Premenopausal abnormal uterine bleeding and risk of endometrial cancer

Abstract

$\textbf{Background}$ Endometrial biopsies are undertaken in premenopausal women with abnormal uterine bleeding but the risk of endometrial cancer or atypical hyperplasia is unclear. $\textbf{Objectives}$ To conduct a systematic literature review to establish the risk of endometrial cancer and atypical hyperplasia in premenopausal women with abnormal uterine bleeding. $\textbf{Search strategy}$ Search of PubMed, Embase and the Cochrane Library from database inception to August 2015. $\textbf{Selection criteria}$ Studies reporting rates of endometrial cancer and/or atypical hyperplasia in women with premenopausal abnormal uterine bleeding. $\textbf{Data collection and analysis}$ Data were independently extracted by two reviewers and cross-checked. For each outcome, the risk and a 95% CI were estimated using logistic regression with robust standard errors to account for clustering by study. $\textbf{Main results}$ Sixty-five articles contributed to the analysis. Risk of endometrial cancer was 0.33% (95% CI 0.23–0.48%, $\textit{n}$ = 29 059; 97 cases) and risk of endometrial cancer or atypical hyperplasia was 1.31% (95% CI 0.96–1.80, $\textit{n}$ = 15 772; 207 cases). Risk of endometrial cancer was lower in women with heavy menstrual bleeding (HMB) (0.11%, 95% CI 0.04–0.32%, $\textit{n}$ = 8352; 9 cases) compared with inter-menstrual bleeding (IMB) (0.52%, 95% CI 0.23–1.16%, $\textit{n}$ = 3109; 14 cases). Of five studies reporting the rate of atypical hyperplasia in women with HMB, none identified any cases. $\textbf{Conclusions}$ The risk of endometrial cancer or atypical hyperplasia in premenopausal women with abnormal uterine bleeding is low. Premenopausal women with abnormal uterine bleeding should first undergo conventional medical management. Where this fails, the presence of IMB and older age may be indicators for further investigation. Further research into the risks associated with age and the cumulative risk of co-morbidities is needed.