Systematic evaluation of endometriosis by transvaginal ultrasound can accurately replace diagnostic laparoscopy, mainly for deep and ovarian endometriosis

Abstract

Abstract STUDY QUESTION What is the sensitivity and the specificity of preoperative transvaginal ultrasound with bowel preparation (TVUS-BP) compared to diagnostic laparoscopy (DL) for the identification of ovarian and deep sites of endometriosis? SUMMARY ANSWER DL was able to detect retrocervical, ovarian, and bladder endometriosis with similar sensitivity and specificity as TVUS-BP, whereas for vaginal and rectosigmoid endometriosis, DL had lower sensitivity and specificity than TVUS-BP. WHAT IS KNOWN ALREADY TVUS-BP is a non-invasive examination with good accuracy for diagnosing ovarian and deep endometriosis. DL is expensive and can lead to surgical complications. STUDY DESIGN, SIZE, DURATION This prospective study included a total of 120 consecutive patients who underwent surgery for suspected endometriosis with preoperative imaging (TVUS-BP), including a video of the laparoscopic procedure, between March 2017 and September 2019. PARTICIPANTS/MATERIALS, SETTING, METHODS Two radiologists performed preoperative TVUS-BP using the same protocol for diagnosing endometriosis. Two surgeons, who were blinded to the results of the preoperative imaging and clinical data, reviewed the surgical videos from the entry of the abdominal cavity until the surgeon finalized a complete and systematic review prior to beginning any dissection (considered as a DL). A data sheet was used by surgeons and radiologists to record the sites and size of disease involvement, the American Society for Reproductive Medicine (ASRM) stage, and the Enzian score. The surgical visualization of endometriosis lesions that were confirmed by histological analysis was the gold standard. MAIN RESULTS AND THE ROLE OF CHANCE DL was able to detect retrocervical, ovarian, and bladder endometriosis with similar sensitivity and specificity as TVUS-BP. DL was not able to detect vaginal endometriosis (sensitivity and specificity 0%): this is compared to a sensitivity and specificity of 85.7% and 99.1%, respectively with the utilization of a preoperative TVUS-BP. In addition, DL was notably poor at detecting rectosigmoid endometriosis, with a sensitivity of 3.7–5.6%, and this compares to 96.3% sensitivity with utilization of a preoperative TVUS (P < 0.001). For the ASRM stage, TVUS-BP results were highly correlated with the degree of endometriosis and pouch of Douglas (POD) obliteration (weighted Kappa of 0.867 and 0.985, respectively). For the Enzian score, there was a substantial correlation between TVUSP-BP and DL for compartment A (weighted Kappa = 0.827), compartment B (weighted Kappa = 0.670), and compartment C (weighted kappa = 0.814). LIMITATIONS, REASONS FOR CAUTION The number of participants included may be a limitation in this study and, as the evaluators were blinded to the physical exam, the DL accuracy could be underestimated. As biopsies of pelvic organs were obtained only if there was a suspicion of endometriosis, the gold standard was not always applicable. This aspect could underestimate the prevalence of lesions and overestimate the sensitivity and the specificity of both the TVUS-BP and the DL. WIDER IMPLICATIONS OF THE FINDINGS Preoperative TVUS-BP was accurate in identifying all sites of ovarian and deep endometriosis that were evaluated. It had significantly higher sensitivity than DL in detecting rectosigmoid endometriosis and predicting intraoperative ASRM staging and the Enzian score. These results suggest that TVUS-BP can replace DL for the diagnosis and treatment planning for patients with ovarian and deep endometriosis. STUDY FUNDING/COMPETING INTEREST(S) The authors declare no source of funding or conflicts of interest. TRIAL REGISTRATION NUMBER N/A