Abstract 3807: Validation of a functional genomics screen based on lentiviral integration site analysis reveals single-gene disruptions enhancing CAR19 T-cell effectiveness in preclinical models

Abstract Anti-CD19 CAR (CAR19) T-cell therapy has revolutionized the treatment for relapsed and refractory B-cell malignancies, showing long-term efficacy and potential cures. However, while many patients, including those with Chronic Lymphocytic Leukemia (CLL), exhibit initial signs of response, nearly two-thirds will eventually relapse. This highlights the critical need to understand treatment failure mechanisms and to identify targets to improve therapies. Our team previously reported an exceptional response in a CLL patient treated with CAR19 T-cells, which coincided with significant expansion of a CD8+ CAR T-cell clone harboring a disruption of the TET2 gene from an integration event of the lentiviral construct. Building on this finding, we conducted a large study analyzing lentiviral vector integration sites (LVIS) in 40 CAR19-treated Acute Lymphoblastic Leukemia (ALL) and CLL patients. Consistent with other studies, our data revealed a correlation between LVIS and therapeutic outcomes. To test whether these vector-mediated gene disruptions influence CAR T-cell growth kinetics and anti-tumor efficacy, we designed a CRISPR library targeting 180 LVIS-associated genes using 4-8 guide RNAs (gRNAs) per gene, including positive and negative controls. Our custom vector was engineered to drive the expression of both the anti-CD19 CAR and gRNAs under the promoters EF1a and human U6, respectively. To generate these modified CAR T-cells, we first transduced healthy donor T-cells at low multiplicity of infection, ensuring single-gene edits per cell, purified for CAR positivity three days post-transduction, and expanded cells for additional 10 days. The impact of each gene edit on CAR T-cell fitness was assessed in our well-established in vitro repeat tumor challenge and an in vivo xenogeneic model using the aggressive leukemia cell line NALM6. In both scenarios, we identified several genes with increased frequency (Log2FC ≥ 0.25, FDR ≤ 0.1) upon Robust Rank Aggregation (RRA) analysis using the MAGeCKFlute package in R, suggesting a potential fitness advantage. To validate the top candidates from our screen, we utilized Cas9-ribonucleoprotein (RNP) complexes with 2 independent gRNAs compared to AAVS1 (safe-targeting) and wild type controls. The targeted disruptions of the genes PCNX1 and PDCD10, among others, exhibited superior proliferative capacity upon our in vitro assay compared to controls. Ongoing studies include in vivo validation, phenotypic analyses of all candidates through spectral flow cytometry, and RNA sequencing to elucidate the biological roles of these genes in T-cell function. This preclinical package has the potential to advance into a Phase 1 clinical trial, where our optimized CAR T-cell product will be tested in CLL and potentially other refractory/relapsed B-cell malignancies. Citation Format: Beatriz Coutinho de Oliveira, Saaurav Bari, Jessica Kerr, Victoria Putzbach, Yee Peng Phoon, Joseph A. Fraietta, Wei-Ting Hwang, Frederic D. Bushman, Jan Joseph Melenhorst. Validation of a functional genomics screen based on lentiviral integration site analysis reveals single-gene disruptions enhancing CAR19 T-cell effectiveness in preclinical models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3807.