Meta-analysis of annual probability of outcome for different cervical disease states

Abstract

Objective: To explore the annual probabilities of outcomes for different cervical disease states. Methods: Cohort studies related to the natural history of cervical cancer were retrieved from PubMed, Embase and China Biomedical Literature Database, and the retrieval time was from the establishment of the database to May 2020. Newcastle-Ottawa scale was used to evaluate the quality of the included literatures. The annual outcome probabilities of different cervical disease states in high-risk human papillomavirus (hrHPV) positive, negative and cervical intraepithelial neoplasia grade 1 (CIN1) population were calculated (95%CI). Random-effects model was used for meta-analysis. Egger's test was used to evaluate publication bias; sensitivity analysis was used to evaluate the robustness of the combined parameters. Meta-regression was used to explore factors associated with the heterogeneity of annual outcome probability. Results: A total of 37 studies were included, including 12, 20 and 15 studies involving hrHPV negative, hrHPV positive and CIN1 population, respectively, with a Newcastle -Ottawa scale (NOS) score of 7.05±1.20. The annual probability (95%CI) of progression to CIN1, CIN2 and CIN3+ in hrHPV-positive population were 0.022 2 (0.014 3, 0.031 0), 0.017 0 (0.012 0, 0.022 0) and 0.016 2 (0.012 6, 0.019 8), respectively. The annual probability (95%CI) of progression to CIN1, CIN2 and CIN3+ in hrHPV-negative population was 0.002 7 (0.000 9, 0.004 6), 0.000 7 (0.000 3, 0.001 1) and 0.000 6 (0.000 3, 0.000 9), respectively. The annual probability (95%CI) of reversal to normal, maintenance of CIN1 status and progression to cervical intraepithelial neoplasia grade 2 or above (CIN2+) in CIN1 population were 0.578 1 (0.369 9, 0.786 3), 0.400 1 (0.167 4, 0.632 9), 0.056 9 (0.034 9, 0.078 9), respectively. Egger's test showed that there was publication bias in the annual outcome probability of hrHPV positive progression to CIN2 and CIN3+ and hrHPV negative progression to CIN2 and CIN1 progression to CIN2+, with t values of 5.50, 2.36, 2.80 and 4.12, respectively (all P values<0.05). Sensitivity analysis showed that when excluding any of the studies, the range of annual probability of progression to CIN1, CIN2 and CIN3+ were 0.016 6-0.024 7, 0.014 9-0.018 9 and 0.013 6-0.017 7 among hrHPV-positive population; 0.002 4-0.003 5, 0.000 6-0.000 9 and 0.000 5-0.000 7 among hrHPV-negative population and the range of annual probability of CIN1 reversal to normal, maintenance as CIN1 and progression to CIN2+ were 0.531 8-0.631 2, 0.321 9-0.443 3, and 0.052 0-0.061 0, respectively. Meta-regression analysis showed that region, population origin, population cytological diagnosis, follow-up time, and NOS score were not associated with the heterogeneity of annual outcome probability (all P values>0.05). Conclusion: The annual outcome probability of different cervical disease states in hrHPV positive population is high, and the CIN1 population only needs close follow-up.