Varnimcabtagene autoleucel (ARI‐0001) in relapsed or refractory mantle cell lymphoma Authors: Nil Albiol, Nuria Martínez‐Cibrián, Amanda Isabel Pérez, Sergio Navarro‐Velázquez, Guilherme Muzzi, Aina Oliver‐Caldes, Marta Español‐Rego, Leticia Alserawan, Mercedes Montoro‐Lorite, Carla Ramos, Pilar Ayora, Ana África Martin‐López, Almudena Cabero, Estefanía Pérez‐López, Alejandro Martín García‐Sancho, Tycho Baumann, José María Sánchez‐Pina, Anabelle Chinea, Paula Fernández‐Caldas, Mariana Bastos‐Oreiro, Mi Kwon, Luis Gerardo Rodríguez‐Lobato, Laura Magnano, Andrea Rivero, Juan G. Correa, Pablo Mozas, María Sánchez‐Castañón, Juan J. Mata‐Molanes, Paola Charry, Jose R. Álamo, Xavier Setoaín, Sonia Rodríguez, Eulalia Olesti, Joaquín Sáez‐Peñataro, Carlos Fernández de Larrea, Álvaro Urbano‐Ispizua, Armando López‐Guillermo, Julio Delgado, E. Azucena González‐Navarro, Manel Juan Published: 2026 ABSTRACT AbstractVarnimcabtagene autoleucel (var‐cel) is an autologous CD19‐directed chimeric antigen receptor (CAR) T‐cell therapy administered using an adaptive, fractionated dosing strategy. We conducted a multicenter retrospective study to evaluate the safety and efficacy of var‐cel in patients with relapsed or refractory (R/R) mantle cell lymphoma (MCL) treated at four Spanish centers, including patients enrolled in the CART19‐BE‐01 trial and a subsequent compassionate use program. Thirty patients underwent leukapheresis; 27 received var‐cel following lymphodepletion and fractionated infusion in three escalating doses (0.1 to 5 × 106 CAR T cells/kg). Any‐grade cytokine release syndrome (CRS) and immune effector cell‐associated neurotoxicity syndrome (ICANS) occurred in 93% and 4%, respectively, with Grade ≥3 CRS and/or ICANS observed in a single patient (4%). The overall response rate among infused patients was 89% (95% CI, 71–98), including 78% complete responses (95% CI, 58–91). Median duration of response and progression‐free survival were 17.1 and 15.4 months, respectively. With a median follow‐up of 21.7 months, median overall survival was 35.2 months. In this multicenter academic experience, var‐cel demonstrated high antitumor activity with low rates of severe CRS and ICANS, supporting its use within a coordinated national network. INTRODUCTION Mantle cell lymphoma (MCL) is an aggressive B‐cell malignancy characterized by a relapsing clinical course.1, 2, 3 Despite the therapeutic advances achieved with Bruton's tyrosine kinase inhibitors (BTKi), 4 patients who relapse after BTKi exposure continue to have poor outcomes, highlighting a critical unmet need for effective and durable treatment strategies. 3 , 5 CD19‐directed chimeric antigen receptor (CAR) T‐cell therapies have significantly reshaped the treatment landscape of relapsed/refractory (R/R) MCL. Brexucabtagene autoleucel (brexu‐cel), an autologous CD28‐costimulated CAR T‐cell product, has demonstrated high response rates and durable remissions in the post‐BTKi setting. 6 , 7 Similarly, lisocabtagene maraleucel (liso‐cel) and tisagenlecleucel (tisa‐cel) have shown encouraging activity, including in high‐risk subgroups such as patients with TP53 aberrations and blastoid morphology. 8 , 9 However, brexu‐cel is associated with substantial toxicity. Rates of Grade ≥3 cytokine release syndrome (CRS) and immune effector cell‐associated neurotoxicity syndrome (ICANS) have been reported at 15% and 31%, respectively,6, 10, 11, 12, 13 with non‐relapse mortality (NRM) approaching 10% n some real‐world series. 10 In contrast, 4‐1BB–based CAR T‐cell products, including liso‐cel and tisa‐cel, have been associated with a more favorable safety profile, with lower incidence and severity of CRS and ICANS. 8 , 9 Varnimcabtagene autoleucel (var‐cel, or ARI‐0001) is an autologous, 4‐1BB–based, CD19‐directed CAR T‐cell therapy developed and manufactured at Hospital Clínic de Barcelona (HCB). It is approved in Spain under the Hospital Exemption pathway for adult patients with R/R B‐cell acute lymphoblastic leukemia (B‐ALL), and represents a hospital‐based alternative for patients with R/R MCL who are ineligible for, or lack access to, commercial CAR T‐cell therapies. 14 Patients were enrolled within the pilot CART19‐BE‐01 trial or under a subsequent compassionate use program (CUP). Here, we report the multicenter outcomes of var‐cel in patients with R/R MCL treated across four Spanish centers. MATERIALS AND METHODS Patient population This multicenter retrospective study included 30 patients with R/R MCL who were referred for var‐cel therapy at four Spanish centers between February 2018 and January 2025. One patient was included in the CART19‐BE‐01 pilot trial (NCT03144583), 15 following the Declaration of Helsinki and approved by the Spanish Medicines Agency (AEMPS) and the Institutional Review Board (IRB)/Ethics Committee (EC) of HCB, while the remaining 29 patients were recruited under the subsequent CUP, following the same eligibility criteria and requiring individual patient authorization from the AEMPS. All patients signed informed consent. Key eligibility criteria included: (i) CD19‐positive MCL with relapse after or refractoriness to at least one BTKi‐containing regimen, in second or subsequent relapse, in patients who were either ineligible for or had relapsed after autologous hematopoietic stem cell transplantation (HCT); (ii) age ≤80 years; (iii) Eastern Cooperative Oncology Group Performance Status (ECOG PS) 0–2; and (iv) absence of suitable approved therapeutic alternatives. Key exclusion criteria included severe renal, hepatic, pulmonary, or cardiac impairment; active immunosuppressive therapy; and active infections requiring systemic treatment. Study design and analysis This multicenter retrospective study evaluated the safety and efficacy of var‐cel in patients with R/R MCL. Safety endpoints included the incidence and severity of CRS and ICANS, both graded according to the ASTCT 2019 criteria 16 ; immune effector cell‐associated hemophagocytic lymphohistiocytosis‐like syndrome (IEC‐HS) was retrospectively analyzed using ASTCT 2023 criteria. 