Efficacy and safety of allogeneic CD19 CAR NK-cell therapy in systemic lupus erythematosus: a case series in China.

# Efficacy and safety of allogeneic CD19 CAR NK-cell therapy in systemic lupus erythematosus: a case series in China *The Lancet 2025* Articles Efficacy and safety of allogeneic CD19 CAR NK-cell therapy in systemic lupus erythematosus: a case series in China Jie Gao*, Mengtao Li*†, Ming Sun*†, Yiyi Yu*, Ruina Kong*, Xia Xu, Suxuan Liu, Qian Chen, Xiaofang Li, Yang Wu, Enshun Xu, Jianmin Yang†, Dongbao Zhao† Summary Lancet 2025; 406: 2968–79 Background Lately, autologous CD19-targeting chimeric antigen receptor (CAR) T cells have shown excellent efficacy Published Online in treatment of autoimmune diseases, but with great safety concerns, such as infections. In this study, we aimed to November 12, 2025 evaluate the safety, tolerability, and efficacy of allogeneic CD19 CAR natural killer (NK)-cell therapy in patients with https://doi.org/10.1016/ relapsed or refractory systemic lupus erythematosus (SLE). S0140-6736(25)01671-X This online publication has Methods In this open-label, single-arm, prospective, first-in-human case series, we evaluated allogeneic CD19 CAR been corrected. The corrected version first appeared at NK-cell therapy in adult patients (aged 18–65 years) with relapsed or refractory SLE at one site in China. Patients who thelancet.com on Dec 18, 2025 had received at least two previous standard systemic therapies and continued to exhibit moderate-to-severe disease See Comment page 2972 activity were eligible for inclusion. This study consisted of schedule escalation and dose escalation, with schedule *Contributed equally escalation from 7 days and dose escalation commencing at 0·75 × 10⁹ CAR NK cells on day 0. All patients received a †Jointly supervised lymphodepleting conditioning regimen with fludarabine (25 mg/m² per day) and cyclophosphamide (300 mg/m² For the Chinese translation of the per day) administered daily from days –5 to –3, followed by three CAR NK-cell infusions within a single treatment abstract see Online for cycle at identical dose levels and inter-infusion intervals. Dose-limiting adverse events were monitored in patients for appendix 1 28 days. The primary endpoints of this study were safety and tolerability, including the incidence of dose-limiting Department of Rheumatology toxicities and adverse events according to National Cancer Institute Common Terminology Criteria for Adverse Events and Immunology, National Key version 5.0. This study was registered with ClinicalTrials.gov (NCT06010472) and follow-up is ongoing. Laboratory of Immunity and Inflammation, Changhai Hospital, Naval Medical Findings 18 patients with relapsed or refractory SLE with moderate-to-severe disease activity were enrolled between University, Shanghai, China Aug 21, 2023, and June 16, 2024. Of the 18 patients, 17 (94%) were female; the median age was 37·5 years (J Gao PhD, Y Yu PhD, (IQR 32·0–39·8), and the median disease duration was 10·5 years (IQR 4·5–14·8). Patients had received at least R Kong PhD, X Xu BS, Q Chen BS, X Li BS, Prof D Zhao PhD); two standard systemic therapies, including biological agents (belimumab and telitacicept) in 14 (78%) of 18 patients, Department of Rheumatology, and plasmapheresis in one patient. Cytokine release syndrome was reported in one (6%) of 18 patients (grade 1). Peking Union Medical College Neurotoxicity and other CAR NK-cell therapy-related severe adverse events were not observed, and there were no Hospital (PUMCH), Peking dose-limiting toxicities. Of the nine patients with more than 12 months’ follow-up, six (67%) attained DORIS Union Medical College and Chinese Academy of Medical remission and lupus low disease activity state. Sciences, National Clinical Research Center for Interpretation This study suggests that allogeneic CAR NK-cell therapy is a potent option for treatment of autoimmune Dermatologic and diseases and indicates that such a therapy might address limitations of current autologous CAR T-cell therapy, Immunologic Diseases (NCRC-DID), Key Laboratory of including manufacturing scale and time, access, safety, and cost. Rheumatology and Clinical Immunology, Ministry of Funding Shanghai Municipal Health Commission, Changhai Hospital Affiliated to Naval Medical University, and Education, Beijing, China National Natural Science Foundation of China. (Prof M Li PhD); Medical Science and Technology Innovation Center, The Affiliated Suzhou Copyright © 2025 Elsevier Ltd. All rights reserved, including those for text and data mining, AI training, and similar Hospital of Nanjing Medical technologies. University, Suzhou Municipal Hospital, Gusu School of Nanjing Medical University, Introduction disease flares, persistent disease activity, and treatment- Suzhou, China Systemic lupus erythematosus (SLE) is a multisystem associated comorbidities pose substantial clinical (Prof M Sun PhD); Rui autoimmune disease characterised by the production of burdens. Considering these challenges, contemporary Therapeutics, Nanjing, China autoantibodies targeting nuclear antigens, deposition of rheumatology practice emphasises dual therapeutic (Y Wu PhD, E Xu PhD, Prof M Sun); Department of immune complexes, and chronic inflammatory priorities: achieving DORIS remission or maintaining Cardiology, Changhai Hospital, manifestations affecting the skin, joints, kidneys, and lupus low disease activity state (LLDAS) constitutes the Naval Medical University, other organ systems.1,2 The past decade has witnessed primary treatment objective, while concurrently Shanghai, China (S Liu PhD); major therapeutic advancements in SLE management, mitigating treatment-related comorbidities through Department of Hematology, Institute of Hematology, including the approval of biologics such as belimumab,3 judicious risk–benefit assessment. This balanced Changhai Hospital, Naval anifrolumab,4 and telitacicept5 (in China only). However, approach has emerged as a critical paradigm in the Medical University, Shanghai, a substantial proportion of patients exhibit suboptimal comprehensive management of SLE.6,7 China (Prof J Yang PhD) responses to biological therapies and do not attain The central pathogenic role of autoreactive B cells in long-term DORIS remission. Furthermore, recurrent SLE has established B-cell depletion as a cornerstone 2968 Articles Correspondence to: Research in context Prof Dongbao Zhao, Department of Rheumatology and Evidence before this study autoimmune disease, but the broad applicability of these Immunology, National Key Autologous chimeric antigen receptor (CAR) T-cell therapy has approaches remains unclear. Laboratory of Immunity and shown the ability to produce meaningful long-term remission Inflammation, Changhai Added value of this study in patients with autoimmune diseases, but barriers impede Hospital, Naval Medical To our knowledge, this study is the first-in-human trial of cord University, Shanghai 