High Ferritin and C-Reactive Protein Levels Result in Poor Clinical Outcome in R/R Large B-Cell Lymphoma Patients Despite Significant Anti-CD19 CAR-T Cell In Vivo Expansion

Introduction: clinical response to CD19-directed chimeric antigen receptor (CAR) T-cell therapy in relapsed/refractory (R/R) large B-cell lymphoma (LBCL) has been correlated with in vivo expansion of Tisa-cel and Axi-cel in clinical trials while mainly monocentric and pivotal studies have explored its role in real life settings. Additionally, although CAR T expander patients (pts) are more likely to respond, expansion cannot not be considered a robust predictive biomarker at the single patient level. In fact 30-40% of patients with significant CAR T in vivo expansion do not respond or relapse.

Aim: we analyzed CAR T cell dynamics in 262 lymphoma pts enrolled in the CART-SIE multicenter prospective observational study (NCT06339255) to correlate expansion and clinical outcome.

Methods: 262 R/R LBCL pts had longitudinal peripheral blood samples and paired clinical data. Pts received either Tisa-cel (n=94) or Axi-cel (n=168) in 13 centers from Dec 2019 to May 2024. Absolute quantification of circulating CAR T cells was performed by flow cytometry with CD19 CAR FMC63 antibody (Miltenyi). Pts who achieved complete or partial response by day 90 were considered responders (RE) while the others were defined as non-responders (NR).

Results: circulating CAR T cells were monitored on days 7, 10, 14, 21 and 30 after infusion and then monthly until disappearance. Median CAR T levels at peak expansion (Cmax) was 35.6 c/µl, median cumulative CAR T levels within the first month (AUC_0-30) was 79.5. Kinetic parameters strongly correlated (Spearman’s: r=0.9772, p<0.0001). Expansion was not affected by the type of infusion product (median Cmax: 34.6 vs 36.4 c/µl in Tisa-cel and Axi-cel, p=ns), but it was positively associated with CRS (p<0.0001) and with response at day 90 [median Cmax: 43 vs 24 c/µl for RE and NR (p=0.0036)], even in multivariable analysis [OR 2.75 (CI 1.47-5.13) p=0.0015]. When pts were divided into expanders and poor-expanders based on the median Cmax, expanders had significantly longer progression-free survival (PFS) (median PFS: not reached vs 135 days, p=0.0046). Both univariable and multivariable Cox models confirmed the positive association of expansion and PFS [HR 0.67 (CI 0.47–0.95) P=0.0233; HR 0.65 (CI 0.45-0.95) P=0.0245]. To understand why expansion does not always result in ­­durable clinical efficacy, we focused on expanders who achieved a response by day 90 (exp-RE, n=77) and expanders who were non-responders (exp-NR, n=42). Exp-RE had significantly lower levels of lactate dehydrogenase (LDH), ferritin and C-reactive protein (CRP) as compared to exp-NR, both pre CAR T infusion [median normalized LDH: 0.68 versus 0.97 for exp-RE and exp-NR (p=0.0022); median ferritin: 362 versus 943 ng/mL for exp-RE and exp-NR (p<0.0001); median CRP: 7 versus 17 mg/L for exp-RE and exp-NR (p=0.0002)], and within the first month after infusion [median AUC_0-30 ferritin: 4888 versus 15536 for exp-RE and exp-NR (p=0.0002); median AUC_0-30 CRP: 196.3 versus 439.8 for exp-RE and exp-NR (p=0.0008)]. CRS incidence was not different in the two groups. No differences in memory phenotypes among infused CAR T products were observed. In addition, bivariable models including interaction, indicated that the effect of ferritin was more prominent in strong expanders and negatively impacted survival (PFS: interaction p=0.0342; overall survival: interaction p=0.0354), whereas CRP tended to impact disease response (interaction p=0.06), and no differences were observed for LDH in strong and poor expanders. We then combined pre CAR T infusion ferritin and CRP levels with expansion and observed that the shortest PFS was present in expanders with both ferritin and CRP levels above the median [median PFS: not reached vs 135 days for exp-low ferritin-low CRP and exp-high ferritin-high CRP (p=0.0012)]. Of note this latter group displayed a median PFS equal to poor-expanders.

Conclusions: to our knowledge, this is the first study evaluating CAR T dynamics in a large prospective cohort of R/R LBCL pts treated with commercial Axi-cel and Tisa-cel. Together with showing that expansion is relevant for response and survival, we have highlighted the essential role of the persisting systemic inflammatory environment that could contribute to limiting the efficacy of expanding CAR T cells. These findings may suggest optimizing bridging strategies to lower systemic inflammation before CAR T infusion to improve responses.