Objectives: To assess the value of MRI and ano-rectal endosonography (ARES) for the diagnosis and surgical prognosis of rectovaginal septum endometriosis and to analyse the surgical management in order to evaluate its functional results and complications.
Patients and methods: Retrospective study of 50 consecutive patients operated for a clinical presumption of endometriosis nodule of the recto vaginal septum. Thirty-nine patients had a MRI, 31 an ARES and 28 both exams. All the patients had a complete dissection of the rectovaginal septum and all lesions were excised.
Results: For the diagnosis of rectovaginal septum endometriosis nodule, MRI results are: sensitivity 73%, specificity 50%, positive predictive value (PPV) 89%, negative predictive value (NPV) 25%; for uterosacral ligaments involvement: sensitivity 84%, specificity 95%, PPV 94%, NPV 86% and for rectal wall infiltration: sensitivity 53%, specificity 82%, PPV 69%, NPV 69%. The ARES results for diagnosis of rectovaginal septum endometriosis nodule are: sensitivity 93%, specificity 100%, PPV 100%, NPV 50% and for rectal wall infiltration: sensitivity 100%, specificity 71%, PPV 81%, NPV 100%. ARES appeared more sensitive than MRI for the detection of rectal wall infiltration (P = 0.002) and for rectovaginal septum endometriosis nodule diagnosis (P = 0.03). Eighty-nine percent of the patients had a coelioscopy in first intention and 15 laparoconversions were performed, 11 in order to perform a digestive resection: 45 nodules were found. In 43cases the nodule was excised, associated to 19 digestive resections, 30 colpectomys, and 22 uterosacral ligaments resections. Three patients required an additional surgical treatment by Hartman's procedure with Mickulicz's drainage for peritonitis. Forty-one nodules were endometriosis nodules: the two other cases were fibrosis nodules. Thirty-three patients were interviewed about the evolution of their pains over a mean history of 20 months: 90% of the patients were satisfied with the management results.
Discussion and conclusions: Our data support the efficiency of MRI for rectovaginal septum endometriosis nodule and uterosacral ligaments involvement diagnosis; accord ARES to rectovaginal septum endometriosis nodule diagnosis and its reliability in establishing a diagnosis of rectal wall involvement. The surgical cure of rectovaginal septum nodules without digestive infiltration is performed by coelioscopic or coelio-vaginal procedure, but in case of associated digestive affliction, laparotomy is actually the standard procedure in order to achieve a complete cure of the lesions. Complications, in particular peritonitis, are not frequent. Our data support the efficiency of radical surgical treatment for the improvement of pain symptoms. Results on fertility seem to be satisfactory, but complication risks suggest being careful in this indication. Clinical examination during a catamenial period is essential in order to evoke the diagnosis. MRI yields a complete map of the sub-peritoneal and peritoneal lesions and ARES allows for the diagnosis of an infiltration of the rectal wall. Pre-operative association of those two exams is actually indispensable for the surgical management of those patients, which consists of complete excision of endometriosical lesions and is efficient at treating pain symptoms and fertility. Complications are rare but severe, therefore, justifying a cure in specialised centres.