G-CSF receptor is a signaling node linking thrombo-inflammation to vaso-occlusion in sickle cell disease: Target for a new phase 2 trial

Abstract Introduction Granulocyte-colony stimulating factor (G-CSF) administration is known to induce severe vaso-occlusive crisis (VOC) in patients with sickle cell disease (SCD). G-CSF and neutrophils rise acutely during VOC, whereas an elevated circulating neutrophil count at steady state is a prognostic factor of SCD severity. G-CSF is induced by many inflammatory triggers, including lipopolysaccharides (LPS), tumor necrosis factor alpha (TNFα) and toll-like receptor 4 activation, which provoke VOC in SCD mice. Importantly, G-CSF induces neutrophil extracellular traps (NETs) formation and activates platelets and monocytes, leading to endothelial cell activation and thrombo-inflammation in SCD. We report the first preclinical evidence validating G-CSF receptor (G-CSFR) blockade as a therapeutic target to prevent VOC in SCD. Methods Townes knock-in humanized SCD mice were exposed to hypoxia-reoxygenation (HR) or intravenously (IV) challenged with oxy-hemoglobin (oxy-Hb), LPS, or G-CSF to induce VOC. Some mice were pretreated IV with VR81, an anti-mouse G-CSFR monoclonal antibody (mAb), prior to VOC induction. Blood flow and NETs formation were assessed by intravital microscopy of the skin and lungs. Liver and lung tissues were examined for inflammation by Western blots and immunofluorescence staining, respectively. Gene expression analyses were performed on transcriptomic datasets from the blood of SCD patients with/without VOC and healthy humans with/without treatment with G-CSF and/or CSL324, an anti-human G-CSFR mAb under development. Results A single IV dose of VR81 in mice had a half-life of ~7 days. Full receptor occupancy in neutrophils was immediate, maintained for at least 1 week, and confirmed by the complete inhibition of phosphorylated STAT3, the signal transduction mediator of G-CSFR signaling. G-CSF administered IV to SCD mice dose-dependently induced microvascular vaso-occlusion in the skin. Vaso-occlusion in the skin was prevented by VR81 administration IV 1 hour or 7 days prior to VOC induction with G-CSF, HR, or oxy-Hb. Importantly, VR81 resolved skin vaso-occlusion in SCD mice when administered 40 minutes after VOC induction. Biomarker analysis revealed several effects caused by G-CSFR blockade. In the liver, expression of pro-inflammatory endothelial adhesion molecules VCAM-1, ICAM-1, and E-selectin and nuclear factor-ĸB-phospho-p65 were markedly decreased, while expression of anti-inflammatory heme oxygenase-1 and nuclear localization of nuclear factor erythroid 2-related factor 2 were increased. Intravital microscopy in the lungs revealed that VR81 pretreatment also prevented neutrophil-platelet aggregate-dependent lung vaso-occlusion in oxy-Hb challenged SCD mice, with a reduced number of NETs within the lung and blood circulation. In the lungs, the number of neutrophils, the expression of pro-adhesive Weibel-Palade body P-selectin, and von Willebrand factor were decreased. Long-term administration of VR81 once per week for 5 months reduced signs of chronic organ injury. The weight of the right ventricle was normalized, suggesting a possible reduction in pulmonary hypertension. The white blood cell count was also reduced. Transcriptomic analyses support the translatability of G-CSFR blockade to SCD patients. We previously published a gene signature of differentially expressed genes in blood after G-CSF administration to healthy humans. In SCD patients, this G-CSF gene signature was significantly enriched during VOC vs. baseline (p-value < 10-3) when analyzed using microarray and RNA-Seq datasets. Conclusions Literature and our data strongly support the critical role of G-CSF in thrombo-inflammatory pathobiology in VOC. Because G-CSF activates multiple inflammatory pathways that trigger vaso-occlusion, G-CSFR is a suitable target for drug intervention to prevent VOC. Our data suggest multiple anti-inflammatory effects in SCD, including prevention of NET formation, which provide a plausible mechanism of action for G-CSFR blockade. The known biology of NETs aligns with a role for G-CSFR in SCD. The NETs formation and their rise during VOC have been confirmed from the literature for patients with SCD. CSL324 was safe and well tolerated in repeat-dose toxicological studies in non-human primates and has completed a phase 1 trial, with satisfactory safety (some risk of transient neutropenia) and pharmacokinetic profiles. A phase 2 trial in patients with SCD will seek clinical proof of concept for VOC prevention.