Immune remodeling enables sustained EBV clearance in B-cell and plasma cell malignancies after anti-CD19/BCMA CAR-T therapy: A post-hoc analysis

Abstract

Abstract Background Epstein-Barr virus (EBV) reactivation has increasingly been recognized as a poor prognostic indicator, as it may progress to severe EBV-associated diseases, including EBV+ lymphomas, EBV-driven HLH, and EBV+ PTLD. Since EBV predominantly infects B lymphocytes, depletion of B cells using anti-CD20 monoclonal antibodies has been adopted to reduce EBV load; however, it fails to eliminate EBV in T or NK cells, which remains a major therapeutic challenge. In this study, we found that anti-CD19/BCMA CAR-T therapies not only cleared EBV-infected B cells but also T/NK cells in B-cell or plasma cell hematologic malignancies, possibly due to immune remodeling during cytokine release syndrome (CRS). Method A post-hoc analysis was conducted of patients with B-cell or plasma cell malignancies who had received CD19- or BCMA-CAR-T cell therapy at our institution (2018–2024; ChiCTR1800015575, ChiCTR1800017404, ChiCTR2100053871, ChiCTR2100052009, NCT04532268). Baseline EBV reactivation status, defined as detectable EBV DNA in fresh whole-blood samples, was evaluated using clinical-use real-time qPCR (Daan Gene) with a detection limit of 2.0×10² copies/ml. Longitudinal EBV DNA levels and clinical outcomes were systematically recorded to evaluate viral clearance dynamics. To further delineate the cellular reactivation of EBV and CAR-T therapy response, cryopreserved baseline PBMCs before lymphodepletion were thawed, with T, B, and NK cell subsets sorted. EBV DNA in each was quantified using research-use-only qPCR (Sansure Biotech) with a detection limit of 4.0×10² copies/ml. Paired PBMC samples (baseline and 2 months post-infusion) were processed for scRNA-seq and TCR-seq (10x Genomics) to explore potential immune remodeling mechanisms. Results Baseline EBV reactivation was identified in 20 patients (B-ALL, n=10; MM, n=6; lymphoma, n=4) receiving CD19 (n=12), BCMA (n=6), or CD19/CD22 (n=2) sequential CAR-T therapy. Median age of this cohort was 56 (range 27–77), with 50.0% having prior transplantation. Baseline plasma EBV DNA levels ranged from 3.65 × 10² to 3.84 × 10⁴ copies/ml. Nineteen patients (95%) achieved undetectable EBV DNA post-CAR-T. All patients experienced CRS, and grade 3–4 CRS occurred in 2 patients (10.5%). The remaining patient (5%) with persistent EBV DNA showed no CAR-T cell expansion, suggesting therapeutic inefficacy. In light of EBV negativity observed, baseline T/B/NK subsets were subsequently analyzed in 17 patients to explore mechanisms. Among 8 with EBV in all subsets, 75% showed the highest levels in B cells, and 25% in T cells, suggesting potential clearance of non-CAR-target EBV-reactivated cells in sustained systemic clearance. During further elucidation of how CAR-T therapy achieved beyond B-cell elimination, scRNA/TCR-seq in 5 patients (B-ALL, n=4; MM, n=1) with paired PBMCs showed decreased overall proportion of exhausted T cells and increased proportion of CD8⁺ T cells with an effector-related phenotypic shift (T_CM to T_EM) post-CAR-T, especially in B-ALL patients. EBV-specific clonotypes expanded after therapy (CALRGGGADGLTF_CASSSGGNEQFF, NA_CASSSGGNEQFF, CAEKTSYDKVIF_CASSESYTQYF), indicating CAR-T-related restored endogenous EBV-specific T cell antiviral responses.Interactome analysis demonstrated that EBV-specific T cells consistently exhibited the strongest interactions with CD8⁺ effector T cells and NK cells, with post-treatment signaling changes favoring enhanced effector function. Mechanistically, these T cells received weaker apoptotic signals (FASLG/TRAIL↓) and engaged in more robust CXCL12-CXCR4 signaling with CD8⁺ effectors, facilitating recruitment to EBV-reactivated sites. Concurrently, NK-derived LTA-TNFRSF14 signaling increased, supporting T cell survival and memory. Myeloid cells shifted from a pro-apoptotic state (FASLG/TRAIL↑) to a costimulatory and supportive state (CD40↑, LIGHT↑, GAS6↑, IL6↑). These immune remodeling events during CRS fostered a permissive microenvironment that enhanced EBV-specific T cell function and contributed to durable EBV clearance. Conclusion CAR-T therapy targeting B or plasma cells effectively cleared EBV reactivation in hematologic malignancies. Beyond B-cell depletion, CAR-T induced systemic EBV clearance, likely mediated by immune remodeling during CRS. Enhanced EBV-specific T cell responses and supportive cellular interactions suggest CAR-T as a promising strategy to overcome EBV-related complications in this population.