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)

The therapeutic approach to relapsed or refractory (R/R) diffuse large B-cell lymphoma (DLBCL) in the United States has shifted with the advent of chimeric antigen receptor T-cell (CAR-T) therapies and bispecific antibodies (BsAbs). The National Comprehensive Cancer Network (NCCN) now recommends CAR-T for patients relapsing within 12 months of first-line treatment, with autologous stem cell transplant (ASCT) for later relapses. BsAbs are recommended as monotherapy in third-line and later settings, and in combination with gemcitabine/oxaliplatin in second line CAR-T/ASCT ineligible patients. Recently, a phase III trial with mosunetuzumab plus polatuzumab vedotin in this same indication showed strong progression-free and overall survival, and response rates (Westin et al. 2025). However, the definition of “ASCT ineligibility” is not standardized, and many such patients remain candidates for CAR-T. Despite its efficacy and NCCN recommendation, many eligible patients are not receiving CAR-T due to logistical barriers, clinical challenges, and misunderstandings of eligibility (Kaltwasser 2022; Perales et al. 2025). To examine the impact of NCCN guideline non-adherence on survival outcomes in CAR-T eligible patients with R/R DLBCL. A patient-level simulation model, informed by clinical trial data, projected long-term survival under NCCN guideline-adherent versus nonadherent treatment pathways (Fig 1). The modeled population, based on clinical opinion and literature estimates, included 11,999 R/R CAR-T eligible patients (Kanas et al. 2022), of whom 70% (8,399) were assumed CAR-T eligible. Scenarios simulated up to 75% misallocation (i.e., patients receiving non-NCCN-recommended therapies), reflecting real-world practice (Perales et al. 2025). Survival outcomes and life expectancy were compared across adherence levels. Sensitivity analyses varied age and misallocation rates. With 75% misallocation, 6,299 CAR-T eligible patients received non-recommended therapy annually, yielding a 5-year overall survival of 42% versus 54–57% with NCCN-concordant care (Fig 2), and resulting in suboptimal outcomes overall (Fig 3). Findings were robust in sensitivity analyses. Findings suggest that patients with R/R DLBCL experience improved survival outcomes when treated with CAR-T therapy, supporting its NCCN recommendation. While adherence is associated with better outcomes, clinical or logistical factors may limit guideline adherence. Nonetheless, the analysis highlights the impact of NCCN guideline deviations for CAR-T eligible patients, underscoring the importance of efforts to minimize misallocation and expand access to optimal therapies. Previously presented at the 67th American Society of Hematology (ASH) Meeting (Locke et al. 2025). © American Society of Hematology (2025). Reused with permission.