Patterns of Hypomethylating Agent and Transfusion Use in Myelodysplastic Syndrome: a Claims Database Study.

Myelodysplastic syndrome (MDS) is characterized by ineffective hematopoiesis, multilineage dysplasia, peripheral cytopenias, and susceptibility to leukemic transformation. Supportive treatments include red blood cell (RBC) and platelet (PLT) transfusions as well as erythrocyte stimulating agents (ESAs) to correct disease-induced cytopenias. Active treatment with hypomethylating agents (HMAs) have been shown in clinical trials to reduce transfusion dependence and leukemic progression among patients with MDS. The purpose of this study is to describe the patterns of care among patients with MDS, and outcomes associated with HMA treatment options.

Data were obtained from the MarketScan® database, which contains pharmacy, medical, hospital, and laboratory claims for several million members, from January 2002 through June 2008. Inclusion criteria included: age 18 or older; at least 2 claims with an ICD-9 diagnosis of MDS in 2006 or later; at least 6 months of pre-diagnosis health plan enrollment; at least 4 months of enrollment after initial HMA treatment; at least 1 complete cycle of decitabine (DAC; 5 treatments) or azacitidine (AZA, 7 treatments). Patients were excluded if they were treated with lenalidomide, had a prior cancer diagnosis, prior treatment with DAC or AZA, or had other isolated cytopenia or myeloproliferative diagnoses. Patients with a diagnosis of acute myelogenous leukemia within the first 28 days of treatment were also excluded. Descriptive statistics characterized patient demographics, including time between diagnosis and treatment, days per cycle, use of ESA, and number of treatment cycles. Logistic regression assessed predictors of HMA treatment using age at MDS diagnosis, gender, Charlson Comorbidity Index (CCI), and calendar year as predictors. Poisson regression compared the risk of RBC or PLT transfusion between DAC and AZA, controlling for age, gender, CCI, treatment cycles, time to initiating treatment, and follow-up duration.

2525 patients met full inclusion criteria (51% female), of whom 95.4% received no HMA treatment. Logistic regression revealed that females were less likely to receive HMA therapy (OR 0.486, p<0.001). There was no significant difference in follow-up duration between the DAC and AZA groups. Over 50% of HMA-treated patients received 4 or more treatment cycles, with no significant difference between DAC (n=37) and AZA (n=60). Mean (SD) days from MDS diagnosis to first treatment was 93.7 (101.4) for DAC vs. 50.8 (73.4) for AZA (p=0.03). Median treatment days per cycle were 4.86 for DAC and 5.00 for AZA (p>0.05). Mean (SD) days to discontinuation of RBC/PLT transfusion was 15.8 (48.3) for DAC and 70.1 (136.1) for AZA (p<0.05). The RBC/PLT transfusion rate was lower for DAC than for AZA (0.06 vs 0.17 per month, p<0.05). Poisson regression found a significantly lower likelihood of RBC/PLT transfusion in the DAC group (OR=0.206, p<0.05). Use of ESA did not differ between treatments (p>0.05).

Only a small portion of MDS patients receive HMA treatments, with females less likely to receive drug therapy. Initiation of decitabine occurs later than azacitadine after MDS diagnosis for unclear reasons. Decitabine is associated with lower rates of RBC/PLT transfusion and shorter time to discontinuation of transfusions, consistent with a prompt time to clinical response. Further research is needed to clarify optimal initiation timing for HMA treatment to maximize therapeutic benefits.

Hatoum:Eisai Inc.: Research Funding. Lin:Eisai Inc.: Research Funding. Buchner:Eisai Inc.: Employment. Kim:Eisai Inc.: Employment.