Efficient manufacturing of CAR-T cells from whole blood: a scalable approach to reduce costs and enhance accessibility in cancer therapy

Abstract

Background Chimeric antigen receptor T (CAR-T) cells have significantly advanced the treatment of cancers such as leukemia and lymphoma. Traditionally, T cells are collected from patients through leukapheresis, an expensive and potentially invasive process that requires specialized equipment and trained personnel. Although whole blood collections are much more technically straightforward, whole blood starting material has not been widely utilized for clinical CAR-T cell manufacturing, in part due to lack of manufacturing processes designed for use in a good manufacturing practice (GMP) environment. Collecting cellular starting material from whole blood without leukapheresis could reduce manufacturing complexity and cost, thereby improving accessibility to CAR-T cell therapy. Methods Whole blood samples were collected from eight healthy donors and one pediatric B-cell acute lymphoblastic leukemia (B-ALL) patient. These samples were processed using the Sepax C-Pro (Cytiva) instrument to isolate mononuclear cells (MNCs) via density gradient separation. CAR-T cells were then manufactured from the isolated MNCs using a GMP-compliant 7-day protocol, whereby T cells were activated with anti-CD3 and IL-2, transduced with GMP lentiviral vector encoding a CD19/CD22 bispecific CAR, and expanded in gas permeable cell culture bags. The resulting CAR-T cells were then evaluated for their phenotypic and functional properties using flow cytometry, cytokine release and cytotoxicity assays. Results From an average 77.7 mL of whole blood from healthy donors (range = 29–96 mL), we isolated an average of 42.2 × 10 6 CD3⁺ T cells (range 7.3–63.0) postprocessing. CAR-T cell cultures were initiated from thaw using 1–10 × 10 6 starting CD3 + T cells, yielding a median T cell number of 105 × 10 6 cells on day 7 (range 61–188 × 10 6 ). We observed 66 ± 11% mean transduction efficiency and produced a mean of 77.4 × 10 6 transduced CAR-T cells (range 30.8–143.5 × 10 6 ). Similar results were obtained when using a blood sample (28mL) obtained from a patient with relapsed B-ALL who had received recent chemotherapy. Conclusions Therapeutically relevant doses of CD19/CD22 CAR-T cells can be successfully manufactured from whole blood. On average, 80 mL of whole blood yields enough CAR-T cells to create a single dose for a pediatric patient (50 kg) at a dosage of 1 × 10 6 CAR-T cells/kg. For larger patients, scaling up is straightforward by collecting a larger blood volume. This method also demonstrates a cost-effective approach to T cell activation and expansion which, alongside a more straightforward collection of whole blood, makes it more widely accessible especially for middle- and low-income countries. By reducing costs and labor, this strategy has the potential to significantly expand global access to CAR-T cell therapy.