17 Immune effector cell‐associated hematotoxicity (ICAHT) was assessed according to EHA/EBMT consensus grading. 18 Efficacy was assessed according to the Lugano criteria, 19 including positron emission tomography (PET)/computed tomography (CT) imaging and, when clinically indicated, bone marrow aspirate/biopsy and/or cerebrospinal fluid (CSF) evaluation in patients with bone marrow and/or central nervous system (CNS) involvement. Measurable residual disease (MRD) was assessed by flow cytometry with a sensitivity of 1 in 10,000 (10−4; Supplementary p. 1). Efficacy endpoints included overall response rate (ORR), complete response (CR) rate, progression‐free survival (PFS), duration of response (DOR), overall survival (OS), and cumulative incidence of B‐cell recovery (BCR). PFS was defined as the time from var‐cel infusion to disease progression, death, or last follow‐up; DOR was defined among responders as the time from first documented response to relapse, death, or last follow‐up. OS was defined as the time from var‐cel infusion to death or last follow‐up. NRM was estimated at 1 year using cumulative incidence from var‐cel infusion, with disease relapse as a competing event. BCR was defined as the time from absolute B‐cell aplasia (<1 peripheral blood B‐cell/μL) to subsequent recovery (≥1 B‐cell/μL). POD12 and POD24 were defined as progression of disease within 12 and 24 months, respectively, from the start of first‐line therapy. Patients were classified as POD12 or POD24 if lymphoma progression or relapse occurred within the corresponding time frame. Time‐to‐event outcomes were estimated using the Kaplan–Meier method and competing‐risk endpoints using cumulative incidence functions. To evaluate the impact of achieving CR on OS while minimizing immortal time bias, we used a time‐to‐event approach in which CR was modeled as a time‐dependent covariate. Effect estimates were obtained with a time‐dependent Cox model, and results were visualized using Simon–Makuch curves to reflect changes in response status over time. Analyses were primarily descriptive. Statistical analyses were performed using R, version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria). The data cut‐off was August 2025, after all patients had a minimum follow‐up of 3 months unless relapse or death occurred earlier. CAR T‐cell production, administration, and monitoring Eligible patients underwent leukapheresis, and var‐cel was manufactured as previously described. 20 Var‐cel production was performed in a GMP‐compliant facility using validated procedures and predefined release specifications, under regulatory authorization and oversight by the AEMPS. The target dose of manufactured var‐cel was set at 5 × 106 CAR T cells/kg, which was administered as three increasing doses of 0.5, 1.5, and 3 × 106 CAR T cells/kg. Prior to var‐cel infusion, patients underwent intravenous lymphodepleting chemotherapy with fludarabine (30 mg/m2/day) and cyclophosphamide (300 mg/m2/day) on Days −6, −5, and −4. The fludarabine dose was adjusted to 80% or 50% for patients with a reduced glomerular filtration rate (GFR) of 45–60 or 30–45 mL/min/1.73 m2, respectively. Lymphodepletion was followed by intra‐patient dose escalation with consecutive infusions of the available var‐cel fractions separated by at least 24 h (Figure S1). A safety gate was established before each fraction infusion, requiring the absence of clinically relevant active toxicities that could be exacerbated by additional CAR T‐cell infusions, including systemic infections, fever, CRS, ICANS, or IEC‐HS. Var‐cel expansion and persistence were assessed in peripheral blood by quantifying the CAR transgene using quantitative polymerase chain reaction (qPCR), as previously described. 15 Monitoring was performed weekly for the first 6 weeks, every 2 weeks until month 3, monthly up to 1 year, and quarterly thereafter. RESULTS Var‐cel manufacturing and bridging therapy All 30 patients underwent leukapheresis. Var‐cel was successfully manufactured for 29/30 patients (97%), and 26/29 (90%) met all prespecified release criteria. Median production time was 8 days (range, 6–11), and the median manufactured dose was 5.5 × 106 CAR T cells/kg (range, 1–10). Three products were released as out‐of‐specification (OOS): one with transduction efficiency below the predefined threshold (15% CAR+ T cells) but within total cell dose specifications, and two with manufactured cell doses below the intended target (1.0 and 2.5 × 106 CAR T cells/kg). Bridging therapy was administered to 28/30 patients (93%). Among these, 14 patients (50%) received BTKi‐based regimens, 8 (29%) received chemotherapy, and 3 (11%) received a combination of BTKi and chemotherapy; the remaining patients received alternative approaches (Table S1). Details on bridging treatment response and disease dynamics are provided in Table S2 and Figure S2. Lymphodepletion and var‐cel administration Var‐cel was administered to 27/30 patients (90%). The median vein‐to‐vein time was 29 days (interquartile range [IQR], 26–43). Three patients did not receive var‐cel due to manufacturing failure (n = 1), death due to rapidly progressive disease (n = 1), and septic shock during lymphodepletion (n = 1; Figure 1). All patients receiving var‐cel underwent lymphodepleting chemotherapy; two required fludarabine dose reduction due to reduced GFR (Table S3). CONSORT diagram of patients with mantle cell lymphoma included in the study. CRS, cytokine release syndrome. Var‐cel administration was completed over a median of 3 days (range, 0–9) from first to last infusion. All three planned fractions were delivered to 24/27 patients (89%) (Figure S3). Three patients (11%) did not receive all planned fractions due to CRS, including one Grade 4 event, one Grade 2 event, and one persistent Grade 1 CRS associated with worsening of lymphoma‐related autoimmune hemolytic anemia. Characteristics of the infused population Baseline characteristics of infused patients (n = 27) are summarized in Table 1. Median age at inclusion was 60 years (range, 47–74), and 9 patients (33%) were female. High‐risk disease features were frequent. Among patients with available data, blastoid or pleomorphic morphology was present in 11/27 (41%), del(17p) and/or TP53 mutation in 9/20 (45%), and Ki‐67 ≥ 30% in 13/16 (81%); TP53 status and Ki‐67 percentage were unavailable in a subset of patients (Table 1). Extranodal involvement was detected in 19/27 (70%) patients, with bulky disease (≥5 cm) in 8/27 (30%), elevated lactate dehydrogenase (LDH) in 11/27 (41%), CNS involvement in 3/27 (11%), and a high CAR‐HEMATOTOX score was observed in 5/27 (19%) patients. Table 1Baseline characteristics of treated patients.Characteristic n = 27Age (years), median (range)60 (47–74)Female sex, n (%)9 (33)ECOG PS at screening, n (%)019 (70)17 (26)21 (4)sMIPI, n (%) ‐Low 12 (44) ‐Intermediate 10 (37) ‐High 5 (19)MCL subtype, n (%) ‐Classical 16 (59) ‐Blastoid 9 (33) ‐Pleomorphic 2 (8)Disease burden ‐Extranodal disease at screening, n (%) 19 (70) ‐Bulky disease at screening, n (%) 8 (30) ‐Bone marrow involvement at screening, n (%) 13 (50) ‐CNS disease at screening, n (%) 3 (11) ‐LDH at lymphodepletion (U/L), median (IQR) ◦High LDH, n (%) 195 (158–251) 11 (41) Cytogenetics, n (%) ‐ TP53 mutations 6/16 (38) ‐17p deletion 4/19 (21) ‐Complex karyotype 4/11 (36)Ki‐67 proliferation index, median % (IQR)60 (35–90)≥30%, n (%)13/16 (81)≥50%, n (%)11/16 (69)Prior lines of therapy, median (range)4 (2–9) ‐≥3 prior treatment lines, n (%) 20 (74)Previous HCT, n (%) ‐Autologous 16 (59) ‐Allogeneic 8 (30) ‐Both 5 (19)Previous BTKi, n (%)a 27 (100)Covalent BTKi27 (100) ‐Ibrutinib 26 (96) ‐Acalabrutinib 5 (19) ‐Zanubrutinib 1 (4)Non‐covalent BTKi5 (19) ‐Pirtobrutinib 3 (11) ‐Nemtabrutinib 2 (7)Previous BCL2i, n (%)3 (11)POD12, n (%)13 (48)POD24, n (%)19 (70)Time from diagnosis to apheresis (years), median (IQR)3.6 (2.0–8.4)Abbreviations: BTKi, Bruton's tyrosine kinase inhibitors; CNS, central nervous system; ECOG PS, Eastern Cooperative Oncology Group Performance Status; HCT, hematopoietic stem cell transplantation; IQR, interquartile range; LDH, lactate dehydrogenase; MCL, mantle cell lymphoma; POD12/24, disease progression within 12/24 months of initiation of frontline therapy; sMIPI, simplified mantle cell lymphoma international prognostic index. a Percents exceed 100 because some patients received more than one BTKi.John Wiley & Sons, Ltd. Baseline characteristics of treated patients. ‐Low ‐Low Low ‐Intermediate ‐Intermediate Intermediate ‐High ‐High High ‐Classical ‐Classical Classical ‐Blastoid ‐Blastoid Blastoid ‐Pleomorphic ‐Pleomorphic Pleomorphic ‐Extranodal disease at screening, n (%) ‐Extranodal disease at screening, n (%) Extranodal disease at screening, n (%) ‐Bulky disease at screening, n (%) ‐Bulky disease at screening, n (%) Bulky disease at screening, n (%) ‐Bone marrow involvement at screening, n (%) ‐Bone marrow involvement at screening, n (%) Bone marrow involvement at screening, n (%) ‐CNS disease at screening, n (%) ‐CNS disease at screening, n (%) CNS disease at screening, n (%) ‐LDH at lymphodepletion (U/L), median (IQR) ◦High LDH, n (%) ‐LDH at lymphodepletion (U/L), median (IQR) ◦High LDH, n (%) LDH at lymphodepletion (U/L), median (IQR) ◦High LDH, n (%) ◦High LDH, n (%) High LDH, n (%) 195 (158–251) 11 (41) ‐ TP53 mutations ‐ TP53 mutations TP53 mutations ‐17p deletion ‐17p deletion 17p deletion ‐Complex karyotype ‐Complex karyotype Complex karyotype ‐≥3 prior treatment lines, n (%) ‐≥3 prior treatment lines, n (%) ≥3 prior treatment lines, n (%) ‐Autologous ‐Autologous Autologous ‐Allogeneic ‐Allogeneic Allogeneic ‐Both ‐Both Both ‐Ibrutinib ‐Ibrutinib Ibrutinib ‐Acalabrutinib ‐Acalabrutinib Acalabrutinib ‐Zanubrutinib ‐Zanubrutinib Zanubrutinib ‐Pirtobrutinib ‐Pirtobrutinib Pirtobrutinib ‐Nemtabrutinib ‐Nemtabrutinib Nemtabrutinib Abbreviations: BTKi, Bruton's tyrosine kinase inhibitors; CNS, central nervous system; ECOG PS, Eastern Cooperative Oncology Group Performance