CAR T-cell therapy as a definitive consolidation for older adults with B-ALL in first complete remission∗† Authors: Ibrahim Aldoss, Lior Goldberg, Jianying Zhang, Andrew Artz, Mary C Clark, Min Guan, Ruby Espinosa, Thiruvaimozhi Abimannan, Jiaqi Wu, Bryan Garcia, Eric R Haas, Khyatiben V Pathak, Brooke Lovell, Patrick Pirrotte, Vaibhav Agrawal, Neguine Sanani, Stephanie Kasten, Dileshni Tilakawardane, Jamie R Wagner, Jinny Paul, Haoyue Shan, Samer Khaled, Lihua E Budde, Paul Koller, Anthony S Stein, Vinod Pullarkat, Amandeep Salhotra, Ahmed Aribi, Guido Marcucci, Xiuli Wang, Stephen J Forman Published: 2026 ABSTRACT Key Points•First-line CAR T-cell therapy for adults with ALL in CR was safe and tolerable and potentially extended the duration of remission.•Clinical samples of patients in remission and patients with relapsed ALL had distinct inflammatory and immunometabolic profiles. Introduction Patients aged ≥55 years who are diagnosed with B-cell acute lymphoblastic leukemia (B-ALL) experience poor outcomes due to adverse-risk disease biology and limited tolerance of curative therapies.1,2 With new frontline approaches with or without immunotherapy, this population is increasingly able to achieve negative measurable residual disease (MRD-negative) first complete remission (CR1) (70%-95%).3, 4, 5, 6 However, older adults remain at high risk of leukemia relapse and death in remission with traditional and immunotherapy-based consolidation strategies,3,7 including allogeneic hematopoietic cell transplantation (alloHCT).3,8,9 CD19-directed chimeric antigen receptor (CAR) T-cell therapy is highly effective for relapsed or refractory (R/R) B-ALL and can induce deep remissions,10,11 which may be durable without further treatment.11,12 We previously reported the favorable safety and activity of our memory-enriched CD19 CAR T-cell platform13,14 in which we selected naïve and memory T-cell populations for transduction with our investigational CD19 CAR construct14 that includes a CD28 costimulatory domain and truncated human epidermal growth factor receptor for tracking.13,14 In a phase 1/2 clinical trial (ClinicalTrials.gov identifier: NCT02146924) in the setting of R/R B-ALL,13 our CD19 CAR T cells induced 87% CR rate; however, the durability of response was improved by alloHCT consolidation, suggesting that CAR T cells alone may not be sufficient for durable remission in the R/R setting. Patients with R/R leukemia are often heavily pretreated, which may negatively affect the quality of autologous CAR T cells and the potential for durable remission. Moreover, high preinfusion leukemia burden is associated with inferior outcomes and the development of severe CAR T-cell–related toxicity.11,15,16 Older adults, who are more vulnerable to treatment-related toxicity, may benefit from CAR T-cell therapy while in CR1. This strategy could lead to durable remission while avoiding morbidity and mortality associated with intensive consolidation therapies and alloHCT. We conducted a phase 1 clinical trial to evaluate the safety, feasibility, and duration of remission of a single dose of investigational CD19 CAR T cells as consolidation in older adults with B-ALL in CR1. We compared clinical samples of patients with R/R B-ALL and patients in CR1 who received CAR T-cell products generated using the same CAR construct and manufacturing platform13 to identify potential mechanistic insights. Methods This was a single-center phase 1 study registered at ClinicalTrials.gov as NCT05707273. The City of Hope Institutional Review Board approved the study protocol, which was conducted in accordance with the principles of the Declaration of Helsinki. All patients provided written informed consent. Patients aged ≥55 years with B-ALL in CR1, regardless of MRD status, following any frontline therapy (induction with or without reinduction therapy) and without plans for alloHCT consolidation were eligible. Patients with previous alloHCT were excluded. The intent was to enroll patients as soon as they achieved CR1; however, there were no restrictions on the timing of enrollment after remission. Patients underwent leukapheresis for CAR T-cell manufacture. At least 4 intrathecal chemotherapy treatments were required before CAR T-cell infusion because postinfusion intrathecal therapy was prohibited. CD19-targeted immunotherapy was allowed before but was prohibited after leukapheresis. Patients received lymphodepletion (LD) (fludarabine [30 mg/m2] and cyclophosphamide [500 mg/m2] daily for 3 days) before a single infusion of CD19 CAR T cells (200 × 106 CAR+ cells).13 