Accuracy of magnetic resonance in deeply infiltrating endometriosis: a systematic review and meta-analysis

Arch Gynecol Obstet. 2015 Mar;291(3):611-21. doi: 10.1007/s00404-014-3470-7. Epub 2014 Oct 7.

Abstract

Objective: To estimate the accuracy of pelvic magnetic resonance imaging (MRI) in the diagnosis of deeply infiltrating endometriosis (DIE).

Methods: A comprehensive search of the Medline, Pubmed, Lilacs, Scopus, Embase, Cochrane Central Register of Controlled Trials (CENTRAL), Biomed Central, and ISI Web of Science databases was conducted from January 1990 to December 2013. The medical subject headings (MeSHs) and text words "deep endometriosis", "deeply infiltrating endometriosis", "DIE", "magnetic resonance", and "MRI" were searched. Studies that compared the parameters of pelvic MRIs with those of paraffin-embedded sections for the diagnosis of DIE were included.

Results: Twenty studies were analyzed, which included 1,819 women. Pooled sensitivity and specificity were calculated across eight subgroups: for all sites, these were 0.83 and 0.90, respectively; for the bladder, 0.64 and 0.98, respectively; for the intestine, 0.84 and 0.97, respectively; for the pouch of Douglas, 0.89 and 0.94, respectively; for the rectosigmoid, 0.83 and 0.88, respectively; for the rectovaginal, 0.77 and 0.95, respectively; for the uterosacral ligaments, 0.85 and 0.80, respectively; and for the vagina and the posterior vaginal fornix, 0.82 and 0.82, respectively.

Conclusion: In summary, pelvic MRI is a useful preoperative test for predicting the diagnosis of multiple sites of deep infiltrating endometriosis.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't
  • Review
  • Systematic Review

MeSH terms

  • Abdominal Cavity / pathology*
  • Adnexa Uteri / pathology*
  • Endometriosis / pathology*
  • Female
  • Humans
  • Magnetic Resonance Imaging / methods*
  • Pelvis / pathology*
  • Predictive Value of Tests
  • Preoperative Care / methods
  • Sensitivity and Specificity
  • Vagina / pathology