The 2026 European Hematology Association (EHA) Annual Congress was held in Stockholm, Sweden, from June 11 to 14, 2026. Featuring highly anticipated plenary sessions, the presidential symposium, specialized seminars, and interactive poster sessions, EHA 2026 provided a comprehensive exchange platform for medical professionals worldwide.

CAR-T cell therapy has significantly improved the prognosis for patients with relapsed/refractory large B-cell lymphoma (R/R LBCL). Based on the results of the ZUMA-1 and ZUMA-7 trials, the indication for Axicabtagene ciloleucel (Axi-cel) has expanded from third-line and beyond (3L+) to second-line (2L) treatment. Clinical efficacy data spanning the ZUMA-1 ^[1,2], ZUMA-7 ^[3,4], and ZUMA-12 ^[5,6] studies consistently demonstrate that earlier intervention with Axi-cel yields superior response rates and survival benefits. Compared to later-line treatments, 2L administration of Axi-cel not only achieves a higher single-run manufacturing success rate but also delivers a two-fold increase in median naive T-cell composition compared to 3L+ patients, translating into significantly enhanced CAR-T cell quality ^[7].

Two large-scale real-world studies (PF966 ^[8] and PF1026 ^[9]) presented at this year's EHA Congress further reinforce the efficacy and safety advantages of 2L Axi-cel. Among them, Poster 1 (PF966) evaluated the impact of prior lines of therapy on the risk of disease progression after CAR-T cell therapy and overall survival post-treatment failure (post-failure OS, OS2) within an Italian real-world setting. Poster 2 (PF1026) compared the early toxicity patterns (CRS and ICANS) of Axi-cel in Spanish DLBCL patients—stratified by 2L versus 3L+ groups—focusing on onset time, severity, duration, and interventions.

Poster 1: PF966^[8]

^{Impact of prior lines of therapy on Progression Risk and post-failure Overall Survival following Axicabtagene ciloleucel or Lisocabtagene maraleucel CAR T-cells therapy}

^Methodology

^{CART-SIE is a prospective observational study of patients (pts) treated with anti-CD19 CAR T-cells at 22 Italian centers. A total of 627 pts were treated: 604 received Axicel, and 23 Lisocel. Time to progression was calculated from infusion to first documented progression and analyzed using a competing-risks framework, with death as the competing event. Cumulative incidence functions were estimated, and between-group comparisons were performed using univariable and multivariable Fine and Gray models. Six-month progression estimates were derived from the final multivariable model across covariate combinations to identify factors associated with early relapse/progression. Post-failure overall survival (OS2) was measured from relapse/progression to death and estimated using the Kaplan–Meier method.} 

^{Study Results}

^{Baseline Characteristics}

^{Of 627 pts, 221 (35%) received CAR T-cells in the 2L, 282 (45%) in 3rd (3L), and 124 (20%) in 4th or later line (4L+). Patient characteristics differed significantly across groups for the following factors: 1) High-grade B-cell lymphomas were more common in 4L+ and 3L than in 2L (20% vs 17% vs 9%, p<0.001); 2) transformed DLBCL was more frequent in 4L+ and 3L than in 2L (35% vs 17% vs 6.8%, p<0.001); 3) as expected, a greater proportion of pts with primary refractory disease received CAR T-cells in 2L than in 3L or 4L+ (67% vs 21% vs 16%, p<0.001); 4) CAR-Hematotox was more frequently high in 4L+ and 3L than in 2L (70% vs 66% vs 48%, p<0.001). No significant differences were found in stage, IPI, CRP, ferritin, LDH levels, extranodal sites, use of bridging therapy, or time from apheresis to infusion.}

Efficacy Outcomes

At a median follow-up of 12.5 months (IQR: 6-23), 260 of 627 (41%) progressed. The estimated cumulative 6- and 24-month progression risk was 35% and 47% in 2L, 38% and 50% in 3L, and 43% and 56% in 4L+. Predicted cumulative incidence of PD within 6 months, extracted from a multivariable model, identified late line of therapy, advanced stage, no response to bridging, and bulky disease as significant factors. The difference in the probability of PD within 6 months was more pronounced between 4L+ and 3L (13.9%) and between 4L and 2L (20.7%). 

