Endometriosis and risk of ovarian and endometrial cancers in a large prospective cohort of U.S. nurses
Menée aux Etats-Unis à partir des données portant sur 199 134 participantes, cette étude de cohorte prospective évalue l'association entre une endométriose et le risque de cancer de l'ovaire (durée de suivi : 18 ans, 228 cas de cancer de l'ovaire, 116 cancer de l'endomètre)
Purpose : Endometriosis is associated with ovarian cancer, but the relation with endometrial cancer is unclear. Prior studies generally were retrospective and had potential limitations, including use of self-reported endometriosis, failure to account for delays between symptom onset and endometriosis diagnosis, and changes in risk factors post-endometriosis diagnosis. We evaluated whether these limitations obscured a weak association with endometrial cancer and the extent to which these limitations impacted associations with ovarian cancer. Methods : Cox proportional hazards regression models were used to assess associations between endometriosis and cancer risk, evaluating the impacts of self-reported vs. laparoscopically confirmed endometriosis, delayed diagnosis, and post-endometriosis diagnosis changes in risk factor exposures on relative risk estimates. Results : Over 18 years of follow-up, we identified 228 ovarian and 166 endometrial cancers among 102,025 and 97,109 eligible women, respectively. Self-reported endometriosis was associated with ovarian cancer [relative risk (RR): 1.81; 95% confidence interval (CI): 1.26–2.58]; this association was stronger for laparoscopically confirmed endometriosis (HR: 2.14; 95% CI 1.45–3.15). No association was observed with endometrial cancer (self-report RR: 0.78; 95% CI 0.42–1.44; laparoscopic-confirmation RR: 0.76; 95% CI 0.35–1.64). Accounting for diagnosis delays or post-endometriosis diagnosis changes in risk factors had a little impact. Conclusions : This study adds to the evidence that endometriosis is not strongly linked to endometrial cancer risk and that the association with ovarian cancer is robust to misclassification, diagnostic delay, and changes in exposures post-endometriosis diagnosis. Our analysis suggests that confounding and misclassification do not obscure a weak association for endometrial cancer risk, although our results should be replicated.