Editor's Note: This ICML conference brought together top lymphoma research experts and clinicians from around the world to extensively discuss the latest advances in the treatment of diffuse large B-cell lymphoma (DLBCL). Anastasios Stathis focused on microenvironmental predictors of chimeric antigen receptor T-cell (CAR-T-cell) therapy, immune evasion mechanisms, and precise treatment strategies combining circulating tumor DNA (ctDNA) with PET imaging, offering more promising individualized treatment options for DLBCL patients.DLBCL Microenvironment Subtyping and Prediction of CAR-T-cell Therapy
DLBCL, an aggressive B-cell non-Hodgkin lymphoma, can achieve long-term remission in some patients through standard treatments like R-CHOP, but approximately 30% of patients still experience treatment failure or relapse. For relapsed/refractory DLBCL, CAR-T cell therapy has shown significant efficacy, yet more than half of patients do not benefit from it. Therefore, identifying predictive biomarkers for CAR-T-cell therapy is crucial. At this conference, a new study delved into three main “prototypes” (lymphomaps) of the DLBCL tumor microenvironment (TME) and their association with CAR-T-cell therapy response. These three prototypes include: Normal Lymph Node-like environment: Characterized by functional rather than exhausted T cells, typically exhibiting a “hot” immune microenvironment. T-cell Exhausted environment: Contains a large number of T cells, but these T cells display an exhausted phenotype and impaired function. Stromal Component environment: Contains fewer T cells and can be considered a “cold” immune microenvironment. Research results indicate that DLBCL patients with a “Normal Lymph Node-like” microenvironment show the best response to CAR-T-cell therapy, while patients with “Exhausted” or “Fibroblast-Monocyte” microenvironments are less likely to achieve complete remission. This aligns with the results from the axicabtagene ciloleucel (axi-cel) treatment arm in the ZUMA-7 trial, further confirming the critical role of TME in predicting CAR-T-cell efficacy. This microenvironment subtyping method provides a valuable predictive biomarker for screening DLBCL patients who might benefit from CAR-T-cell therapy. This finding is highly consistent with previous research data. For example, the Wild Cornell team previously described similar lymphoma subgroups, including germinal center-like, inflammatory, and mesenchymal environments. Data from the University of Chicago also indicated that germinal center-type DLBCL patients with a “hot” microenvironment are most likely to benefit from bispecific antibodies or CAR-T-cell therapy. These studies collectively emphasize the importance of activated T cells and lymph node-like or germinal center-like phenotypes for CAR-T-cell efficacy.
Potential Targets and Strategies to Optimize CAR-T-cell Efficacy
Despite the significant success of CAR-T-cell therapy, its biggest hurdle lies in patients failing to achieve complete remission. Experts point out that current research focuses on how to improve the response rate of CAR-T-cell therapy and further enhance the prognosis of responding patients. Dr. Ruella, in his review, categorized the reasons for CAR-T-cell therapy failure into three main types: CAR-T-cell dysfunction (e.g., insufficient expansion and persistence), intrinsic tumor resistance (e.g., tumor antigen loss), and immunosuppressive microenvironment. The conference report particularly emphasized the importance of the immunosuppressive microenvironment, including restricted CAR-T-cell infiltration into lymphoma areas, the presence of other inhibitory immune cells, suppression of CAR-T-cell function by soluble cytokines and ligands, and intercellular competition. The research further revealed that immune activation or over-activation (leading to exhaustion or monocyte-macrophage expansion) is associated with poor prognosis. In this process, specific cytokines, such as gamma interferon and transforming growth factor-beta (TGF-beta), were found to be closely related to these immunosuppressive phenomena. These cytokines may become potential therapeutic targets for improving the efficacy of CAR-T-cell therapy and other cell therapies in the future. Previously published data from the Stanford group corroborated this view. Their study analyzed biopsy samples from patients who relapsed after CAR-T-cell therapy and found that patients with a large number of CAR-T cells within the tumor but significant inflammation at relapse had a worse prognosis. Similarly, gamma interferon and TGF-beta were identified as key cytokines associated with CAR-T-cell therapy failure. This suggests that by modulating these cytokines, it may be possible to alleviate TME immunosuppression, thereby enhancing the efficacy and persistence of CAR-T cells.
