Abstract
Background: Ovarian cancer is a rare disease with a poor prognosis. Type 2 Diabetes Mellitus (T2D) is associated with increased cancer risk and mortality. Metformin is a common first-line treatment for T2D that has been shown to have multiple anti-cancer effects. Repurposing metformin for ovarian cancer treatment is attractive, but associations between T2D, metformin use, and ovarian cancer survival have been evaluated in only a few studies to date.
Methods: Electronic medical record (EMR) data was abstracted for tumor registry confirmed ovarian and fallopian tube cancer cases from the Vanderbilt University Medical Center; MedEx was used to determine metformin use. Associations with T2D or metformin were evaluated with chi-square tests or Student's t-tests. Associations with overall survival were evaluated using Cox proportional hazards regression to compute Hazard Ratios (HR) and corresponding 95% Confidence Intervals (CI). Associations with and without adjustment for age, race, stage of disease, and histologic subtype were modeled. Kaplan-Meir functions were used to visualize survival functions; differences were evaluated via the log-rank test. For 50 subjects, we used natural language processing assisted manual review to assess metformin use; a Kappa coefficient was used to compare this to MedEx captured use.
Results: EMR data was available for 414 Tumor Registry confirmed cases; of these, 372 (89.9%) were invasive epithelial ovarian cancers. Of the 372 cases, 332 were non T2D ovarian cancer cases of which 47 (14.2%) were from minority population(s) (non-Caucasian). There were a total of 40 T2D ovarian cancer cases of which 7 (17.5%) were from minority population(s). There were more high stage (III & IV) cancers among T2D than non-T2D OC cases (P=0.032). T2D cases on metformin were taller and weighed more than T2D cases without metformin use (P=0.034 and P=0.044, respectively). Associations with overall survival were suggestive but non-significant for T2D (HR: 0.72, 95% CI: 0.47-1.08) and metformin use, either compared to T2D cases without metformin use (HR: 0.56, 95% CI: 0.23-1.39) or compared to all ovarian cancer cases (HR: 0.53, 95% CI: 0.27-1.05). MedEx captured and manually curated metformin use was highly concordant (Kappa=0.96).
Conclusions: We found that ovarian cancer cases with metformin use had marginally better overall survival than both T2D cases not taking metformin and all ovarian cancer cases. While not significant, these associations indicate that metformin may have utility clinically repurposed as a treatment for ovarian cancer. Finally, high agreement between MedEx captured and natural language processing assisted manually curated metformin use indicates that MedEx captured metformin use can be used in larger EMR studies.
Citation Format: Leshaun Clayton, Spencer Keene, Samantha P. Stansel, Min Jiang, Melinda Aldrich, Hua Xu, Jeremy Warner, Joshua Denny, Dineo Khabele, Alicia Beeghly-Fadiel. Assessing metformin use and ovarian cancer survival from electronic medical records. [abstract]. In: Proceedings of the Eighth AACR Conference on The Science of Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; Nov 13-16, 2015; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2016;25(3 Suppl):Abstract nr C19.
RSS Feeds