Treatment Patterns and Healthcare Costs in Newly Diagnosed Patients with Chronic Myeloid Leukemia Receiving Dasatinib or Nilotinib As First-Line Therapy in the United States: A Retrospective Claims Database Analysis

Background:To date, no studies have directly compared the treatment patterns and economic outcomes of dasatinib and nilotinib as 1st-line therapy for Philadelphia chromosome positive chronic myeloid leukemia in chronic phase (Ph+ CML-CP). This study aimed to address this gap using retrospective claims data.

Method: Newly diagnosed adult pts with CML who received dasatinib or nilotinib as 1st-line therapy were identified from 2 large US administrative claims databases (Q1/2010 - Q4/2014). Pts with continuous health plan coverage for ≥6 months before (baseline period) and ≥1 month after the 1st-line TKI initiation (index date) were classified as dasatinib or nilotinib pts based on the 1st-line TKI. Pts were excluded from the study if they were enrolled in a clinical trial, had a stem-cell transplant, received chemotherapy (except hydroxyurea), had a metastatic cancer/solid tumor or end stage renal disease during the baseline period. Dose adjustment (dose decrease/increase of ≥20mg for dasatinib and ≥100mg for nilotinib vs the initial dose) were measured from the index date to treatment discontinuation (i.e., treatment gap of ≥60 days) or end of follow-up. Treatment adherence was measured using proportion of days covered (PDC) during a 6- and 12-month period following index date among pts with continuous health plan coverage during these periods. Treatment discontinuation, switching to another TKI, and healthcare resource utilization (HRU) and costs (pharmacy+medical costs; per person per month: PPPM; USD 2014) were measured from index date to end of follow-up. Outcomes were compared between cohorts using multivariate regression analyses adjusting for potential confounding factors identified at baseline/index date (age, gender, CML complexity, CCI, starting dose, and health plan type).

Results: A total of 604 dasatinib and 418 nilotinib pts met the sample selection criteria with average follow-up of 17 months in both cohorts. Pts in dasatinib and nilotinib cohorts did not significantly differ in age (mean=51 vs 52, respectively; p=.114), sex (46% female in both cohorts; p=.834), and comorbidities (mean CCI=1.4 vs 1.5, respectively; p=.600). In dasatinib pts, 91% started with 100mg/day, 3% with <100mg/day, and 6% with 140mg/day. In nilotinib pts, 76% started with 600mg/day, 16% with 800mg/day, and 8% <600mg/day. Dasatinib pts had a higher risk of dose decrease compared to nilotinib pts (crude rate: dasatinib=19%, nilotinib=15%; adjusted hazard ratio [HR]=1.66; p=.002). There was no significant difference in risk of dose increase between cohorts (crude rate: dasatinib= 3%, nilotinib= 4%; adjusted HR=0.76; p=.423). Dasatinib and nilotinib pts had similar adherence levels during the first 6-month (mean PDC: dasatinib =87% [N=481], nilotinib=86% [N=333]; adjusted difference= -0.03%; p=.981) and 12-month periods (mean PDC: dasatinib=78% [N=343], nilotinib=78% [249]; adjusted difference = -0.22%; p=.880). There was no significant difference in risk of treatment discontinuation between cohorts (crude rate: 39% in both cohorts; adjusted HR=1.10; p=.372). Dasatinib pts had higher risk of switching compared to nilotinib pts (crude rate: dasatinib=13%, nilotinib=9%; adjusted HR=1.62; p=.019). Pts had similar HRU; average number of inpatient days, emergency room visits, and outpatient visits PPPM were 0.13 vs 0.16 (adjusted incidence rate ratio [IRR]=1.03; p=.930), 0.06 vs 0.04 (adjusted IRR=1.26; p=.197), and 1.91 vs 1.95 (adjusted IRR=1.01; p=.842) for dasatinib vs nilotinib pts, respectively. Dasatinib pts incurred higher healthcare costs by $749 PPPM compared to nilotinib pts (adjusted incremental costs; p=.044; average healthcare costs PPPM: dasatinib=$10,341, nilotinib=$9,611).

Conclusions: Based on retrospective claims data, this study shows that dasatinib was associated with an increased risk of dose decrease despite majority of pts starting on the recommended dose, an increase risk of switching to another TKI, and higher total healthcare costs compared to nilotinib pts. No difference in adherence, discontinuation, and HRU was found. The study showed the association between nilotinib versus dasatinib as first-line therapy in CML patients and outcomes, not causation. The proportion of CML patients in accelerated phase or blast crisis by cohort at the first-line TKI initiation and its impact on results is unknown but was assumed to be marginal.