Modifiable risk factors associated with long-term survival in women with serous ovarian cancer: a National Cancer Database study.

ABSTRACT

OBJECTIVE: To identify patient, clinical and hospital factors associated with long-term survival (≥10 years) in women with serous ovarian cancer. METHODS: This National Cancer Database cohort study included women with stage II-IV serous ovarian cancer. Multivariate logistic regression models were used to examine the association of long-term survival with patient (race, insurance, location, household income, education, distance traveled), clinical (age, comorbidities, stage, grade, primary treatment) and hospital factors (region, institution, hospital volume ≥20). RESULTS: Of the 4640 women identified, 12% (n=561) experienced long-term survival. Median overall survival was 41 months (95% CI 39 to 42). The odds of long-term survival were lower for women with public or no insurance (adjusted OR 0.71, 95% CI 0.55 to 0.92), age ≥75 years (0.33, 0.22 to 0.50), any comorbidities (0.70, 0.54 to 0.92), higher stage (stage III 0.31, 0.25 to 0.41; stage IV 0.16, 0.12 to 0.22), and moderately/poorly differentiated, undifferentiated, or tumors of unknown grade (moderately/poorly differentiated 0.30, 0.20 to 0.47; undifferentiated 0.28, 0.17 to 0.47; unknown 0.30, 0.18 to 0.50). The odds of long-term survival among women who were publicly insured were lower with neoadjuvant chemotherapy (0.13, 0.04 to 0.044) and higher with optimal cytoreduction (2.24, 1.49 to 3.36). Among women who were privately insured, the odds of long-term survival were higher with optimal cytoreduction (1.99, 1.46 to 2.70) and unaffected by neoadjuvant chemotherapy. CONCLUSIONS: While immutable clinical factors such as age, stage, and grade are associated with long-term survival in women with serous ovarian cancer, modifiable factors, such as insurance type, optimal cytoreductive status, and neoadjuvant chemotherapy provide an opportunity for targeted improvement in care with potential to affect long-term patient outcomes.