# paper.pdf ## Metadata - PDFFormatVersion=1.7 - Language=en-US - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - CreationDate=D:20260725071953Z - Creator=Adobe InDesign 18.3 (Windows) - Custom.CrossMarkDomains[1]=springer.com - Custom.CrossMarkDomains[2]=springerlink.com - Custom.CrossmarkDomainExclusive=true - Custom.CrossmarkMajorVersionDate=2010-04-23 - Custom.doi=10.1186/s13046-026-03785-5 - ModDate=D:20260728152133+02'00' - Producer=Adobe PDF Library 17.0; modified using iText® 5.5.13.4 ©2000-2024 iText Group NV (SPRINGER SBM; licensed version) - Trapped=False - crossmark:crossmarkdomains=springer.comspringerlink.com - pdfx:crossmarkdomains=springer.comspringerlink.com ## Contents ### Page 1 © The Author(s) 2026. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) andthe source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in thisarticle are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not includedin the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you willneed to obtain permission directly from the copyright holder. To view a copy of this licence, visit. TheCreative Commons Public Domain Dedication waiver () applies to the data made availablein this article, unless otherwise stated in a credit line to the data. †Robert J. Oldham and Linda Mårtensson contributed equally to thispaper. †Stephen A. Beers, Ingrid Teige, Mark S. Cragg and Björn Frendéusshare senior authorship. *Correspondence: Björn Frendéus Full list of author information is available at the end of the article Abstract Background Fc-gamma receptors (FcγRs) regulate IgG antibody activity, and Fc-engineering is a proven methodto improve the efficacy of tumor-targeting antibodies. Here, we explore tailored FcγR blockade to enhance thetherapeutic efficacy and tolerability of immune checkpoint-blocking (ICB) antibodies. Methods Mechanistically matched murine surrogate and human lead FcγR-blocking and immune checkpoint-blocking antibodies were used to study whether tailored FcγR-blockade, targeting FcγRIIB selectively or all FcγRs,can enhance the efficacy and overcome resistance to immune checkpoint therapy in vivo and in vitro. Mechanisticstudies were performed with clinical reagents, including ipilimumab, nivolumab, pembrolizumab, and humanFcγRIIB-selective (BI-1607) and pan-FcγR-blocking (BI-1206) antibodies, using human cells and transgenic animals withclinically relevant expression of immune checkpoint receptors. Results We demonstrate that FcγRIIB-selective and pan-FcγR-blocking antibodies increase the in vivo efficacy ofαCTLA-4 and αPD-1 antibodies, respectively. FcγRIIB-selective antibody enhancement of αCTLA-4 was associatedwith increased intratumoral Treg depletion, myeloid reprogramming, interferon-γ and CXCL10-induction, andincreased activated effector CD8+ T cells, correlating with higher activating-to-inhibitory (A:I) FcγR engagementratios. Conversely, pan-FcγR blockade protected αPD-1-coated T cells from macrophage phagocytosis, increasingintratumoral activated CD8+ T cells by decreasing activating and inhibitory FcγRs. Conclusions Our studies provide in vivo proof of concept that tailored FcγR blockade enhances immune checkpointtherapy and overcomes resistance through mechanistically distinct pathways. Clinical trials with tailored humanFcγRIIB-blocking antibodies are ongoing. Keywords Fc gamma receptors, Immune checkpoint blockade, Antibody Immunotherapy, Drug Resistance Tailored FcγR blockade enhances immunecheckpoint therapy and overcomes resistance Robert J. Oldham1†, Linda Mårtensson2†, Monika Semmrich2, Niyaz Yoosuf2, Petra Holmkvist2, Lara V. Graham1,Martin C. Taylor1, Kirstie L. S. Cleary1, Mona Yazdani1, Josephine F. Buckingham1, Ali Roghanian1, Ingrid Karlsson2,Stephen A. Beers1†, Ingrid Teige2†, Mark S. Cragg1† and Björn Frendéus1,2*†[ ![http://orcid.org/0000-0003-3177-8686]() ](http://orcid.org/0000-0003-3177-8686) Oldham et al. Journal of Experimental & Clinical Cancer Research (2026) 45:167 [ ![http://crossmark.crossref.org/dialog/?doi=10.1186/s13046-026-03785-5&domain=pdf&date_stamp=2026-7-25]() ](http://crossmark.crossref.org/dialog/?doi=10.1186/s13046-026-03785-5&domain=pdf&date_stamp=2026-7-25) ![BioMed