Sequential glofitamab and consolidative ASCT plus CAR-T therapy achieved durable remission in relapsed/refractory DLBCL

To the Editor:

A middle-aged male was initially diagnosed with stage I_A diffuse large B-cell lymphoma (DLBCL) following a biopsy of a neck mass in 2014. After 3 cycles of rituximab plus cyclophosphamide, pirarubicin, vincristine, and prednisone (R-CHOP) chemotherapy, the patient underwent involved-field radiotherapy. The subsequent response evaluation demonstrated complete remission, and the patient successfully completed the planned treatment. In December 2023, the patient presented with an intermittent fever accompanied by bilateral pleural effusion. Atypical cells were identified using pleural fluid analysis. Whole-body positron emission tomography/computed tomography (PET-CT) revealed multiple hypermetabolic lymphadenopathies, diffuse pleural and peritoneal thickening with increased metabolic activity, and multifocal bone marrow hypermetabolism. Bone marrow cytology demonstrated that primitive and immature lymphocytes accounted for 10% of nucleated cells. The patient was ultimately diagnosed with stage IV_B DLBCL (relapsed), a non–germinal center B-cell (non-GCB) subtype, with dual expression of BCL-2 and BCL-6, and an International Prognostic Index (IPI) score of 4. On February 7, 2024, during treatment with a rituximab plus cyclophosphamide, vincristine, and prednisone (R-CVP) regimen, the patient developed severe hypersensitivity reactions and myelosuppression. On February 28, the treatment regimen was switched to polatuzumab vedotin (Pola) combined with vincristine and dexamethasone. However, the hematological parameters did not improve.

Given the patient’s intolerance to conventional chemoimmunotherapy, persistent hematological toxicity, and insufficient disease control, a chemotherapy-sparing immunotherapeutic strategy was considered appropriate. Therefore, the patient was treated with glofitamab, a CD20×CD3 bispecific antibody (BsAb), to redirect T cell-mediated cytotoxicity against CD20-positive lymphoma cells while avoiding further intensive cytotoxic chemotherapy. Although glofitamab-based treatment initially achieved only a partial response, continued glofitamab therapy, with a brief intervention using a combination of polatuzumab vedotin and subsequent addition of orelabrutinib, was associated with a gradual deepening of the response. Importantly, PET-CT evaluation on June 6, 2024, demonstrated complete remission.

Considering the patient’s relapsed stage IV_B disease, high-risk clinical features, and the need for durable disease control after achieving remission, autologous stem cell transplantation (ASCT) combined with chimeric antigen receptor T (CAR-T) cell therapy was planned as a consolidative strategy. Furthermore, CD19-, CD20-, and CD22-targeted CAR-T cells were selected to broaden antigen coverage and potentially reduce the risk of relapse driven by single-antigen loss or antigen heterogeneity. Peripheral blood T-cells and stem cells were collected on May 16 and 21, respectively. Notably, T-cell collection and subsequent production of CD19-, CD20-, and CD22-targeted CAR-T cells were successfully completed despite prior exposure to glofitamab. Pre-transplant conditioning with Carmustine, Etoposide, Cytarabine, Melphalan, and Fludarabine (BEAMF regimen) was initiated on June 24, followed by autologous stem cell infusion of 8.6 × 10⁶/kg on June 30. CD19-, CD20-, and CD22-targeted CAR-T cells were sequentially infused on July 3, 4, and 6. Notably, complete remission was confirmed before CAR-T cell infusion, and CAR-T therapy was administered as a post-remission consolidation rather than as the primary treatment responsible for achieving remission.