200433, widespread adoption in terms of limited manufacturing blood or peripheral blood-derived CAR natural killer (NK)-cell China constraints and treatment accessibility, safety profiles requiring dongbaozhao@163.com therapy in autoimmune disease. Allogeneic CD19 CAR NK cells intensive monitoring, and prohibitive manufacturing costs. were tolerable in patients with relapsed or refractory systemic These challenges might be addressed by standardised lupus erythematosus and had favourable safety profiles with allogeneic CAR-bearing cellular therapies. We searched PubMed minimal cytokine release syndrome and no immune effector on Feb 17, 2025, restricting to clinical trials published in any cell-associated neurotoxicity syndrome or other severe adverse language and on any date, using the terms “autoimmune events observed, allowing administration with off-the-shelf disease” AND “allogeneic” AND “chimeric antigen receptor”. availability. Meanwhile, durable remissions were observed in Only one publication on allogeneic CD19-targeted CAR T-cell this trial, indicating meaningful efficacy. therapy in one patient with severe myositis and two patients with systemic sclerosis was identified. This product was Implications of all the available evidence characterised by universal, healthy-donor-derived T cells, Development of allogeneic cell therapies carries distinct genetic engineering of multiple genes (HLA-A, HLA-B, CIITA, advantages over autologous sources in autoimmune disease. TRAC, and PDCD1) with CRISPR-Cas9, and lentivirus Cord blood or peripheral blood-derived CAR NK-cell therapies transduction of CAR. Furthermore, the allogeneic CD19- are actively being studied in haematological malignancies and targeted universal CAR T cells showed tolerability, and achieved autoimmune diseases. Our trial indicates that this approach deep remission in patients with severe myositis and systemic might address challenges with autologous CAR T-cell therapies, sclerosis. This study supports the potential for allogeneic cell particularly in terms of safety files, manufacturing scalability therapeutics to achieve meaningful efficacy in patients with constraints, and improving global treatment accessibility. therapeutic strategy, exemplified by anti-CD20 exhibit intrinsic biological advantages for allogeneic monoclonal antibodies (eg, rituximab).8 However, the applications. Unlike αβT cells, NK-cell activation does clinical efficacy of antibody-mediated B-cell depletion not induce graft-versus-host disease,14,15 enabling direct remains suboptimal, likely due to anatomical limitations therapeutic infusion of donor-derived cells without in eradicating autoreactive B-cell reservoirs within necessitating genetic modification and its associated lymphoid organs and inflammatory niches, coupled with risks of genomic instability.16 NK cells execute potent the persistence of CD20-negative plasma cells that evade cytotoxicity against pathological targets—including anti-CD20 targeting.9 Of notable scientific promise, the infected, malignant, and senescent cells—through paradigm-shifting work by Georg Schett’s group on three principal mechanisms: (1) perforin and granzyme CD19-targeted autologous CAR T-cell therapy in SLE and secretion, (2) death receptor-mediated apoptosis other autoimmune diseases has demonstrated that deep (eg, FAS–FASL and TRAIL–TRAILR pathways), and depletion of CD19+ B-cell lineages—including auto- (3) antibody-dependent cellular cytotoxicity. These reactive clones and tissue-resident precursors—enables attributes position NK cells as promising candidates for immune homeostasis reconstitution in patients with off-the-shelf, allogeneic cell therapies with enhanced SLE. In their study, all eight patients with SLE met safety profiles and scalable production potential.17,18 In LLDAS and DORIS remission criteria at 6 months this pioneering study, we report both short-term and following autologous CAR T-cell therapy, and remained long-term safety and efficacy outcomes of allogeneic free of SLE disease activity through 29 months of CD19-targeted CAR NK-cell therapy in 18 patients with extended follow-up.10 This therapeutic strategy achieves relapsed or refractory SLE. durable drug-free remission in treated SLE cohorts, thereby establishing a novel benchmark for sustained Methods immunological reset in autoimmune therapeutics.10–12 Study design and participants Nevertheless, critical limitations hinder the widespread This open-label, single-arm, prospective, first-in-human clinical adoption of autologous CAR T-cell therapy for case series assessed the safety, efficacy, and autoimmune diseases, including limited manufacturing pharmacokinetics of allogeneic CD19-targeted CAR constraints and treatment accessibility, elevated risks of NK cells in adult patients (aged 18–65 years) with relapsed cytokine release syndrome (CRS) and infectious or refractory SLE and was done at one site in China. This complications, and prohibitive costs.13 Natural killer study protocol was approved by the institutional review (NK) cells, innate immune effectors distinct from board of Shanghai Changhai Hospital Ethics Committee αβT cells by their lack of T-cell receptor expression, (CHEC2023-174). The study protocol and details of Articles specific information on patient enrolment and eligibility with fludarabine (25 mg/m²) and cyclophosphamide See Online for appendix 2 criteria can be found in appendix 2 (pp 16–18). (300 mg/m²) intravenously on days –5 to –3 before CAR In this study, refractory or relapsing SLE was NK-cell infusion. Then, all enrolled patients received operationally defined as follows: sustained disease one treatment cycle of three sequential CAR NK-cell activity (Systemic Lupus Erythematosus Disease Activity infusions at identical dose levels (0·75 × 10⁹ cells [dose Index 2000 [SLEDAI-2K] ≥8) despite at least 6 months of level 1], 1·5 × 10⁹ cells [dose level 2], 3·0 × 10⁹ cells [dose standard therapy comprising adequate-dose gluco- level 3], or 4·5 × 10⁹ cells [dose level 4]) and inter-infusion corticoids combined with at least two distinct intervals (7 days, 5 days, or 3 days). The CD19 CAR NK cells disease-modifying anti rheumatic drugs, including, but were administered within 1 h post-thawing. Additional not limited to, mycophenolate mofetil (1·5–2 g/day), details on management of pre-existing medications cyclophosphamide (0·5–1 g/m² per month), azathioprine (including glucocorticoids) and on CAR NK-cell infusion (1–2 mg/kg per day), ciclosporin (3–5 mg/kg per day), are provided in appendix 2 (p 7). Quantitative real-time tacrolimus (2–4 mg/day), methotrexate (10–15 mg/week), PCR was used to monitor the pharmacokinetic changes of leflunomide (20 mg/day), hydroxychloroquine CAR NK cells in patients’ peripheral blood. (400 mg/day), telitacicept (160 mg/week), belimumab Dose-limiting adverse events and adverse events (10 mg/kg per month), and rituximab (375 mg/m² per according