Postoperative levonorgestrel-releasing intrauterine system versus oral contraceptives after gonadotropin-releasing hormone agonist treatment for preventing endometrioma recurrence.

OBJECTIVE

Although the levonorgestrel-releasing intrauterine system (LNG-IUS) is effective in reducing the recurrence of endometriosis-associated pain, its efficacy in preventing endometrioma recurrence is questionable. We compared the efficacy of postoperative use of LNG-IUS with oral contraceptives (OC) for preventing endometrioma recurrence.

DESIGN

A retrospective cohort study.

SETTING

Medical university hospital.

POPULATION: Ninety-nine women with endometriomas.

METHODS

A chart review was performed of women of reproductive age who had undergone laparoscopic surgery for endometrioma followed by three cycles of gonadotropin-releasing hormone agonist (leuprolide acetate) treatment. Women were categorized into two groups: a group that had postoperative LNG-IUS placement (n = 42) and a group that received postoperative, cyclic, low-dose, monophasic, OCs (n = 57). Main outcome measures. Endometrioma recurrence was analyzed according to several clinical variables and postoperative treatment modalities.

RESULTS

During the follow-up period (median 17 months), recurrent endometriomas were detected in eight women (8.1%). Patients with LNG-IUS had a recurrence rate of 4.8% (2/42), whereas women receiving OC had a recurrence rate of 10.5% (6/57). Cumulative recurrence-free survival assessment revealed that mean disease-free survival times for both groups were similar, but that for LNG-IUS was slightly longer than that for OC, with statistical significance (34.4 ± 1.0 months, 95% confidence interval 32.3–36.5, vs. 33.4 ± 1.3 months, 95% confidence interval 30.8–36.0, p = 0.045). Univariate analysis revealed a hazard ratio of 0.178 (95% confidence interval 0.029–1.075) (p = 0.060) for postoperative LNG-IUS use and endometrioma recurrence. However, for the multivariate regression analysis, only postoperative serum CA 125 levels were significantly associated with endometrioma recurrence (hazard ratio 1.012, p = 0.010).

CONCLUSIONS

Postoperative LNG-IUS use seemed to be comparable to the use of cyclic OC in preventing endometrioma recurrence.