AUTHORS
Ahmed Salem, BPharm, Caroline Horne, DO2, Myra Robinson3, Hang Zeng, MS4, Payal Desai, MD5
AFFILIATIONS
1 Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States
2 Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States
3 Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States, Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States
4 Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States, Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States
5 Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States, Atrium Health Levine Cancer Institute, Charlotte, North Carolina, United States
ABSTRACT
Introduction:
Voxelotor was approved for the treatment of sickle cell disease (SCD) to improve hemoglobin levels and reduce hemolysis. Following its voluntary market withdrawal in 2024, limited data have been published regarding clinical outcomes and healthcare utilization after treatment discontinuation. Characterizing healthcare utilization following voxelotor cessation may provide real-world insight into disease burden during treatment transitions. We evaluated longitudinal patterns of healthcare utilization before, during, and after voxelotor therapy in a large real-world SCD cohort.
Methods:
This IRB-approved retrospective observational study included patients with SCD prescribed voxelotor at Levine Cancer Institute and Levine Children's Hospital between January 2020, and August 2024. Healthcare utilization was assessed across four periods: pre-treatment [1 year (yr) prior], during treatment, short-term post-discontinuation (ST; ≤30 days), and long-term post-discontinuation (LT; ≤1 yr or until death or data cutoff on February 24, 2026).
Outcomes included annualized emergency department (ED) encounters, inpatient (IP) admissions, IP days, acute infusion center visits requiring intravenous opioid administration, and red blood cell transfusions (simple and exchange). Poisson regression models for repeated measures with an offset for period duration and Pearson scale adjustment for overdispersion were used to estimate least-squares mean (LSMean) utilization. Bonferroni-adjusted pairwise comparisons were performed. High-utilizer rates (>10 annualized encounters) were evaluated descriptively.
Results:
A total of 142 patients were included. Median age at voxelotor initiation was 26 years (range 4–73); 64% were female, 93% were Black/African American, and 90% had HbSS genotype.
Annualized ED encounters, IP encounters, and IP days increased numerically during and after voxelotor therapy but did not differ significantly across study periods with Bonferroni adjustment for multiple comparisons. LSMean annualized ED encounter rates were 2.2, 2.3, 3.5, and 2.5 visits/yr during the pre-treatment, treatment, ST follow-up, and LT follow-up periods, respectively (P>0.05). Corresponding LSMean annualized IP encounter rates were 1.4, 1., 2.0, and 1.9 encounters/yr, while annualized IP days were 7.5, 8.8, 9.2, and 9.5 days/yr.
In contrast, acute infusion visits increased significantly following discontinuation. Compared with pre-treatment utilization (1.0/yr), LSMean infusion visits increased to 3.2/yr during ST follow-up and 3.0/yr during LT follow-up (adjusted P<0.001 and P=0.002, ). Compared with during-treatment utilization (1.4/yr), acute infusion visits also remained significantly higher during ST and LT follow-up (P=0.004 and P=0.01, respectively).
Simple transfusions increased from 0.6/yr pre-treatment to 2.4/yr during ST follow-up and 1.7/yr during LT follow-up (both adjusted P<0.001). Significant increases were also observed relative to during-treatment (1.0/yr) utilization (P<0.01 and P=0.02, respectively). Exchange transfusions were significantly higher after discontinuation (ST 0.7/yr, LT 0.8/yr) than during treatment (during 0.2/yr; ST P=0.02; LT P<0.001), although differences from pre-treatment were not significant.
Descriptively, the highest proportions of high utilizers were observed during the ST post-discontinuation period compared to during treatment across ED encounters (18% vs 5%), IP encounters (14% vs 2%), acute infusions (20% vs 6%), exchange transfusions (4% vs 0%), and simple transfusions (15% vs 2%).
Conclusions:
In this real-world cohort of patients with SCD, healthcare utilization was not significantly reduced during voxelotor therapy, and ED visits, hospitalizations, and inpatient days did not increase significantly following discontinuation. In contrast, acute infusion visits and transfusion utilization increased significantly after voxelotor discontinuation, particularly during the early post-discontinuation period. These findings suggest increased reliance on supportive care following treatment withdrawal and raise the hypothesis that voxelotor discontinuation may be associated with increased acute healthcare utilization in some patients. Close clinical monitoring and alternative disease-management strategies after discontinuation may be important considerations for future therapies with similar mechanisms of action, or for voxelotor should it ever be reconsidered for clinical use.
DISCLOSURES
Ahmed Salem, BPharm: No relevant conflicts of interest to declare.