Can We Diagnose Early Endometriosis with Ultrasound Rather than Laparoscopy?

Abstract

Various societies have guideline statements on the role of imaging in diagnosing endometriosis. ACOG states that transvaginal ultrasound (TVUS) is the “preferred imaging technique when assessing endometriosis and/or deep endometriosis of the rectum or rectovaginal septum.” The American Society of Reproductive Medicine (ASRM) differs. It has stated, for both TVUS and MRI, “that imaging modalities have not been found to increase diagnostic accuracy.” Perhaps the Canadian Society (SOGC) stance, in my opinion, is the most reasonable. It states TVUS is “the first line investigational tool for suspected endometriosis,” whereas MRI, “could be required if deep endometriosis is suspected.” Hudelist et al. reported a sensitivity for endometriosis of approximately 30% with pelvic examination alone, but 96% with pelvic examination plus TVUS. However, this is utilizing TVUS as an anatomic image, which previously had been the mainstay of how it was being performed. Now we need to employ “dynamic imaging.” Careful examination of the work by Okaro et al. underscores this point. They evaluated 120 women who had chronic pelvic pain. What they called “hard markers” (ovarian endometriomas, hydrosalpinges) had a 100% correlation with laparoscopic findings (24 of 24 cases). “Soft markers” defined as reduced ovarian mobility, site-specific pelvic tenderness, or the presence of loculated peritoneal fluid were predictive of pelvic pathology in 37/51 (73%) of women. Many of these women, if examined by a typical anatomic ultrasound survey, would have been classified as “normal.” This underscores the increasing importance of the use of dynamic scanning in patients with chronic pelvic pain. These data further suggest an empiric course of treatment may be appropriate as 61 of 75 (81%) of the women evaluated by transvaginal ultrasound had such a need for treatment confirmed laparoscopically. All who perform or order transvaginal ultrasound should do or expect to be done a dynamic assessment in real-time looking for tenderness, decreased mobility, and non-anatomically located ovaries. When adhesions exist, ovaries may appear anatomically normal in size and appearance but not necessarily in location. While not diagnostic of adhesions, when coupled with pain on movement with a vaginal probe, or a lack of mobility of such ovaries, one’s index suspicion for pathology like endometriosis should be greatly increased.