Our investigational CD19 CAR T cells were generated by transducing selected naïve and memory T cells with a second-generation CD19-targeting FMC63-based CAR that includes a CD28 costimulatory domain (for details, refer to Aldoss et al and Wang et al13,14). Patients were monitored for toxicity and MRD relapse every 3 months for 2 years. Patients received no additional therapy except for patients with Philadelphia chromosome–positive (Ph+) disease, who could initiate maintenance with tyrosine kinase inhibitors (TKIs) 60 days after infusion. Study design This study implemented a traditional “3 + 3 rules” design with a 6-patient safety lead-in and a 12-patient expansion cohort. The starting dose level 1 (DL1) was 200 × 106 CAR+ T cells, which was the recommended phase 2 dosing determined in our phase 1/2 study in R/R B-ALL.13 If there were 0 to 1 dose-limiting toxicities (DLTs) in 6 evaluable safety lead-in patients, then DL1 was declared the recommended phase 2 dosing after a full review of toxicity, response, and correlative data, and 12 patients were enrolled on the expansion cohort. If there were ≥2 DLTs in the safety lead-in cohort, then a de-escalation dose (DL −1 [100 × 106 CAR+ T cells]) was evaluated. The primary objective was to assess safety and tolerability by the evaluation of toxicities as graded using Common Terminology Criteria for Adverse Events version 5.0. Cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) were graded according to consensus criteria.17 The DLT evaluation period was the start of protocol therapy through day 28 after infusion. DLT was defined as any of the following events occurring within the DLT evaluation period, assessed as definitely, probably, or possibly related to CAR T-cell infusion, and not included as protocol-specified expected “allowable” adverse events associated with infusion or expansion of CAR T cells: (1) any grade ≥3 nonhematological toxicity (including CRS and ICANS) not related to underlying leukemia, (2) any grade ≥3 autoimmune toxicity, (3) any grade 3 toxicity involving any vital organ system, (4) grade ≥4 hematological adverse event not related to underlying leukemia and lasting >42 days, and (5) any grade 5 toxicity not related to underlying leukemia or preexisting conditions. Disease assessments were performed before LD, 28 ± 4 days after CAR T-cell infusion, and then every 3 months for 2 years. MRD was evaluated using flow cytometry, clonoSEQ Assay (Adaptive Biotechnologies, Seattle, WA) when feasible, and polymerase chain reaction (PCR) for BCR::ABL1 in patients with Ph+ disease. The study was considered feasible if ≥50% of the patients who provided consent underwent leukapheresis, had sufficient CAR T cells manufactured, and received their CAR T-cell product. Event-free survival (EFS) was estimated from CAR T-cell infusion to death or disease progression/relapse or last follow-up (censored if free of event), whichever came first. The definition of relapse included MRD and extramedullary relapse. MRD relapse was defined as detectable leukemic cells at ≥0.01% in morphological remission bone marrow by PCR, next-generation sequencing, or multicolor flow cytometry. Overall survival (OS) was defined as the time from CAR T-cell infusion to death or at last follow-up (censored), whichever came first. Statistical analyses Observed toxicities (worst grade) were summarized by frequency (count/percentage) for type, severity, attribution, time of onset, and reversibility or outcome. Demographic variables, baseline characteristics, primary and secondary diagnoses (molecular phenotypes), and previous and concomitant therapies were summarized using descriptive statistics (median, range, and count/percentage). Feasibility, MRD negativity, and MRD or central nervous system (CNS) relapse were assessed by corresponding frequency (count and/or percentage). EFS and OS were assessed by the Kaplan-Meier product-limit method and Brookmeyer-Crowley 95% confidence interval (CI). One-year EFS and OS rates were estimated accordingly. Geriatric assessment Patients underwent the following assessments at baseline and day 100 after infusion: 5-component physical frailty phenotype (Frailty),18 Montreal Cognitive Assessment (MoCA),19 and physical functioning quality of life by the Medical Outcomes Study Physical Health Score.20 Data were summarized by median and range. Paired change from baseline to day 100 was assessed by Wilcoxon signed-rank tests (MoCA and Medical Outcomes Study Physical Health Score) or Stuart-Maxwell test (Frailty). A P value of ≤ .05 was considered statistically