204 - Improvement in Radiomic Features Following Bridging Therapy is Prognostic for CD19 Chimeric Antigen Receptor (CAR) T-Cell Therapy Outcome

H. G. Hubbeling¹, D. Leithner², E. A. Silverman³, J. Flynn⁴, S. Devlin⁴, G. L. Shah⁵, B. Fregonese³, A. Bedmutha², A. Boardman⁵, P. B. Dahi⁵, R. J. Lin⁵, J. H. Park⁵, M. Scordo⁶, G. Salles⁵, J. Yahalom³, M. L. Palomba⁵, H. Schoder², M. A. Perales⁵, R. Shouval⁵, and B. S. Imber⁷; ¹Radiation Oncology, Univeristy of Pennsylvania, Philadelphia, PA, ²Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY, ³Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, ⁴Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, ⁵Memorial Sloan Kettering Cancer Center, New York, NY, ⁶Department of Medicine, Bone Marrow Transplant Division, Memorial Sloan Kettering Cancer Center, New York, NY, ⁷Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY

Purpose/Objective(s): Greater disease burden is associated with poor outcomes after CAR T. Bridging therapy (BT) is widely used between apheresis and CAR T infusion. We hypothesized that the dynamics of radiomic cytoreduction during the bridging period would be prognostic. Materials/

Methods: Patients with large B-cell lymphoma (LBCL) treated with CD19 CAR T from 2016–2022 were stratified into 5 BT cohorts: 1) no BT 2) radiotherapy (RT) 3) systemic therapy (ST) 4) ST+RT and 5) steroids alone. All patients had a pre apheresis PET. Patients who received BT also had a repeat PET post BT but pre-CAR T infusion. MTV for all scans was analyzed using a semi-automated method with SUV4 threshold. Progression free (PFS) and overall survival (OS) from CAR T infusion were estimated by Kaplan Meier; multivariable analysis was performed using proportional hazards. Patients were stratified by pre-BT disease burden using an absolute MTV cutpoint of 65.4cc established by a maximally selected log-rank statistic for PFS. “High” and “low” MTV were defined as MTV above or below this cutpoint, respectively. To quantify the impact of effective cytoreduction during BT, we then created 4 BT MTV risk groups: a) “pLow” with persistently low MTV pre and post BT, b) “pHigh” with persistently high MTV pre and post BT, c) “Rising” with baseline low MTV which increased to high post BT and d) “Improved” with baseline high MTV which decreased to low post BT.

Results: 191 patients with LBCL (79%), high grade BCL (17%) or primary mediastinal BCL (4%) received CAR T (53% axicabtagene, 22% tisagenlecleucel, 25% lisocabtagene). 47 (25%) received no BT, 104 (54%) had ST, 30 (16%) had RT, 5 (3%) had ST+RT, and 5 (3%) steroids alone. Of the 144 patients who received BT, 56% had any degree of quantitative cytoreduction post-BT and only 49% achieved at least 50% MTV reduction. With median follow-up of 21.6 months post CAR T infusion, 12mo PFS was 65% for RT (CI: 47–90%), 50% for no BT (CI: 37–67%), 39% for ST (CI: 30–50%). Our established MTV cutpoint of 64.5cc and was significantly associated with PFS (median 20 vs. 2.7m, p<0.0001) and OS (unreached vs. 11m, p<0.0001). MTV dynamics during BT were further prognostic with ‘Improved’ patients having significantly better outcomes vs. the ‘pHigh’ patients with PFS of 11 vs. 2.0m and unreached vs. 7.2m OS (p<0.0001 for both). On multivariate analysis, MTV trajectory across the bridging period remained significantly associated with PFS (p<0.001); importantly, there was no significant difference in PFS between Improved and pLow patients (HR for Improved: 2.74, CI: 0.82-9.18).

Conclusion: In our real-world experience, BT was able to reduce disease burden in approximately half of patients. We demonstrate that effective BT can enable initially high disease burden patients to achieve outcomes comparable to low disease burden patients, suggesting BT can convert patients from high to low risk pre-CAR T.