Yoganand Balagurunathan ^1* Zhouping Wei ¹ Jin Qi ¹ Zachary Thompson ¹ Erin Dean ² Hong Lu ¹ Saran Vardhanabhuti ³ Salvatore Corallo ¹ Jung W Choi ¹ Jenny J Kim ³ Mike Mattie ³ Michael Jain ¹ Fredrick L Locke ¹

• ¹ Moffitt Cancer Center, Tampa, United States
• ² University of Florida, Gainesville, Florida, United States
• ³ Kite Pharma (United States), El Segundo, California, United States

Background: Relapsed and refractory Diffuse large B-cell lymphoma (DLBCL) can be successfully treated with axicabtagene ciloleucel (axi-cel), a CD19 directed autologous chimeric antigen receptor T-cell therapy (CAR-T). Diagnostic image based features could help identify the patients who would clinically respond to this advanced immunotherapy. Purpose: To establish a radiomic image feature-based signature derived from PET/CT, including metabolic tumor burden, which could predict a durable response to CAR-T cell therapy in refractory/relapsed DLBCL. Methods: A retrospective review of 155 patients with relapsed/refractory DLBCL treated with axi-cel CAR-T cell therapy. Patients’ disease involvement was evaluated based on nodal or extranodal sites. A sub-cohort of these patients was assessed that had at least one nodal lesion (n=124), while an overlapping sub-cohort (n=94) had at least one extra-nodal lesion. The lesion regions were characterized using 306 quantitative imaging metrics for PET images and CT images independently. Principal component (PC) analysis was performed to reduce the dimensionality in feature-based functional categories: size (n=38), shape (n=9), and texture (n=259). The selected features were used to build prediction models for survival at one year and tested for prognosis to overall/progression free survival (OS/PFS)using Kaplan-Meier (KM) plot. Results: Shape-PC features of the largest extra-nodal lesion on PET were predictive of 1-year survival (AUC 0.68 [0.43,0.94]) and prognostic of OS/PFS (p<0.018). MTV was an independent predictor with an AUC of 0.74 [0.58, 0.87]. Combining these features improved the predictor performance (AUC of 0.78 [0.7, 0.87]). Additionally, Shape-PC features were unrelated to total MTV (Spearman’s ρ of 0.359, p≤ 0.001). Conclusion: Our study finds shape based radiomic features on PET imaging are predictive of treatment outcome (1-year survival) and prognostic of overall survival. We also found non-size based radiomic predictors that had comparable performance to MTV and provided complementary information to improve the predictability of treatment outcomes.