3732 - A Multimodal Analysis of Neurotoxicity Following Chimeric Antigen Receptor T-Cell Therapy Using Deep Learning-Based MRI
Presenter(s)
A. Zhu1, J. Rudie2, R. Saluja3, P. Sanghvi4, J. A. Hattangadi-Gluth5, C. Costello6, D. Koura1, A. Goodman1, A. Hamdan1, A. R. Jeong1, E. Ball1, J. S. Kohli7, C. McDonald7, W. Pearse8, M. Choi8, B. Heyman8, E. Reid8, D. Tzachanis1, and K. R. Tringale5; 1UC San Diego, La Jolla, CA, 2Department of Radiology, University of California San Diego, La Jolla, CA, 3Cornell University, Ithaca, NY, 4Assistant Dean for Admissions and Director, Learning Environment, UCSD SOM, La Jolla, CA, 5Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA, 6University of California San Diego, San Diego, CA, 7Department of Psychiatry, University of California San Diego, La Jolla, CA, 8Department of Medicine, University of California San Diego, La Jolla, CA
Purpose/Objective(s): Immune effector cell-associated neurotoxicity syndrome (ICANS) is a complication of chimeric antigen receptor T-cell therapy (CAR-T), yet risk factors and quantitative neuroimaging criteria remain incompletely characterized. Bridging radiotherapy (BRT) does not appear to increase the risk of ICANS, and some suggest it may even lower the incidence of severe cytokine release syndrome (CRS). We implemented a novel application of a deep learning (DL)-based MRI approach alongside clinical and lab biomarkers to better characterize neurotoxicity after CAR T-cell therapy. We hypothesize that BRT would not increase the risk for ICANS nor deep learning-based MRI changes.
Materials/Methods: We analyzed all patients with non-Hodgkin lymphoma (NHL) or acute lymphoblastic leukemia (ALL) who received commercial CAR-T at a single institution from 2018-2024. ICANS was graded per the American Society for Transplantation and Cellular Therapy. Bridging RT (BRT) was defined as within 1 month pre-CAR-T. Pertinent labs were evaluated pre-RT, post-RT, and prior to infusion. Available post-CAR-T brain MRIs were processed with a 3D U-Net convolutional neural network to quantify T2 FLAIR hyperintensity volumetrics within 6-months post-CAR-T. Independent sample t-tests and Mann Whitney U tests were performed.
Results: Of 163 patients (89% NHL, 11% ALL) prior to CAR-T, 52 had intrathecal chemotherapy, 27 had high dose methotrexate, 24 had prior CNS disease, and 29 had prior RT (24 extracranial only, 5 CNS-directed). Most (106, 65%) received axi-cel (34 tisa-cel, 23 brexu-cel). Most experienced CRS (133, 82%) and ICANS occurred in 73 (45%) at a median of 7 days post-infusion (34, 47% grade 3-4). DL-derived FLAIR was higher in patients with ICANS (24 vs 9 cm3, p=0.02) but not impacted by receipt of BRT (p=0.44; nor by RT within 1 year pre-CAR-T, p=0.97). 12 patients received BRT (median dose 28 Gy [IQR 20-30] in 10 fractions [IQR 6-12]); 7 developed ICANS (1, 14% grade 3-4). Higher BRT dose was associated with a lower risk of ICANS (mean difference 10 Gy, p=0.06). Post-BRT reduction in absolute neutrophils was associated with lower risk of ICANS (reduction by 6.8 for no ICANS vs 0.3 x 1000/m3 for ICANS; p=0.04). Elevated LDH pre-CAR-T was associated with ICANS (mean difference 116 U/L, p=0.01).
Conclusion: Here, we demonstrate a novel application of DL-based MRI quantification of ICANS post-CAR T-cell therapy. This metric may have potential as a quantitative biomarker of ICANS, and we show encouraging data that BRT does not increase post-CAR-T FLAIR nor risk of ICANS. We also demonstrate the potential for RT-induced modulation of neutrophils and LDH, which may have differential implications for risk of ICANS. These findings warrant further investigation and have informed a prospective study, including standardized brain MRI pre- and post-CAR-T, to develop a comprehensive phenotype of neurotoxicity following CAR T-cell therapy.