Metformin use in diabetic patients with multiple myeloma (MM): A SEER-Medicare study.
e18127 Background: Metformin, an oral hypoglycemic agent, has been linked to favorable outcomes among patients (pts) with cancer (Coyle et al. Ann Oncol 2016). Wu et al. demonstrated better overall survival (OS) with metformin use in pts with mm and Diabetes Mellitus (DM) (Br J Cancer 2014). They also found that patient with steroid-induced DM [SID] had worse outcomes than pts with pre-existing DM [PDM]. However, Keller et al. did not find these associations in a study of VA mm pts (Blood 2015 126:4502). We evaluated the impact of DM and metformin use on outcomes of pts with mm in the SEER-Medicare database. Methods: We used SEER-Medicare claims from 2007-2013 for pts with newly diagnosed MM. DM was identified using ICD-9 codes, antidiabetic medications by Medicare part D prescription claims. Multivariate analysis was performed by Cox proportional hazard modelling to evaluate (1) Association of DM type (PDM or SID) with mm outcomes and (2) Impact of metformin on outcomes of pts with mm and DM. Covariates included age, gender, transplant, CKD (chronic kidney disease), CCI (Charlson Comorbidity Index), and use of other antidiabetic medications. Potential smoldering mm cases in the dataset were controlled for using a pre-established algorithm (Fakhri, et al,Blood 2016 128:2348) Results: Out of 4708 mm pts, 45% had DM. Among the pts with DM, 28% were on metformin. Out of the 2112 pts with DM, 20% had SID. Median follow up was 26 months. We found no association between PDM (aHR 1.04, 95% CI 0.96-1.12) or SID (aHR 1.03, 0.91-1.17) with OS when compared to non-diabetic pts. Metformin use, female gender, transplant and CCI < 5 were significantly associated with improved OS. Metformin use was associated with an 18% decrease in mortality (aHR 0.82, 0.72-0.93, p = 0.002); controlling for CKD did not impact the association (aHR 0.83 (0.73- 0.94, p = 0.005). Other antidiabetic medications including insulin were not associated with OS. Conclusions: DM (PDM or SID) and insulin use were not associated with poorer outcomes. Metformin use was associated with improved OS among pts with mm and DM in the SEER-Medicare database. The impact of metformin on outcomes in mm pts should be further investigated in prospective clinical trials.