AUTHORS
Sarah A Cotteta, Nicholas Giacobbe2, Connor Holden3, Brooklyn L McDonald4, Tanja Cupac5, Wayne Green6, Adam Campbell7, Omar Banda, PhD8, Mohamad-Gabriel Alameh, PhD9, Stephan A Grupp, MD, PhD10, Yongping Wang, MD, PhD11, Eugene Khandros, MD, PhD12, Stephan Kadauke, MD, PhD13
AFFILIATIONS
1 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
2 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
3 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
4 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
5 Miltenyi Biotec, Gaithersburg, Maryland, United States
6 Miltenyi Biotec, Gaithersburg, Maryland, United States
7 Miltenyi Biotec, Gaithersburg, Maryland, United States
8 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
9 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, United States
10 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States
11 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, United States
12 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, United States
13 Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, United States, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, United States
ABSTRACT
INTRODUCTION: Autologous gene therapies that induce fetal hemoglobin (HbF) are transformative in sickle cell disease (SCD) but reach few patients, limited by centralized manufacturing, cost, and stem cell losses in processing. Automation of ex vivo hematopoietic stem cell (HSC) editing has lagged behind CAR T because HSC manufacturing must absorb contingencies automated programs traditionally cannot: CD34+ yield varies widely between collections, many patients need more than one apheresis cycle to reach dose, and whether to culture, formulate, or hold is often decided during the run. The usual workarounds, splitting a product across instruments or reverting to open processing, add cost and risk. We report a single custom application (CAP) that performs CD34+ selection, in-chamber lipid nanoparticle (LNP) transfection, culture, and formulation on one closed automated system (CliniMACS Prodigy, Miltenyi Biotec) and resolves those contingencies within the program. To our knowledge this is the first program for ex vivo LNP delivery to HSCs on this platform.
METHODS: The CAP was co-developed with Miltenyi Biotec, runs on one instrument and one tubing set per collection cycle, and accepts up to 200 x 109 total nucleated cells and 2 x 109 CD34+ cells from mobilized apheresis. When a collection yields too few CD34+ cells, the run can end after enrichment so the cells can be cryopreserved. A later collection is then run through the same program, with the thawed cells combined with the freshly selected fraction in the chamber before transfection. Fresh, frozen, and combined starting material are therefore handled by one program, and because pooling precedes transfection, a patient needing more than one collection still receives a single manufacturing round: one transfection, one product, and one round of release testing. Enriched cells are retained in the culture chamber and transfected with LNPs co-delivering ABE8.8 mRNA and a single guide RNA targeting the HBG1/HBG2 promoter to recreate a hereditary persistence of fetal hemoglobin variant that disrupts repressor binding and creates an activator site. Culture and washes run from a programmable activity matrix, and the product is formulated and harvested automatically. Apart from manual CD34 labeling of the starting material, the process is closed and requires no electroporation or ISO 7 cleanroom. The first full-scale run processed a healthy-donor mobilized apheresis product end to end through the enrichment and culture path. Editing was quantified by Sanger sequencing, HbF by HPLC and flow cytometry after erythroid differentiation, and hematopoietic potential by colony-forming assay. A parallel small-scale arm compared 0.4 and 1.0 micrograms LNP per 106 cells.
RESULTS: On-target A-to-G conversion in the final product was 88% at the primary target adenine and 46% at the secondary target adenine, with 6% and 9% bystander editing at two adjacent adenines and none in untreated cells; the profile was unchanged after continued culture (82%, 44%, 8%, 9%). Full-scale editing fell within the range of the small-scale arm (primary target 79% and 95% at 0.4 and 1.0 micrograms per 106 cells), with no loss of editing efficiency at scale. After erythroid differentiation, HbF was 52% of total hemoglobin by HPLC compared with 7.6% for untreated cells, and F-cell frequency rose from 50.0% to 99.0%. Colony-forming assays of the final product showed multilineage colony formation. Viability was high, sterility was negative, and endotoxin was within specification. Genome-wide off-target analysis of this guide RNA by in silico prediction and iGUIDE sequencing showed minimal activity with no edits in known oncogenes.
CONCLUSIONS: A single automated program manufactured a full-scale, base-edited CD34+ product with editing efficiency and HbF induction comparable to bench scale, without electroporation or a second instrument. The design addresses practical obstacles that have limited automation of ex vivo HSC editing, is retargetable to other single-nucleotide indications by changing only the guide RNA, and will be made available to collaborating centers so they can manufacture their own products, as we have done for point-of-care CAR T. Additional engineering runs, NBSGW xenotransplantation, and off-target analysis are ongoing toward an IND; full release testing and process data from three engineering runs will be reported at the meeting.
DISCLOSURES
Stephan Kadauke, MD, PhD: Miltenyi Biotec Inc - Honoraria; Synthego - Research Funding; Lonza - Research Funding