Reoperations Following Surgical Treatment of Ovarian Endometriosis: Incidence, Risk Factors, and Surgical Approaches

BACKGROUND: Ovarian endometriosis is one of the most frequently diagnosed endometriosis subtypes. Surgical treatment may be required in selected cases, and indications for surgery have evolved over the years. However, recurrence is common and reoperation is likely to compromise ovarian reserve, while long-term data on rates of reoperation and its risk factors remain limited. OBJECTIVE: The aim was to assess long-term reoperation risk and its predictors in surgically verified ovarian endometriosis. STUDY DESIGN: We used Finnish national registers to identify women with first surgically verified diagnosis of ovarian endometriosis who underwent an ovarian procedure at the index surgery between 1987 and 2012 (n = 13,912). Follow-up continued until 2019. Subsequent operations were identified using predefined procedural codes at plausible endometriosis sites, with a concurrent diagnosis of endometriosis or adenomyosis, excluding procedures within 30 days. We classified all procedures as ovarian, peritoneal, deep infiltrating endometriosis, or other procedures, and compared women with a single versus repeat surgeries (overall and ovarian) and across index eras (ICD-9 1987-1995; ICD-10 1996-2012). Cumulative reoperation risk was estimated using Kaplan-Meier analyses, and associated factors were evaluated using Cox regression. RESULTS: Median follow-up was 20.7 years (IQR 14.8-27.2). The 30-year cumulative risks of any and ovarian reoperation were 17.3% and 16.5%, respectively. Median time to first reoperation was 3.4 years (IQR 1.4-7.2), but subsequent reoperations occurred sooner (median <2 years). Reoperated women had fewer births (61.6% vs 70.7%, P<.001) and more infertility (51.5% vs 17.1%, P<.001) than those with a single surgery. Reoperation was more common during the ICD-10 than ICD-9 era (19.0% vs 12.4%, P<.001) and occurred sooner (median 3.1 vs 4.5 years, P<.001). Despite restriction ovarian endometriosis as the sole diagnosis at index surgery, peritoneal (3.6%) and deep infiltrating endometriosis (1.8%) procedures were recorded. At subsequent surgeries, all endometriosis and adenomyosis diagnosis were permitted, and the proportion of deep infiltrating endometriosis procedures increased to 13.4% by the fourth or subsequent operation. In multivariable Cox regression, young age, family history of endometriosis, and primary endometriosis diagnosis were associated with increased risks of both any and ovarian reoperations, whereas hysterectomy before or at index surgery reduced the risks. Infertility before index surgery was associated with any reoperation. Compared with oophorectomy/ies, enucleation/resection (aHR 1.43, 95% CI 1.28-1.60) and other ovarian procedures (aHR 2.20, 95% CI 1.67-2.89) increased the risk of any reoperation; for ovarian reoperation, increased risk was observed only with other ovarian procedures (aHR 1.68, 95% Cl 1.23-2.31). Index surgery performed during ICD-10 era was associated with reduced ovarian reoperation risk. CONCLUSIONS: Approximately one in six women required reoperation after surgery for ovarian endometriosis. Risks of any and ovarian reoperation were higher among younger women, those with a family history of endometriosis, and a primary diagnosis of endometriosis. Reoperation occurred in approximately one-third of women after ovarian-conserving procedures compared with one in ten women after oophorectomy. These findings highlight the surgical burden of ovarian endometriosis and support individualized, fertility-conscious management when balancing ovarian preservation against risk of recurrence.