Post-infusion, the patient developed recurrent fever (maximum temperature 39.3°C) accompanied by hypoxemia, which improved with low-flow nasal cannula oxygen supplementation. No central nervous system (CNS) symptoms were observed. No clinically significant hepatotoxicity or nephrotoxicity was observed during treatment. These findings were consistent with those of grade 1 cytokine release syndrome (CRS); no immune effector cell–associated neurotoxicity syndrome (ICANS) was observed. Following anti-infective therapy and hematopoietic support, the patient achieved neutrophil and platelet engraftment on day +11 and day +14 post-transplant, respectively. The patient was maintained on long-term methylprednisolone (5 mg/d). Subsequently, the patient developed severe nausea, vomiting, and exfoliative dermatitis, and the dose of orelabrutinib was reduced to 50 mg/d. After symptom resolution, the dose was increased to 100 mg/d until December 2024. However, gastrointestinal symptoms persisted, and esophagogastroduodenoscopy revealed thickening and narrowing of the lower esophagus. Further, endoscopic ultrasonography suggested lymphomatous involvement. Despite multiple biopsies, no malignant cells were detected. PET-CT was repeated on November 18, 2024; June 4, 2025; and October 15, 2025, demonstrating progressively decreasing metabolic activity at the lesion site. This finding was suggestive of an inflammatory process. The patient remains in complete remission at the time of submission of this report (Fig. 1).

Figure 1.

Timeline of treatments and clinical responses following relapse of DLBCL. After relapse in December 2023, the patient initially received R-CVP and polatuzumab-based therapy before switching to glofitamab. Following partial remission, orelabrutinib was added, and complete remission was confirmed by PET-CT before ASCT. Peripheral blood T-cell and stem cell collection were successfully completed after glofitamab exposure. The patient subsequently underwent BEAMF conditioning, ASCT, and sequential infusion of CD19-, CD20-, and CD22-targeted CAR-T cells as consolidative therapy. The patient has remained in durable complete remission during follow-up. ASCT = autologous stem cell transplantation, BEAMF = carmustine, etoposide, cytarabine, melphalan, and fludarabine, CAR-T = chimeric antigen receptor T cell, Dex = dexamethasone, DLBCL = diffuse large B-cell lymphoma, Pola = polatuzumab vedotin, R-CVP = rituximab, cyclophosphamide, vincristine, prednisone.

BsAbs and CAR-T cell therapy have rapidly reshaped the treatment landscape of B-cell lymphomas, highlighting the need to define rational sequencing strategies for these 2 T-cell-engaging approaches.^1,2 However, currently available evidence is insufficient for the establishment of a universally optimal sequence. Crochet et al³ assessed the impact of prior BsAb exposure on subsequent CAR-T cell therapy in 47 patients with R/R DLBCL who had undergone BsAb therapy and then received CAR-T-cell therapy. Despite prior BsAb exposure, CAR-T cells maintained potent anti-tumor activity in relapsed patients, yielding an objective response rate (ORR) of 85%, a complete response (CR) rate of 43%, and a 1-year overall survival (OS) of 55%. The comparable efficacy of treatment among prior BsAb responders and non-responders indicated that resistance to these T-cell engagers operates through distinct mechanistic pathways. A 1:1 propensity score–matched analysis comparing 42 BsAb-naïve patients revealed no efficacy disadvantage in the BsAb-exposed cohort; however, the ORR was significantly higher than that in the control group. The CR rate, 1-year progression-free survival (PFS), and 1-year OS were also higher, with no increased risk of CRS and ICANS. Iacoboni et al⁴ analyzed the safety and efficacy of CAR-T cell therapy in patients with prior BsAb exposure and found that, upon comparing outcomes with patients who were not previously administered BsAb therapy, prior exposure did not compromise the survival outcomes of subsequent CAR-T cell therapy or increase treatment-related toxicity.

