Purpose: In patients with histopathologically proven or suspected endometriosis with possible involvement of the rectum, endorectal ultrasound was performed to determine the sensitivity and specificity of this method with regard to rectal wall involvement and the impact on the following operation.
Methods: In an historical cohort analysis, 85 females with histopathologically proven or suspected endometriosis with possible involvement of the rectum were treated between 1992 and 2001. Endorectal ultrasound was performed with a 7.5 MHz real-time unit, and results of endorectal ultrasound were compared with intraoperative findings and histopathologic diagnosis of 65 patients undergoing operation. A questionnaire was used to evaluate postoperative signs and symptoms.
Results: Of 65 patients undergoing surgery, 37 underwent laparotomy with 25 resections of the bowel and 28 laparoscopy. In 31 of 32 patients with suspected rectal wall infiltration, preoperative endorectal ultrasound diagnosis was confirmed. In patients in whom endorectal ultrasound showed no rectal wall involvement, histopathology revealed infiltration in one patient, leading to sensitivity of 97 percent and specificity of 97 percent with regard to rectal wall involvement. In terms of the deepness of rectal wall infiltration, endorectal ultrasound had a sensitivity of 76 percent with regard to infiltration of the muscularis propria and 66 percent for infiltration of the submucosa. Operations led to a significant (P < 0.05) reduction of preoperative symptoms by approximately 60 percent.
Conclusions: Endorectal ultrasound is a useful, noninvasive technique for preoperative evaluation of possible rectal wall involvement in endometriosis. Based on the high sensitivity and specificity, recommendation for laparotomy and bowel resection in cases with suspected rectal involvement can be facilitated.