Extranodal disease is associated with shorter progression‐free survival after cd19‐car t‐cell therapy for relapsed/refractory diffuse large b‐cell lymphoma

Introduction: CD19 CAR T cells induce complete remissions in ≥40% of heavily pretreated patients with relapsed/refractory (r/r) diffuse large B-cell lymphoma (DLBCL). However, a significant proportion of patients progress early after therapy. Thus, to prospectively identify patients most suitable for currently available commercial CAR T-cell products, we assessed the predictive value of clinical and laboratory parameters for PFS in r/r DLBCL patients treated with Axicabtagene Ciloleucel or Tisagenlecleucel at our institution. Methods: Patient, disease, and treatment characteristics of r/r DLBCL patients treated with CD19-CAR T cells were retrospectively assessed. Pre-defined patient and lymphoma characteristics known to confer adverse outcomes in DLBCL were evaluated for their association with PFS by univariable log-rank tests, as well as multivariable stepwise Cox regression analyses. Characteristics encompassed age, refractoriness to first-line therapy, no. of prior therapy lines, tumor volume, presence of bulky disease, presence of extranodal disease (END), Ann Arbor stage, IPI, LDH, and ECOG, all assessed at time of lymphodepletion. Results: As of November 2020, 35 patients have been transfused. Median age was 60 years (range 19-82). ECOG was 0-1 in 23, and 2-3 in twelve patients at CAR T-cell transfusion. Only 3 of the 35 transfused patients (9%) would have met the inclusion criteria of the pivotal clinical trials at all eligibility assessment timepoints (enrolment, apheresis, and time of lymphodepletion). CRS occurred in 32/35 patients (80% CRS °1-2, 11% °3). 16 patients (46%) experienced ICANS (29% °1-2, 14% °3-4, and 3% °5). Response assessment after three months was available for 34/35 patients. Objective response rate was 53%, with complete remission in 15 (44%) and partial remission in three patients (9%). In univariable analyses, apart from primary refractory disease (p = 0.79), all pre-selected characteristics were suggestive to be associated with shorter PFS (p = 0.002-0.061). We therefore included these variables (p < 0.1) in multivariable Cox regression analyses. In the resulting model, presence of END was associated with inferior PFS (Hazard ratio [HR] 11.3, p = 0.02), adjusted for prior therapy lines (HR 1.3, p = 0.06) and ECOG > 1 (HR 2.3, p = 0.12). Other variables did not improve the prognostic validity of the model. Conclusions: We provide real-world evidence that CD19-CAR T cells induce remissions in a substantial proportion of r/r DLBCL patients, despite very frequent non-eligibility for pivotal clinical trials in our patient cohort. Extranodal disease at transfusion was associated with significantly shorter PFS both in uni- and multivariable analyses. To address END-associated resistance mechanisms, we have set up a comprehensive translational research program with specific focus on the nodal vs. extranodal lymphoma microenvironment. Keywords: Aggressive B-cell non-Hodgkin lymphoma, Cellular therapies Conflicts of interests pertinent to the abstract V. Bücklein Consultant or advisory role: Pfizer, Amgen, Kite Gilead Research funding: Novartis, Celgene, Kite Gilead