cautious about using early stage as a surrogate outcome in screening trials. Meanwhile the bioresource provides a unique opportunity to evaluate early cancer detection tests. Funding: Long-term follow-up UKCTOCS (2015-2020) - National Institute for Health and Care Research (NIHR HTA grant 16/46/01), Cancer Research UK, and The Eve Appeal. UKCTOCS (2001-2014) - Medical Research Council (MRC) (G9901012/G0801228), Cancer Research UK (C1479/A2884), and the UK Department of Health, with additional support from The Eve Appeal. Researchers at UCL were supported by the NIHR UCL Hospitals Biomedical Research Centre and by MRC Clinical Trials Unit at UCL core funding (MR_UU_12023).