to the Common Terminology Criteria for week for four doses). Key inclusion criteria were Adverse Events (CTCAE) version 5 were monitored in treatment failure (sustained moderate-to-severe activity patients for 28 days. A comprehensive safety surveillance [SLEDAI-2K ≥8] or recurrent flares after 6 months of protocol was implemented across all dose cohorts therapy) and glucocorticoid dependence (inability to (0·75–4·5 × 10⁹ cells) with scheduled monitoring of maintain glucocorticoid tapering). treatment-emergent adverse events during two critical Written informed consent was obtained from all windows: the acute phase (0–28 days post-infusion) and participants in strict accordance with the Center for Drug subacute phase (4–52 weeks post-infusion). Evaluation Good Clinical Practice guidelines for cellular Emerging insights into autoimmune pathogenesis and therapy trials. This study was registered with clinical successes of CAR T-cell therapies in immune ClinicalTrials.gov (NCT06010472) and follow-up is resetting provide a rationale for targeting pathogenic ongoing. B-cell populations to restore immune homeostasis and prolong remission in refractory SLE.19 Pathogenic B-cell Procedures depletion establishes a critical foundation for immune The CAR NK cells were manufactured in the Current resetting and facilitating tissue repair in SLE. To Good Manufacturing Practice-concordant facility of Rui longitudinally characterise this immunological Therapeutics (Nanjing, China) under grade A and B reprogramming, we analysed the distribution of different cleanroom conditions. Detailed information about CD19 B-cell subsets and the clonal diversity of immunoglobulin CAR NK-cell manufacturing and therapy procedures is heavy chain within the B-cell receptor (BCR) repertoire at provided in appendix 2 (pp 2–5). In brief, NK cells were predefined intervals: baseline and 3, 6, 9, and 12 months isolated from healthy, HLA-mismatched donors’ cord post-CAR NK-cell infusion. blood mononuclear cells or peripheral blood mononuclear cells, and NK cells were activated and Outcomes transduced with CAR gene-carrying retrovirus and The primary endpoints of this study were the safety and expanded for 16 days. The CD19 CAR NK-cell products tolerability of CD19 CAR NK cells in patients with were stored and shipped cryopreserved in vapour phase relapsed or refractory SLE, including the incidence of liquid nitrogen (≤–130°C). The final CAR NK-cell product dose-limiting toxicities, CRS, immune effector cell- release was conducted according to established cell associated neurotoxicity syndrome (ICANS), and other therapy product testing criteria. The quality-released adverse events. CRS, ICANS, and other potential adverse CAR NK cells were transported to the clinical site via a events were monitored every day during the first 15 days, cryogenic cold chain system and preserved in liquid and every 7–28 days after the last CAR NK-cell infusion. nitrogen storage. The cytotoxicity of CD19 CAR NK cells The testing of CRS and ICANS were performed following against CD19-positive tumour cells and B cells was the American Society for Transplantation and Cellular determined by a luciferase-based in-vitro cytotoxicity Therapy consensus criteria.20 More detailed information assay. A NALM-6-based xenograft model and tissue about adverse event assessment is described in haematoxylin and eosin staining were used to evaluate appendix 2 (pp 19–21). the in-vivo cytotoxicity and off-target organ safety of the The secondary endpoint was response rate assessed by CD19 CAR NK cells (appendix 2 pp 8–9). the investigator, defined as the proportion of patients This study consisted of schedule escalation and dose who attained SLE Responder Index-4 (SRI-4), SRI-6, escalation, with schedule escalation from 7 days and dose SRI-8, LLDAS, 21 and DORIS 20217 remission. Response escalation commencing at 0·75 × 10⁹ CAR NK cells on assessments were done by investigators every month day 0. All patients received conditioning chemotherapy from baseline to 6 months and every 3 months thereafter 2970 Articles tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP 71 51 31 11 9 7 5 3 1 scitsiretcarahc tneitaP 83 93 34 04 93 73 83 03 32 sraey ,egA elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elameF elaM elameF elameF xeS 8·1 11 71 11 51 6 51 41 6 sraey ,noitarud esaesiD 8 41 01 41 61 41 82 21 41 *erocs K2-IADELS enilesaB seidobitnaotuA ANDsd ·· ·· ANDsd ANDsd ·· ANDsd ANDsd ANDsd ANDsd ANDsd ANDsd ·· ANDsd ANDsd ANDsd ·· ·· daeL CUN ,mS ·· ·· CUN ·· ·· ·· mS ·· ·· ·· ·· CUN mS ·· ·· mS ·· dael-oC ,PNRr ,PNRn ,PNRn ,25-oR ,25-oR ,25-oR 25-oR ,25-oR ASS ·· ASS ,25-oR ,25-oR ,ASS ,BSS ,ASS ,25-oR ·· ,25-oR ,ASS rehtO ASS ASS ,PNRr ASS BSS BSS ,ASS ASS BSS ,ASS ASS 25-oR 07-LCS ASS ASS 25-oR tnemevlovni nagrO – – – + + – + – + + – – – + + + – – lasocum ro nikS + – + – – – – – – + – + + + – + – + yendiK + – + – – – – – – – – + – + – – – + airunietorP – – – – – – – – – – – + – + – – – + airutameaH – – + – – – – – – + – – + + – + – – airuyP – – – – – – – – – – – – – – – – – – gnuL – – – – – – – – – – – – – – – – – – traeH – – – – – + – – + – + + + – + – + – worram enoB – + – + + + + – – + + – + – + – + + stnioj dna selcsuM – – – – – – – + – – – – – + – – – – ralucsaV tnemtaerT + + + + + + + + + + + + + + + + + + sdiocitrococulG – + + + + + + + + + – + + + + + + – eniuqorolhcyxordyH + + – + + – + + – – + + + + + – – + etalonehpocyM + – – – – – – – – – – – – – – – – – enirpoihtazA + – + + – – – – – – – + – + + + – + edimahpsohpolcyC + + + – – – – – + – – – – – – – – – sumilorcaT + – – – – – – – – – – – – – + – – – niropsolciC – – – – – – – – – – – – – – – – – + bamixutiR – – + + + – + – – – – + + + + + + + bamumileB – + + + – + – – – – + – – – – – – – tpecicatileT – – – – – – – – – – – – – – – – – – gIVI + – – – – – + – – + – – + + – + – + †rehtO supul cimetsys=ELS .nietorpoelcunobir lamosobir=PNRr .emosoelcun=CUN .nietorpoelcunobir raelcun=PNRn .nilubolgonummi suonevartni=gIVI .AND dednarts-elbuod=ANDsd .ecnesba setacidni lobmys – eht dna ecneserp setacidni lobmys + ehT serocs rehgih htiw ,501 ot morf segnar erocs K2-IADELS* .B negitna detaler-emordnys-s’nergöjS=BSS .A negitna detaler-emordnys-s’nergöjS=ASS .htimS=mS .0002 xednI ytivitcA esaesiD susotamehtyrE supuL cimetsyS=k2-IADELS .susotamehtyre .edimodilaht deviecer dna ,21 ,9 ,5 stneitap dna ;etaxertohtem deviecer 6 tneitap ;biniticafot deviecer tneitap ;siserehpamsalp deviecer 1 tneitaP† .ytivitca esaesid retaerg gnitacidni enilesab ta ELS htiw stneitap fo scitsiretcarahC :1 elbaT Articles until disease progression following CD19-targeted CAR NK-cell therapy, according to the Physician Global Assessment and disease activity evaluation using the SLEDAI-2K score22 and the British Isles Lupus Assessment Group (BILAG) 2004 index. CAR DNA copy number, lymphocyte counts, double-stranded DNA (dsDNA) antibody and ANA titres, proteinuria, creatine kinase, and complement C3 and C4 