Study of Real-World Treatment Patterns in Patients with Sickle Cell Disease Using US Medicaid Claims Data

Introduction: Sickle cell disease (SCD) is a group of autosomal, codominant, inheritable diseases caused by a single mutation in the gene encoding the beta subunit of hemoglobin (HBB). SCD affects millions of people globally (CDC.gov SCD Data, 2024). Currently approved SCD drugs in the US include hydroxyurea (HU), voxelotor, crizanlizumab, and L-glutamine. Real-world evidence (RWE) of treatment patterns in patients prescribed these SCD drugs is limited. Our study aimed to address this knowledge gap among those enrolled in Medicaid.

Methods: We conducted a retrospective cohort study of treatment patterns using Medicaid claims data in the US. Patients were included if they had: ≥1 SCD inpatient or 2 outpatient diagnoses dated ≥30 days apart between Jan 1, 2016 and Dec 31, 2021; newly initiated any of the above specified SCD treatments on or after Dec 1, 2019; and completed a ≥6-month (mo) washout period prior to treatment initiation. We assessed the proportions of medication usage and discontinuations, average dose, treatment duration, medication adherence, and medication persistence.

Medication adherence was assessed as proportion of days covered (PDC = [sum of days on treatment] - [days covered of the last fill] ÷ [sum of days in the follow-up period × 100]) or a modified medication possession ratio (mMPR = [days' supply for all refills excluding last refill] ÷ [sum of days in the follow-up period × 100]) ≥80%. All analyses were descriptive. This study was institutional review board exempt.

Results: Of 75,343 patients with SCD in the Medicaid database, 7461 met the inclusion criteria. The median (IQR) age of patients was 20.6 (12.7-31.4) y and 42% were aged <18 y. Most patients were female (54%) and 65% were Black. In the 6 mo prior to initiating treatment, 52% visited a hematology specialist and 48% had hospital visits (average hospital stay, 5.5 days). Opioids (63%) and NSAIDs (42%) were most frequently prescribed. Among patients with a Charlson Comorbidity Index, 89% had a score of ≤1. Patients initiated treatment on HU (n=5144), voxelotor (n=1080), crizanlizumab (n=522), L-glutamine (n=283), and combination therapy with and without HU (n=404 and n=28, respectively).

Over 25 mo (from Dec 1, 2019-Dec 31, 2021), patients who initiated voxelotor or crizanlizumab had a longer average treatment duration (172.6 and 171.7 days, respectively) than those initiating L-glutamine or HU (128.9 and 115.9 days). In the same period, discontinuation rates by drug initiated were: crizanlizumab (66%), voxelotor (67%), L-glutamine (77%) and HU (78%). Adherence to crizanlizumab was 85% using mean PDC and 94% using mean mMPR. Adherence to HU, voxelotor, and L-glutamine was 65, 69, and 66%, using mean PDC, respectively and 83, 87, and 85% using mean mMPR. Persistence at 3 mo was highest for patients who initiated voxelotor (55%) and crizanlizumab (53%) and lower in those who initiated L-glutamine (41%) and HU (34%). The average daily dose of voxelotor, crizanlizumab, L-glutamine, and HU was 1386.2, 328.3 (on intravenous dosing days), 19,853.0, and 967.8 mg, respectively. The proportion of patients receiving SCD treatment(s) changed over time (Dec 2019-May 2020 to Jun-Dec 2021) for HU (84.5 to 69.7%, respectively), voxelotor (6.7 to 8.8%), crizanlizumab (4.3 to 8.1%), L-glutamine (3.8 to 4.5%), and combination therapy (<1.0 to 6.9%). Most initiators of HU, voxelotor, crizanlizumab and L-glutamine discontinued after first-line (1L) therapy. Most crizanlizumab initiators in 2L therapy were prior HU initiators, most HU and most L-glutamine initiators in 2L therapy were prior voxelotor initiators, and most voxelotor initiators in 2L therapy were prior L-glutamine or prior HU initiators.

Conclusions: Our RWE study of treatment patterns in patients initiating SCD disease-modifying treatments helps to address an identified evidence gap. We showed that patients initiating treatment with voxelotor and crizanlizumab had longer treatment durations and lower discontinuation rates than those who initiated HU and L-glutamine. Treatment adherence and persistence was suboptimal for all assessed SCD treatments, and approximately 75% of patients discontinued SCD treatment after initiating it on or after Dec 2019. Future studies are needed to evaluate the relationship of SCD treatments with adherence and health outcomes to support the development of strategies to improve adherence and persistence.

Sun:Pfizer: Current Employment. Gu:Pfizer: Current Employment, Current equity holder in publicly-traded company. Goswami:Pfizer: Current Employment, Current equity holder in publicly-traded company. Ye:Pfizer: Current Employment. Colavecchia:Pfizer: Current Employment, Current equity holder in publicly-traded company. Purdie:Pfizer: Current Employment, Current equity holder in publicly-traded company. Zhou:Genesis Research Group: Current Employment. Silver:Genesis Research Group: Current Employment. Huang:Pfizer: Ended employment in the past 24 months. Schachterle:Pfizer: Current Employment, Current equity holder in publicly-traded company.