Central Logo]() ### Page 2 Introduction Immune checkpoint blockade (ICB) with antibodies tar-geting PD-1, PD-L1, or CTLA-4 has transformed cancersurvival and is used to treat more than forty cancer types[]. Despite this, many patients fail to respond or developresistance during therapy []. Combining αPD-1 andαCTLA-4 increases efficacy, but dose-dependent toxic-ity limits the use of available αCTLA-4 antibody regi-mens []. Identifying and overcoming mechanisms thatmediate resistance to αPD-(L)1 and developing means todeliver safer, efficacious αCTLA-4-based therapies couldimprove survival and bring the clinical benefits of ICB tocurrently non-responding cancer types and patients. Fc gamma receptors (FcγR) regulate IgG antibodyactivity, and their importance for tumor-targeting anti-bodies is well documented [–]. Several activatingFcγRs (FcγRI, IIA, and IIIa) and a single inhibitory FcγR(FcγRIIB) collectively regulate FcγR+ immune effectorcell-mediated responses during antibody therapy [,].A high ratio of activating to inhibitory FcγRs promoteseffector cell activation and downstream cellular events,including the depletion of antibody-coated target cells[]. Consistent with the positive and negative regulationof IgG-induced effector cell responses by activating andinhibitory FcγRs, we previously demonstrated the thera-peutic potential of blocking the inhibitory FcγRIIB toimprove antibody therapy. Antagonistic FcγRIIB antibod-ies enhanced the activity of rituximab and trastuzumaband overcame rituximab resistance in vivo [], with clini-cal investigation ongoing (NCT03571568). Emerging data suggest that FcγRs also influencethe effectiveness of immune-modulatory antibodies,including those directed against CTLA-4 and PD-(L)1(reviewed in [,]). These antibodies were createdto block inhibitory signals in CD8+ T cells, effectivelyreleasing the brakes on the immune system to elimi-nate cancer cells. This process does not rely on a role forFcγRs. Surprisingly, as seen with tumor-directed anti-bodies, Vargas et al. reported that FcγRs play a part inthe therapeutic action of the CTLA-4-specific antibody Graphical Abstract ![inline-fig-Figa]() FcγRIIB-selective blockade enhances αCTLA-4-mediated Treg depletion, myeloid reprogramming, and effector CD8+ Tcell activation by increasing the activating-to-inhibitory (A:I) FcγR engagement ratio. Conversely, pan-FcγR blockade protects αPD-1-coated T cells from macrophage phagocytosis and from αPD-1removal via trogocytosis, thereby increasing intratumoral activated CD8+ T cells by decreasing activating andinhibitory FcγRs. ### Page 3 ipilimumab []. Dahan et al. discovered that αPD-1and αPD-L1 antibodies have opposing requirements forFcγR engagement in preclinical models []. Since then,we and others have found that FcγRs may enhance theeffectiveness of αCTLA-4 and other immune-modula-tory antibodies through various mechanisms, includingTreg depletion [,], improved T cell–APC immunesynapse formation [–], and reprogramming tumor-associated macrophages [,]. On the other hand,some studies indicate a harmful role for FcγRs in theefficacy of αPD-1 antibodies, with conflicting reports onthe roles of activating and inhibitory FcγRs [,–].Importantly, how FcγRs regulate the activities of indi-vidual ICB antibodies remains only partly understood,which limits rational optimization. Additionally, whileFc-engineering is a proven approach to boost therapeuticantibody efficacy, creating new Fc-modified antibodiesfrom existing approved drugs with refined FcγR engage-ment would be a long and expensive process. Therefore,we have focused on an alternative approach: combiningcurrent, approved immune checkpoint therapies withspecific FcγR blockade. Here, we examined the impact of FcγR blockade onthe therapeutic effectiveness of αCTLA-4 and αPD-1antibodies using FcγRIIB-specific antibodies with Fcconstant domains capable of binding multiple FcγRs(hIgG1 ‘BI-1206’ or mIgG2a) or genetically engineeredto silence FcγR engagement (hIgG1N297Q ‘BI-1607’ ormIgG1N297A). We further