concentrations were measured at the indicated follow-up timepoints. Detailed information about CAR NK-cell pharmacokinetics, leukocyte subpopulation, immunoglobulin testing, flow cytometry B-cell subtype analysis, and BCR immuno- globulin heavy chain repertoire sequencing are provided in appendix 2 (pp 7–8). Statistical analysis This clinical trial was designed as a small-sample, single- arm study, and no formal sample size calculation was performed. Descriptive statistics were used to analyse specific parameters from baseline and follow-up data at various timepoints. All analyses were conducted using GraphPad Prism software (version 8.0). Role of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Results CD19-targeting CAR NK cells showed strong cytotoxicity to CD19+ B cells in preclinical models. In our foundational studies, we developed CD19-targeted CAR NK cells incorporating signalling domains (4-1BB co-stimulatory motif and CD3ζ activation module) with membrane- bound IL-15 for sustained persistence (appendix 2 pp 10–11). In-vitro cytotoxicity assays demonstrated potent specific lysis against both malignant CD19+ B-cell lines (NALM-6, Raji) and primary CD19+ B cells isolated from peripheral blood mononuclear cells of patients with SLE (n=3 donors; appendix 2 pp 10–11). The therapeutic efficacy was further validated in an NSG murine xenograft model engrafted with systemic NALM-6-luc cells, where CAR NK-cell infusion (5 × 10⁶ cells per mouse) achieved complete B-cell aplasia (bio- luminescence signal reduction >99% by day 14) in seven of ten mice and greatly prolonged median survival (appendix 2 pp 10–11). Comprehensive safety evaluation confirmed absence of off-target cytotoxicity. A cohort of 18 patients (one [6%] male, 17 [94%] female) with relapsed or refractory SLE meeting predefined inclusion criteria underwent allogeneic CD19-targeted CAR NK-cell therapy between Aug 21, 2023, and June 16, 2024 (appendix 2 p 12). Baseline demographic and clinical characteristics are summarised in table 1. The study population had a median age of 37·5 years (IQR 32·0–39·8; range 19–48) and median disease durat ion of 10·5 years (IQR 4·5–14·8; range 1–17). All 2972 tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP 71 51 31 11 9 7 5 3 1 3 3 3 3 3 3 3 7 7 syad ,lavretni noisufnI 2 4 3 4 4 4 4 3 1 *level esoD 7 7 9 11 21 21 21 51 81 shtnom ,pu-wolloF †ypareht llec-KN RAC-tsop erocs K2-IADELS 8 41 01 41 61 41 82 21 41 shtnom AN 0 2 4 4 5 3 4 3 shtnom AN AN AN AN AN AN AN AN AN 5 2 4 3 shtnom 8 0 4 4 4 5 0 4 2 pu-wollof tsaL ‡ypareht llec-KN RAC-tsop sdiocitrococulG gm gm gm gm gm gm gm gm gm 57·3 gm gm gm 5·7 gm gm gm gm gm 5·7 gm shtnom AN gm gm gm 57·3 gm gm gm gm gm 57·3 gm gm gm gm gm 5·7 gm gm gm gm shtnom AN AN AN AN AN AN AN AN AN gm gm gm 57·3 gm gm gm gm gm gm shtnom gm 5·7 gm gm gm 57·3 gm gm gm gm gm 57·3 gm gm gm 57·3 gm gm gm gm gm 57·3 gm pu-wollof tsaL :4 level esod ;sllec ⁹01 × 0·3 :3 level esod ;sllec ⁹01 × 5·1 :2 level esod ;sllec ⁹01 × 57·0 :1 level esoD* .0002 xednI ytivitcA esaesiD susotamehtyrE supuL cimetsyS=K2-IADELS .rellik larutan=KN .elbaliava ton=AN .rotpecer negitna ciremihc=RAC .yad rep tnelaviuqe-enosinderP‡ .ytivitca esaesid retaerg gnitacidni serocs rehgih htiw ,501 ot 0 morf segnar erocs K2-IADELS† .sllec ⁹01 × 5·4 semoctuo dna ,pu-wollof ,noitamrofni ypareht llec-KN RAC fo yrammuS :2 elbaT Articles part i ci pants had previously received high-dose optimisation during the study period, as shown in table 2. glucocorticoids combined with at least two immuno- The specific days on which patients at each dose level suppressive agents. Previous biological therapies included received their three infusions are shown in appendix 2 belimumab (n=11) and telitacicept (n=5), with two patients (p 12). Manufacturing protocols and quality control receiving both agents. At enrolment, all patients exhibited parameters for CAR NK-cell products—including donor active disease manifestations defined by SLEDAI-2K selection criteria, NK-cell purity (>95%), post-thaw scores of 8 or higher (range 8–28) and at least one severe viability (>80%), and CAR transduction efficiency organ involvement according to BILAG 2004 criteria. (mean 68%)—are comprehensively described in Complete pretreatment therapeutic regimens are detailed appendix 2 (pp 2–5). in table 1. Participants received three sequential infusions Disease activity was longitudinally monitored in all of allogeneic CD19 CAR NK cells across four dose levels: patients following CAR-NK cell therapy using SRI, dose level 1 (0·75 × 10⁹ cells; n=1), dose level 2 LLDAS, and DORIS (2021) remission criteria. At final (1·5 × 10⁹ cells; n=1), dose level 3 (3·0 × 10⁹ cells; n=4), and follow-up, 13 (72%) of 18 patients attained DORIS dose level 4 (4·5 × 10⁹ cells; n=12). Inter-infusion intervals remission and 14 (78%) had attained LLDAS. Serial were modified according to the lymphocyte recovery and assessments revealed progressive clinical improvement. pharmacokinetics of CAR NK cells through protocol At 6 months post-treatment, all evaluable patients A B 6 months 12 months Patient 1 Patient 2 SRI-4 Patient 3 Patient 4 Patient 5 Patient 6 SRI-6 Patient 7 Patient 8 Patient 9 Patient 10 SRI-8 Patient 11 Patient 12 Dose level 1 Patient 13 Dose level 2 LLDAS Patient 14 Dose level 3 Patient 15 Dose level 4 Patient 16 On follow-up Patient 17 DORIS remission DORIS remission (2021) Patient 18 LLDAS 0 1 2 3 4 5 6 7 8 9101112131415161718192021222324 Time from CAR NK-cell infusion (months) C 28 20 12 4 0 3 6 9 12 15 18 18 18 17 14 9 3 1 Figure 1: Clinical efficacy of allogeneic CD19 CAR NK-cell therapy (A) Follow-up and outcomes in patients with SLE after CD19 CAR-NK cell therapy (n=18). The coloured bars represent specific dose level groups: dose level 1 (0·75 × 10⁹ CAR NK cells); dose level 2 (1·5 × 10⁹ CAR NK cells); dose level 3 (3·0 × 10⁹ CAR NK cells); and dose level 4 (4·5 × 10⁹ CAR NK cells). The red dashed vertical line indicates the 12-month follow-up timepoint. (B) SRI response, LLDAS, and DORIS remission rates of patients who received CAR NK-cell therapy at 6-month and 12-month follow-up timepoints. (C–E) Outcomes of CD19 CAR NK-cell therapy in patients with SLE, including SLEDAI-2K scores (ranges from 0 to 105; increasing scores indicate more disease activity), PGA scores (ranges from 0 to 3; 0 indicates no disease activity, 3 indicates the most severe disease activity), and levels of urinary protein excretion at 3-month, 6-month, 9-month, 12-month, 15-month, and 18-month follow-up timepoints. For urinary protein excretion, the dashed horizontal line indicates the upper limit of the normal range. The protein-to-creatinine ratio was calculated with urinary protein measured in milligrams and urinary creatinine in grams. Each circle represents data from a single patient at a specific timepoint. Grey circles denote baseline data; circles of the same colour, with gradation from light to dark shades, represent data