Status; HCT, hematopoietic stem cell transplantation; IQR, interquartile range; LDH, lactate dehydrogenase; MCL, mantle cell lymphoma; POD12/24, disease progression within 12/24 months of initiation of frontline therapy; sMIPI, simplified mantle cell lymphoma international prognostic index. Percents exceed 100 because some patients received more than one BTKi. Patients were heavily pretreated, with a median of 4 prior lines of therapy (range, 2–9). Prior autologous and allogeneic HCT had been performed in 16/27 (59%) and 8/27 (30%) patients, respectively. All patients had been exposed to at least one covalent BTKi, and 14/27 (52%) were BTKi‐refractory. Five patients (19%) had also received a non‐covalent BTKi, and all were refractory to this class. Safety Any‐grade CRS occurred in 93% of patients (n = 25/27; 95% CI, 77–99) with a median onset time of 2 days (range, 0–7) and a median duration of 3 days (range, 1–19; Table 2). Grade ≥3 CRS was observed in only one patient (4%), who experienced a Grade 4 event. Tocilizumab was used in 12 patients (44%) and 4 (15%) received corticosteroids. Any‐grade ICANS occurred in only one patient (4%), corresponding to a Grade 3 event. Notably, both Grade ≥3 CRS and ICANS were observed in the same patient. This patient had refractory pleural and peritoneal disease at infusion and developed early CRS that rapidly escalated to Grade 4 CRS with concomitant Grade 4 IEC‐HS, later complicated by colonic perforation with severe peritonitis and septicemia. A PET/CT showed an ongoing partial metabolic response. However, on Day +27 she developed progressive neurological decline, clinically managed as suspected Grade 3 ICANS without improvement despite escalated dexamethasone, and deceased on Day +36 despite maximal supportive care. Autopsy revealed previously unrecognized disseminated cerebral toxoplasmosis. Accordingly, this was the only case of NRM and the only possible ICANS event in the cohort; however, the neurological syndrome could not be conclusively attributed to ICANS, cerebral toxoplasmosis, or a combination of both. Accordingly, at 1 year, the cumulative incidence of NRM was 4.2% (95% CI, 0.28–18). Table 2Immune‐related adverse events.EventVar‐cel–treated cohort (n = 27)CRS, any grade, n (%)25 (93)Grade 121 (78)Grade 23 (11)Grade 30Grade 41 (4)Time to onset, days, median (range)2 (0–7)Duration, days, median (range)3 (1–19)Tocilizumab use (CRS), n (%)12 (44)Corticosteroid use (CRS), n (%)4 (15)ICANS, any grade, n (%)1 (4)Grade 10Grade 20Grade 31 (4)Grade 40Time to onset, days, median (range)16 (16–16)Duration, days, median (range)20 (20–20)Corticosteroid use (ICANS), n (%)1 (4)IEC‐HS, n (%)1 (4)Grade 41 (4)Early ICAHT, any grade, n (%)26 (96)Grade 18 (30)Grade 29 (33)Grade 37 (26)Grade 42 (7)Late ICAHT, any grade, n (%)18 (67)Grade 16 (22)Grade 26 (22)Grade 35 (19)Grade 41 (4)Abbreviations: CRS, cytokine release syndrome; ICAHT, immune effector cell‐associated hematotoxicity; ICANS, immune effector cell‐associated neurotoxicity syndrome; IEC‐HS, immune effector cell‐associated hemophagocytic lymphohistiocytosis‐like syndrome.John Wiley & Sons, Ltd. Immune‐related adverse events. Abbreviations: CRS, cytokine release syndrome; ICAHT, immune effector cell‐associated hematotoxicity; ICANS, immune effector cell‐associated neurotoxicity syndrome; IEC‐HS, immune effector cell‐associated hemophagocytic lymphohistiocytosis‐like syndrome. Severe (Grade ≥3) early and late ICAHT occurred in 9/27 (33%) and 6/27 (22%) patients, respectively (Table 2 and Figures S4–S6). Grade 4 thrombocytopenia was observed in 8/27 (30%) patients (95% CI, 14–50), with a median duration of 15 days (IQR, 9–58). Prolonged Grade ≥3 neutropenia, thrombocytopenia, or anemia at Day 90 was documented in 9% of evaluable patients (Figures S7–S9). During long‐term follow‐up, any‐grade and Grade ≥3 infections occurred in 12/27 (44%) and 7/27 (26%) patients, respectively, with a median time to event of 151 days (range, 3–870) (Tables S4–S6). In 11/27 patients (41%), immunoglobulin replacement therapy was used as secondary prophylaxis. Efficacy With a median follow‐up of 21.7 months (IQR, 8.9–25.8), the best ORR was 89% (95% CI, 71–98), with 78% (95% CI, 58–91) achieving a CR (Table 3, Figure 2). Among six patients with a partial response (PR) at first assessment, four converted to CR after a median of 71 days (range, 61–155). Among patients with known bone marrow involvement (n = 13), CR with undetectable MRD by flow cytometry (10−4) was achieved in 100% of cases (95% CI, 81–100). Table 3Efficacy outcomes.Treated population (n = 27)ORR, % (95% CI)89 (71–98)CR, % (95% CI)78 (58–91)PR, % (95% CI)11 (2–29)Median follow‐up, months (IQR)21.7 (8.9–25.8)DOR, median, months (95% CI)17.1 (8.5–NR)12‐Month rate, % (95% CI)60 (43–82)18‐Month rate, % (95% CI)42 (26–67)PFS, median, months (95% CI)15.4 (6.7–NR)12‐Month rate, % (95% CI)59 (43–81)18‐Month rate, % (95% CI)47 (31–71)OS, median, months (95% CI)35.2 (35.2–NR)12‐Month rate, % (95% CI)77 (62–95)18‐Month rate, % (95% CI)73 (57–92)Abbreviations: CR, complete response; DOR, duration of response; IQR, interquartile range; NR, not reached; ORR, overall response rate; OS, overall survival; PFS, progression‐free survival; PR, partial response.John Wiley & Sons, Ltd. Efficacy outcomes. Abbreviations: CR, complete response; DOR, duration of response; IQR, interquartile