Restoration of Normal Blood Flow in Mouse Models of Sickle Cell Vaso-Occlusion Following Intravenous or Subcutaneous Administration of a Highly Potent E-Selectin Specific Inhibitor

Background: In sickle cell disease (SCD), recurrent episodes of vaso-occlusive crisis (VOC) contribute to morbidity and accelerated mortality. Analysis of the RESET Phase 3 study demonstrated that administration of the pan-selectin inhibitor, rivipansel, early in the course of VOC was associated with efficacy outcomes in favor of rivipansel, suggesting that reversal of vaso-occlusion may be achieved by treatment early in VOC. While both P- and L-selectin likely contribute to the underlying inflammatory cascade leading to VOC, previous studies have shown that E-selectin expression on the vascular luminal surface appears to be crucial for adhesion of sickled (SS)-RBC and leukocytes to the endothelium. GMI-1687 is an innovative potent E-selectin specific antagonist that demonstrates high bioavailability following subcutaneous (SC) administration. In the current investigation, we report on in vitro and in vivo studies evaluating the activity of GMI-1687 in SCD mouse models of VOC.

Methods: Townes transgenic mice and nude mice transfused with human SS-RBC were used as in vivo models of SCD. VOC was initiated by TNF injection in both models. Adhesion to inflamed venules and development of VOC was recorded using intravital microscopy at sites of window chamber implants. The in vivo activity of E-selectin inhibition with GMI-1687 was assessed following IV injection in the nude mouse and Townes transgenic models, and SC injection in the nude mouse model of SCD.

Results: In an initial study, IV treatment with saline or GMI-1687 was initiated 30 min post infusion of SS-RBC (t₃₀) into TNF-treated mice, a time during which sickle cells are adherent to inflamed venules and VOC has initiated, and repeated 30 minutes later (t₆₀). As expected, mice treated with saline alone showed marked adhesion of human SS-RBC to venules (61% occlusion of the total vessels recorded at t₃₀) with evident blood stasis. Under these conditions, GMI-1687 at 40 μg/kg significantly reversed SS-RBC adhesion and VOC at both t₃₀ and t₆₀ (p=0.002 and 0.0073, respectively, compared to saline treatment). This led to restoration of blood flow in 85% of the venules in GMI-1687 treated mice vs. 18% of vessels with normal blood flow in the control saline treated group. Because of reduced sickle cell adhesion and VOC, the number of circulating SS-RBC increased by approximately 6-fold in mice treated with GMI-1687. Similar anti-adhesive and anti-VOC activities of GMI-1687 treatment were obtained in the Townes transgenic model of SCD where following IV administration of the E-selectin antagonist, adhesion of sickled RBCs was significantly diminished and as a result, the percentage of the total occluded vessel segments was reduced by 80-90%. We next assessed the activity of SC administration with 40 μg/kg GMI-1687 on human SS-RBC adhesion and development of VOC in the nude mouse model. Similar to the IV experiments, GMI-1687 initiated at t₃₀ and t₆₀ post SS-RBC infusion significantly reduced SS-RBC adhesion in inflamed vessels (69% reduction at t₃₀, and 63% at t₆₀, p=0.0001). The reduction in SS-RBC adhesion led to a lower incidence of VOC (2% compared to 30% of vessels in mice injected with saline), and a 5.1-fold improvement (p=0.0315) in the number of circulating human SS-RBC. It is noteworthy that the effective dose of GMI-1687 used in these IV and SC studies resulting in the restoration of normal blood flow was approximately 500-fold less than that reported for rivipansel.

Conclusions: These data demonstrate that the administration of the potent and E-selectin specific compound GMI-1687 was effective in restoring blood flow in two mouse models of SCD, and support the development of GMI-1687 formulated for SC use and self-administration with the potential for early intervention of VOC.