Document resource
Background Primary central nervous system lymphoma (PCNSL) is an aggressive B-cell malignancy confined to the CNS (brain, cerebrospinal fluid, spinal cord, and/or eye), with limited treatment options upon relapse. While CD19-directed CAR-T cell therapies have revolutionized the treatment of systemic B-cell malignancies, their application in PCNSL has been limited by neurotoxicity concerns. Recent studies have demonstrated safety and feasibility; however, relapse remains common, underscoring the need for more durable and CNS-adapted CAR-T cell products.Methods We evaluated a panel of second-generation CD19 CAR constructs, including ITAM-tuned ‘1XX’ variants, using a 3D spheroid model of PCNSL. The (SJ25C1)1XX-CAR format showed superior antitumor activity. To enhance CAR-T cell persistence and reduce exhaustion, we used CRISPR/Cas9 to knockout SUV39H1, a histone methyltransferase that regulates H3K9 trimethylation and limits memory T cell differentiation. CAR-T cells expressing the (SJ25C1)1XX-CAR and lacking SUV39H1 were evaluated in vitro and in vivo.Results SUV39H1-deficient (SJ25C1)1XX CAR-T cells showed increased expression of memory-associated genes and reduced expression of dysfunction and exhaustion markers in vitro. In an orthotopic PCNSL xenograft model, these dual-engineered T cells achieved superior tumor control and extended survival compared to conventional second-generation CAR-T cells. When compared to non-edited (SJ25C1)1XX-CAR-T cells, survival outcomes were similar, likely reflecting the high baseline efficacy and absence of relapse in this model. Nevertheless, ex vivo single-cell RNA sequencing and spectral flow cytometry confirmed that SUV39H1 knockout enhanced memory features and reduced exhaustion signatures. To assess long-term functional persistence, we developed a vitreoretinal lymphoma rechallenge model that allows delayed tumor re-exposure following initial remission.Conclusions These data support the clinical translation of a next-generation cell therapy for PCNSL based on epigenetically reprogrammed, signaling-optimized CAR-T cells. We are currently developing a GMP-compatible process incorporating the RQR8 marker/suicide switch for enrichment and rituximab-mediated elimination in the event of unexpected toxicity, in preparation for a future phase I/II clinical trial.