range; NR, not reached; ORR, overall response rate; OS, overall survival; PFS, progression‐free survival; PR, partial response. Best response. Responses were durable in most cases, with a median DOR of 17.1 months (95% CI, 8.5–not reached [NR]) and an estimated 12‐month DOR rate of 60% (95% CI, 43–82) (Table 3, Figure 3A). The median PFS was 15.4 months (95% CI, 6.7–NR), with an estimated 12‐month PFS rate of 59% (95% CI, 43–81) (Figure 3B). A plateau‐like tail was observed in both PFS and DOR beyond 18 months, with no additional relapse or progression events recorded thereafter (Figure 3A,B). Kaplan–Meier curves for duration of response (A), progression‐free survival (B), and overall survival (C). Simon–Makuch plot for overall survival in patients after complete response (CR) and without CR status (D). At data cut‐off, 15/27 patients (56%) had experienced death, relapse, or progression, while 12/27 (44%) remained alive and in CR. The median OS for the treated population was 35.2 months, with an estimated 12‐month OS rate of 77% (95% CI, 62–95) (Figure 3C). Achieving CR was associated with a substantially lower risk of death in time‐dependent analyses (hazard ratio [HR] 0.06, 95% CI 0.01–0.35; P = 0.0019). Median OS was 7.4 months while patients were without CR (95% CI 5.6–NR) and was not reached after CR (95% CI 35.1–NR) (Figure 3D). Var‐cel kinetics and B‐cell aplasia Peripheral blood qPCR data for var‐cel expansion and persistence were available in 21/27 infused patients. Robust expansion was measured, with a median time to peak (Tmax) of 21 days (IQR, 14–34), and a median peak transgene level (Cmax) of 2.62 (IQR, 1.68–4.21) copies per ng of genomic DNA. Var‐cel persistence in peripheral blood was detected for a median of 12 months (95% CI, 7.8–NR), with ongoing detection up to 32 months (Figure 4A,B). Chimeric antigen receptor (CAR) T‐cell expansion by quantitative polymerase chain reaction (qPCR) quantification in peripheral blood (A), CAR T‐cell survival (qPCR) in peripheral blood (B), cumulative incidence (CI) of B‐cell recovery in peripheral blood (C). All infused patients achieved absolute B‐cell aplasia, with a median time of 6 days (IQR, 0–9). BCR was subsequently observed, with a median time to recovery of 8.3 months (95% CI, 4.4–NR) (Figure 4C). At relapse or progression, 6/15 (40%) evaluable patients had previously experienced BCR, and 6/14 (43%) had loss of CAR T‐cell detection in peripheral blood. Persistent CD19 expression in MCL cells was confirmed in 9/15 (60%) cases, whereas 1/15 (6%) was CD19‐negative; in the remaining 5/15 (33%), CD19 expression was not assessed. Predictors of response and subgroup analyses Var‐cel efficacy appeared broadly comparable across clinically relevant subgroups, including patients with elevated LDH, TP53 aberrations, high Ki‐67 expression, simplified MCL International Prognostic Index (sMIPI) categories, extranodal involvement, number of prior lines, prior allogeneic HCT, and POD12/24 status. In contrast, patients with blastoid morphology (n = 9) had inferior outcomes compared with non‐blastoid variants, including shorter DOR (median 2.9 months vs. NR; 12‐month DOR 38% vs. 70%, P = 0.001), PFS (median 4 months vs. NR; 12‐month PFS 33% vs. 72%, P = 0.003), and OS (median 10 months vs. NR, 12‐month OS 38% vs. 94%, P = 0.002) (Figures S10–S19). An exploratory univariate Cox analysis was performed to identify factors associated with survival outcomes (Table S7). The presence of B symptoms at screening and larger maximum tumor size were associated with inferior outcomes. Other variables, including age, sex, ECOG PS, prior therapies, baseline SUVmax, and hematologic toxicity risk, were not significantly associated with PFS in this cohort. Overall, these findings suggest that adverse tumor biology (i.e., blastoid morphology) and higher disease burden may be key determinants of outcome following var‐cel therapy (Table S7; Figures S10–S12). DISCUSSION In this multicenter cohort of heavily pretreated patients with R/R MCL, varnimcabtagene autoleucel achieved high response rates (ORR/CR 89%/78%) and durable disease control (median DOR of 17.1 months), with very low rates of Grade ≥3 CRS and ICANS (4%). These findings support var‐cel as a feasible therapeutic option with a favorable safety profile and encouraging efficacy in a real‐world academic setting. Overall, outcomes are consistent with the range reported in pivotal and real‐world experiences of commercial CD19‐directed CAR T‐cell therapies in MCL.6, 7, 8, 9, 11, 12, 13 Despite the introduction of novel targeted agents, most patients with R/R MCL ultimately relapse. Although brexu‐cel has been available for R/R MCL since 2020 in the United States and the European Union, and liso‐cel has recently received EU approval (Nov 2025), access to commercial CAR T‐cell therapies remains uneven across Europe and is frequently constrained by financial and regulatory barriers. 