Ultrasonographic Findings Indirectly Predicting Parametrial Involvement in Patients with Deep Endometriosis: The ULTRA-PARAMETRENDO I Study

Study objective: To evaluate ultrasonographic findings as a first-line imaging tool to indirectly predict the presence of parametrial endometriosis (PE) in women with suspected deep endometriosis (DE) undergoing surgery. Design: Retrospective analysis of a prospectively collected database (ULTRA-PARAMETRENDO I study; NCT05239871). Setting: Referral center for DE. Patients: Consecutive patients undergoing laparoscopic surgery for DE. Interventions: Preoperative transvaginal ultrasonography was done according to the International Deep Endometriosis Analysis consensus statement. A stepwise forward regression analysis was performed considering the simultaneous presence of DE nodules and the following ultrasonographic indirect signs of DE: diffuse adenomyosis, endometriomas, ovary fixed to the lateral pelvic wall or the uterine wall, absence of anterior/posterior sliding sign, and hydronephrosis. The gold standard for the presence of PE was surgery with histologic confirmation. Measurements and main results: Of 1079 patients, 212 had a surgical diagnosis of PE (left: 18.5%; right: 17.0%; bilateral: 15.9%). The obtained prediction model (χ2 = 222.530; p <.001) for PE included, as independent indirect DE signs presence of hydronephrosis (odds ratio [OR] = 14.5; p = .002), complete absence of posterior sliding sign (OR = 3.3; p <.001), presence of multiple endometriomas per ovary (OR = 3.0; p = .001), and ovary fixation to the uterine wall (OR = 2.4; p <.001); as independent concomitant DE nodules, presence of uterosacral nodules with the largest diameter >10 mm (OR = 3.2; p <.001), presence of rectal endometriosis with the largest diameter >25 mm (OR = 2.3; p = .004), and rectovaginal septum infiltration (OR = 2.3; p = .003). The optimal diagnostic balance was obtained considering at least 2 concomitant DE nodules and at least 1 indirect DE sign (area under the curve 0.75; 95% confidence interval, 0.72-0.79). Conclusion: Specific indirect ultrasonographic findings should raise suspicion of PE in women undergoing preoperative assessment for DE. The suspicion of parametrial invasion may be critical to address patients to expert leading centers, where proper diagnosis and surgical treatment for PE can be performed.