Impact of holding therapy prior to apheresis on outcomes following CD19 CAR-T cell therapy in large B-cell lymphoma

Abstract

Abstract Background: Patients with relapsed or refractory large B-cell lymphoma (LBCL) often experience delays between CAR-T decision and leukapheresis due to logistical or clinical limitations.While bridging therapy—administered between apheresis and lymphodepletion—has been well studied, there are limited data on “holding therapy,” defined as any systemic therapy or radiotherapy administered prior to apheresis to stabilize disease. To address this knowledge gap, we evaluated the impact of holding therapy on outcomes following CD19-directed CAR-T cell therapy. Methods: We conducted a retrospective study of adults with LBCL who underwent leukapheresis for CD19 CAR-T (2022–2025) at our center. CAR-T decision was defined as the earliest documentation in the electronic medical record indicating intent to proceed with CAR-T, including referral from an outside center, treating physician's decision, or multidisciplinary discussion following relapse. Holding therapy was defined as any systemic therapy or radiotherapy (excluding steroids) administered after CAR-T decision and before apheresis. Patients were categorized by treatment exposure: holding only, bridging only, both, or neither. A separate salvage cohort included patients initially intended for autologous HCT who transitioned to CAR-T. Primary endpoints were disease status at apheresis and progression-free survival (PFS) following CAR-T infusion. Secondary endpoints included overall survival (OS) and best response within 1-year post-infusion. Cox proportional hazards models were constructed to evaluate associations between clinical variables and survival outcomes. Results: A total of 159 patients were included (57% liso-cel, 38% axi-cel, 4.4% tisa-cel); 82% received CAR-T as second-line therapy. Regarding treatment exposure, 52% received both holding and bridging therapy (n=82), 35% bridging only (n=55), 5.7% holding only (n=9), 5% salvage therapy with bridging (n=8), 0.6% salvage only (n=1), and 2.5% received neither (n=4). Overall, 58% received holding therapy and 92% received bridging. Among those who received holding therapy, the most common regimens were chemotherapy-based with 51% (n=46), including GemOx, ICE, or bendamustine with rituximab. Polatuzumab vedotin (Pola) was used in combination with rituximab (23%, n=21) or rituximab and chemotherapy (15%, n=14). Other regimens less represented included ibrutinib (n=8), checkpoint inhibitors (n=2) and glofitamab (n=1). Radiotherapy alone was administered as holding in 20% of cases. Among those who received bridging therapy alone, 80% (n=44) received chemotherapy-based regimens, 9% (n=5) received chemotherapy plus Pola-R, and 9% (n=5) received Pola-R without chemotherapy. Median time from holding therapy start to apheresis was 31 days (IQR 35.5). From apheresis to infusion: 39 days (IQR 10.8) with holding/bridging, 35 (IQR 9.0) with holding, and 40 (IQR 13.5) with bridging. At apheresis, the overall response rate among patients receiving holding therapy was 38%. Best response within 1-year post-CAR-T was higher in the holding/bridging group (CR/PR 94%) compared to bridging only (75%) (p=0.0039). One-year PFS was 65% for holding/bridging, 71% for holding only, and 52% for bridging only, log-rank p=0.06, and 1-year OS rates of 78%, 71%, and 78%, p=0.6, respectively. In univariable analyses, elevated lactate dehydrogenase (LDH) prior to lymphodepletion (HR 1.98, p=0.01, FDR=0.051), Karnofsky performance status (KPS) <90 (HR 2.07, p=0.019, FDR=0.078), and higher SUVmax on PET prior to infusion, after holding or bridging (HR 1.03, p=0.002, FDR=0.017) were associated with inferior PFS. CAR-T product, age, transformed histologies, and number of prior treatment lines were not significantly associated with outcomes. Conclusion: In this real-world study, holding therapy prior to apheresis did not compromise the efficacy of CD19 CAR-T therapy in LBCL. Holding, often followed by bridging, was frequently used to maintain disease control and was associated with favorable responses when combined with bridging. Survival outcomes were comparable across groups, with prognosis primarily driven by biological risk factors (LDH prior to lymphodepletion, KPS, metabolic activity). These findings support the feasibility of holding strategies during the pre-apheresis interval, especially in patients needing disease control while awaiting apheresis.