from different follow-up timepoints. CAR=chimeric antigen receptor. LLDAS=lupus low disease activity state. NK=natural killer. PGA=Physician Global Assessment. SLE=systemic lupus erythematosus. SLEDAI-2K=Systemic Lupus Erythematosus Disease Activity Index 2000. SRI=SLE Responder Index. erocs K2-IADELS D 3 1 0 3 6 9 12 15 18 Time from CAR NK-cell infusion (months) Number of 18 18 17 14 9 3 1 patients erocs AGP E 5000 3000 1000 0 3 6 12 15 18 Time from CAR NK-cell infusion (months) 18 17 17 16 16 14 9 3 1 oitar eninitaerc-ot-nietorp enirU )g/gm( Total 17 9 Response 17 9 Response rate 100% 100% Total 17 9 Response 16 9 Response rate 94% 100% Total 17 9 Response 12 7 Response rate 71% 78% Total 17 9 Response 13 6 Response rate 77% 67% Total 17 9 Response 10 6 Response rate 59% 67% 4 5 9 Time from CAR NK-cell infusion (months) Articles (17 of 17) had attained SRI-4 response, with 16 (94%) of eight (50%) of 16 patients and partial recovery 17 reaching SRI-6 thresholds. Furthermore, 13 patients (≥25% increase) in 11 (61%) patients. Anti-dsDNA titres (76%) maintained LLDAS status and ten (59%) fulfilled decreased in nine (69%) of 13 seropositive patients DORIS remission criteria. This therapeutic effect (appendix 2 p 14), SSA antibody titres were decreased in persisted through 12-month follow-up, with 100% SRI-4 two (14%) of 14 patients, Ro52 antibody titres were and SRI-6 response rates (nine of nine patients), LLDAS decreased in three (27%) of 11 patients, and RNP antibody attainment in six (67%) patients, and DORIS remission titres were decreased in one (50%) of two patients. Of the in six (67%) patients (figure 1A, B). Notably, both five patients with proteinuria, proteinuria resolved SLEDAI-2K scores and Physician Global Assessment completely in four (80%) patients by month 3, with the demonstrated substantial reductions by month 3 post- remaining patient demonstrating more than treatment. Remarkably, 14 patients had SLEDAI-2K 40% reduction in urinary protein excretion (figure 1E). scores of 4 or below (including three with complete Long-term safety monitoring (median follow-up disease resolution [SLEDAI=0]) during this period 11 months [IQR 9·0–12·0]) showed no disease flares or (figure 1C, D). BILAG 2004 analysis confirmed treatment-related adverse events. All patients successfully comprehensive disease control, with scores of D or E in discontinued all immunosuppressants (eg, mycophenolate all organ domains, in patients meeting LLDAS or DORIS mofetil) and antimalarials (eg, hydroxychloroquine), and criteria at 6 months. Immunological profiling revealed tapered glucocorticoids to maintenance doses of 7·5 mg normalisation of complement C3 concentrations in prednisone-equivalent per day or lower, except for 5000 3000 1000 0 1 3 4 5 6 7 81011141517192128 18181416114176 3 5143 1 2163 18 18 18 17 18 18 18 18 17 17 13 9 18141313 817 5 3 2 1517181714 9 0 1 3 4 5 6 7 8 1011141521 Figure 2: Pharmacokinetics, pharmacodynamics, and safety profile of CAR NK-cell therapy In every graph in this figure, each circle represents data from an individual patient at a single timepoint, with distinct colours indicating different follow-up timepoints. (A) Quantitation of circulating CAR NK cells among total peripheral blood mononuclear cells at various days after CAR NK-cell treatment in 18 patients by quantitative real-time PCR. (B) Counts of CD19+ B cells in peripheral blood of 18 patients with SLE at baseline and various days and months of follow-up after CAR NK-cell therapy. The dashed horizontal line represents the normal reference value. (C) IgG concentrations of 18 patients with SLE at baseline and various days and months of follow-up after CAR NK-cell therapy. (D–F) IL-6 concentrations, leukocyte counts, and platelet counts of 18 patients with SLE before and after lymphodepletion with cyclophosphamide and fludarabine, and treatment with CAR NK cells at baseline and follow-up. For the leukocyte counts, red circles denote data for patient 2 and purple triangles denote data for patient 13, as both of their leukocyte counts remained below the normal reference range by the last follow-up. For the platelet counts, light red circles denote data for patient 8 and patient 13, as both of their platelet counts remained below the normal reference range by the last follow-up. CAR=chimeric antigen receptor. NK=natural killer. SLE=systemic lupus erythematosus. 2974 AND gµ rep seipoc RAC Time from CAR NK-cell infusion (days) Number of patients 120 60 0 181812 18 3 1415 3 4 4 2 3 17 )Lm/gp( noitartnecnoc 6-LI 1000 600 200 0 Time from CAR NK-cell infusion (days) Number of patients Lµ rep sllec B +91DC 20 10 0 Baseline 1 day 7 day 1 s 4 day 2 s 8 d 2 a m ys on 3 t h m s on 4 t m hs on 5 t h m s on 6 t h m s on 9 t h m s o 1 n 2 t h m s onths )L/g( noitartnecnoc GgI Baseline 1 da 3 y day 4 s day 6 s day 7 s day 8 s day 9 s da 1 y 4 s da 2 y 1 s d 1 a m ys 3 o n m th o 6 n t m hs o 9 n m th 1 o s 2 n t m h o s nths 10 500 8 400 6 300 4 200 2 Patient 2 100 Patient 13 Patient 13 0 0 Patient 8 0 1 2 3 4 5 6 9 12 0 1 2 3 4 5 6 9 12 Time from CAR NK-cell infusion (months) Time from CAR NK-cell infusion (months) 18 18 18 18 18 17 17 14 9 18 18 18 18 18 17 17 14 9 )L/⁹01×( tnuoc etycokueL )L/⁹01×( tnuoc teletalP A B C Time from CAR NK-cell infusion Time from CAR NK-cell infusion D E F Patient 7 Articles patient 8 (table 2), who maintained a prednisone dose of 10 mg/day due to refractory thrombocytopenia secondary to long-term haematological involvement; the patient exhibited incomplete platelet recovery following CAR NK-cell therapy. In contrast to autologous CAR T cells demonstrating robust post-infusion expansion, our previous clinical trials in haematological malignancies as well as other clinical studies revealed a distinct pharmacokinetic profile of allogeneic CAR NK cells.23 Quantitative real-time PCR analysis demonstrated transient peripheral blood CAR NK-cell detection, with pharmacokinetic peaks occurring in distinct temporal patterns: six (33%) of 18 patients reached maximal concentration (C ) at day 1, max while seven (39%) and five (28%) reached C at days 3–4 max and day 7 post-infusion, respectively (figure 2A). Profound B-cell depletion (CD19+ cells <5 cells per μL, lower limit of quantification) was observed in 16 (89%) of 18 patients, with residual CD19+ cells (5–10 cells per μL) persisting in two patients. B-cell reconstitution kinetics showed two distinct patterns: 12 (67%) of 18 patients demonstrated early recovery (≥50% baseline B-cell counts) by month 2, whereas the remaining six (33%) patients exhibited delayed recovery initiating at months 3–4 (median time to recovery 3 months [IQR 2–5]; figure 2B). Immunoglobulin monitoring revealed clinically stable profiles: serum IgG concentrations showed transient reduction post-infusion but remained above 5 g/L in all patients. Critically, no patients required immunoglobulin replacement therapy throughout follow-up (figure 