Disease Control Affects Outcomes of Patients with Relapsed Diffuse Large B Cell Lymphoma Treated with CD19 Directed CAR T-Cell Therapy

Introduction: CAR T-cell therapy against CD19 (CART-19) has become the standard of care for pts with primary refractory or relapsed diffuse large B-cell lymphoma (DLBCL) in the second and third line settings. The use of CART-19 therapy improves clinical outcomes compared to standard chemotherapy-based salvage regimens. However, DLBCL pts with disease progression following CART-19 therapy have poor overall survival (OS). Identifying biomarkers associated with improved outcomes can inform the optimal timing for CART-19 implementation. We report the impact of disease control at the time of CART-19 therapy on clinical outcomes.

Methods: We conducted a retrospective study of relapsed/refractory (r/r) DLBCL treated at our institution with commercial CART-19 products between 4/2018 and 2/2023. Clinical, pathologic, and pharmacologic (type of bridging, lymphodepletion) characteristics were collected. PET/CT imaging prior to CART-19 therapy was used to determine pretreatment disease status (CR, PR or chemo-resistant disease). Treatment outcomes included best response to CART-19 therapy, frequency and severity of adverse events (CRS or ICANS), PFS and OS. Associations with clinical outcomes were evaluated using logistic regression models by Firth's method. Odds ratios [OR] and 95% CI were obtained. OS and PFS were summarized via Kaplan-Meier method; estimates of median 6- and 12-m rates were obtained with 95% CI. Associations with OS and PFS were evaluated using Cox regression models. Analyses were conducted in SAS v9.4; significance level 0.05.

Results: 94 pts were included in this study. Median age was 64 yrs [range 22-81]; 66% were male; 90% Caucasian. Comorbidity index [assessed by HCT-CI] was 2+ in 51.7% pts. Median time from CART-19 discussion to re-infusion was 3.1 m [range 0.2-80 m]. Based on PET/CT scan assessment prior to CART-19 therapy, 9.6% pts achieved CR to bridging, 19% had PR, 64% were resistant (SD or PD). High risk IPI score was present in 23.4%. 14.9% pts had prior ASCT. FluCy lymphodepletion was given in 83% pts. Best response to CART-19 therapy: CR 45.7%, PR 12.8%. PD and SD observed in 28.7% and 3.2% of pts respectively. At median follow up of 36.5 m, median PFS was 6.7 m [95% CI 3.8-18.4]; median OS was 16.0 m [95% CI 9.8-30.6]. Any grade CRS occurred in 61.7%; Grade 3+ CRS 8.5%. Any grade ICANS occurred in 45.7%; Grade 3+ ICANS 16%. Second cancers occurred in 9.6%; 4 pts developed MDS. Adequate disease control at the time of CART-19 therapy was associated with improved ORR (P=0.028). Lack of disease control determined by PET/CT at the time of CART-19 infusion correlated unfavorably with OS and PFS. Pts with chemoresistant disease at the time of CART-19 therapy had a median survival of 7.2 m [HR 2.07 (0.74-5.79)] compared to pts in CR or PR at the time of CART-19 (median survival 27.9 m and 26.8 m respectively, p=0.023). Any grade ICANS [HR 2.46 (1.43-4.23) p=0.001], and ICANS grade 3+ [HR 3.38 (1.82-6.26) p=<0.001] were associated with an increased risk of death. On analysis of associations with PFS, lack of response to CART-19 therapy [SD/NR HR 15.05 (3.92-57.74) p=<0.001], any grade ICANS [HR 1.74 (1.06-2.87) p=0.029], and ICANS grade 3+ [HR 2.52 (1.38-4.59) p=0.003] were associated with increased risk for progression. Grade 3+ ICANS was adversely correlated with PFS in CR pts [HR 3.81 (1.22-11.90) p=0.021] and CR + PR pts. Several factors were associated with increased risk for toxicities. Lack of disease control at the time of CART-19 therapy, [OR 3.30 (0.79-13.79) p=0.003], development of ICANS, any grade [OR 4.08 (1.64-10.14) P=0.003] or grade 3+ ICANS [OR 7.71 (1.28-46.27) p=0.025] correlated with the development of any grade CRS. ICANS of any grade [OR 6.92 (1.12-42.79) p=0.037] and ICANS grade 3+ [OR 6.57 (1.50-28.77) p=0.013] correlated with Grade 3+ CRS. Lack of disease control status [OR 4.18 (0.85-20.55) p=0.011], any grade CRS [OR 4.08 (1.64-10.15) p=0.003], and grade 3+ CRS [OR 6.92 (1.03-46.64) p=0.047] were associated with an increased incidence and severity of ICANS.

Conclusions: Disease control assessed by functional imaging at the time of CART-19 therapy for r/r DLBCL may be an independent predictor of CART-19 therapy outcomes. In this cohort, adequate disease control prior to CART-19 therapy was associated with improved ORR, PFS and OS. Optimizing disease control prior to cellular therapy may ultimately improve clinical outcomes while mitigating cellular therapy related toxicity.