Highly efficient collection and manufacture of autologous HSC gene therapy cell product for patients with sickle cell disease using a lentiviral vector containing a shmir targeting BCL11a

Abstract

Abstract Achieving an adequate dose of genetically modified hematopoietic stem cells (HSCs) for gene therapy in patients with sickle cell disease (SCD) remains a challenge due to limitations related to stem cell mobilization using plerixafor alone, reduced apheresis collection efficiency, and losses during ex vivo cell manipulation. To date, 39 SCD patients were enrolled in NHLBI-funded Phase I (NCT03282656; PMC7962145) and NHLBI/CIRM-funded multi-site Phase II (NCT05353647) gene therapy trials to receive autologous CD34+ HSCs transduced with a lentiviral vector encoding a short hairpin RNA embedded in a microRNA (shmiR) targeting BCL11A. To collect HSCs as the starting material for drug product manufacture (minimum 4×10⁶ cells/kg), apheresis (minimum 4 blood volumes or up to 8 hours) was performed within 3 hours of daily plerixafor for 2 consecutive days. A third day of collection was used to generate a back-up product as needed. Mean mobilized, pre-apheresis peripheral blood (PB) CD34+ count was 40 cells/μL (range 7–126). Among 35 evaluable patients (4 excluded: 1 pending release, 3 withdrew), 74% (26/35) reached target CD34+ collection in a single mobilization cycle (mean 2.6 procedures, range 2–6). Manufacturing was achieved with a mean of 11.92×10⁶ CD34+ cells/kg collected resulting in a drug product of 7.04×10⁶ CD34+ cells/kg and a net cell recovery from apheresis to drug product of 62% (range 38–91%). Drug products were successfully generated for all patients with a mean product vector copy number of 4.04 copies/cell (range 1-7). The time interval from first collection cycle to completion of product testing was a median of 39 days (mean 56) for all patients, and a median of 37 days (mean 38) for those collected in 1 cycle. Preparative transfusions were given before collections to bridge patients after stopping hydroxyurea or to mitigate stress erythropoiesis. We examined the impact of preparative transfusion on mobilization and apheresis efficiency in 16 evaluable patients at a single site. All underwent ≥ 3 months of preparative transfusion with a HbS target of ≤30% before mobilization using simple transfusion or automated red cell exchange. The mean HbS% at mobilization was 10.8% (range 2.9–21.5%). Overall, the preparative transfusion regimen reduced PB reticulocyte percentage by 50% to a mean of 7.2% (range 3.7–12.3), indicating decreased stress erythropoiesis. Resting PB CD34+ counts before and after 3-months of transfusion were 10.8 cells/μL (range 2–21) and 6.2 cells/μL (range 1–13), respectively. A total of 39 mobilization/collection procedures were performed in 16 patients with a mean post-plerixafor CD34+ count of 43 cells/μL (range 13–85) with collections performed between the two darkest color preferences of the manufacturer. Cell collection preference was monitored using real-time intraprocedural CD34+ cell sampling. A mean of 4.3 total blood volumes was processed (range 2.5–6.4), with mean collection efficiency (CE) (CD34+ cells collected ÷ [pre-apheresis CD34+ count × total blood processed]) of 48% (range 8.2–107.1). The mean reticulocyte percent in patients with CE <30% (11.4%, range 2.5 –18.9) was significantly higher than in those with CE >30% (6.2%, range 1.8–11.9, p=0.0026). Two patients had a history of delayed hemolytic transfusion reactions and/or multiple RBC alloantibodies before study entry that limited preparative transfusions to reach a HbS of 30% for 3 months. They were successfully collected, and products were generated after simple transfusions with a single limited-volume red cell exchange immediately prior to collection. The success of this study was driven by optimizing apheresis strategies, including suppression of stress erythropoiesis through transfusion, real-time instrument adjustments, and efficient manufacturing that allowed a collection target of nearly half the CD34+ cells recommended in the FDA-approved gene product. Strikingly, the highly efficient ex vivo manipulation platform described here enables successful mobilization and manufacturing of gene therapy products for SCD in several months, easing the path to autologous gene therapy for SCD.