High complete response rates, low dropout rate, and low-grade toxicities in patients with relapsed/refractory diffuse large B-cell lymphoma (DLBCL) receiving GLPG5101, a fresh, early memory-enriched CAR T-cell therapy with a 7-day vein-to-vein time: Results from the ATALANTA-1 DLBCL cohort

Abstract

Abstract Introduction:CAR T-cell therapies have revolutionized treatment for relapsed/refractory (R/R) DLBCL; however, high rates of progression and relapse have been observed, and their widespread use is limited by high toxicity rates, long vein-to-vein (VTV) times, and limited access. Anti-CD19 CAR T-cell therapies offering short VTV times and preservation of early phenotype can improve efficacy and safety outcomes in patients (pts) with aggressive lymphomas. Here, we present updated results for pts with R/R DLBCL from the ATALANTA-1 study of GLPG5101, a fresh, early memory-enriched phenotype autologous CD19 CAR T-cell therapy manufactured using a decentralized cell therapy platform. Methods: ATALANTA-1 (CTIS: 2022-502661-23-00; NCT06561425) is a Phase 1/2 study of GLPG5101 in pts with R/R non-Hodgkin lymphoma. Pts who were refractory or relapsed after ≥1 lines of therapy were enrolled. Primary objectives were safety and establishment of the recommended Phase 2 dose in Phase 1, and efficacy in Phase 2. Safety, pharmacokinetics, T-cell phenotypes, and efficacy data pooled across planned dose levels (DLs) are reported for pts with R/R DLBCL: DL1, 35–50×106; DL2, 85–110×106; DL3, 200–250×106CAR T cells. Results: As of 21 April 2025, 24 pts with DLBCL underwent leukapheresis; 1 pt discontinued before lymphodepletion: a dropout rate of 1/24 (4%). Decentralized manufacturing enabled all 23 pts to receive a fresh product with a median (range) VTV time of 7 (7–21) days. None of the pts required bridging chemotherapy. In the efficacy analysis set (EAS, n=22), median (range) pt age was 69 (25–79) years; 5/22 (23%) pts had high-risk International Prognostic Index scores at study entry and 16/22 (73%) had intermediate risk. 20/22 (91%) pts had Ann Arbor disease Stage III–IV; 4/22 (18%) had MYC and BCL2 and/or BCL6 rearrangements (double/triple hit); 12/22 (55%) were refractory to the last line of therapy. Median (range) number of prior systemic lines of therapy was 1 (1–6), and 10/22 (45%) pts received GLPG5101 CAR T-cell therapy in the third or later line (3L+) setting. Among 12 pts infused in 2L, 8 (67%) were primary refractory. During the treatment period (≤14 weeks post-infusion), most Grade (G) ≥3 treatment-emergent adverse events (TEAEs) were hematologic. CRS was reported in 41% (9/22) of pts, all G1/G2 apart from a single G3 event. ICANS was reported in 27% (6/22) of pts, all G1 apart from a single G3 event. There were 7 deaths: 3 due to progressive disease, 2 considered attributable to TEAEs during the treatment period (intracranial hemorrhage, n=1; intra-abdominal hemorrhage, n=1), and 2 considered attributable to AEs during the follow-up period (hemophagocytic lymphohistiocytosis due to a second primary malignancy, 22.9 months post-infusion while in complete response [CR], n=1; COVID-19, n=1). At data cutoff, median (95% CI) follow-up duration was 15.1 (4.6, 23.2) months. The objective response rate (ORR) was 75% (18/24) in the intention-to-treat population and 82% (18/22) in the EAS. The CR rate (CRR) was 58% and 64%, respectively. Notably, for 3L+ pts, the ORR was 90% (9/10), with a CRR of 80% (8/10). The Kaplan–Meier (KM)-estimated 12-month duration of response (DOR) rate for the overall and 3L+ pt population was 50% and 63%, respectively. Compared with the starting material, the proportion of early memory phenotype CD8+CAR T cells increased significantly in the final product, and increased moderately within CD4+ CAR T cells, in 89% (17/19) of evaluable pts. These increases drove a corresponding reduction of terminally differentiated effectors. CAR T cells were detected in peripheral blood in >50% of evaluated pts until ≥18 months post-infusion. Conclusions: GLPG5101, a CAR T-cell product with an early memory-enriched T-cell phenotype, demonstrated encouragingly high clinical activity and a favorable safety profile, as well as long-term persistence, in pts with R/R DLBCL. Decentralized manufacturing resulted in a fresh product infusion in all pts, with a median VTV time of 7 days, and a low dropout rate (4%). Among 3L+ pts (n=10), the ORR and CRR were 90% and 80% respectively, with a KM-estimated 12-month DOR of 63%. High-grade toxicities were infrequent, with only 1 case each of G3 CRS and ICANS. These results suggest GLPG5101 addresses current unmet needs of durable efficacy, low toxicity, and rapid VTV time, and thus support continued development of GLPG5101 for pts with R/R DLBCL.