Background: How long can CAR T-cell therapy remain effective?

In long-term follow-up of 38 patients with B-cell non-Hodgkin lymphoma treated with CD19-targeted chimeric antigen receptor T-cell (CAR T-cell) therapy, remission lasted up to 10 years in some patients. To interpret the findings, it is important to distinguish an endpoint based on lymphoma relapse or lymphoma-related death from survival measures that include deaths from all causes.

CD19-targeted CAR T-cell therapy is a standard treatment for relapsed or refractory B-cell non-Hodgkin lymphoma, but uncertainty has remained about long-term outcomes and the possibility of cure. The patients in this study received CTL019, now called tisagenlecleucel, after multiple prior treatments. The therapy uses patients’ own T cells, engineered to express a receptor that recognizes CD19 and carries a 4-1BB costimulatory signal.

The analysis included 24 patients with large B-cell lymphoma and 14 with follicular lymphoma. The data cutoff was October 1, 2025; median follow-up was 10.1 years, with a range of 7.9–11.5 years. The provided abstract does not allow confirmation of a comparator group or the method of randomization.

Key findings: Lymphoma-free survival differs from overall survival

No relapse occurred more than 5.4 years after treatment during the observation period. Estimated 10-year lymphoma-free survival was 32% in large B-cell lymphoma (95% confidence interval, 14–51%) and 47% in follicular lymphoma (95% confidence interval, 20–71%). The study defined lymphoma-free survival as the time from infusion to relapse or lymphoma-related death. These figures should therefore not be interpreted as the proportion of patients alive 10 years later.

Ten-year progression-free survival, which includes deaths from all causes, was 17% in large B-cell lymphoma (95% confidence interval, 5–34%) and 29% in follicular lymphoma (95% confidence interval, 9–52%). In the same order, reported 10-year overall survival was 17% (95% confidence interval, 5–34%) and 50% (95% confidence interval, 23–72%). Given the small patient groups and wide confidence intervals, it is difficult to conclude that treatment was more effective in one lymphoma type than the other.

Long-term safety findings included persistent grade 2 or 3 neutropenia in 2 patients (5%). No late anemia or thrombocytopenia was observed. Second primary cancers, distinct from the original lymphoma, occurred in 9 patients; the estimated 10-year cumulative incidence was 21%. This estimate should be distinguished from the simple proportion of 9 out of 38 patients. The 10-year non-relapse mortality rate was reported as 18%, or 14% when COVID-19-related deaths were excluded.

Meaning and outlook: Evidence of long-term remission, with uncertainties remaining

The investigators interpreted the findings as showing remission lasting up to 10 years in about one-third of patients with large B-cell lymphoma and nearly half of those with follicular lymphoma after a single tisagenlecleucel infusion. The results support the possibility that responses can persist for a long time in some patients. They do not rule out relapse after the observation period or establish a cure.

Persistence of the CAR transgene showed an association with long-term response. B-cell aplasia, or B-cell deficiency, persisted in 44% of patients with long-term responses. This observation alone does not establish a causal mechanism for long-term response or validate a marker for predicting an individual patient’s prognosis.

The evidence reviewed for this article is limited to the bibliographic record and abstract. The types of second primary cancers, their relationship to treatment, and detailed analysis conditions could not be confirmed. Further evidence is needed before applying the findings to other CAR T-cell products or patient groups. The key findings and endpoint definitions are available in the NEJM article abstract indexed in Europe PMC.