ctDNA and PET Imaging for Precision Guidance in First-line DLBCL Treatment
In addition to optimizing CAR-T-cell therapy, this conference also explored precision strategies for first-line DLBCL treatment. Professor Anastasios Stathis from Bellinzona, Switzerland, presented the preliminary results of the SAKK 38/19 Phase II trial, which aims to evaluate the feasibility of ctDNA and PET imaging-guided treatment for untreated DLBCL patients. Professor Stathis noted that while DLBCL can be cured, the overall survival rate with the R-CHOP regimen has not seen breakthroughs for decades. Molecular analysis helps identify specific DLBCL subtypes, for example, patients with MYD88 L265P and/or CD79B mutations may benefit from the BTK inhibitor acalabrutinib. Simultaneously, combining ctDNA with PET response can further refine risk stratification, enabling more personalized treatment. The SAKK 38/19 trial aims to evaluate the safety and efficacy of acalabrutinib combined with R-CHOP as first-line treatment for DLBCL NOS (not otherwise specified) and to explore PET and ctDNA-guided treatment escalation or de-escalation strategies. The main hypotheses of the study include: Acalabrutinib R-CHOP can improve progression-free survival (PFS) in patients with MYD88 L265P and/or CD79B mutations. For wild-type patients who are PET and ctDNA positive after two cycles of R-CHOP, escalating treatment to acalabrutinib R-CHOP can improve efficacy. For wild-type patients who are PET and ctDNA negative after two cycles of R-CHOP, de-escalating treatment to a total of four cycles of R-CHOP plus two cycles of rituximab will not reduce efficacy. The study enrolled treatment-naive DLBCL NOS patients who had quantifiable ctDNA at baseline. Professor Stathis stated that the trial demonstrated the feasibility of the PET/ctDNA-guided approach in a multicenter setting. The median turnaround time from sample collection to ctDNA results was 9 days, and the median time from informed consent to treatment initiation was 15 days, clearly demonstrating the operational success of this strategy. Baseline ctDNA was detectable in the vast majority of patients (88%), and ctDNA burden was statistically correlated with EP score, bulky disease, B symptoms, metabolic tumor volume (MTV), and total lesion glycolysis (TLG). After two cycles of treatment, most patients achieved a ctDNA-negative state. High baseline MTV and TLG were found to predict positive PET scans at the end of treatment. Professor Stathis concluded that although the number of patients in cohorts B and C (treatment escalation and de-escalation) is currently limited and serves only for hypothesis generation, the SAKK 38/19 trial laid the foundation for introducing precise, response-adaptive strategies in first-line DLBCL treatment. Follow-up will continue to evaluate the efficacy of acalabrutinib in mutated patients.
Expert Opinions and Clinical Application Prospects
The latest research advances presented at the ILRC 2025 conference indicate that DLBCL treatment is moving from a “one-size-fits-all” approach towards deep individualization and precision. On one hand, a deeper understanding of the tumor microenvironment allows us to better predict patient response to CAR-T-cell therapy. Patients with a “Normal Lymph Node-like” microenvironment are most likely to benefit, providing an important predictive biomarker for selecting suitable CAR-T-cell therapy patients in clinical practice. Concurrently, the elucidation of the immunosuppressive microenvironment, especially the roles of gamma interferon and TGF-beta, points the way for developing new combination treatment strategies to overcome CAR-T-cell resistance. On the other hand, the results of the SAKK 38/19 trial reported by Professor Stathis demonstrate the immense potential of combining ctDNA and PET imaging to individualize treatment intensity in first-line DLBCL. This response-adaptive strategy, based on liquid biopsy and imaging assessment, is expected to provide de-escalated treatment for low-risk patients without sacrificing efficacy, thereby reducing side effects; simultaneously, it aims to offer more intensified treatment for high-risk patients, increasing cure rates. Overall, these innovative advances not only bring more options to the clinical diagnosis and treatment of DLBCL but also highlight the significant contributions of global research teams in tumor immunology and precision medicine. In the future, DLBCL treatment will increasingly focus on biomarker-driven patient selection, the combination of novel immunotherapies and cell therapies, and the application of cutting-edge technologies like liquid biopsy in comprehensive management, allowing more patients to achieve optimal treatment outcomes. Submitted by / Interview source: Oncology Frontier – Oncology Express
