Impact of late immune effector cell associated hematotoxicity (ICAHT) and prolonged neutropenia after CAR-T therapy on infection and survival outcomes.

Abstract

7041 Background: Prolonged hematologic toxicity is a common side effect of chimeric antigen receptor T-cell (CART) therapy. The EHA/EBMT panel developed the late ICAHT grading which incorporates depth of neutropenia after Day 30. Recently, the early ICAHT grade (D 0-30) was associated with infection (infx) and survival outcomes. Here, we evaluate the impact of late ICAHT and neutropenia on survival and infx outcomes after CD19 CART. Methods: We included two cohorts of adult patients (pts) with large B cell lymphoma (LBCL), mantle cell lymphoma (MCL), and follicular lymphoma (FL) who received commercial CD19 CART. Late was graded based on depth of neutropenia after D30 (G1 ≤ 1.5; G2 ≤ 1.0; G3 ≤ 0.5; G4 ≤ 0.1 G/L) and pts were censored for relapse, progression, or loss of follow up. Infxs after D30 were graded per CTCAE as mild G ≤ 2 or severe G ≥ 3. Cohort 1 included pts from 8 US medical centers for survival outcomes. Only data for severe infx was collected from D30-D100 and longer term infx data was not available. Landmark survival analysis at D+100 was performed using the Kaplan-Meier method and multivariate cox proportional hazards. To assess long term late infx outcomes, cohort 2 included pts at Vanderbilt with granular cytopenia and infx data (D30-365). Results: Cohort 1 included 277 pts with LBCL of whom 143 received axi-cel, 61 tisa-cel, and 72 liso-cel with median follow up 395 days (range 30-1763). By D100, there were 111 (40%) without, 89 (41%) with mild to moderate (G1-G2), and 53 (19%) with severe (G3-G4) late ICAHT. There were 13 pts (6%) with severe infx which was not associated with severe ICAHT (p=0.68). In multivariate landmark survival analysis using previously published covariates, severe late ICAHT was not associated with PFS (HR 1.64, p=0.11) or OS (HR 1.54, p=0.17); however, D100 neutropenia (ANC ≤ 1000) was associated with inferior PFS (HR 2.00, p=0.002) and OS (HR 1.95, p=0.003). NRM was 11% and associated with D100 neutropenia (p=0.015) but not late ICAHT (p=0.15). Cohort 2 had 43 pts (35 LBCL, 7 MCL, 1 FL) with median follow up 483 days (range 233-2133). 34 pts received axi-cel, 2 tisa-cel, and 7 brexu-cel. By 1 yr, there were 10 (23%) without, 27 (63%) with mild to moderate, and 6 (14%) with severe late ICAHT. There were 52 infxs by 1 yr and 16 pts (37%) had no infx. Median time to first infx after D30 was 165 days (range 31-365). There were 31 (60%) mild and 21 (40%) severe infx. IVIG was used in 28 (65%) and G-CSF in 27 (62%). Only 1 pt with severe ICAHT had severe infx. Late ICAHT and neutropenia were not associated with severe infection in logistic regression. Conclusions: These data did not show an association between EHA/EBMT late ICAHT grading and infx after D30 or survival. A detailed reporting of T and B cell reconstitution data may provide clues to better predict late infx. Persistent neutropenia at D100 was associated with inferior PFS and OS which needs validation in future studies.