In patients with R/R DLBCL who progress after CAR-T cell therapy, CD20×CD3 BsAbs therapy continues to demonstrate significant anti-tumor activity.^5,6 Therefore, sequential consolidation therapy with BsAbs following CAR-T cell therapy may be a novel therapeutic strategy for high-risk patients. In preclinical studies, Brinkmann et al⁷ demonstrated that BsAb consolidation effectively enhanced the immunotherapeutic efficacy of CD19-directed CAR-T cell therapy. In vitro experiments revealed that following the co-culture of CAR-T cells with lymphoma cells, BsAbs significantly enhanced tumoricidal activity by promoting the expansion of both endogenous T cells and CAR-T cells. In vivo experiments further showed that the sequential treatment group exhibited a marked increase in endogenous T cells and CAR-T cells (particularly CD8 + subsets), with significantly prolonged survival compared to the monotherapy group.⁷ Sequential administration of CD19-directed CAR-T cells and CD20-targeted BsAb therapy demonstrated synergistic efficacy against B-cell malignancies. This approach enhances tumor clearance by activating CAR-T and endogenous T cells, thereby addressing two major limitations of CAR-T cell monotherapy: restricted in vivo expansion and antigen escape. Kato et al⁸ further elucidated the underlying mechanism through which BsAbs augment CAR-T cell efficacy, demonstrating that pre-activated bystander CAR⁻ CD8⁺ memory T cells—generated during the CAR-T cell manufacturing process—can persist long-term in vivo and establish immune memory reserves. These cells can subsequently be recruited and clonally expanded during BsAb therapy and serve as critical effector cells that mediate tumor eradication.⁸ These studies elucidate the pivotal role of memory T cells in sequential therapy combining CAR-T cells and BsAbs, and provide a theoretical foundation for optimizing immunotherapeutic strategies for B-cell lymphoma.

A key observation in the present case was that prior glofitamab exposure did not impair subsequent T cell collection or CAR-T cell production. Both BsAbs and CAR-T cells rely on functional T cells, raising concerns that prolonged CD3-mediated stimulation may compromise T cell fitness and affect the feasibility of cellular therapy. CD3-engaging BsAbs induce potent T cell activation and tumor-directed cytotoxicity,⁹ but sustained stimulation may theoretically promote T cell exhaustion. Importantly, emerging evidence suggests that BsAb-induced T cell dysfunction is not necessarily permanent and may be partially reversible after treatment-free intervals.¹⁰ Consistent with this concept, our patient underwent successful peripheral blood T cell collection following glofitamab-based therapy, and CD19-, CD20-, and CD22-targeted CAR-T products were successfully produced and infused. Furthermore, recent clinical studies have demonstrated that prior BsAb exposure does not significantly compromise the efficacy or safety of subsequent CAR-T therapy in large B-cell lymphoma.^3,4 A recent meta-analysis by Sorin et al¹¹ also reported favorable outcomes for CAR-T therapy administered after BsAb exposure. Collectively, these findings support the feasibility of CAR-T consolidation after BsAb-induced remission, although prospective studies are needed to define the optimal sequencing strategy and clarify the roles of T-cell fitness, washout interval, and production feasibility.

Importantly, this case differs from the conventional “bridging-to-CAR-T” paradigm. The patient achieved CR following glofitamab-based therapy before undergoing ASCT and CAR-T cell infusion. Therefore, CAR-T cell therapy was administered as post-remission consolidation rather than as a salvage treatment intended to achieve remission. In this context, the potential role of BsAbs extends beyond tumor debulking and may include optimization of the immune milieu before cellular therapy. Although the precise mechanisms remain incompletely understood, this case suggests that prior BsAb exposure does not preclude the successful collection of functional T cells or durable disease control following triple-target CAR-T therapy. Further studies are needed to determine whether patients who achieve remission after BsAb therapy derive additional benefits from consolidative CAR-T strategies.

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China (Nos. 82070213, 82370196) and Joint Fund for Innovation and Development of Hubei Provincial Natural Science Foundation (2025AFD788).

The figures were created with Adobe Illustrator.

AUTHOR CONTRIBUTIONS

S.Y. and J.W. were the major contributors to the writing of the manuscript and contributed equally to the study. Y.X. and M.Z. supervised the writing of the article and provided the critical review. All the authors have read and approved the final manuscript.