Prognostic impact of CAR-T cell expansion and PD1 expression in patients with DLBCL treated with axicabtagene ciloleucel Free

Introduction. Chimeric antigen receptor T-cell (CAR-T) therapy targeting CD19 has transformed the treatment landscape for patients with relapsed or refractory diffuse large B-cell lymphoma (DLBCL). Among approved products, axicabtagene ciloleucel (axi-cel) has demonstrated high response rates in pivotal trials and real-world settings. Nevertheless, the clinical course after CAR-T infusion remains variable, with a substantial proportion of patients experiencing early progression or limited durability of response. This heterogeneity underscores the urgent need for reliable prognostic markers that can predict treatment outcomes and guide post-infusion monitoring or early intervention strategies. While several clinical and biological factors have been studied, the prognostic significance of CAR-T cell expansion dynamics and phenotypic characteristics after infusion has yet to be clearly established. The aim of this study was to assess whether specific parameters of CAR-T cell behavior in peripheral blood — i.e., expansion dynamics and circulating CAR-T subpopulation composition — could correlate with progression risk and survival in a cohort of patients treated with axi-cel.

Methods. This retrospective study included adult patients with DLBCL who received axi-cel as third-line therapy between 2022 and 2025 at a single institution. Peripheral blood samples were collected on days 3, 7, 14, 30, and 90 after infusion. Flow cytometry was performed using a DxFlex cytometer (Beckman Coulter, USA) to quantify CAR-T cells and assess expression of CD3, CD4, CD8, PD1, CD19, CD56, and CD45. CAR-T cells were detected using a recombinant CD19 protein (Acro Biosystems, USA). Data were analyzed using Kaluza software. Cumulative incidence (CI) was estimated using Gray's test to assess progression, non-relapse mortality (NRM), cytokine release syndrome (CRS), and immune effector cell-associated neurotoxicity syndrome (ICANS), considering competing risks. Kaplan-Meier estimates were used to evaluate progression-free survival (PFS) and overall survival (OS). Statistical significance was set at p<0.05.

Results. The cohort included 84 patients, comprised 45 men (53.6%), with 60 years of median age (range, 18–83). With a median follow-up of 24.3 months (range: 4–64.4), 2-year PFS and OS for the entire cohort were 39.6% (95% CI: 30.0–52.2%) and 47.8% (95% CI: 37.3–61.3%), respectively. The CI of disease progression was 47.1% at 12 months and 52.4% at 24 months. The CI of NRM was 6.2% at 6 months, 7.1% at 12 months, and 8.3% at 24 months. At day 28 post-infusion, the CI of CRS was 95.2% for any grade and 2.4% for grade ≥3, while the CI of ICANS was 44.0% for any grade and 21.4% for grade ≥3. CAR-T expansion peaked on day 7 in 64.3% of patients and on day 14 in 35.7%. The median peak CAR-T cell expansion was 44.4 cells/μL (range, 0–1926.6). Patients with a peak CAR-T cell expansion ≥30 cells/μL had significantly improved 2-year PFS (46.7% vs. 29.1%; log-rank p = 0.035) and OS (59.5% vs. 31.3%; log-rank p = 0.012), compared to those with lower expansion. A CD8⁺ CAR-T cell proportion ≥55% at peak was associated with superior 90-day PFS (77.6% vs. 53.3%, log-rank p = 0.023). Moreover, a high PD1^hi expression ≥85% on CD8⁺ CAR-T cells at day 14 was associated with a lower CI of progression at both 12 months (32.4% vs. 64.1%) and 24 months (35.9% vs. 67.8%) compared to lower PD1^hi expression (p = 0.024).

Conclusions. These findings suggest that the magnitude of CAR-T cell expansion, the predominance of CD8⁺ subsets, and PD1^hi elevated expression are promising prognostic markers in DLBCL patients treated with axi-cel in third line. Immune profiling on days 7 and 14 post-infusion appears particularly informative and may be feasible for integration into routine clinical practice. Multivariable analysis is ongoing to confirm the independence of these associations.