14 In Spain, brexu‐cel became available in 2024 under restrictive eligibility criteria, and liso‐cel is not yet available. In response to these access gaps, a network of Spanish academic centers led by HCB established an integrated pathway for hospital‐based advanced therapies. Under Europe's Hospital Exemption pathway, var‐cel obtained AEMPS authorization for adults with R/R B‐ALL, and compassionate use mechanisms enabled treatment of other B‐cell malignancies, including MCL. 14 The efficacy of var‐cel appears broadly comparable to that of commercial products, although such comparisons should be interpreted cautiously considering differences in patient selection and disease characteristics. While the median age of our cohort was somewhat younger, we included patients with prior allo‐HCT and CNS involvement, and they had received a median of four prior lines of therapy (vs. three in other product trials). In contrast, other high‐risk disease features—including blastoid morphology, TP53 mutations, and high Ki‐67 (≥30%)—were found at frequencies that were comparable to, or in some cases higher than, those reported in trials of other products.11, 12, 13 In a European multicenter real‐world series of brexu‐cel (n = 33), ORR/CR were 91%/79% with 12‐month PFS/OS rates of 51%/61%, broadly comparable to var‐cel's 12‐month estimates (PFS 58%, OS 76%). 11 As previously noted, the longer PFS observed in ZUMA‐2 (median 25.8 months, with a median follow‐up of 35.6 months) may reflect favorable patient selection, including the absence of post‐allogeneic HCT cases. 6 , 11 In the TRANSCEND‐MCL, liso‐cel achieved ORR/CR in 86%/74%, with a median PFS of approximately 15.5 months, similar to that observed in our cohort and with comparable median follow‐up times (16.1 vs. 21.7 months, respectively). 8 The TARMAC study combining investigational tisa‐cel (CTL019) with peri‐infusion ibrutinib reported high CR rates of 80% and encouraging 12‐month PFS/OS rates of 75%/100%, although the cohort was enriched for BTKi‐naïve and non‐blastoid patients. 9 Bridging therapy was administered in most patients (28/30), with all infused patients receiving some form of bridging therapy, in contrast to the ZUMA‐2 trial and TRANSCEND‐MCL trial, in which only 37% and 66% of patients received bridging, respectively. Furthermore, in the ZUMA‐2 trial, only BTKi and steroids were permitted, while TRANSCEND allowed chemotherapy. 6 , 8 Although bridging therapy in our cohort may have facilitated access to CAR T‐cell infusion and reduced tumor burden in a subset of patients, the majority still had progressive or stable disease at the time of var‐cel infusion (Figure S2). Regarding OS, the 3‐year ZUMA‐2 update reported a median OS of 46.6 months for brexu‐cel, whereas TRANSCEND‐MCL reported a median OS of 18.2 months for liso‐cel. 7 , 8 In this context, var‐cel showed a median OS of 35.2 months, intermediate between these datasets. Safety signals with var‐cel were favorable, with a low rate of 1‐year NRM (4.2%). Brexu‐cel, the only commercially available CAR T‐cell therapy for MCL in Spain, has been associated with high rates of CRS and ICANS incidence of approximately 90% and 60%, respectively, including severe (Grade ≥3) CRS and ICANS events in 15% and 31%, respectively, 6 , 11 that often require prolonged corticosteroid exposure, which may contribute to the risk of morbidity and mortality. 21 In contrast, ICANS was rare in our var‐cel cohort (a single case), consistent with previous experiences using this product,15, 22, 23, 24 and severe CRS was uncommon as well (the same single case). This pattern is consistent with a possible contribution of the gated, fractionated dosing strategy, which may allow early toxicity detection and management without compromising CAR T‐cell exposure. Given the current limited availability of lower toxicity commercial options like liso‐cel in Spain, var‐cel may represent a relevant alternative for patients with R/R MCL who are ineligible for or lack access to approved products. Cytopenias and long‐term B‐cell aplasia are directly linked to infections and NRM. 10 , 25 Although the ICAHT grading was not applied in the ZUMA‐2 and TRANSCEND‐MCL trials, real‐world studies show that severe early ICAHT is quite prevalent in MCL (28% with brexu‐cel; 33% in our cohort). 25 Regarding long‐term cytopenia, we reported an incidence of 22% of severe late ICAHT. In comparison, the TRANSCEND‐MCL trial reported 24% of Grade ≥3 neutropenia not resolved by Day 29 after liso‐cel infusion. 8 The TRANSCEND‐MCL trial reported a 15% rate of severe infections, while 26% of individuals in our cohort experienced them, and 38% in ZUMA‐2, although differences in reporting periods should be considered. 6 , 8 Timely CAR T‐cell delivery remains critical in aggressive MCL. 8 , 13 Three patients did not reach var‐cel infusion, including one due to rapid progression, underscoring the importance of short vein‐to‐vein intervals. With a median of 29 days from apheresis to infusion, hospital‐based manufacturing resulted in competitive timelines. Although additional options after BTKi failure exist, indirect comparative evidence has favored CAR T‐cell therapy over historical standards. A propensity‐matched analysis demonstrated an OS advantage for brexu‐cel over historical regimens (SCHOLAR‐2). 26 In this line, brexu‐cel outperformed pirtobrutinib as well in a matching‐adjusted indirect comparison, with superior ORR/CR (93%/70% vs. 57%/19%, respectively) and PFS (29.3 vs. 6.9 months), despite a worse toxicity profile. 