Effect of erythropoiesis-stimulating agent policy decisions on off-label use in myelodysplastic syndromes

Erythropoiesis-stimulating agents (ESAs) are widely used to treat anemia associated with myelodysplastic syndromes (MDS) as an off-label indication. In early 2007, the U.S. Food and Drug Administration (FDA) released safety alerts and mandated label changes, and the Centers for Medicare & Medicaid Services (CMS) implemented a National Coverage Determination (NCD) in August 2007, dramatically restricting ESA coverage based on specific clinical parameters in non-MDS patients. We sought to determine the effect on ESA use in MDS, examining both treatment initiation and concordance with guidelines designed to target patients most likely to benefit from therapy.

We determined receipt of ESA within 6 months of diagnosis. For ESA recipients, we operationalized three National Comprehensive Cancer Network guidelines: serum erythropoietin determination before ESA initiation, transfusion-independent at ESA initiation, and initial ESA treatment episode of >= 8 weeks. Logistic regression models tested the effect of time (half-year increments pre-post the August '07 CMS NCD implementation), controlling for demographics and health status.

17,491 (61.1%) of 28,627 beneficiaries with MDS received ESAs. ESA use increased prior to the reference period (Jan.-July 2007), but declined beginning in August 2007, the date of NCD implementation (marginal probability =-0.05, p-value<0.01). Concordance with treatment guidelines changed during the observation period, with increased rates of serum erythropoietin levels, but declined in the other two guidelines.

These results suggest a mixed pattern of change in the face of the FDA safety warnings and CMS NCD in MDS and highlight the importance of monitoring for unintended consequences of policy changes.