2C). Notably, CRS, ICANS, and other CAR NK-cell-specific toxicities were absent in patients receiving dose levels 1–3 (n=7). A single grade 1 CRS event (fever only) occurred in patient 7 (dose level 4 cohort, 4·5 × 10⁹ cells), manifesting within 24 h post-infusion and resolving spontaneously within 48 h (figure 2D). This patient was administered one dose of 162 mg tocilizumab; the CRS subsequently resolved, and body temperature returned to normal. Although patient 6 exhibited fever, this was due to a bacterial upper respiratory tract infection after lymphodepleting preconditioning and was not associated with CAR NK-cell therapy, as IL-6 concentrations were not increased (figure 2D and appendix 2 p 15). No neurotoxicity, haematopoietic suppression, or dose-limiting toxicities were observed in any cohort (table 3). Longitudinal monitoring demonstrated stable haemodynamic parameters (blood pressure: 110–130/70–85 mm Hg), stable inflammatory markers (C-reactive protein <10 mg/L), IgG above 5 g/L in all patients, and preserved leukocyte homeostasis (appendix 2 p 15). Notably, CAR NK-cell administration did not induce lymphodepletion, with CD4+ and CD8+ T-cell subsets and NK-cell populations maintaining physiological reconstitution patterns (appendix 2 p 15). Longitudinal immunophenotyping revealed profound B-cell repertoire remodelling post-CAR NK-cell therapy tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP tneitaP 71 51 31 11 9 7 5 3 1 – – – – – – – – – – – + – – – – – – SRC – – – – – – – – – – – – – – – – – – SNACI – + – – – – – – + + + + + – + + – – eugitaF – – – – – – – – – – – + + – – – – – reveF – – – – – – – – + – + + + – + – – – etiteppa desaerceD – – – – + – – – – – + + – – – – – – gnitimov dna aesuaN – – – – – – – – – – – – – – – – – – aeohrraiD – + – – – – – – – – – – – – – – – – ssenizziD – – + – – + – – – – – – – – + – – – desaercni TLA – – – – – + – – – – – – – – – – + – desaercni TSA – – – – – – – – – – – – – – – – – – yticixot worram enoB – – – – – – – – – – – – – – – – – – GgI woL – – – – – – – – – – – + – – – – – – tnemtaert bamuzilicoT ot detaler saw hcihw ,7 tneitap ni )revef detaicossa dna( SRC eht morf trapa tnemtaert edimahpsohpolcyc–enibaradufl ot detaler eb ot deredisnoc erew stneve esrevda detsil llA .ecnesba setacidni lobmys – eht dna ecneserp setacidni lobmys + ehT .esarefsnartonima eninala=TLA .noitcejni cimredopyh yb gm bamuzilicot deviecer tneitap eht dna ,1 edarg saw 7 tneitap ni SRC ehT .noitcefni tcart yrotaripser reppu lairetcab ot detaler saw hcihw ,6 tneitap ni revef eht dna ,ypareht llec-KN RAC .susotamehtyre supul cimetsys=ELS .rellik larutan=KN .emordnys yticixotoruen detaicossa-llec enummi=SNACI .emordnys esaeler enikotyc=SRC .rotpecer negitna ciremihc=RAC .esarefsnartonima etatrapsa=TSA ELS htiw stneitap ni tnemtaert edimahpsohpolcyc–enibaradufl dna ypareht llec-KN RAC 91DC fo ytefas mret-trohS :3 elbaT Articles 90 70 50 30 10 Baseline months mon L t a h s s t follo w-up Figure 3: Reconstitution of B cells and BCR heavy chain distribution and diversity after CD19 CAR NK-cell treatment (A) Quantification of the percentage of naive B cells (CD21+CD27−), memory B cells (CD21+CD27+), activated memory B cells (CD11c+), and plasmablasts (CD20−CD38+) between baseline (before CAR NK-cell infusion) and B-cell reconstitution at 3 months (n=18), 6 months (n=17), and the last follow-up timepoint (n=14) analysed by flow cytometry. Each circle represents data from an individual patient at a single timepoint, with distinct colours indicating different follow-up timepoints. (B–C) Assessment of heavy chain distribution and diversity in BCRs in 17 patients by high-throughput RNA sequencing at baseline and following CAR NK-cell therapy (6-month and last follow-up timepoints). BCR=B-cell receptor. CAR=chimeric antigen receptor. NK=natural killer. 2976 −72DC+12DC tnecreP +72DC+12DC tnecreP 90 70 50 30 10 Baseline months mon L t a h s s t follo w-up Time from CAR NK-cell Time from CAR NK-cell Time from CAR NK-cell Time from CAR NK-cell infusion infusion infusion infusion +c11DC−12DC+02DC tnecreP +83DC+72DC−02DC tnecreP A 60 20 16 40 30 12 8 4 0 0 Baseline months mon L t a h s s t follo w-up Baseline months mon L t a h s s t follo w-up Number of patients 18 18 17 14 18 18 17 14 18 18 17 14 18 18 17 14 B 80 40 0 C )%( noitubirtsid niahc yvaeh RCB Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Bas e l m in o e nths Time from CAR NK-cell infusion 600000 400000 200000 ytisrevid niahc yvaeh RCB Naive B cells Memory B cells Activated memory B cells Plasmablasts IgM IgD IgG IgA Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 La B s a t s f e o l l i l n o e w-u L p a B st a s f e o l l i l n o e w-u L p a B st a s fo el l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B st a s fo el l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B st a s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-u L p a B st a s f e o l l i l n o e w-u L p a B s a t s f e o l l i l n o e w-up Time from CAR NK-cell infusion Articles (figure 3A). Naive B-cell subsets dominated peripheral established SLE efficacy. However, a single administration reconstitution (up to 96%), concomitant with drastically of low-dose cyclophosphamide is not sufficient to induce reduced CD21+CD27+ memory B cells. Early-phase long-term durable remission in these patients with reductions in CD27+CD38+ plasmablasts and SLE- severe SLE, which is supported by evidence that associated CD11c+ activated memory B cells were treatments including cyclophosphamide or observed, although transient subset fluctuations mycophenolate mofetil and belimumab before CAR occurred in four (29%) of 14 patients at 6-month follow- NK-cell therapy failed to control the symptoms of disease up. Furthermore, BCR deep sequencing demonstrated in these patients. Tur and colleagues recently reported isotype class-switch reversal in ten (71%) of 14 patients, that CD19 CAR T-cell therapy-induced remission with dominant heavy-chain isotypes shifting from (≥12 months) in autoimmune diseases correlates with IgG/IgA to IgD/IgM after CAR NK-cell therapy tissue-resident B-cell depletion.25 Analogously, CAR (figure 3B). Clonal architecture analysis revealed two key NK cells demonstrated comparable lymphoid homing transitions: progressive contraction of dominant clones capacity, suggesting that their deep B-cell depletion and polyclonal repertoire expansion (figure 3C). capacity underlies the observed durable responses. Another critical insight from allogeneic CAR NK-cell Discussion therapy is the transient in-vivo persistence of infused Engineered autologous CAR T-cell therapy has shown cells (up to 14 days), consistent with the intrinsic substantial promise in the treatment of autoimmune biological properties of NK cells, including their limited diseases, particularly SLE, but notable limitations lifespan.26,27 Despite the absence of clonal expansion, hindering clinical translation and broad implementation