Old problems, new tools: ZR2 in older patients with DLBCL Free

In this issue of Blood, Xu et al¹ present a phase 2 study investigating the chemo-free regimen of zanubrutinib, rituximab, and lenalidomide (ZR2) in newly-diagnosed older (≥75 years) patients with diffuse large B-cell lymphoma (DLBCL) unfit for anthracycline-based immunochemotherapy. They demonstrate that this regimen is effective and well-tolerated, and that responses were associated with specific immune profiles.

The median age at diagnosis of DLBCL is 65 years, with ∼30% of cases occurring in patients >75 years, a proportion expected to rise with an aging population. For fit older patients without cardiac comorbidities, the standard of care remains 6 cycles of R-CHOP-21. However, both efficacy and tolerability for this immunochemotherapeutic regimen decline with advancing age and frailty.² In the oldest and most vulnerable patients, treatment-related mortality exceeds 10%, and necessary dose reductions are associated with inferior outcomes. In unfit or very elderly patients (≥80 years), a dose-reduced regimen (R-miniCHOP) is commonly employed, although outcomes remain modest with 2-year overall survival (OS) rates near 59% and considerable treatment-related mortality.³ In patients unable to receive anthracyclines, substitutions such as etoposide offer limited benefit, and many frail patients derive minimal advantage from chemotherapy.

Compared with younger patients, DLBCL in older adults appears biologically distinct, more often exhibiting the ABC subtype, Epstein-Barr virus-positive disease, age-related mutations, and an immunosuppressive tumor microenvironment (TME) characterized by T-cell exhaustion and senescence.^4,5 Meanwhile, increasing evidence highlights the TME as a key determinant of treatment response. Using a multiomic approach, Tumuluru et al⁶ identified a “germinal center B cell (GCB)-hot” TME, enriched for GCB-cell of origin (COO) and immune-related gene signatures, as predictive of favorable response to the bispecific antibody mosunetuzumab (CD20×CD3). Similarly, Li et al⁷ described a “lymph node archetype” TME, with intact T-cell populations and supportive stroma as favorable for CD19 CAR T-cell therapy, whereas a fibroblast- and macrophage-rich, T-cell–depleted “cancer-associated fibroblasts and macrophages (FMAC) archetype” correlated with inferior outcomes to frontline R-chemotherapy. These analyses are especially interesting for the development of chemo-free regimens for older patients.

Currently, several first-line chemo-free regimens in older patients with DLBCL are being evaluated. Mosunetuzumab in combination with the antibody-drug conjugate polatuzumab-vedotin results in an overall response rate (ORR) of 80% and complete response (CR) rate of 61%.⁸ In the phase 2, single-arm FIL_ReRi study, lenalidomide plus rituximab in patients with DLBCL ≥70 years was feasible, but responses were moderate (ORR, 50.8%; CR, 27.7%).⁹ Yet, prospective data on chemo-free regimens for DLBCL remain sparse. The ZR2 regimen investigated by Xu et al represents a focused effort to meet this need.

The rationale for the ZR2 regimen lies in the combination of 3 agents with different mechanisms: a Bruton tyrosine kinase inhibitor targeting B-cell receptor signaling, the immunomodulatory drug lenalidomide with both antitumor (downregulating the NF-κB pathway) and immune-stimulating properties, and the anti-CD20 monoclonal antibody rituximab. This single-center, phase 2 trial by Xu et al, enrolled 40 treatment-naive patients with DLBCL with a median age of 78 years (range 75-91). Patients were unfit or frail as defined by comprehensive geriatric assessment (CGA) or refused systemic chemotherapy (n = 6) and received 6 cycles of ZR2. The CR rate was 65%. The 2-year progression-free survival (PFS) and OS were 67.1% (95% confidence interval [CI], 50.1-79.4) and 82.4% (95% CI, 66.5-91.2), respectively. These results compare favorably to historical data with R-miniCHOP (2-year PFS and OS of 47% and 59%, respectively)³ or ibrutinib plus R-miniCHOP (2-year PFS and OS of 60% and 68%, respectively).¹⁰ ZR2 demonstrated an acceptable safety profile. Grade 3 to 4 neutropenia occurred in 35% of patients; other adverse events were modest.

The study by Xu et al adds to the ongoing development of chemotherapy-free frontline approaches for older and frail patients with DLBCL. Although encouraging, the clinical results should be interpreted with caution given single-arm design, limited sample size (including some fit older patients who refused chemotherapy), and inherent heterogeneity in the translational data. Nonetheless, ZR2’s outpatient-friendly, chemotherapy-free profile with efficacy that does not appear inferior to R-miniCHOP and manageable toxicity make it a strong candidate for further evaluation. A randomized trial evaluating ZR2 vs R-miniCHOP in unfit or frail older patients with newly-diagnosed DLBCL (NCT05179733) is currently underway.

An important contribution of this study is the immunologic profiling to elucidate response mechanisms. Single-cell RNA sequencing of lymphoma tissue in a small subset of patients and peripheral T-cell receptor clonotyping after therapy revealed that ZR2 responses correlated with features of an activated TME, including high HLA class I and class II expression, enrichment for type 1 conventional dendritic cells, and subsequent clonal expansion of CD8⁺ T cells in the microenvironment of responders. Moreover, long-term persistence of tumor-reactive clones in patients with durable remissions was observed, suggesting durable antilymphoma immunity. These findings align with emerging literature emphasizing the prognostic and therapeutic relevance of the TME. Notably, response depth was not associated with genetic subtype, cell of origin, International Prognostic Index score, or “double expressor” status. Although this may reflect the small sample size, it also indicates that immune contexture, rather than genomic classification, may better predict response in this setting.

Despite progress, key challenges remain in identifying the optimal chemo-free strategy for older patients with DLBCL. Although Eastern Cooperative Oncology Group performance status and CGA are commonly used to assess frailty, more precise frailty assessment alongside treatment impact on quality of life are needed to improve patient selection in future chemo-free trials. Moreover, responses to ZR2 and bispecific antibodies appear to be associated with an activated TME, yet the TME in older individuals is often shaped by immune senescence. Although CAR T-cell therapy seems to overcome an immune senescent TME, it remains unfeasible for the very old or frail. Whether the senescent TME can be modulated to enhance responsiveness to therapies such as ZR2 or bispecifics warrants further investigation. In parallel, the development of clinically applicable biomarkers to define TME phenotypes predictive of response is essential. With its integration of immunologic profiling in an older cohort, the study by Xu et al, alongside the work of Tumuluru and Li, contributes to this effort.^6,7 Although the observed differences are modest, this work underscores the importance of embedding translational research within clinical trials to refine response prediction and guide the integration of chemo-free strategies in DLBCL.

Conflict-of-interest disclosure: M.E.D.C. received research support from AbbVie, Genmab, Bristol Myers Squibb, and Gilead; and has an advisory function at AbbVie, Novartis, and Incyte. A.V.d.J. declares no competing financial interests.