We assessed the evidence about management of uterine fibroids. Specifically, we sought to determine effectiveness of interventions, risks of harm, and whether individual or fibroid characteristics influence outcomes.
We searched MEDLINE® via PubMed® and Embase® to identify publications, as well as reviewed the reference lists of included studies.
We included studies published in English from January 1985 to September 2016. We identified randomized clinical trials to assess outcomes and harms of interventions. We used data from trials in a meta-analysis to estimate probability and timing of subsequent interventions for fibroids based on initial type of intervention. To describe risk of unrecognized leiomyosarcoma, we included studies that allowed calculation of prevalence of leiomyosarcoma discovered at the time of surgery for masses believed to be fibroids. We also identified publications that indicated operative approaches to removal of leiomyosarcoma tissue and built models to estimate survival. We extracted data, assessed risk of bias, and rated the strength of evidence for informing care.
Of 97 included randomized trials, 43 studies assessed medications, 28 assessed procedures, and 37 assessed surgeries. Gonadotropin-releasing hormone (GnRH) agonists, mifepristone, and ulipristal reduced fibroid size and improved fibroid-related symptoms, including bleeding and quality of life (moderate strength of evidence [SOE] except quality of life for GnRH agonist [low SOE]). Several other medications have promise but are not supported by sufficient evidence. Uterine artery embolization (UAE) (high SOE) as well as high intensity focused ultrasound (low SOE) are effective for decreasing fibroid size/volume. Few other outcomes are well investigated for high intensity focused ultrasound. UAE studies reported improved outcomes for bleeding (moderate SOE), and quality of life (moderate SOE). Myomectomy and hysterectomy improved q uality of life (both low SOE). Few well-conducted trials directly compared different treatment options. No studies were designed to evaluate expectant management, and evidence is insufficient to guide clinical care. Subsequent intervention ranged from 0 to 44 percent in studies that followed women after initial fibroid treatment. At 2-year followup, subsequent intervention rates were lowest for initial medical management and higher for UAE and myomectomy, especially among younger women. No individual characteristics of women or their fibroids were definitely associated with likelihood of intervention benefits or patient satisfaction. These findings were limited by the number and size of available studies. Using data from 160 studies, we estimated that among 10,000 women having surgery for presumed fibroids, between 0 and 13 will have a leiomyosarcoma detected. Of the surgical approaches, the 5-year survival after leiomyosarcoma diagnosis was 30 percent with power morcellation (95% Bayesian credible interval [BCI]: 13% to 61%), 59 percent with scalpel morcellation (BCI: 33% to 84%), and 60 percent with intact removal (BCI: 24% to 98%).
A range of interventions are effective for reducing fibroid size and improving symptoms. Some medications and procedures also improve quality of life. Few studies directly compare interventions. The risk of encountering a leiomyosarcoma at the time of fibroid surgery is low, and the method of fibroid removal may influence survival. Evidence to guide choice of intervention is likely best when applied in the context of individual patient needs and preferences.
