Rheological Impact of GBT1118 Cessation in a Sickle Mouse Model

Abstract

In sickle cell disease (SCD), higher whole blood viscosity is a risk factor for vaso-occlusive crisis, avascular necrosis, and proliferative retinopathy. Blood viscosity is strongly impacted by hemoglobin (Hb) levels and red blood cell (RBC) deformability. Voxelotor is a hemoglobin S (HbS) polymerization inhibitor with anti-sickling properties that increases the Hb affinity for oxygen, thereby reducing HbS polymerization. In clinical trials, voxelotor increased Hb by an average of 1g/dl, creating concern that this rise in Hb could increase viscosity, particularly when the drug was cleared. To investigate this potential rebound hyperviscosity effect, we treated SCD mice with GBT1118, a voxelotor analog, and stopped the treatment to determine the effect on blood viscosity and RBC deformability under a range of oxygen concentrations. GBT1118 treatment increased Hb, improved RBC deformability by increasing the elongation index under normoxic (EI_max) and hypoxic conditions (EI_min), and decreased the point of sickling (PoS) without increasing blood viscosity. The anti-sickling effects and improvement of RBC deformability balanced the effect of increased Hb such that there was no increase in blood viscosity. Forty-eight hours after ceasing GBT1118, Hb declined from the rise induced by treatment, viscosity did not increase, and EI_min remained elevated compared to control animals. Hb and PoS were not different from control animals, suggesting a return to native oxygen affinity and clearance of the drug. RBC deformability did not return to baseline, suggesting some residual rheological improvement. These data suggest that concerns regarding viscosity rise above pre-treatment levels upon sudden cessation of voxelotor are not warranted.

Keywords: GBT1118; deformability; hemorheology; red cell; rheological biomarkers; sickle cell disease; viscosity; voxelotor.

Figure 1

GBT1118 produces Hb rise and drug occupancy comparable to GBT440 in clinical trials. (A) Expected rise and fall of total hemoglobin with GBT1118 treatment and its withdrawal. Ctl n=14, on treatment n=20, 24h post n=10, 48h post n=12, and 72h post n=12. (B) 30% drug occupancy achieved with GBT1118. Ctl n=15, on treatment n=20, 24h post n=10, 48h post n=12, and 72h post n=12. ^*p<0.05; ^**p=0.051; ^***p<0.0001; and ^****p<0.0001.

Figure 2

Viscosity and HVR measurements on and off GBT1118. (A) At a shear of 45s⁻¹, modeling in viscosity from baseline. Ctl n=6, on treatment n=3, 24h post n=4, and 48h post n=7. (B) At a shear of 225s⁻¹, modeling arterial circulation, viscosity on drug declines from baseline on treatment, and 24h off drug. There is no significant rise in viscosity 48h off drug compared to baseline. Ctl n=4, on treatment n=3, 24h post n=4, and 48h post n=5. (C) At an HVR at 45s⁻¹, oxygen carrying increased 24h off drug. Ctl n=4, on treatment n=4, 24h post n=3, and 48h post n=5. (D) At an HVR at 225s⁻¹, oxygen carrying capacity increased on drug and 24h off drug. There was no significant decline in HVR 48h off drug compared to baseline at 225s⁻¹ or 45s⁻¹ HVR. Ctl n=4, on treatment n=3, 24h post n=3, and 48h post n=6. ^***p<0.0001; ^**p<0.0051; and ^*p<0.05. HVR, hematocrit-to-viscosity ratio.

Figure 3

Deformability and point of sickling measurement using oxygen gradient ektacytometry on and off GBT1118. (A) RBC deformability when deoxygenated rises on drug and declines but remains above baseline up to 48h off drug. Ctl n=5, on treatment n=9, 24h post n=5, 48h post n=3, and 72h post n=7. (B) RBC deformability when oxygenated rises on drug. Ctl n=7, on treatment n=9, 24h post n=5, and 72h post n=10. (C) Point of sickling drops on drug and returns to baseline 48h drug. Ctl n=7, on treatment n=9, 24h post n=5, 48h post n=4, and 72h post n=10. ^***p<0.0001; ^**p<0.05. EI_max, maximum elongation index; EI_min, minimum elongation index; PoS, point of sickling; and RBC, red blood cell.