significant. Correlative methods Peripheral blood (PB) and cerebrospinal fluid (CSF) were processed as described.14 Cytokines were analyzed with the Human 30-Plex panel (Invitrogen, Carlsbad, CA) for the Luminex platform. CAR+ cells in PB were detected by quantitative PCR analysis of the WPRE transgene encoded by the lentiviral vector. CAR+ T cells in CSF were identified by detection of epidermal growth factor receptor by flow cytometry, and absolute CAR T-cell numbers were calculated based on the percentage of CAR+ cells and total CSF cell count. Correlative studies were assessed for over-time trend by graphical statistics and linear mixed models or 2-sample t tests when applicable. CyTOF Metal isotope–conjugated monoclonal antibodies (supplemental Data 1) were generated using standard procedures (Maxpar Antibody Labeling User Guide PRD002). Samples were stained per the manufacturer’s instructions (Maxpar Phospho-Protein Staining PRD016) and acquired on a Helios mass cytometer (Standard BioTools, Boston, MA). Data were normalized using bead-based time normalization per the manufacturer’s instructions, cleaned using the Gaussian parameter discrimination method (supplemental Figure 1), and manually debarcoded in OMIQ (www.omiq.ai). Files (.fcs) were reuploaded to OMIQ as single cells and manually gated to resolve CAR+ T cells. Data were batch-corrected in OMIQ with CyCombine21 using standard parameters. Statistical comparisons were made in GraphPad Prism 10.2.1 (GraphPad Software, San Diego, CA; www.graphpad.com) using multiple unpaired t tests with Welch correction. Volcano plots were created in R using tidyverse packages. Metabolomics Untargeted metabolomics was performed as described.22 Metabolites from 1.0 × 106 (CR1) or 3.75 × 106 (R/R) CAR T cells were extracted using solvents (water:methanol:acetonitrile, 1:2.25:0.75 volume per volume per volume) spiked with isotopically labeled internal standards. Extracts were normalized to cell count. Metabolites were vacuum concentrated, separated using hydrophilic interaction chromatography and reversed-phase chromatography on a Vanquish Flex UHPLC, and analyzed on an Orbitrap Eclipse mass spectrometer in positive and negative modes simultaneously. Raw data were analyzed on Compound Discoverer 3.4 (Thermo Fisher Scientific, San Jose, CA) using HMDB, KEGG, mzCloud, and LipidSearch databases and a hand-curated metabolite library. Compound annotations were manually verified. Systematic error removal using random forest was used for normalization,23 and only compounds in the pool quality control with an area under 25% relative standard deviation, mass error under ±5 ppm, peak rating threshold of 4.5, and <0.2 minutes retention time drift were analyzed. Data were further normalized using the vsn package (version 3.74.024) in R (version 4.4.0). Differences between R/R and CR1 were assessed using Welch t test; P values were adjusted for multiple testing using the Benjamini-Hochberg false discovery rate method. A P value of < .05 was considered statistically significant. Redundancy within and across chromatographic methods was addressed by retaining metabolites with the highest signal intensity in pooled quality control samples. Endogenous metabolites were included for pathway analysis using Ingenuity Pathway Analysis25 (Qiagen, Hilden, Germany). Seahorse XF96 extracellular flux Mitochondrial oxygen consumption rate and extracellular acidification rate were measured using the Seahorse Bioscience XF96 Extracellular Flux Analyzer (Agilent, Santa Clara, CA). A total of 150 000 CAR T cells per well were seeded in Seahorse XFe96 PDL cell culture microplates (103799-100; Agilent) in Seahorse XF RPMI medium (103576-100; Agilent) with 10mM glucose, 2mM glutamine, and 1mM sodium pyruvate. Oxygen consumption rate was measured following sequential injections of 2.5μM oligomycin A, 2.5μM BAM15, and 0.5μM rotenone/antimycin A (Seahorse XF T Cell Metabolic Profiling Kit, 103772-100; Agilent) per the manufacturer’s instructions. Results Patient characteristics We obtained consent from 22 patients between March 2023 and February 2025, of whom 20 underwent leukapheresis and 18 received LD and CAR T-cell infusion (feasibility rate, 18/20 [90%]). There were no manufacturing failures. Two patients were ineligible for leukapheresis, and 2 patients who underwent leukapheresis were not eligible to proceed to LD. Because there were no DLTs in the safety lead-in cohort, we opened the expansion cohort and treated a total of 18 patients at DL1. Figure 1 depicts the trial schema and CONSORT diagram. Table 1 and supplemental