Estimating the survival impact of not receiving CAR T-cell (CAR T) therapy when eligible in patients with relapsed or refractory (R/R) diffuse large B-cell lymphoma (DLBCL) in the United States (US)

Abstract Introduction: The therapeutic approach to relapsed or refractory (R/R) diffuse large B-cell lymphoma (DLBCL) in the United States has shifted with the emergence of chimeric antigen receptor T-cell (CAR T) therapies and bispecific antibodies (BsAbs). As a result, the National Comprehensive Cancer Network (NCCN) now recommends BsAbs monotherapy for use in third-line and later settings of R/R DLBCL, and recommends CAR T for patients who relapse within 12 months following first-line treatment, with high-dose chemotherapy with autologous stem cell transplant (ASCT) reserved for later relapses (12 months) (National Comprehensive Cancer Network Inc. 2025). NCCN guidelines also now recommends the use of glofitamab, in combination with gemcitabine and oxaliplatin (GemOx), for patients ineligible for CAR T and ASCT in R/R 2L DLBCL. In the recent phase III trial involving patients with R/R DLBCL who were ineligible for ASCT, another BsAb mosunetuzumab, in combination with polatuzumab vedotin (mosun-pola), led to significantly higher overall response rates and showed a trend toward improvement in both progression-free and overall survival compared to rituximab in combination with GemOx (Westin et al. ICML 2025). The definition of “ineligible for ASCT” is not standardized, and many of these patients do meet eligibility for CAR T. Despite the favorable efficacy, many patients who meet eligibility criteria for CAR T are not receiving it in line with NCCN guidelines, due to a range of logistical and clinical challenges and potentially a general misunderstanding or lack of clarity around CAR T eligibility (Kaltwasser Target Oncol 2022; Perales et al. Transplant Cell Ther 2025). Objective: To examine the impact of non-adherence to NCCN treatment guidelines on survival outcomes in patients with R/R DLBCL who are eligible for CAR T-cell therapy. Methods: We constructed a patient-level simulation model, utilizing clinical trial data across the treatment pathway in R/R DLBCL, to project long-term survival for patients eligible for CAR T-cell therapy according to NCCN recommendations (National Comprehensive Cancer Network, Version 2.2025). The population size of the model is based on clinical opinion and literature estimates; it assumes that 70% of the 11,999 treatment eligible R/R patients (Kanas et al. 2022) are CAR T eligible. The model simulated a series of scenarios in which up to 75% of eligible patients were misallocated (i.e., received therapies not recommended by NCCN guidelines), reflecting real-world treatment patterns (Perales et al., Transplant Cell Ther 2025). Survival outcomes and life expectancy were extrapolated and compared between the scenario of full adherence to NCCN guidelines and partial adherence (up to 75% misallocation). Sensitivity analyses were performed to evaluate the effects of varying degrees of treatment misallocation. Results: Of the estimated 8,399 CAR T-eligible patients experiencing relapse after first-line therapy annually in the US, applying the maximum 75% misallocation rate considered, resulted in as many as 6,299 patients projected to receive non-recommended treatments. Those treated outside guideline-recommended pathways were projected to have a 5-year overall survival rate from initiation of 2L therapy of 42%, compared to 54-57% among patients managed according to guidelines. This misallocation corresponded to an average reduction in life expectancy of 25 months per patient, leading to an estimated 783 additional deaths within five years. In sensitivity analyses assuming a lower misallocation rate of 15%, the number of misallocated patients was projected to be 1,260, with approximately 157 excess deaths over the same period. Conclusions: Our findings suggest that patients with relapsed/refractory DLBCL experience improved survival outcomes when treated with CAR T-cell therapy, reinforcing its designation as a category 1 recommendation by the NCCN. While adherence to these guidelines is associated with better outcomes, clinical or logistical factors may understandably limit the ability to follow recommended treatment pathways at times. Nonetheless, the analysis highlights the impact of deviations from NCCN recommendations for CAR T-eligible patients, underscoring the importance of efforts to minimize misallocation and improve access to optimal therapies.