Effects of utero-ovarian anastomoses on clinical outcomes and repeat intervention rates after uterine artery embolization

PURPOSE: To assess the prevalence of anastomoses between uterine and ovarian arteries on angiography and their association with clinical outcome after uterine artery embolization (UAE). MATERIALS AND METHODS: Consecutive patients who underwent UAE for symptomatic uterine leiomyomata according to a uniform embolization technique by a single operator at a single institution were included in the study. Patients' angiographic images, pre- and postoperative magnetic resonance (MR) images, and clinical symptom evaluations were reviewed. MR imaging was performed 6 months after UAE, and clinical evaluation with symptom severity score (SSS) measurement was performed at 6-month and yearly intervals afterward. Leiomyomata volume change, SSS, and repeat intervention rates were compared for patients with and without anastomoses between uterine and ovarian arteries on angiography. RESULTS: Of 288 consecutive patients in the study (mean age, 43.8 +/- 6.4 y), patent anastomoses between the uterine and ovarian arteries were detected in 116 patients (40.3%) by angiography. On follow-up, there was a mean leiomyomata volume reduction of 50.4% and an improvement in mean transformed SSS of 38.2 points. In patients with anastomoses, mean leiomyomata volume reduction was 49.5% and mean transformed SSS improvement was 38.1 points. In patients without anastomoses, mean leiomyomata volume reduction was 50.4% and mean transformed SSS improvement was 38.4 points. At a mean follow-up of 21.5 months, 16 patients (5.6%) elected to undergo further therapy for residual symptoms, including seven hysterectomies, four myomectomies, and five repeat UAE procedures. There were statistically significant differences in repeat intervention rates between the two groups: 14 patients with anastomoses (12.1%) underwent five hysterectomies, four myomectomies, and five repeat UAE procedures, whereas two patients without anastomoses (1.2%) elected to undergo hysterectomy (P < .0001). CONCLUSION: Anastomoses between the uterine artery and ovarian artery were demonstrated on angiography in 40.3% of 288 consecutive patients studied. Although the overall repeat intervention rate after initial UAE is only 5.6%, UAE in patients with anastomoses is associated with a significantly higher incidence of repeat intervention than UAE in patients without anastomoses.