Table 1 show the characteristics of patients who received CAR T cells. The median age was 64 years (range, 55-79). Fifteen patients (83%) received blinatumomab during induction and/or consolidation, and 2 (11%) received frontline inotuzumab ozogamicin (InO). The median time from diagnosis to leukapheresis was 4.7 (range, 2.4-9) months. Although our intent was to enroll patients as soon as they achieved CR1, enrollment timing was influenced by leukapheresis scheduling, referral timing to the study, and the study pause after the safety lead-in cohort. All patients were MRD-negative (<0.01%) by flow cytometry before LD. Trial schema and CONSORT diagram. (A) Schematic of the phase 1 clinical trial design. After providing consent, patients underwent apheresis for the collection of PB nuclear cells for CAR T-cell manufacture. After product release, patients received LD chemotherapy followed by a single infusion of CD19 CAR T cells. Patients were followed for toxicity and response. Biological samples were collected at the indicated time points for correlative analyses. (B) CONSORT flow diagram of all patients who consented to this study. Table 1.Patient demographic and characteristicsVariablesMedian (range) or n (%)Median age (range), y64 (55-79)Sex Male11 (61) Female7 (39)Race/ethnicity Non-Hispanic White8 (44) Hispanic6 (33) African American1 (6) Asian3 (17)Bone marrow blasts at diagnosis, median (range), %90% (47%-95%)CNS status at diagnosis CNS-117 (94) CNS-21 (6) CNS-30NCCN cytogenetic and molecular risk Standard risk9 (50) Poor risk9 (50)Disease genetics Ph+6 (33) NOS5 (28) Hypodiploidy/TP53m2 (11) Ph-like2 (11) KMT2Ar1 (6) ZNF384::EP3001 (6) TCF3::PBX11 (6)Prior blinatumomab15 (83)Prior inotuzumab ozogamicin2 (11)Number of treatment cycles to achieve CR116 (89) 12 (11) 2Frontline TKI in patients with Ph+ disease Ponatinib3 Dasatinib3Bridging therapy with systemic intermediate dose methotrexate ± TKI15 (83)Median time from diagnosis to leukapheresis (range), mo4.7 (2.4-9)Number of pre-LD IT chemotherapy (range)7 (5-10)Negative pre-LD MRD status18 (100)ECOG at enrollment 05 (28) 112 (67) Unknown1 (6)ECOG, Eastern Cooperative Oncology Group; IT, intrathecal chemotherapy; KMT2Ar, KMT2A-rearrangement; NCCN, National Comprehensive Cancer Network; NOS, not otherwise specified; TP53m, TP53 mutations. Patient demographic and characteristics ECOG, Eastern Cooperative Oncology Group; IT, intrathecal chemotherapy; KMT2Ar, KMT2A-rearrangement; NCCN, National Comprehensive Cancer Network; NOS, not otherwise specified; TP53m, TP53 mutations. Safety There were no DLTs and there were no deaths in the study at the time of data cutoff. Thirteen (72%) patients developed transient grade 1 CRS, all of which resolved with tocilizumab with or without corticosteroids. There was no ICANS of any grade or grade ≥2 CRS. Grade ≥3 treatment-emergent adverse events were neutropenia (100%), lymphopenia (94%), leukopenia (94%), anemia (39%), thrombocytopenia (28%), and febrile neutropenia (11%). Figure 2A depicts all treatment-emergent adverse events. Median time for absolute neutrophil count ≥1000 × 103/μL and platelet count >100 × 103/μL after infusion were 11 (range, 6-86) and 5 (range, 0-461) days, respectively. Clinical outcomes. (A) Most common treatment-emergent adverse events of any grade (light blue) and grade ≥3 (dark blue). (B) EFS. (C) OS. Geriatric assessments included MoCA (n = 18) (D), MOS Physical Health Score (n = 18) (E), and Frailty score (n = 18) (F). P values for the comparisons between day 100 and baseline scores were calculated by the Wilcoxon signed-rank tests (D,E) and Stuart-Maxwell test (F). A P value of ≤ .05 was considered statistically significant. MOS, Medical Outcomes Study. Clinical outcomes All 18 (100%) patients remained in MRD-negative CR on day 28 after CAR T-cell infusion (supplemental Table 2). As of 14 November 2025, the median follow-up was 18 (range, 6.4-28.6) months. The estimated 18-month EFS and OS rates were 84% (95% CI, 64.9-100) and 100% (95% CI, 100-100), respectively (Figure 2B-C). One patient with Ph+ B-ALL had molecular MRD relapse 6 months after infusion while on dasatinib maintenance, detected by PCR for BCR::ABL1 and clonoSEQ (MRD-negative by flow cytometry). This patient was switched to ponatinib, achieved complete molecular remission, underwent alloHCT, and remained alive and in remission. Another patient with Ph-like B-ALL developed myeloid lineage switch 18 months after infusion. He achieved CR after decitabine and venetoclax, underwent alloHCT, and remained in remission. A third patient had