The OS2 was 29.5% (27.5% in 2L, 33% in 3L, and 23.5% in 4L+). Among progressing pts, there was no difference in the use of salvage therapy between those that failed CAR T-cells in 2L and those in 3L or 4L+ (p=ns). The use of bispecific antibodies, alone or in combination, was higher in pts treated with CAR T-cells in 2L (p=0.0009), but they have not yet affected survival. 

Post Relapse Overall Survival based on lines of therapy before CAR-T Cell Infusion (OS2)

Poster 2: PF1026^[9]

Early Toxicity of Axicabtagene Ciloleucel by Line of Therapy in DLBCL: A Real-World Comparison of 2L vs ≥3L Treatment.

Methodology

This retrospective, multicenter study on behalf of the GETH-TC (Spanish Group of Stem Transplantation and Cell Therapy) group compared consecutive DLBCL patients receiving axi-cel in second line from April 2024 (following regulatory approval in Spain) with a previously published historical cohort from our group treated between 2018 and 2021 in third or later line (Bailén et al. Transpl Cell Ther 2022). Patients were grouped by line of therapy at infusion: 2L vs ≥3L. CRS and ICANS were graded per ASTCT consensus criteria. Endpoints included incidence (any grade; grade ≥3), time-to-onset (days from infusion), duration (days), ICU admission, and treatment (tocilizumab, corticosteroids, anakinra if applicable). Comparative analyses used χ²/Fisher for categorical and Mann-Whitney/t-test for continuous variables. 

Study Results

Baseline Characteristics

^{A total of N=413 patients were included (2L n=184, ≥3L n=229). Baseline characteristics are summarized in Table 1. Briefly, the 2L cohort showed a better ECOG performance status at infusion, lower R-IPI, lower prior ASCT exposure, lower rates of bulky disease and elevated LDH, and more frequent use of bridging therapy. No differences were observed in the median time from apheresis to infusion between groups.} 

^{Safety Outcomes}

^{Comparing 2L vs ≥3L, any-grade CRS occurred in 95% vs 92%, with grade ≥3 CRS in 4% vs 8% (p=0.002). Median time-to-onset was 2 days in both groups and median duration 4 vs 5 days (p=0.02). Tocilizumab was administered in 61% vs 62% patients (median doses 1.5 vs 2, p=0.03), and corticosteroids were administered in 48% vs 32% (p<0.001). 
Any-grade ICANS occurred in 47% vs 52%, with grade ≥3 ICANS in 12.5% vs 21% (p<0.001). Median time-to-onset was 6 days in both groups and median duration 3 vs 4 days. Corticosteroids for ICANS were used in 35% vs 43% (p<0.001).
ICU admission related to toxicity occurred in 21% vs 24%. No differences were observed in non-relapse mortality (NRM).}

References:
1. Neelapu SS, et al. N Engl J Med. 2017 Dec 28;377(26):2531-2544.
2. Neelapu SS, et al. Blood. 2023 May 11;141(19):2307-2315. 
3. Locke FL, et al. N Engl J Med. 2022 Feb 17;386(7):640-654. 
4. Westin JR, et al. N Engl J Med. 2023 Jul 13;389(2):148-157. 
5. Neelapu SS, et al. Nat Med. 2022 Apr;28(4):735-742. 
6.Chavez JC, et al. Blood. 2025 May 15;145(20):2303-2311.
7. Jason Westin, et al. 2024 EHA. Abstract# P1425.
8. Anna Dodero, et al. 2026EHA. Abstract# PF966.
9. Rebeca Bailen, et al. 2026EHA. Abstract# PF1026.