To compare the rate and risk of ovarian cancer in patients with endometriosis or pelvic inflammatory disease (PID). A nationwide population cohort research compared the risk of ovarian cancer in 135,236 age-matched comparison females, 114,726 PID patients, and 20,510 endometriosis patients out of 982,495 females between 1 January 2002 and 31 December 2014 and ended on the date of confirmation of ovarian cancer, death, or 31 December 2014. In order to reduce the unbalanced characteristics, propensity score matching (PSM) was performed for 20,478 females in each subgroup. The incidence rate (per 100,000 person-years) of ovarian cancer was 8.74 (95% CI, 7.16-10.66) in comparison, 9.26 (7.54-11.39) in PID, and 28.73 (21.07-39.16) in endometriosis cohorts. The adjusted hazard ratio (aHR) of ovarian cancer was 1.17 (p = 0.296) in PID and 3.12 (p < 0.001) in endometriosis cohorts, compared with the comparison cohort in full cohort, using the multiple Cox regression model. The aHR of ovarian cancer was 0.83 (p = 0.650) in PID and 3.03 (p = 0.001) in endometriosis cohorts, compared with the comparison cohort after performing PSM. In the full cohort and PSM population, the cumulative incidence rate of ovarian cancer was significantly higher in patients with endometriosis than in those with PID or in the comparison cohort (p < 0.001 and p < 0.001). In conclusion, after considering the differences in the impacts of exposure to endometriosis or PID, patients with endometriosis were more likely to develop ovarian cancer.
Keywords: endometriosis; nationwide population cohort; ovarian cancer; pelvic inflammatory disease; propensity score matching.