Erythropoietin stimulating agent dose and use patterns in cancer patients

Abstract

19547 Background: Patients with cancer who develop anemia are often treated with erythropoietin-stimulating agents (ESAs), such as darbepoetin alfa (DA) or epoetin alfa (EA). This study compared baseline characteristics, patterns of ESA use, and dosing among cancer patients as observed in a large managed care claims database. Methods: Medical and pharmacy claims from a U.S. managed care database were used to identify 15,007 unique episodes of care (7,238 with DA, 7,769 with EA) among cancer patients between January 2004 and January 2006. Episodes were defined as the time over which consecutive ESA claims occurred with =42 day gap between claims, plus a duration of clinical benefit (DCB) based on the median days between consecutive doses for each product. Six sensitivity analyses were performed that examined the robustness of findings to study methods and assumptions. Results: The mean (± SD) number of administrations per episode was 3.7 (± 4.1) for DA and 5.3 (± 6.4) for EA. The mean ESA episode duration was 55.5 (± 64.7) days for DA and 59.2 (± 59.2) days for EA and median times between consecutive claims was 15 and 8 days, respectively. The estimated mean weekly dose, including the DCB, was 105 (± 56) mcg for DA and 34,242 (± 28,174) U for EA. Average weekly doses were found to be sensitive to the inclusion/exclusion of the duration of clinical benefit (dose ratio 1:326 vs 1:237, respectively). Compared to EA episodes, DA episodes were associated with younger age (75% <65 years, for DA vs. 69% <65 for EA), more platinum containing chemotherapy (31.9% DA vs. 28.4% EA (p<.001), more classes of chemotherapy used (77% with one or more classes (DA) vs. 65% with one or more classes (EA) (p<.001)) and higher comorbidity burden, as measured by the Deyo-Charlson index (54.3% with a score =8 (DA) vs. 46.6% (EA) (p<.001). Conclusions: Patterns of ESA use differed significantly between DA and EA. Additionally, patients receiving DA differed significantly in baseline characteristics compared to patients receiving EA. Adjustments for patterns of use are necessary to appropriately compare the use of the two ESA’s. [Table: see text]