therapeutic efficacy was maintained through potent of this modality remain, such as manufacturing cytotoxic activity, achieving deep B-cell depletion followed constraints, restricted accessibility, toxicity profile, and by naive B-cell reconstitution (≥88% by month 3) in most prohibitive costs.24 The findings of this first-in-human patients. This contrasts with CAR T-cell therapies, where study support the potential for allogeneic CAR NK-cell prolonged persistence drives sustained B-cell aplasia, therapies derived from healthy donors to address these increasing infection risks. Although discontinuation of barriers. high-dose glucocorticoids and immunosuppressants This study provides preliminary evidence supporting triggered immune reconstitution that might paradoxically the therapeutic potential of allogeneic CD19-targeted accelerate allogeneic CAR NK-cell clearance, rapid CAR NK-cell therapy in patients with relapsed or clinical improvement was nevertheless observed refractory SLE, demonstrating sustained efficacy, following CAR NK-cell infusion. Meanwhile, longitudinal favourable tolerability and safety, and transient cellular flow cytometry demonstrated sustained naive B-cell persistence. Notably, CAR NK-cell pharmacokinetics and predominance with concomitant decrease of memory B-cell depletion dynamics—characterised by naive B-cell B cells. Considering all patients were resistant to several dominance in reconstituted populations—remained immunosuppressive therapies, our data indicate that consistent across heterogeneous disease subtypes and CD19 CAR NK cells could deeply deplete memory B cells previous therapies. However, dose–response relation- and plasmablasts in the peripheral and might directly ships remain undefined due to limited cohort sizes of target tissue-infiltrating B cells, which is in contrast to dose levels 1 and 2 (n=1 per group). Furthermore, we the CD20 monoclonal antibody rituximab that could only adjusted the infusion intervals of CAR NK cells based on eliminate CD20+ B cells from circulation.28 This their post-infusion pharmacokinetic profiles and the observation is corroborated by comprehensive analysis of lymphocyte reconstitution kinetics in patients with SLE. BCR repertoire characteristics, demonstrating a distinct Infusion interval optimisations revealed that shortened class-switching transition from IgA/IgG to IgM/IgD infusion intervals (7-day vs 3-day regimens) maintained isotypes after CAR NK-cell treatment. Notably, the therapeutic safety and efficacy while improving treatment immunophenotypic evolution manifests as concomitant accessibility through reduced hospitalisation duration. reductions in both singleton clones and high-frequency Importantly, all patients sustained disease remission or clones, concurrent with expansion of total clonal diversity control (Physician Global Assessment <1) with and overall repertoire breadth following CAR NK-cell glucocorticoid doses of 5 mg/day or lower up to final therapy. Accordingly, we propose that CAR NK-cell- follow-up (except for patients 13 and 18). mediated transient cytotoxic elimination of circulating The sustained clinical remission (median 11 months) and tissue-resident autoreactive B-cell clones achieves: observed in patients with SLE following three-dose (1) immediate disruption of the pathogenic cascade CD19-targeted CAR NK-cell therapy suggests that through depletion of antigen-experienced memory B-cell durable disease control could be achievable through reservoirs, and (2) sustained central tolerance complete B-cell reconstitution without relapse. All re-establishment via regenerative lymphopoiesis patients received fludarabine–cyclophosphamide generating polyclonal, self-tolerant naive B-cell lymphodepletion, a regimen potentially contributing to repertoires, ultimately manifesting as durable immune early symptom resolution given cyclophosphamide’s tolerance and disease remission. Articles NK cells demonstrate attenuated cytokine secretory CAR-NK therapy, it is critical to note its differential (especially IL-6) capacity compared with T lymphocytes, efficacy profile compared with autologous CAR T-cell while maintaining potent innate antimicrobial effector therapy in SLE management. CAR NK-cell therapy functions including direct cytolysis of virus-infected demonstrated slower kinetics in achieving dsDNA cells, granulocyte recruitment, and pathogen antibody seroconversion and did not induce DORIS clearance.29,30 This biological profile translates to superior remission in all patients. The aforementioned safety outcomes in CD19 CAR NK-cell therapy, with only limitations of this study will be systematically addressed one of 18 patients developing grade 1 CRS (CTCAE in future prospective registry trials, designed to establish version 5.0; transient fever <39°C resolving within 48 h), evidence-based dosing parameters and longitudinally and zero instances of neurotoxicity (ICANS) or evaluate the sustained efficacy and safety of CAR NK-cell haematological toxicity and other CAR NK-cell treatment- therapy in SLE. related adverse effects. Notably, the absence of Contributors hypogammaglobulinaemia (IgG >5 g/L in all patients) JG, ML, MS, YY, and RK contributed equally. JG, JY, MS, ML, and DZ correlated with zero treatment-emergent severe designed the treatments and analyses. YY, RK, XX, SL, QC, and XL monitored the patients. YY, RK, and XX collected clinical data. YW, EX, infections (no respiratory or urinary tract infections over and MS did molecular analyses. JG, MS, ML, and DZ wrote the 6-month follow-up) contrasts sharply with the manuscript. JG, YY, MS, and DZ directly accessed and verified the data. 40–93% infection rates reported in cohorts treated with All authors had full access to the data in the study and had final autologous CD19 CAR T-cell therapy,10,12 potentially responsibility for the decision to submit for publication. attributable to preserved innate immune competence by Declaration of interests infused NK cells retaining functional pathogen JG has received grants from the Shanghai Municipal Health Commission and Changhai Hospital Affiliated to Naval Medical recognition receptors (eg, NKG2D, CD16) critical for University. MS has received grants from the National Natural Science antiviral defence and timely B-cell reconstruction. This Foundation of China. YW, EX, and MS are employees of Rui safety advantage is clinically significant given that Therapeutics. All other authors declare no competing interests. infectious complications accounted for 51% of all non- Data sharing relapse mortality (mantle cell lymphoma, 11%; multiple To request access to the de-identified study data, please contact the myeloma, 8·0%) events following autologous CAR T-cell corresponding author. Requests will be reviewed and written applications from investigators with the academic capability to therapy in lymphoma and multiple myeloma.31 undertake the work proposed will be