Impact of Erythropoiesis Stimulating Agents on Survival and Risk of Venous Thromboembolism Among Patients with Colorectal Cancer Receiving Chemotherapy

Abstract

Objectives: To describe utilization of erythropoiesis stimulating agents (ESAs) over time in the VA; to compare characteristics of ESA users with non-users among solid tumor patients receiving chemotherapy; and to evaluate the effect of ESA use on survival and risk of venous thromboembolism among colorectal cancer patients undergoing chemotherapy. Methods: This study is a nonconcurrent prospective cohort study using the 2006 - 2008 Veterans Health Administration administrative data. There were two components to the study sample; 28,868 solid tumor patientseceiving chemotherapy to describe ESA utilization over time and 2,462 colorectal cancer patients receiving chemotherapy to estimate the effect of ESA on mortality and time to venous thromboembolism. A conventional multivariate adjusted and propensity score weighted Cox proportional hazard regressions were performed. Results: Overall, 10% of solid tumor patients receiving chemotherapy in the cohort used ESAs at some time during the follow-up period. The proportion of ESA users among solid tumor patients decreased by half during the 2-year period. Use of ESA's during chemotherapy treatment was associated with a 2-fold increase in mortality. The estimated hazard ratios of ESA use for mortality among colorectal cancer patients receiving chemotherapy were 1.957 (95% CI = 1.522 - 2.517) in a conventional multivariate model and 2.045 (95% CI = 1.799 -2.326) in a propensity score weighted model. Conclusion: This population-based observational study demonstrated that the use of ESAs declined over time in the VA healthcare system between July 2006 and August 2008 and showed survival was decreased with the combination use of chemotherapy plus ESAs compared with chemotherapy alone. While gaps in the available data elements, particularly cancer staging, and other potential biases inherent in observational studies must be considered when interpreting these results, the study's findings are consistent with the clinical trials at other cancer sites and the meta-analyses which combined cancer sites suggesting that the decreases in survival are not site dependent. This provides important insight regarding a fair balance of risks and benefits for the management of anemia by the combination use of chemotherapy and ESAs in colorectal cancer population.