morphological CD19+ relapse ∼2 years after infusion. All other patients remained in MRD-negative CR at last follow-up without undergoing alloHCT consolidation. Notably, the first patient treated on study who had Ph− B-ALL remained in MRD-negative CR by clonoSEQ >28 months after infusion without additional therapy. Of the 6 patients with Ph+ B-ALL, all started TKI maintenance 60 days after infusion (ponatinib [n = 3], dasatinib [n = 2], and nilotinib [n = 1]). Geriatric trajectories For geriatric assessments, the median (range) scores at baseline vs day 100 showed preserved cognition by MoCA (26 [21-29] vs 26 [22-30]; P = .170; Figure 2D), improved function by Patient Reported Physical Health (P = .005; Figure 2E), and no decline in frailty by Frailty score (1 [0-4] vs 0 [0-2]; P = .319; Figure 2F). Objective measures of physical performance were similar at baseline vs day 100, including average 4-m walk speed (1.02 [0.41-1.64] vs 0.93 [0.45-1.41] m/s; P = .369) and grip strength (24.30 [13.67-57.67] vs 27.58 [11.27-58.10] kg; P = .799). Expansion and persistence CAR T cells expanded in all patients, peaking in the PB on days 7 to 14, with a median of 52 464 copies per μg of genomic DNA (95% CI, 52464-173695) (Figure 3A; supplemental Figure 2). Most patients had B-cell recovery by 6 months after infusion (Figure 3B). We detected CAR T cells in the CSF of all 17 patients who had lumbar puncture on day 28 (mean absolute = [2.92 ± 6.2] ×103/ml of CSF; Figure 3C). For 1 patient with TP53 mutation and CNS-1 status who achieved MRD-negative CR1 after blinatumomab induction, CAR T cells remained detectable in the CSF 1 year after infusion. All 6 patients with Ph+ B-ALL had low CAR T cells detected before initiating TKIs (supplemental Figure 3). Correlative studies. (A) Individual cellular kinetic profiles of CAR T cells in the PB assessed by qPCR and measured as copies per nanogram of genomic DNA (n = 18 patients). (B) Individual levels of CD19+ B cells in the PB assessed by flow cytometry as a surrogate for CAR T-cell functionality (n = 18 patients). (C) Levels of CAR T cells in the CSF on day 28 after infusion assessed by flow cytometry (n = 17 patients). (D) Cellular kinetics of CAR T cells in the PB according to cohort (n = 59 patients). (E) Serum cytokine profile during the first 28 days of CAR T-cell infusion according to cohort (n = 59 patients). (D-E) The colored lines indicate local polynomial regression fitted values and the surrounding shading represents the 95% CI. P values were calculated using 2-tailed unpaired Student t tests. Significance values: ns; ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. IL-2Rα, interleukin-2 receptor α; MCP-1, monocyte chemoattractant protein 1; MIP-1α, macrophage inflammatory protein 1α; ns, not significant; qPCR, quantitative PCR. Comparison of clinical samples from patients with R/R B-ALL and CR1 To gain insight into the biology of first-line CAR T-cell therapy, we compared clinical samples of patients with R/R B-ALL from our historical phase 1/2 trial (ClinicalTrials.gov identifier: NCT02146924)13 and patients treated in CR1. Both trials tested an identical CAR T-cell manufacturing platform. The R/R cohort failed a median of 3 previous lines of therapy, and at the time of LD, 80% had bone marrow blasts ≥5% and/or active extramedullary disease (supplemental Table 3). CAR T cells expanded in both cohorts with modestly different kinetics (Figure 3D), possibly due to differences in antigen-mediated expansion in patients with active disease vs patients in CR1. We next investigated inflammatory mediators in the PB (supplemental Figure 4). Patients with R/R B-ALL had significantly higher levels of inflammatory markers than patients in CR1 both before and after infusion, including soluble interleukin-2 receptor α, interferon gamma (IFN-γ), and IFN-γ–induced chemokines (MIG/CXCL9, IP-10/CXCL19, monocyte chemoattractant protein 1/CCL2, and macrophage inflammatory protein 1α/CCL3) (Figure 3E). Consistent with a higher rate of CRS in patients with R/R B-ALL,13 interleukin-6 levels were significantly higher at the time of peak expansion compared with the CR1 cohort (supplemental Figure 4). Overall, this demonstrates that patients in CR1 had less inflammation before and after CAR T-cell infusion compared with patients with R/R B-ALL. We then applied an integrative approach to interrogate the metabolic fitness26 and immunometabolic properties of CAR T-cell drug products (DP) (Figure 4A). Using mass cytometry by time of flight (CyTOF), we measured key metabolic transporters and enzymes as