Voxelotor Safety and Effectiveness in Patients With Sickle Cell Disease in the RETRO and PROSPECT US Registries
Nirmish Shah, Biree Andemariam, William Ershler, Edward Donnell Ivy, Deepika S Darbari, Alan Anderson, Parul Rai, Ted Wun, Ching-Ray Yu, Michelle XuAbstract
Background
Two US registry studies (RETRO and PROSPECT) evaluated real-world clinical and safety outcomes for patients with sickle cell disease (SCD) receiving voxelotor as part of their usual clinical care. The PROSPECT study was discontinued early in September 2024, after Pfizer withdrew voxelotor due to emerging preliminary data from 4 studies (including RETRO and PROSPECT) showing a potential imbalance in vaso-occlusive crises (VOCs) and an imbalance in fatal events in HOPE Kids 2 and RESOLVE studies (mostly in Sub-Saharan Africa). Preliminary data from an exploratory analysis of RETRO and PROSPECT suggested a potential increase in acute pain crisis (a surrogate for VOC) after initiation of voxelotor. The final assessment of RETRO and PROSPECT data following final database lock is presented.
Methods
After initial US approval of voxelotor in 2019, the RETRO registry (NCT04930328) collected retrospective data for participants with SCD treated with voxelotor for at least 2 wk and aged ≥12 yr at 9 US sites from Q4 2019 to Q1 2022. The PROSPECT registry (NCT04930445) collected retrospective and prospective data for participants with SCD treated with voxelotor aged ≥4 yr at 24 US sites starting in Q1 2022. Both registries collected data from healthcare records 1 yr pre-voxelotor initiation. RETRO data were collected retrospectively for ∼1 yr post-voxelotor treatment and PROSPECT data were planned to be collected prospectively for up to 5 yr post-voxelotor initiation. Safety analysis set included participants who met inclusion criteria and took at least one dose of voxelotor.
Results
RETRO enrolled 216 participants of mean (SD) age 33.1 (14.2) yr. PROSPECT enrolled 265 participants (including 19 from RETRO), 260 of whom received a dose of voxelotor before the study was discontinued. The mean (SD) age of participants in PROSPECT was 32.0 (14.8) yr: 33/265 (12.5%) were 12–< 18 yr, and 23/265 (8.7%) were 4–< 12 yr. The mean (SD) change in hemoglobin (Hb) from baseline ranged between 0.6 (1.6) and 0.8 (1.5) g/dL in RETRO and between 0.4 (1.5) and 0.7 (1.5) g/dL (through 48 months) in PROSPECT. The mean (SD) maximum change in Hb from baseline in RETRO was 1.3 (1.6) g/dL (range, 7.8 [1.5] to 9.2 [2.0] g/dL; n = 179) and in PROSPECT it was 2.0 (1.6) g/dL (range, 7.8 [1.4] to 9.8 [1.8] g/dL; n = 251). Voxelotor treatment also reduced markers of hemolysis in both studies. In both registries, acute pain crisis was the most common SCD-related adverse event, with annualized incidence rates (number of events/participant/yr) pre-voxelotor and post-voxelotor initiation of 1.33 and 1.54 in RETRO and 4.78 and 3.15 in PROSPECT, respectively. Overall, there were 5 fatal events in RETRO (ages 42–69 yr at first dose of voxelotor) and 6 in PROSPECT (ages 20–56 yr at first dose) none of which were considered related to voxelotor treatment per investigators (Table).
Conclusions
The results of these 2 real-world studies were similar, despite limitations including differences in study designs and data collection methods for RETRO and PROSPECT, the limitations inherent in registry studies, as well as disruptions imposed on healthcare utilization by COVID-19 which may have impacted the reporting of acute pain crisis in RETRO and PROSPECT. Consistent with previous clinical and real-world studies, treatment with voxelotor in practice increased Hb and decreased markers of hemolysis. Furthermore, there was no evidence of voxelotor treatment leading to an increased frequency of acute pain crisis, new safety signals, or treatment-related deaths.