7025 Background: ctDNA is a minimally invasive method for detecting MRD in various malignancies. MRD negativity (MRD neg ) has been linked to improved PFS and durable disease control, including after liso-cel treatment in R/R large B-cell lymphoma. The prognostic value of MRD after chimeric antigen receptor (CAR) T cell therapy in R/R FL has not been established. We report an exploratory MRD analysis from TRANSCEND FL (NCT04245839), in which pts had high rates of deep and durable responses after liso-cel. Methods: Of 103 efficacy-evaluable pts with third-line or later (3L+) FL, 89 (86%) had ctDNA samples, of whom 90% (80/89) had evaluable ctDNA and efficacy data after liso-cel infusion. Tumor-derived phased variants were primarily identified from screening (baseline) plasma or tumor tissue when plasma was insufficient, using PhasED-Seq. Post-treatment ctDNA-MRD was assessed at months (M) 1 and 3 after infusion for all pts, and at M24 for ongoing responders. MRD positivity (MRD pos ) was defined as ctDNA above the assay detection threshold (lower limit of detection of 0.7 parts per million, with sample-specific sensitivity dependent on DNA input). Results: MRD neg was achieved by 89% (71/80) of evaluable pts from TRANSCEND FL. Among pts with a best overall response of CR, 93% (71/76) of them achieved MRD neg . ctDNA clearance increased over time, with MRD neg rates of 68% (52/76) at M1, 83% (63/67) at M3, and 97% (57/59) at M24. In contrast, pts with PR, SD, or PD (n=4) never achieved MRD neg . Baseline ctDNA levels were not associated with PFS after liso-cel treatment. MRD neg after infusion was associated with longer PFS compared with MRD pos ( P <0.001). In addition, MRD status at M3 (HR, 7.0 95% CI, 3.0–16.2) showed a stronger correlation with PFS than at M1 (HR, 3.0 95% CI, 1.3–6.8), with 36-M PFS rates of 80% for MRD neg (n=63) versus 38% for MRD pos (n=13) pts. Multivariate analysis controlling for PET findings confirmed the independent prognostic value of MRD for PFS. Importantly, PFS was longer for pts with both CR by PET and MRD neg at M3 (n=62) compared to pts with CR and MRD pos (n=9; HR, 5.0, 95% CI, 1.8-13.4), with 36-M PFS rates of 81% versus 56%, respectively, suggesting that combining MRD with PET may provide additional prognostic information for long-term outcomes. Conclusions: Liso-cel induced deep molecular responses in pts with 3L+ FL with most pts (89%) achieving MRD neg after infusion. Although baseline ctDNA levels were not associated with PFS, MRD neg after infusion was significantly associated with improved PFS, highlighting the therapeutic benefit of liso-cel regardless of tumor burden before treatment. These findings support the potential of ctDNA-based MRD as a prognostic biomarker in FL and underscore that liso-cel is a key treatment option capable of inducing MRD neg and durable clinical outcomes in pts with 3L+ R/R FL. Clinical trial information: NCT04245839 .
Avilion et al. (2026) studied this question.