well as phenotypic and functional proteins at the single-cell proteomic level27 (Figure 4B; supplemental Data 1). CAR T cells from patients with R/R B-ALL had significantly elevated hexokinase, glucose-6-phosphate dehydrogenase, lysosomal acid lipase, carnitine palmitoyltransferase 1A (CPT1A), citrate synthase, and ATP5A levels (Figure 4C), suggesting hypermetabolism and the preferential use of the pentose phosphate pathway (PPP) rather than glycolysis. Moreover, these data suggest that CAR T cells from patients with R/R ALL released intracellular fatty acids for conversion into long-chain fatty acids and acyl-carnitines by CPT1A for shuttling into the mitochondrial matrix for tricarboxylic acid cycle activity and oxidative phosphorylation (OXPHOS). Comparison of the immunometabolic features of preinfusion CAR T-cell DP of R/R and first remission cohorts. (A) Schematic of the experimental design. (B) Overview of the single-cell immunometabolic approach. (C) Volcano plot showing differentially expressed proteins in CD3+ CAR+ T-cell products (n = 22 patients). (D) Bar plot of statistically significant enriched metabolic pathways in CAR T-cell DP (P < .05) (n = 21 patients). (E) Volcano plot of differentially expressed abundant metabolites in CAR T-cell DP (P < .05) (n = 21 patients). (F) Mitochondrial OCR plots (mean ± SEM) and quantification of basal respiration (G), maximal respiration (H), spare respiratory capacity (I), and ATP production (J). (K) Quantification of basal glycolysis (n = 8 patients, 5-6 biological replicates per patient). P values were calculated using 2-tailed unpaired Student t tests. ECAR, extracellular acidification rate; ETC, electron transport chain; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; LAL, lysosomal acid lipase; LDHA, lactate dehydrogenase; mTOR, mechanistic target of rapamycin; OCR, oxygen consumption rate; SCD1, stearoyl-CoA desaturase-1; SEM, standard error of the mean; TCA, tricarboxylic acid; TIGIT, T cell immunoglobulin and ITIM (immunoreceptor tyrosine-based inhibitory motif) domain. Figure 4A created with biorender.com. Goldberg L. (2026) https://biorender.com/6ju2o52. We then performed untargeted metabolomic profiling via mass spectrometry, which yielded 962 annotated metabolites (2350 total compounds; supplemental Data 2A), including 719 endogenous metabolites (supplemental Data 2B). Comparison of R/R and CR1 CAR T-cell DP identified 402 differentially abundant metabolites (P < .05), with 186 metabolites lower and 216 metabolites higher in R/R compared with CR1 (supplemental Data 2B). Pathway enrichment analysis revealed a pronounced hypermetabolic phenotype of R/R CAR T-cell DP (Figure 4D), including the upregulation of pathways involved in pyruvate metabolism, nucleotide biosynthesis, fatty acyl–coenzyme A biosynthesis, glutamate/glutamine metabolism, and branched-chain amino acid catabolism (supplemental Data 2C). As for individual metabolites, R/R CAR T-cell DP had significantly lower glucose and elevated glycolytic intermediates (glucose-6-phosphate, glyceraldehyde-3-phosphate, and pyruvate) (Figure 4E), which aligns with the elevated hexokinase and glucose-6-phosphate dehydrogenase observed by CyTOF (Figure 4C). Lactate levels were significantly lower in R/R CAR T-cell DP, suggesting that glucose carbon may be diverted from aerobic glycolysis toward mitochondrial oxidative metabolism and the PPP. Supporting this interpretation, R/R CAR T-cell DP had significantly elevated levels of nucleotides and nucleotide phosphates (adenine, guanine, cytosine, AMP, and ADP) (Figure 4E), consistent with upregulated PPP activity. R/R CAR T-cell DP also had more long-chain acyl-carnitines and tricarboxylic acid cycle intermediates, including malate, flavin adenine dinucleotide, ADP, and AMP, and reduced levels of triglycerides, diglycerides, and monoglycerides (Figure 4E). These metabolic changes are consistent with the CyTOF data demonstrating the upregulation of key enzymes involved in lipid catabolism and mitochondrial metabolism, including lysosomal acid lipase, CPT1A, and citrate synthase (Figure 4C). Finally, we measured metabolic activity (extracellular flux) of the DP using the Seahorse metabolic functional assay. Consistent with the single-cell proteomic and metabolomic data, we observed significantly higher basal and maximal OXPHOS and ATP production in CAR T-cell DP of patients with R/R B-ALL, without substantial change in basal glycolysis compared with those from patients in CR1 (Figures 4F-H, J-K). Moreover, the DP of patients in CR1 showed higher spare respiratory