OC19.01: Diagnosis of deep endometriosis recurrence: ultrasonography has better performance than magnetic resonance imaging

Transvaginal ultrasonography (TVS) and magnetic resonance imaging (MRI) have similar performance in diagnosing deep endometriosis (DE). The diagnosis of DE recurrence may be more challenging because of the presence of postoperative fibrosis and adhesions (such as those causing obliteration of the pouch of Douglas). This study compared the performance of TVS and MRI in diagnosing DE recurrence. This prospective study enrolled women who underwent a second surgery after surgical excision of deep endometriosis because of suspicion of DE recurrence. TVS was performed by an experienced ultrasonographer. MRI was performed within two months from TVS by a radiologist blinded to the results of TVS. Laparoscopy was performed within four months from MRI; all DE lesions were excised. The findings of TVS and MRI were compared with histological results. The study included 123 women. The median interval between the first and second surgery was 7 years (range, 2-21 years). 86 patients had a histological diagnosis of DE after the second surgery. 63 women underwent a hysterectomy, 36 excisions of ovarian endometriomas, 18 unilateral adnexectomies, and 7 bilateral adnexectomies. The McNemar's test demonstrated that TVS has higher performance than MRI in diagnosis recurrence of deep endometriosis (p = 0.035). The performance of MRI was: accuracy 79.7% (95%CI, 71.5-86.4), sensitivity 86.1% (76.9-92.6), specificity 64.9% (47.5-79.8), positive predictive value (PPV) 85.1% (78.5-89.9), negative predictive value (NPV) 66.7% (52.9-78.1), positive likelihood ratio (LR+) 2.45 (1.57-3.83), negative likelihood ratio (LR-) 0.22 (0.12-0.38). The performance of TVS was: accuracy 87.0% (79.7-92.4), sensitivity 87.2% (78.3-93.4), specificity 86.5 (71.2-95.5), PPV 93.8% (86.9-97.2), NPV 74.4 (62.3-83.7), LR+ 6.45 (2.84-14.64), and LR- 0.15 (0.08-0.26). TVS is superior to MRI in diagnosing postoperative recurrence of endometriosis. MRI has poor specificity in diagnosing DE recurrence, possibly because of postoperative fibrosis.