27 This study has limitations inherent to its retrospective design and modest sample size, which may have led to potential selection bias toward highly selected advanced/high‐risk disease, limiting causal inference and the precision of tail estimates on survival curves. Response assessment was not centralized, although all centers adhered to Lugano criteria for PET/CT evaluations. Also, missing data for TP53 aberrations and Ki‐67 in a subset of patients, as well as the small sample size of patients with blastoid morphology, restrict the interpretation of these biomarkers and high‐risk subgroups. These subgroup analyses were only exploratory, and larger studies are required to confirm their role in var‐cel efficacy. Nonetheless, the inclusion of consecutive patients within a unified academic network, standardized manufacturing and dosing, and structured follow‐up and response assessment support the validity of these real‐world observations. Although smaller than pivotal multicenter clinical trials, this cohort is representative of real‐world R/R MCL patients and provides a national appraisal of a public‐health hospital‐based CAR T‐cell product delivered across four academic centers. Finally, cost and health‐system sustainability are increasingly important. The reported price of var‐cel (€89,270)—approximately one‐third of current brexu‐cel pricing in the EU—suggests that hospital‐based CAR T‐cell platforms have the potential to expand access within publicly funded healthcare systems. 14 As CAR T‐cell therapy moves toward earlier treatment lines, health systems will require effective yet affordable strategies, prioritizing cost‐effectiveness. Ongoing academic initiatives, including a head‐to‐head trial comparing commercially available axicabtagene ciloleucel with point‐of‐care ARI‐0001 cells for R/R diffuse large B‐cell lymphoma in The Netherlands (NCT05641428) and the multinational European trial for var‐cel in adults with R/R B‐ALL (CARTALLEU), will seek the first centralized commercialization authorization for a hospital‐based manufactured CAR T‐cell product. In summary, var‐cel delivered substantial and durable anti‐lymphoma activity with a favorable safety profile in heavily pretreated patients with R/R MCL treated within a multicenter academic network under a public universal healthcare system. We hope that these outcomes, together with emerging comparative evidence, support further prospective evaluation of var‐cel and hospital‐based CAR T‐cell platforms, including formal cost‐effectiveness analyses and comparative trials, to improve equitable access to cellular therapies. AUTHOR CONTRIBUTIONS Nil Albiol: Writing—original draft preparation (lead); conceptualization (equal); investigation (equal); resources (equal); writing—review and editing (equal); formal analysis (supporting). Nuria Martínez‐Cibrián: Writing—original draft preparation (equal); conceptualization (equal); investigation (equal); resources (equal); writing—review and editing (equal). Amanda Isabel Pérez: Writing—original draft preparation (equal); conceptualization (equal); investigation (equal); writing—review and editing (equal). Sergio Navarro‐Velázquez: Investigation (equal); resources (equal); writing—review and editing (equal). Guillherme Muzzi: Investigation (equal); writing—review and editing (equal). Aina Oliver‐Caldes: Investigation (equal); resources (equal); writing—review and editing (equal). Marta Español‐Rego: Investigation (equal); resources (equal); writing—review and editing (equal). Leticia Alserawan: Investigation (equal); resources (equal); writing—review and editing (equal). Mercedes Montoro‐Lorite: Investigation (supporting); resources (supporting); writing—review and editing (equal). Carla Ramos: Investigation (supporting); resources (supporting); writing—review and editing (equal). Pilar Ayora: Investigation (supporting); resources (supporting); writing—review and editing (equal). Ana África Martin‐López: Investigation (equal); resources (equal); writing—review and editing (equal). Almudena Cabero: Investigation (equal); resources (equal); writing—review and editing (equal). Estefanía Pérez‐López: Investigation (equal); resources (equal); writing—review and editing (equal). Alejandro Martín García‐Sancho: Investigation (equal); resources (equal); writing—review and editing (equal). Tycho Baumann: Investigation (equal); resources (equal); writing—review and editing (equal). José María Sánchez‐Pina: Investigation (equal); resources (equal); writing—review and editing (equal). Anabelle Chinea: Investigation (equal); resources (equal); writing—review and editing (equal). Paula Fernández‐Caldas: Investigation (equal); resources (equal); writing—review and editing (equal). Mariana Bastos‐Oreiro: Investigation (equal); resources (equal); writing—review and editing (equal). Mi Kwon: Investigation (equal); resources (equal); writing—review and editing (equal). Luis Gerardo Rodríguez‐Lobato: Investigation (supporting); resources (supporting); writing—review and editing (equal). Laura Magnano: Investigation (supporting); resources (supporting); writing—review and editing (equal). Andrea Rivero: Investigation (supporting); resources (supporting); writing—review and editing (equal). Juan G. Correa: Investigation (supporting); resources (supporting); writing—review and editing (equal). Pablo Mozas: Investigation (supporting); resources (supporting); writing—review and editing (equal). María Sánchez Castañón: Investigation (supporting); resources (supporting); writing—review and editing (equal). Juan J. Mata‐Molanes: Investigation (supporting); resources (supporting); writing—review and editing (equal). Paola Charry: Investigation (supporting); resources (supporting); writing—review and editing (equal). Jose R. Álamo: Investigation (supporting); resources (supporting); writing—review and editing (equal). Xavier Setoaín: Investigation (supporting); resources (supporting); writing—review and editing (equal). Sonia Rodríguez: Investigation (supporting); resources (supporting); writing—review and editing (equal). Eulalia Olesti: Investigation (supporting); resources (supporting); writing—review and editing (equal). Joaquín Sáez‐Peñataro: Investigation (supporting); resources (supporting); writing—review and editing (equal). Carlos Fernández de Larrea: Project administration (equal): investigation (supporting); resources (supporting); writing—review and editing (equal). Álvaro Urbano‐Ispizua: Funding acquisition (equal); project administration (supporting); investigation (supporting); resources (supporting); writing—review and editing (equal). Armando López‐Guillermo: Investigation (supporting); resources (supporting); writing—review and editing (equal). Julio Delgado: Funding acquisition (lead); project administration (lead); supervision (equal); investigation (supporting); resources (supporting); writing—review and editing (equal). E. Azucena González‐Navarro: Investigation (supporting); resources (supporting); writing—review and editing (equal). Manel Juan: Project administration (equal); investigation (supporting); resources (supporting); writing—review and editing (equal). Eva Giné: Conceptualization (equal); project administration (equal); supervision (equal); investigation (supporting); resources (supporting); writing—review and editing (equal). Valentín Ortiz‐Maldonado: Conceptualization (lead); formal analysis (lead); methodology (lead); supervision (lead); visualization (lead); writing—review and editing (lead); investigation (equal), project administration (equal), resources (equal). CLINICAL TRIAL REGISTRATION The CART19‐BE‐01 trial was registered at clinicaltrials.gov (NCT03144583). Patients treated with var‐cel under CUP were individually authorized by the Spanish Medicines Agency. CONFLICT OF INTEREST STATEMENT Nil Albiol: received honoraria and travel grants from Kite/Gilead. Nuria Martínez‐Cibrián: reports consultant or advisory role (Kite Gilead), travel grants (Pfizer). Alejandro Martín García‐Sancho: honoraria and/or consulting fees from AbbVie, AstraZeneca, BeiGene, BMS, Genmab, Gilead/Kite, GSK, Ideogen, Incyte, Janssen, Kyowa Kirin, Lilly, Miltenyi, Regeneron, Roche, Sobi, and Takeda. Tycho Baumann: honoraria and consultancy from AstraZeneca, BeOne, Johnson & Johnson, Kite‐Gilead, and Lilly. José María Sánchez‐Pina: reports honoraria and travel grants from Kite/Gilead, Johnson & Johnson, and Pfizer. Mi Kwon: reports on advisory roles for Sanofi and AbbVie, and speaker fees from Pfizer, Gilead, and Takeda. Luis Gerardo Rodríguez‐Lobato: honoraria and travel grants from Johnson & Johnson, Amgen, BMS, GSK, Menarini, and Sanofi. Juan G. Correa: received honoraria and travel grants from AstraZeneca and AbbVie. Pablo Mozas: received honoraria from AstraZeneca, BeOne, Johnson & Johnson, Lilly, Takeda, Kyowa Kirin, and Roche. Xavier Setoaín: received honoraria from General Electric and Kite‐Gilead. Carlos Fernández de Larrea: advisory boards for Johnson & Johnson, BMS, Amgen, Pfizer, Sanofi, BeiGene, GSK, Roche, and Menarini; honoraria from Amgen, J&J, BMS, GSK, Sanofi, Pfizer, BeiGene, Oncopeptides, and Miltenyi; and grants from BMS, Johnson & Johnson, Amgen, and GSK. Álvaro Urbano‐Ispizua: reports consultant or advisory role for Kite Gilead, Celgene, and Miltenyi, and travel grants from Kite Gilead and Celgene. Manel Juan: reports no economic conflict of interest except for the intellectual interest for varnimcabtagene autoleucel as a producer. Eva Giné: honoraria and research funding from Janssen and Lilly, and honoraria only from Gilead and AstraZeneca. Valentín Ortiz‐Maldonado: consultant or advisory role for Pfizer and BMS. Sergio Navarro‐Velázquez, Amanda Isabel Pérez, Guilherme Muzzi, Jose R. Álamo, Leticia Alserawan, Mercedes Montoro‐Lorite, Carla Ramos, Pilar Ayora, Ana África Martin‐López, Almudena Cabero, Estefanía Pérez‐López, Anabelle Chinea, Paula Fernández‐Caldas, Mariana Bastos‐Oreiro, Laura Magnano, Andrea Rivero, María Sánchez‐Castañón, Juan J. Mata‐Molanes, Paola Charry, Sonia Rodríguez, Eulalia Olesti, Joaquín Sáez‐Peñataro, Armando López‐Guillermo, Julio Delgado, and E. Azucena González‐Navarro report no conflicts of interest. ETHICS STATEMENT The CART19‐BE‐01 trial was approved by the Institutional Review Board (IRB)/Ethics Committee (EC) of Hospital Clinic de Barcelona and the Spanish Medicines Agency (AEMPS). The analysis of patients included in the compassionate use program was approved by the IRB/EC. FUNDING This study was funded by CatSalut, Fundació La Caixa, Projecte ARI, and grants co‐financed by the Instituto de Salud Carlos III and Fondo Europeo de Desarrollo Regional (FEDER) PICI14/122, PI13/676, PIE13/33, and PI18/775. Patients referred as compassionate use from other Spanish regions besides Catalonia had their therapy funded by each Regional Health Authority. PATIENT CONSENT All patients gave informed consent for var‐cel treatment and research. Supporting information Supporting Information.