The risk of malignancy in uterine polyps : a systematic review and meta-analysis

BACKGROUND: Endometrial polyps are a common condition. The risk of malignancy has often led to an overtreatment with high health care costs and huge psychological distress. OBJECTIVE: We conducted a systematic review and a meta-analysis in order to estimate the prevalence of premalignant and malignant lesions in women undergoing hysteroscopic polypectomy. DATA SOURCE: We developed the search using PubMed/MEDLINE and EMBASE databases to identify papers published between 2000- January 2019. The research strategy used on Pubmed was: ("polyps" (MESH) OR "endometrial polyp*") AND ("malignancy" OR "cancer" OR "histopathology" OR "hysteroscopy" OR "ultrasound", OR "sonohysterography"). The same search was modified for EMBASE. STUDY ELIGIBILITY: We included all observational retrospective and prospective studies and studies were selected for the review if they met following inclusion criteria: pre-operative diagnosis of benign-looking endometrial polyps at ultrasound examination or at hysteroscopy, excision of endometrial polyps via surgical hysteroscopy, histopathological diagnosis of benign polyps, or hyperplasia without atypia, or premalignancy (atypical hyperplasia) or malignancy (endometrial cancer). Moreover, studies were included if number or percentage of subjects with and without malignancy was provided and if they reported data about menopausal and/or bleeding status. We excluded data presented exclusively as abstracts in national and international meetings, or case report or review articles that did not include original data and papers published in other than English language. Our primary outcome was the prevalence of endometrial premalignant or malignant polyps in the total series, among premenopausal and postmenopausal women and among women with or without abnormal bleeding and then in subgroup analysis according to study design, diagnostic method, study region and calendar year of publication. RESULTS: A total of 51 studies reporting data on 35,345 women were included in this review. The prevalence of malignant polyps was 2.73% (95% CI 2.57-2.91) with very high heterogeneity among studies. The rates were lower for premenopausal women (1.12%) than post-menopausal ones (4.93%) and the difference was statistically significant (chi-square = 397.21. p < .0001). The risk of malignancy was higher among symptomatic (5.14%) than asymptomatic ones (1.89%) (chi-square = 133.13 p < .001). We observed higher rate of malignant polyps in prospective studies. In the meta-analysis selecting 10 prospective studies the random pooled estimate was 5.88 (95% CI: 4.06-7.97) with heterogeneity among studies (heterogeneity chi square = 17.55 P = .025) whereas in retrospective studies the random pooled estimate was 2.94 (95% CI:2.24-3.71) with high heterogeneity among studies (P < .001). This finding can be due to more strict diagnostic criteria in prospective studies. CONCLUSION: Symptomatic vaginal bleeding and postmenopausal status in women with endometrial polyps increased the risk of malignancy. This finding could be an useful evidence to select patients who need to undergo hysteroscopic resection of endometrial polyps and women to whom, instead, an expectant management can be offer.