Radiation Therapy and Monocyte Activation in Large B-Cell Lymphoma Patients Treated with CAR T-Cell Therapy

Introduction. While the majority of large B-cell lymphoma (LBCL) patients receiving autologous anti-CD19 chimeric antigen receptor T-cell therapy (CART) require bridging treatment between leukapheresis and initiation of lymphodepleting chemotherapy, no optimal regimen has been identified. We previously demonstrated that radiation therapy (RT) is an effective bridging strategy and may be preferred over systemic therapy in select patients. However, the biological effects driving the benefits of peri-CART RT remain unknown. Methods. Peripheral blood samples were collected before initiation and after completion of RT in LBCL patients treated with standard of care CART. RT was delivered to involved sites of disease peri-CART. Bulk RNA sequencing was performed on all samples for identification of immune cell subsets, and immune signatures and portraits. Kassandra cellular deconvolution calculation was used to quantify changes in immune cell composition. Immunotypes (G1-G5) were defined as described by Dyikanov et al. (Cancer Cell, 2024). Response was assessed according to 2014 Lugano criteria. Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) were graded according to ASTCT criteria. To assess changes in the parameters before and after RT, a paired Wilcoxon test was used. The Mann-Whitney test was used to compare the difference between groups of responders and nonresponders on day 30, groups with and without high CRS, and groups with and without high ICANS. Results. Ten patients with matched peripheral blood samples were included in the study: 9 were treated with axicabtagene ciloleucel and 1 with lisocabtagene maraleucel. Eight patients were treated with RT between leukapheresis and CART infusion, and 2 were treated 6 and 9 weeks after CART infusion to areas of limited residual disease. Median RT dose was 22 Gy (range 16-40 Gy) delivered over a median of 9 fractions (range 4-16). Seven patients were treated with RT targeting all sites of active disease and 3 were treated focally to select sites. In samples collected after completion of RT, a significant increase in monocytes (area under the curve [AUC] 0.86, p=0.002), dendritic cells (AUC 0.76, p=0.05) and conventional dendritic cells (AUC 0.61, p=0.05) was observed when compared to pre-RT samples. A significant increase in regulatory/cytotoxic NK ratio (AUC 0.66, p=0.03) and a significant decrease in T cells (AUC 0.85, p=0.002), Th1 and Th17 in particular, was also observed. When custom blood myeloid-related signatures were analyzed, a significant increase in genes associated with monocyte proliferation (POFUT1, AP1S2, SLC46A2, CSF1R, MS4A7, MS4A14; AUC 0.81, p=0.05) and monocyte cytotoxicity (CD33, TNFSF13, MS4A6A, LILRA1, CSF1R, SLC46A2, MS4A7, MS4A14, AP1S2, ALDH1A1; AUC 0.84, p=0.03) was observed after RT. Finally, when analyzing custom blood immunotypes, incorporating 34 functional immune populations and 5 distinct immune profiling subtypes, a significant increase in the G5-suppressive immunotype (enriched in monocytes and granulocytes) was observed after RT with 5 (50%) patients with G5 immunotype post-RT compared to 1 patient pre-RT (AUC 0.76, p=0.02). Patients with the G5-suppressive immunotype after RT were more likely to develop grade 0-1 CRS than grade 2-3 and to experience a complete response on day 30 (AUC 0.94, p=0.09); no association between the post-RT G5-suppressive immunotype and ICANS was observed. Discussion. RT may increase circulating monocytes, and monocyte proliferation and cytotoxicity in CART-treated LBCL patients, with shift to an immunotype associated with favorable response and toxicity outcomes. Further study and validation with larger cohorts is needed. To develop optimal bridging strategy and improve outcomes after CART,studies aimed at further investigating the interaction between RT and myeloid cells are needed.