Rheological effects of voxelotor in SCD: A double-edged sword?

Abstract

Abstract Introduction. Blood viscosity—mainly determined by hematocrit (Ht) and red blood cell (RBC) rheological properties—is considered a potential contributor to vaso-occlusive crisis (VOC) risk in sickle cell disease (SCD). The HEMOPROVE study (NCT05199766) is a Phase II, open-label, single-arm trial assessing the effects of Voxelotor (1500 mg daily for 48 weeks) in SCD patients. This analysis aimed to evaluate the 12-month impact of Voxelotor on blood rheology and its key determinants Methods. Patients with SS or S-beta0 thalassemia had to be more than one month from a vaso-occlusive crisis and three months from a transfusion. Patients under hydroxyurea (HU) or angiotensin-converting enzyme inhibitors (ACEi) had to have received a stable dose for at least three months. Patients were secondarily excluded from the M6 and/or M12 analyses if they had received more than 3 RBC units during the study period. Blood viscosity was measured at 22.5 and 225 s-1 using a cone/plate viscosimeter. Hematocrit-to-viscosity ratio (HVR), linked to better microvascular oxygenation, was calculated as hematocrit/blood viscosity at 225 s-1. RBC aggregation, aggregate strength and deformability (EImax) were measured using an ektacytometer/aggregometer (LORRCA), with an oxygen gradient (to obtain the point of sickling (PoS) and minimum deformability, EImin). Data are presented as median [IQR]. Correlations used Spearman tests, comparisons employed Wilcoxon tests, paired t-test, or ANOVA. Results. Twenty-nine patients were enrolled in the study. Six discontinued prematurely: four due to transfusions, one following voluntary withdrawal, and one due to poor venous access. Additionally, one patient died from sepsis with multiorgan failure, and four others discontinued after Voxelotor was withdrawn by Pfizer. A total of 19 patients (10 females and 9 males; mean age: 42 years [range: 36–48]) were evaluable at M0, M6, and M12. Fifty-three percent of patients were treated with hydroxyurea (HU). Hemoglobin (Hb) levels increased from 7.3 g/dL [6.6–7.8] at baseline (M0) to 9 g/dL [8.2–10.2] after 12 months of treatment (M12) (p < 0.0001), while hematocrit (Ht) rose from 21% [19–23] to 26% [23–30] (p < 0.0001). Voxelotor significantly increased blood viscosity at high shear rate (225 s⁻¹), from 4.6 cP [4.1–5.3] at M0 to 5.8 cP [5.2–6.3] at M12 (p = 0.0002), but not at low shear rate (22.5 s⁻¹). This viscosity increase was solely attributable to the rise in Ht, as evidenced by a mild improvement in the hematocrit-to-viscosity ratio (HVR). Rheological improvements were demonstrated by increased RBC deformability in both oxygenated (EImax: 0.45 at M0 vs. 0.52 at M12, p = 0.0003) and deoxygenated conditions (EImin: 0.08 at M0 vs. 0.10 at M12, p = 0.047). RBC aggregation index was restored to normal (53.6% at M0 vs. 61.3% at M12, p = 0.004), and the shear rate required for RBC disaggregation decreased (700 s⁻¹ at M0 vs. 600 s⁻¹ at M12, p = 0.04). These changes were associated with reduced RBC sickling, as indicated by a negative correlation between Hb and PoS changes under treatment (r = –0.53, p = 0.037). Four patients (including three on HU) experienced a vaso-occlusive crisis (VOC) during follow-up, three of which were complicated by acute chest syndrome. These patients had a trend toward higher baseline reticulocyte counts (291 vs. 198 ×10⁹/L, p = 0.07), significantly higher viscosity at low shear rate at M6 (15.3 vs. 12 cP at 22.5 s⁻¹, p = 0.04), and lacked HVR improvement at M6 compared to the rest of the cohort (p = 0.047). Conclusions. Voxelotor improves red blood cell deformability and aggregation, contributing to reduced sickling in SCD patients. However, despite these rheological benefits, the associated increase in hematocrit leads to elevated blood viscosity, particularly at low shear rates. This viscosity rise may contribute to VOC risk, underscoring the importance of viscosity monitoring in future therapeutic studies targeting HbS polymerization