Location Matters: Likelihood of Superficial Endometriosis by Anatomic Site

Abstract

<h3>Study Objective</h3> Compare anatomic site of visual and histopathologic endometriosis. <h3>Design</h3> Prospective clinical study. <h3>Setting</h3> Tertiary-care, academic medical center. <h3>Patients or Participants</h3> Surgical evaluation of 77 patients for suspected endometriosis by high-volume endometriosis surgeons. <h3>Interventions</h3> Documentation of visually present or absent endometriosis by anatomic site (left pelvic sidewall, left uterosacral ligament, right pelvic sidewall, right uterosacral ligament, posterior cul-de-sac, and bladder) followed by complete pelvic peritonectomy and pathologic evaluation. <h3>Measurements and Main Results</h3> Among all specimens with histopathology of endometriosis, the rates by anatomic site were 33.6% posterior cul-de-sac, 24.2% left pelvic sidewall, 18.4% right pelvic sidewall, 9.8% bladder, 7.4% left uterosacral ligament, and 6.6% right uterosacral ligament. For anatomic sites with visually suspected endometriosis, histopathologic diagnosis was confirmed in 53.4% posterior cul-de-sac, 49.4% left pelvic sidewall, 41.4% right pelvic sidewall, 60.9% bladder, 50% left uterosacral ligament, and 46.2% right uterosacral ligament (p<0.001). Sensitivity by anatomic site was 72.1% posterior cul-de-sac, 67.7% left pelvic sidewall, 61.7% right pelvic sidewall, 56.0% bladder, 84.2% left uterosacral ligament, and 70.6% right uterosacral ligament. Bladder had the highest positive predictive value (60.9%) while right pelvic sidewall had the lowest (41.4%). Specificity was consistent between anatomic sites (74.8% posterior cul-de-sac, 74.6% left pelvic sidewall, 77.7% right pelvic sidewall, 92.8% bladder, 72.4% left uterosacral ligament, 76.7% right uterosacral ligament). Negative predictive value was also consistent between anatomic sites with the highest for left uterosacral ligament (93.3%) and lowest for left pelvic sidewall (86.3%). <h3>Conclusion</h3> Endometriosis was most frequently found in the posterior cul-de-sac and least frequently on the uterosacral ligaments. Correlation between visual and histopathologic diagnosis was 41.4%-60.9%, indicating that visually detecting endometriosis intraoperatively is frequently inaccurate and normal-appearing peritoneum does not rule out disease. Complete peritonectomy may be warranted to prevent missing disease, but if not preferred, removal of the posterior cul-de-sac and left pelvic sidewall should be considered given these locations had the highest rates of endometriosis.