Late immune effector cell-associated hematotoxicity (ICAHT) in patients with relapsed refractory large B-cell lymphoma receiving anti-CD19 chimeric antigen receptor T-cell therapy

Abstract

Abstract Background: Cytopenias are common following chimeric antigen receptor (CAR) T-cell therapy in patients (pts) with relapsed refractory (R/R) large B-cell lymphoma (LBCL) and are associated with increased non-relapse mortality (Wudhikarn 2020). The CAR HEMATOTOX (CAR-HT) score (Rejeski et al, 2021) incorporating baseline blood counts, CRP and ferritin, is a validated tool for predicting cytopenias. Although early cytopenias up to day 30 are expected, late cytopenias beyond day 30 are less well understood. We therefore focused on describing the trajectory of late cytopenias to 1 year and survival outcomes in this report. Methods: This was a single-centre, retrospective study in pts with R/R LBCL who received anti-CD19 CAR T-cell therapy from June 2020 to April 2024. Our primary objective was to evaluate the frequency of persistent cytopenias (beyond day 30) in R/R LBCL pts, with secondary objective to validate the CAR-HT score in predicting late cytopenias at 3, 6 and 12 months and survival outcomes in these pts. Pts with early disease progression within 60 days of CAR T infusion were excluded. Pts were censored at time of death or progression requiring systemic therapy, or last follow up. Cytopenias were graded by CTCAEv.5 and neutropenia was also graded by late immune effector cell-associated hematotoxicity (ICAHT) grading as described by EHA/EBMT (Rejeski 2022). Results: We identified 139 pts who received axi-cel or tisa-cel: 113 were included in our cytopenia analysis; 26 excluded for early progression, death, or initiation of subsequent therapy. Eighty-eight (78%) pts received axi-cel, and 25 (22%) received tisa-cel for treatment of R/R diffuse LBCL/high grade B-cell lymphoma (n=72), transformed LBCL (n=37), or primary mediastinal B cell lymphoma (n=4); 46% (n=52) had primary refractory disease. Median prior lines of systemic treatment were 2 (range 2-6), 36% (n=41) prior autologous stem cell transplant (ASCT), and 28% (n=32) prior bendamustine. Pre-infusion high CAR-HT score (≥2) was observed in 57% (64/112); low score (0-1) in 43% (48/112). At 1 month post CAR T infusion, 98% (111/113) pts had any grade cytopenias in at least 1 lineage: anemia 94.6% (≥ Grade(G)3 11.5%), neutropenia 62.8% (≥G3 36%), thrombocytopenia 82% (≥G3 24%). At 3 months (mos.), 88% (88/101) pts had any grade cytopenias: anemia 80% (≥G3 9%), neutropenia 46% (≥G3 19%), thrombocytopenia 47% (≥G3 17%). Between day 30 and 100 post-infusion, 35% (35/101) of pts required at least 1 dose of GCSF while, 9% and 6% of pts required at least one blood and platelet transfusion support, respectively. Amongst pts alive and in remission, ≥G3 anemia was observed in 5% (4/79) at 6 mos and 3% (2/68) at 12 mos, ≥G3 thrombocytopenia persisted in 11% (9/79) and 1.5% (1/68), respectively. At 3-, 6-, and 12-mos post CAR T, any grade ICAHT was observed in 38% (39/101; ≥ G3 6%), 24% (19/79; ≥ G3 5%) and 13% (9/68; ≥G3 1.5%), respectively. Of the 19 pts with any grade ICAHT at 6 mos, 1 progressed, 2 developed AML, 2 had persistent neutropenia, 3 died from infections, and the remainder recovered their counts by 1 year. A high CAR-HT score was the only factor predictive of any grade ICAHT (p=0.020) or ≥ G3 ICAHT (p=0.036) at 3 months. No association was found with age, gender, CAR product, median number of prior therapies, prior ASCT or bendamustine exposure. At 6 mos, ICAHT was associated with both high CAR-HT score (p=0.0074) and high number of prior therapy lines (p=0.02). However, the CAR-HT score did not predict ICAHT at 1 year (p=0.28) as numbers were small. At median follow up of 21 mos (2-57) 32 patients died: 19 from disease progression, 6 from infection, 3 from AML/MDS, 4 unknown/unrelated causes. Median OS was not reached, while median PFS was 21 mos (95% CI 8.94-NR). Patients with high CAR-HT score had lower OS and PFS than low score patients, but this was not statistically significant [2-year OS 63% (95% CI 49-75) vs 82% (95% CI 65-91) p=0.1033; 2-year PFS 42% (95% CI 30-55) vs 55% (95% CI 40-68) p=0.55)]. Conclusion: Late cytopenias remain persistent toxicity after CAR T-cell therapy in RR LBCL. Although, the majority of our pts recovered counts by 1 year, a notable proportion develop adverse outcomes including late disease progression or MDS/AML. In this report, we validated the CAR-HT score in predicting ICAHT at 3 and 6 months, supporting its potential utility for resource planning and closer monitoring for these negative outcomes.