capacity (Figure 4I), suggesting higher metabolic fitness. Collectively, these analyses revealed a hypermetabolic state in R/R CAR T-cell DP characterized by lipid mobilization and fatty acid oxidation to fuel OXPHOS as well as nucleotide synthesis by the PPP, supporting the higher energetic and biosynthetic demands of these cells. Discussion We demonstrated the safety of administering CAR T cells as consolidation therapy in older patients with B-ALL in MRD-negative CR1. Treatment was well tolerated, resulting in grade 1 CRS but no ICANS of any grade or high-grade CRS. These findings compare favorably to our previous study conducted in patients with R/R B-ALL using the same CAR T-cell product and dose, in which 74% of patients experienced grade ≥2 CRS and 78% developed any grade of neurotoxicity (17% with grade ≥3).13 This highlights the promising potential of early CAR T-cell intervention as a safer and more tolerable strategy. This approach also preserved both the physical and cognitive function of patients. Moreover, the single infusion consolidation approach eliminated the need for prolonged or repeated consolidation treatments, which can lead to death in remission, even in the era of immunotherapy.3 No deaths in remission have been observed to date, although extended follow-up is needed. CAR T-cell therapy was feasible for maintaining MRD-negative CR, including in patients with high-risk disease biology, although longer follow-up is needed to confirm this observation. With the introduction of modern frontline approaches, with or without immunotherapy, achieving MRD-negative CR1 is increasingly attainable in older adults with B-ALL, with low rates of induction death. Reported MRD-negative CR1 rates range from 70% to 95% in older adults treated with contemporary targeted and reduced-intensity regimens, including mini–hyper-CVD (cyclophosphamide, dexamethasone, vincristine, alternating methotrexate and cytarabine) + InO with or without blinatumomab, EWALL-INO regimen, GMALL-INO regimen, Alliance 041703 regimen, mini-hyper-CVD + venetoclax, and the Golden Gate study regimen.3, 4, 5, 6,28 However, sustaining MRD negativity and preventing death in remission, often related to the toxicity of ongoing therapy, remains a major challenge. Other studies are evaluating the use of CAR T-cell therapy as consolidation in various settings, including Ph+ ALL, Ph− ALL (ClinicalTrials.gov identifiers: NCT07400029 and NCT06287229), and MRD+ CR1 (ClinicalTrials.gov identifier: NCT06144606), with encouraging early reports.29,30 We permitted TKI maintenance for patients with Ph+ B-ALL, similar to the widely adopted posttransplantation approach.31 We detected CAR T cells at low levels in patients with Ph+ B-ALL after initiating TKIs, and there was no apparent negative impact of TKIs on persistence. However, the number of patients and low CAR T-cell levels limit the strength of any conclusions that can be drawn. Nevertheless, this maintenance strategy could be extended to other targeted therapies as maintenance, including menin inhibitors for KMT2A-rearranged ALL.32 Critically, we observed the expansion of CAR T cells after infusion despite the low antigen setting. Expansion in this setting was close to that in the R/R setting with the same product and dose,13 albeit with modestly different kinetics, suggesting that MRD-negative CR does not hinder CAR T-cell expansion. Expansion and persistence of CAR T cells in the CSF is of particular interest considering the propensity of B-ALL to relapse in the CNS, especially in the era of frontline blinatumomab.33, 34, 35 Limited CAR T-cell persistence is likely due to the CD28 costimulatory domain in our CAR construct. Although persistence is critical for cure in the R/R stand-alone setting, robust early expansion may be sufficient in the frontline setting to eradicate residual disease and enable cure. We explored biological differences between patients treated in CR1 and patients with R/R B-ALL, identifying key distinguishing features that support our clinical findings. We observed a hyperinflammatory state in patients with R/R B-ALL, which was characterized by elevated levels of inflammatory mediators before and after infusion and suggested a primed and sustained inflammatory microenvironment. Hyperinflammation is associated with worse outcomes after CD19 CAR T-cell therapy for R/R hematological malignancies.36, 37, 38, 39, 40, 41 Moreover, high IFN-γ and IFN-γ–induced signaling is associated with CD19 CAR T-cell treatment failure38,40,42,43 and toxicity44, 45, 46 through mechanisms that have not been