considered. This study has some limitations, such as the small Acknowledgments sample size and the fact that the biological mechanisms This study was supported by Shanghai Municipal Health Commission accounting for the distinct CAR NK-cell pharmacokinetic (202340061 to JG), Changhai Hospital Affiliated to Naval Medical profiles and variable B-cell reconstitution have yet to be University (GH145-34 to JG), and National Natural Science Foundation determined. Moreover, the methodology for determining of China (82372625 to MS). We thank the patients for providing their consent to publish this information. the precise dosage for individual patients remains unclear, and the unequal sample sizes across groups References 1 Fasano S, Milone A, Nicoletti GF, Isenberg DA, Ciccia F. Precision compromised the assessment of dose–response medicine in systemic lupus erythematosus. Nat Rev Rheumatol relationships. Since only a fraction of patients received 2023; 19: 331–42. cyclophosphamide before fludarabine–cyclophosphamide 2 Hoi A, Igel T, Mok CC, Arnaud L. Systemic lupus erythematosus. Lancet 2024; 403: 2326–38. and CAR NK-cell therapy, it is possible that the 3 Zhang F, Bae SC, Bass D, et al. A pivotal phase III, randomised, fludarabine–cyclophosphamide contributed to the placebo-controlled study of belimumab in patients with systemic observed early clinical benefit. The abbreviated follow-up lupus erythematosus located in China, Japan and South Korea. Ann Rheum Dis 2018; 77: 355–63. duration also necessitates extended observation to 4 Kalunian KC, Furie R, Morand EF, et al. A randomized, placebo- delineate long-term safety and therapeutic outcomes. controlled phase III extension trial of the long-term safety and In conclusion, this first-in-human study showed that tolerability of anifrolumab in active systemic lupus erythematosus. Arthritis Rheumatol 2023; 75: 253–65. allogeneic CD19-targeted CAR NK cells derived from 5 Wu D, Li J, Xu D, et al. Telitacicept in patients with active systemic HLA-mismatched healthy donors were tolerable, with lupus erythematosus: results of a phase 2b, randomised, double- minimal treatment-related adverse events that have blind, placebo-controlled trial. Ann Rheum Dis 2024; 83: 475–87. burdened other effective immunotherapies, and showed 6 Siegel CH, Sammaritano LR. Systemic lupus erythematosus: a review. JAMA 2024; 331: 1480–91. encouraging preliminary efficacy in patients with 7 van Vollenhoven RF, Bertsias G, Doria A, et al. 2021 DORIS relapsed or refractory SLE who had been systematically definition of remission in SLE: final recommendations from pretreated. Notably, the cryopreserved, off-the-shelf an international task force. Lupus Sci Med 2021; 8: e000538. 8 Stockfelt M, Teng YKO, Vital EM. Opportunities and limitations of nature of the allogeneic CD19 CAR NK cells enables B cell depletion approaches in SLE. Nat Rev Rheumatol 2025; immediate availability and discontinuation of high-dose 21: 111–26. glucocorticoids and sustained withdrawal of 9 Tur C, Eckstein M, Velden J, et al. CD19-CAR T-cell therapy induces deep tissue depletion of B cells. Ann Rheum Dis 2025; conventional immunosuppressants, achieving 84: 106–14. resolution of treatment-emergent toxicities associated 10 Müller F, Taubmann J, Bucci L, et al. CD19 CAR T-cell therapy in with long-term intensive immunosuppressive regimens. autoimmune disease - a case series with follow-up. N Engl J Med While acknowledging the therapeutic potential of 2024; 390: 687–700. 2978 Articles 11 Mackensen A, Müller F, Mougiakakos D, et al. Anti-CD19 CAR 22 Gladman DD, Ibañez D, Urowitz MB. Systemic lupus T cell therapy for refractory systemic lupus erythematosus. Nat Med erythematosus disease activity index 2000. J Rheumatol 2002; 2022; 28: 2124–32. 29: 288–91. 12 Wang W, He S, Zhang W, et al. BCMA-CD19 compound CAR T cells 23 Fu Y, Xu Z, Wu C, et al. Genetically modified CD19-targeting IL-15 for systemic lupus erythematosus: a phase 1 open-label clinical trial. secreting NK cells for the treatment of systemic lupus Ann Rheum Dis 2024; 83: 1304–14. erythematosus. Ann Rheum Dis 2025; published online Aug 25. 13 Verdun N, Marks P. Secondary cancers after chimeric antigen https://doi.org/10.1016/j.ard.2025.07.028. receptor T-cell therapy. N Engl J Med 2024; 390: 584–86. 24 Scherlinger M, Nocturne G, Radic M, et al. CAR T-cell therapy in 14 Fang F, Xie S, Chen M, et al. Advances in NK cell production. autoimmune diseases: where are we and where are we going? Cell Mol Immunol 2022; 19: 460–81. Lancet Rheumatol 2025; 7: e434–47. 15 Crinier A, Narni-Mancinelli E, Ugolini S, Vivier E. SnapShot: 25 Tur C, Eckstein M, Velden J, et al. CD19-CAR T-cell therapy induces natural killer cells. Cell 2020; 180: 1280–1280.e1. deep tissue depletion of B cells. Ann Rheum Dis 2024; 84: 106–14. 16 Lamers-Kok N, Panella D, Georgoudaki AM, et al. Natural killer 26 Sun JC, Lanier LL. NK cell development, homeostasis and function: cells in clinical development as non-engineered, engineered, and parallels with CD8+ T cells. Nat Rev Immunol 2011; 11: 645–57. combination therapies. J Hematol Oncol 2022; 15: 164. 27 Perera Molligoda Arachchige AS. Human NK cells: from 17 Liu E, Marin D, Banerjee P, et al. Use of CAR-transduced natural development to effector functions. Innate Immun 2021; 27: 212–29. killer cells in CD19-positive lymphoid tumors. N Engl J Med 2020; 28 Kamburova EG, Koenen HJ, Borgman KJ, ten Berge IJ, Joosten I, 382: 545–53. Hilbrands LB. A single dose of rituximab does not deplete B cells in 18 Marin D, Li Y, Basar R, et al. Safety, efficacy and determinants of secondary lymphoid organs but alters phenotype and function. response of allogeneic CD19-specific CAR-NK cells in CD19+ B cell Am J Transplant 2013; 13: 1503–11. tumors: a phase 1/2 trial. Nat Med 2024; 30: 772–84. 29 Smith DM, Schafer JR, Tullius B, Witkam L, Paust S. Natural killer 19 Ramírez-Valle F, Maranville JC, Roy S, Plenge RM. Sequential cells for antiviral therapy. Sci Transl Med 2023; 15: eabl5278. immunotherapy: towards cures for autoimmunity. 30 Vivier E, Artis D, Colonna M, et al. Innate lymphoid cells: 10 years Nat Rev Drug Discov 2024; 23: 501–24. on. Cell 2018; 174: 1054–66. 20 Lee DW, Santomasso BD, Locke FL, et al. ASTCT consensus 31 Cordas Dos Santos DM, Tix T, Shouval R, et al. A systematic review grading for cytokine release syndrome and neurologic toxicity and meta-analysis of nonrelapse mortality after CAR T cell therapy. associated with immune effector cells. Biol Blood Marrow Transplant Nat Med 2024; 30: 2667–78. 2019; 25: 625–38. 21 Franklyn K, Lau CS, Navarra SV, et al. Definition and initial validation of a Lupus Low Disease Activity State (LLDAS). Ann Rheum Dis 2016; 75: 1615–21. --- [PDF原文](https://sci-net.xyz/storage/8010893/f672bd4feec3a406f986a2964acd99b414f3e8629635149edfbca740f248698f/Efficacy-and-safety-of-allogeneic-CD19-CAR-NK-cell-therapy-in-systemic-lupus-erythematosus.pdf) DOI: 10.1016/S0140-6736(25)01671-X