Primary central nervous system lymphoma (PCNSL) is an aggressive and rare subtype of non-Hodgkin lymphoma. Standard of care treatment includes high-dose methotrexate-based chemotherapy followed by autologous stem cell transplant or whole brain radiation. Despite these treatments, 50% of patients relapse within two years and approximately 15-20% have disease that is refractory to initial therapy. Prognosis of these patients is poor, and there have been no prior FDA-approved therapies for relapsed/refractory (R/R) PCNSL. On February 6, 2026, the FDA approved a label update for axicabtagene ciloleucel (axi-cel; Yescarta) that removed the prior Limitations of Use excluding patients with PCNSL based on the results from a study developed at Dana-Farber Cancer Institute. As a consequence, axi-cel became the first and currently only CD19-directed CAR T-cell therapy for which the label explicitly supports use in R/R PCNSL.
In the past, pivotal trials of anti-CD19 chimeric antigen receptor (CAR) T-cell therapies in R/R systemic large B-cell lymphomas excluded enrolling patients with PCNSL due to the concern of increased risk of immune effector cell-associated neurologic syndrome (ICANS). However, as these therapies were being used commercially, it was evident that patients with concurrent CNS involvement seemed to tolerate them reasonably well.
In order to systematically study the safety and efficacy of axicabtagene ciloleucel in R/R CNS lymphomas, Caron Jacobson, MD, MMSc, medical director of the Immune Effector Cell Therapy Program, and Lakshmi Nayak, MD, director of the Center for CNS Lymphoma, came together to develop an investigator-initiated trial at Dana-Farber. The trial enrolled 18 patients with R/R CNS large B-cell lymphoma — 13 with PCNSL and five with secondary CNS lymphoma (SCNSL). Most patients were heavily pretreated. In this study, bridging therapy was not allowed after consenting, although stable to lowering doses of corticosteroids were allowed. The overall response rate was 89% and the complete response rate was 67%, with comparable CR rates in PCNSL (69%) and SCNSL (75%). The median time to best response was three months. The median duration of response was 13.8 months and median progression-free survival was 14.3 months.
The safety profile is the central question for any CAR T cell product in CNS lymphoma, and the trial data were reassuring with no treatment-limiting toxicities. Cytokine release syndrome (CRS) occurred in 89% of patients, but all events were grade 1-2. ICANS occurred in 44%, with grade ≥3 events in 28%, broadly consistent with axi-cel's established profile in systemic large B-cell lymphoma rather than amplified by CNS disease. Importantly, all toxicities were manageable.
The label update has tremendous practical implications. It opens commercial access to CAR T-cell therapy for an underserved population that had been categorically excluded from registration studies. An important aspect to consider is the appropriate selection of candidates for this treatment. Patients with minimally symptomatic disease in the R/R setting are likely ideal to consider for this therapy. In patients with bulky and symptomatic disease, "bridging" therapy to "debulk the tumor" and improve performance status including neurologic status is crucial before apheresis and cell infusion. Candidates for bridging can include therapies that induce quick responses such as high-dose methotrexate in those that have previously responded to it, Bruton's tyrosine kinase (BTK) inhibitors, or focal radiotherapy. It is also important for oncologists to be aware of another syndrome that is unique to CNS tumors: tumor inflammation-associated neurotoxicity (TIAN). This condition is distinct from CRS and ICANS, represents a localized inflammatory process based on the anatomical location of the CNS lymphoma, and needs to be managed appropriately. Similarly, it is critical to distinguish TIAN from tumor progression.
While progress in PCNSL has been lagging, these data and the FDA label update indicate the beginning of a change in the paradigm for this disease. Next steps include larger, multicenter trials as well as evaluation of potential factors that limit long-term efficacy of cellular therapies such as immune escape and T-cell exhaustion. Studies to overcome the hostile brain tumor microenvironment as well as to improve the delivery of cellular therapies to the CNS are being investigated at Dana-Farber.