fully elucidated.38,47,48 Our data provide rationale for potential interventions, including moving therapy up front to avoid a hyperinflammatory state or decreasing inflammation by blocking IFN-γ before and after infusion. The latter strategy has been successfully applied to treat CAR T-cell–related toxicities without compromising CAR T-cell function and clinical outcome.49, 50, 51, 52 We observed several immunometabolic differences in CAR T cells of patients with R/R B-ALL and patients in CR1, including a hypermetabolic state in the DP from patients with R/R B-ALL. Conversely, DP from patients in CR1 had increased spare respiratory capacity, which is associated with enhanced T-cell persistence and antitumor activity.53,54 A study comparing donor- and patient-derived CAR T cells determined that patient-derived (vs donor-derived) CAR T cells upregulate genes controlling autophagy, glycolysis, and nucleotide metabolism after infusion,55 suggesting that hypermetabolism is a biological process in CAR T cells of heavily treated patients. Moreover, T cells from early apheresis (at first relapse) compared with standard apheresis (second relapse and beyond) in patients treated with CD19 CAR T-cell therapy for non-Hodgkin lymphoma showed more naïve T cells and increased in vitro T-cell functionality.56 These findings underscore biological differences between CAR T cells manufactured from healthy donors/lightly treated patients and heavily treated patients, providing rationale for potential intervention by early apheresis to collect T cells with higher fitness or by pharmacological modulation of CAR T-cell immunometabolism57 before infusion. Our study was limited by small sample size, single-center design, lack of comparator arm, short length of follow-up, and use of investigational CAR T cells. However, our study lays the foundation for larger confirmatory trials evaluating commercially available CAR T-cell products. Although we evaluated this approach in a selected population of older adults, we anticipate that this strategy may be extended to younger adults at high risk for the toxicities of pediatric-inspired regimens and alloHCT, including those with comorbidities.58 The established activity of CAR T cells in sanctuary sites offers added potential to prevent CNS relapse, potentially replacing multiple cycles of blinatumomab in chemotherapy-free regimens and reducing the need for intensive CNS-directed therapy.33,34 Altogether, this approach may significantly shorten the treatment duration for B-ALL and usher in a new era of precision and innovation. Conflict-of-interest disclosure: The authors declare no competing financial interests. Acknowledgments This research was supported by Blood Cancer United grant ACT9025-25 (S.J.F.). Research reported in this publication includes work performed in the Biostatistics and Mathematical Modeling Core, the Analytical Pharmacology Core, the Integrated Mass Spectrometry Core, and the Manufacturing Core, supported by the National Cancer Institute of the National Institutes of Health under award number P30CA033572. Support was provided by the National Gene Vector Biorepository at Indiana University, which is funded through the National Heart, Lung, and Blood Institute contract 75N92019D00018. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Authorship Contribution: I.A., J.Z., A. Artz, D.T., J.R.W., X.W., and S.J.F. contributed to the clinical trial design; I.A., J.Z., A. Artz, V.A., N.S., S. Kasten, S. Khaled, L.E.B., P.K., A.S.S., V.P., A.S., A. Aribi, G.M., and S.J.F. executed the clinical trial; L.G. and X.W. designed and oversaw the correlative studies; L.G., M.G., R.E., T.A., J.W., B.G., E.R.H., K.V.P., B.L., P.P., J.P., H.S., and X.W. performed and analyzed the correlative studies; I.A., L.G., J.Z., M.C.C., and X.W. prepared the manuscript; I.A., X.W., and S.J.F. contributed to the study supervision; and all authors performed analysis and interpretation of the data and read and approved the manuscript. Footnotes I.A. and L.G. contributed equally to this study as joint first authors. X.W. and S.J.F. contributed equally to this study as joint senior authors. Deidentified data sets used and/or analyzed during this study are available from the corresponding author, Ibrahim Adloss (ialdoss@coh.org), on request. The full-text version of this article contains a data supplement. Supplementary Material Supplemental Figures and Tables Supplemental Data 1 Supplemental Data 2 References Associated Data Supplementary Materials Supplemental Figures and Tables Supplemental Data 1 Supplemental Data 2