Voxelotor: Changing the Disease by Changing the Conformation

Sickle cell disease (SCD) is a global problem, estimated to affect more than 100,000 individuals in the United States and millions of people worldwide.¹  SCD results from a single nucleotide substitution in the β-globin chain of adult hemoglobin (HbA), which produces instead sickle hemoglobin (HbS). When tense (deoxygenated Hb, or “T-state” Hb), HbS polymerizes, leading to damage and distortion of the red blood cell (RBC), hemolysis, inflammation, thrombosis, acute vaso-occlusive crisis (VOC), and chronic cumulative vasculopathy.^2,3  Despite the global burden, high-grade morbidity, and early mortality, disease-modifying medications for SCD have been few, with hydroxyurea being the sole U.S. Food and Drug Administration-approved medication for SCD for 20 years, until the recent approval of L-glutamine.³ 

The placebo-controlled randomized phase III HOPE trial, conducted by Dr. Elliott Vichinsky and colleagues, and sponsored by Global Blood Therapeutics, is a refreshing example of the current renewed interest in novel therapeutics in SCD. Voxelotor (GBT440) is an orally bioavailable small molecule that causes a delay in HbS polymerization (central to pathophysiology in SCD) by reversibly binding to α-globin in 30 to 40 percent of Hb molecules/RBC, causing an allosteric conformational change and an increase in oxygen affinity (more oxygenated or “S-state” Hb). This is likely disease modifying, analogous to increased Hb F in patients on hydroxyurea or with an hereditary persistence of fetal Hb.⁴  Drs. William Eaton and H. Franklin Bunn convincingly argued that even small reductions in polymerization of HbS are key in ameliorating the clinical outcomes of SCD.⁵  In murine models of SCD, voxelotor had been shown to increase the half-life of erythrocytes, reduce sickling, and improve anemia.⁶  In a phase I/II clinical trial, voxelotor was overall well tolerated at a dose of 1,000 mg daily, with reduction in anemia and hemolysis.⁷ 

The HOPE trial was an international effort, with a primary endpoint defined as the percentage of patients with an increase in total Hb of 1 gm/dL at 24 weeks, although treatment extended up to 72 weeks. Secondary endpoints included mean change in Hb, annualized incidence rate of VOC, and laboratory markers of hemolysis. Individuals eligible for the study were between the ages of 12 to 65 years. All major SCD subtypes were eligible, but more than 85 percent of patients studied had HbS or HbSβ⁰. Crucially, individuals on hydroxyurea were not excluded. The authors enrolled 274 participants with well-distributed baseline characteristics. Participants were randomly assigned in a 1:1:1 fashion to treatment groups receiving either voxelotor 1,500 mg, voxelotor 900 mg, or placebo.

Key findings in this trial were significant improvement in anemia and a possible improvement in hemolysis, without undue near-term toxicity. In the intention-to-treat analysis, there was a significantly higher percentage of patients with a mean Hb response in the voxelotor 1,500 mg group (51%; 95% CI, 41-61) than in the placebo group (7%; 95% CI, 1-12; p <0.001). The 900 mg dose had an intermediate effect. Secondary endpoints showed a possible decrease in hemolysis (% indirect bilirubin and % reticulocytes decreased significantly with treatment, but no significant reduction in lactate dehydrogenase and absolute reticulocyte count). There was no difference between groups regarding percentage receiving RBC transfusion, and there was a nonsignificant reduction in annualized VOC in voxelotor-treated patients versus the placebo group. Safety data were encouraging as most adverse events reported were grade 1 or 2, with diarrhea and headache being the most common. Adverse events of grade 3 or higher did not differ between groups.

Despite the significant optimism engendered by the HOPE trial, caveats remain that should be fully explored before voxelotor is universally adopted. Safety risks not fully evaluated in the HOPE trial include delayed oxygen delivery to the cerebral cortex in patients at risk for silent cerebral infarction. As Drs. Robert Hebbel and Bo Hedlund point out, physiology is not linear, and an increase in Hb may not translate to an increase in tissue oxygen delivery, especially with agents that affect oxygen affinity.⁸  Erythropoietin levels did not differ between treated and untreated patients in the HOPE trial, suggesting that the renal cortex did not detect a physiological decrease in oxygen delivery in vivo. However, the impact of voxelotor on the cerebral cortex, which is highly dependent on oxygen delivery, cannot be readily assessed via biochemical surrogates and needs to be followed closely in the coming years, optimally with more sensitive neuroimaging.⁹  Additionally, patients with SCD who were not well-represented in the HOPE trial include those with HbSC and HbSβ⁺, as well as HbS patients with total Hb less than 7 g/dL. Caution should be used before extending voxelotor to these subpopulations without further study. In variant SCD especially, more data are needed about blood viscosity,¹⁰  which has been associated with increased risk for VOC and osteonecrosis and is expected to rise in conjunction with hematocrit (Hct).²  Previous viscosity measurements for voxelotor, normalized to a Hct of 30 percent in vitro, showed improvement in viscosity,¹⁰  but this may not reflect in vivo changes at true Hct values. Finally, the effect of voxelotor on additional clinically meaningful endpoints such as quality of life and frequency of VOC will need to be established.