Motixafortide (CXCR4 inhibition) alone and in combination with natalizumab (VLA-4 inhibition) to mobilize hematopoietic stem cells for gene therapy in sickle cell disease: A first-in-human, safety and feasibility study

Abstract

Abstract Autologous hematopoietic stem cell (HSC) based gene therapy for sickle cell disease (SCD) requires mobilizing sufficient peripheral blood (PB) HSCs to generate a gene-modified product. G-CSF is unsafe in SCD and the CXCR4 inhibitor (CXCR4i) plerixafor (P) does not reliably yield optimal HSC numbers. Rapid, reliable G-CSF-free HSC mobilization in SCD is an urgent, unmet need. We previously presented preliminary data on HSC mobilization in SCD with the long-acting CXCR4i motixafortide (M) +/- the monoclonal antibody (mAb) VLA4 inhibitor (VLA4i) natalizumab (N). Here we report the completed results of a first-in-human trial (NCT05618301) along with comprehensive immunophenotypic and transcriptional profiling of CD34+ HSCs in SCD subjects mobilized with P, M and N+M. In parallel, we evaluated HSC mobilization in Townes SCD mice with CXCR4i (P or M) +/- small molecule (sm) or mAb VLA4i. Subjects ≥18 yrs with SCD (SS or Sβ0) were enrolled. RBC exchange occurred ≤72h prior to mobilization. Subjects received M (1.25mg/kg, subcutaneous) followed by 1 blood volume (BV) leukocytapheresis (LP) ~14h post-M. After 8wks, a 2nd HSC mobilization with N (300mg, IV) + M (~32h post-N) followed by 1 BV LP ~14h post-M was completed. Primary endpoint was safety. Secondary endpoints were PB CD34+ cell kinetics and CD34+ cells/kg collected via LP. Comprehensive CD34+ HSC profiling was performed by flow cytometry (FC) and single-cell RNA sequencing (scRNA seq). Townes HbAA (non-SCD) and HbSS (SCD) mice were mobilized with P, M, smVLA4i, P+smVLA4i, M+smVLA4i or M+mAb VLA4i and PB HSCs were enumerated by FC (Lin-Sca1+cKit+ cells/μL) and colony forming units (CFU/mL). Ten subjects were enrolled (median age 29.5 yrs, 50% male, 90% SS). M and N+M were safe and well-tolerated. Common adverse events (AEs) were transient, Grade 1-2 injection site and systemic reactions (pruritis, 90%; tingling/pain, 80%; urticaria, 40%). No Grade 4 AEs, DLTs or complicated vaso-occlusive crises were observed. M mobilized median of 189 CD34+ cells/μl (range 77-690) to PB at 10-14h post-M with median 4.22x106 CD34+ cells/kg collected in a 1 BV LP (predicted 16.9x106 cells/kg in a 4 BV LP). N+M mobilized median of 312 CD34+ cells/μl (range 117-447) at 14h post-M with median 4.89x106 CD34+ cells/kg collected in a 1 BV LP (predicted 19.6x106 cells/kg in a 4 BV LP). Relative to M, N+M mobilized median 1.44-fold higher PB CD34+ cells/μl. In 2 subjects with prior P mobilization, M and N+M led to 2.7-2.8 fold higher PB CD34+ cells/μl and 2.8-3.2 fold higher CD34+ cells/kg. SCD subjects also mobilized 5-10 fold greater HSCs with M, relative to non-SCD subjects from prior studies. Moreover, while all SCD subjects mobilized well, two phenotypic SCD subgroups were identified with distinct mobilization kinetics, “super” (n=4) and “standard” (n=6) mobilizers. M mobilized significantly higher CD34+ HSCs in super vs standard mobilizers (median 481 vs 132 CD34+ cells/μl) (p<0.0001), while with N+M the difference in super vs standard was not significant (p=0.1156). Immunophenotyping of CD34+ cells mobilized with M and N+M revealed pan-mobilization of all HSC subsets, with N+M mobilizing a greater relative % of common lymphoid progenitors (CLPs) and megakaryocytic/erythroid progenitors (MEPs) relative to M (all p≤0.05). Transcriptional profiling by scRNA seq revealed upregulation of >200 genes with N+M vs M, including CXCR4, as well as CLP, erythroid progenitor (ERP) and MEP-associated genes (all p≤0.05). Finally, HSC mobilization in Townes mice mirrors our findings in humans, with M outperforming P, M+smVLA4i outperforming all other regimens, and SCD mice mobilizing 10-fold greater HSCs vs non-SCD mice. Remarkably, HSC transplant of non-SCD marrow into SCD mice (myeloablative cKit-ADC conditioning) reverted the enhanced mobilization phenotype in SCD mice to that of non-SCD mice, indicating enhanced CXCR4i-based mobilization in SCD may track with the hematopoietic system. In conclusion, our first-in-human trial demonstrates the potential of M and N+M as novel G-CSF-free regimens to safely optimize HSC mobilization in SCD (median CD34+ cells/μl: P=73, M=189, N+M=312). Correlative FC and scRNA seq highlight regimen-specific mobilization of unique HSC subsets, including increased CLPs, ERPs and MEPs with N+M. Further mechanistic study of HSC mobilization biology may build upon our finding that Townes SCD mice and SCD humans share an enhanced mobilization phenotype with CXCR4i +/- VLA4i.