characterize these reagentsand offer a mechanistic explanation for their differentabilities to enhance the therapeutic activity of αCTLA-4and αPD-1 antibodies. To maximize αCTLA-4 anti-body effectiveness, selective blockade of FcγRIIB withFc-impaired αFcγRIIB was necessary. Conversely, opti-mal αPD-1 activity required blocking all FcγRs with Fc-competent αFcγRIIB. Our findings suggest strategiesto improve clinically approved αPD-1 antibodies, likepembrolizumab and nivolumab, and to support safe andeffective dual αPD-1 and αCTLA-4 immune checkpointblockade (ICB). Both approaches are currently underclinical trial investigation, with the potential to expandtreatment options to resistant cancers. Materials and methods Generation of antibodies targeting FcγRII, PD-1, PD-L1 andTIGIT αPD-1 (clone RMP1-14), αPD-L1 (clone MPDL3280A,atezolizumab), and αTIGIT (clone 10A07) were clonedonto different mouse IgG-backbones. The antibodieswere expressed in suspension-adapted HEK 293 EBNAcells (Thermo Fisher Scientific) and purified from the cellsupernatants on Mabselect (GE Healthcare). α-FcγRIIb(clone AT130-2) was either produced as described aboveor produced in ExpiCHO-S cells using the ExpiCHOtransient expression system (ThermoFisher) and puri-fied by protein A affinity chromatography using aHiTrap MAbSelect SuRe protein A column (Cytiva), fol-lowed by size exclusion chromatography (SEC) using aHiLoad Superdex 200 pg 16/600 SEC column (Cytiva), ifrequired. All antibodies were determined to be endotoxinlow (< 1 EU/mg) using an Endosafe Portable Test Systemdevice (Charles River Laboratories) and aggregate-free(< 1%) using high-performance liquid chromatography(HPLC). Clinically, relevant α-human PD-1 antibodies(nivolumab and pembrolizumab) were obtained from theUniversity Hospital Southampton pharmacy. Surface plasmon resonance analysis Binding of AT130-2 mIgG2a and mIgG1 N297A tomouse FcγR was assessed by Surface Plasmon Resonanceusing a Biacore T200 instrument (Cytiva). Antibod-ies were immobilized on a CM5 chip (Cytiva) via aminecoupling. Recombinant mouse FcγR (R&D Systems) werediluted in HBS-EP + (Cytiva). A 5-fold dilution series wasused starting at 100 nM for FcγRI and 1000 nM for allother FcγR, these were flowed over the immobilised anti-body at a rate of 30 μl/ml for 2 min (5 min for FcγRI) fol-lowed by 5-min dissociation. The chip was regeneratedbetween samples with glycerol pH1.5. Samples were ana-lysed in Biacore Evaluation software (Cytiva) with back-ground binding to the control flow cell removed. Mice Six to eight weeks-old (weight 17-20 grams) femaleC57BL/6J (RRID: IMSR_RJ:C57BL-6JRJ, RRID: IMSR_CRL:027) or BALB/c mice were either purchased fromJanvier (Saint-Berthevin, France) and maintained at Bio-Invent local facilities or purchased from Charles River(Kent, UK) and bred and maintained in UK Home Officeapproved local facilities. Humanized CTLA-4 mice(B-hCTLA-4) were purchased from Biocytogen (Shang-hai, China) and maintained at BioInvent local facilities. For in vivo experiments performed at BioInvent, allmice were housed and maintained in accordance withSwedish Laboratory Animal Care guidelines and wereperformed in an animal care-accredited facility. Allexperimental animal studies were conducted under theapproval of the Swedish Ethical Committee (approvalnumbers 5.8.18–17196/2018 and 5.8.18–19686/2022)and carried out in accordance with the Federation ofLaboratory Animal Science Associations (FELASA)guidelines. For in vivo experiments performed at Univer-sity of Southampton, all mice were housed in IVC unitsand maintained in accordance with the Animal (Scien-tific Procedures) Act 1986 and experiments were per-formed under approved project licenses (P4D9C89EAand PP5396109), and ethical review by local University of ### Page 4 Southampton animal welfare committee (approval num-bers 69840, 65400). In vivo mouse tumor models The CT26 (CRL-2638) and B16 (CRL-6475) murine can-cer cell lines were obtained from ATCC. The MC38 cellline was kindly gifted by Dr Sjef Verbeek. Cells were cul-tured according to providers’ instructions and logarith-mic growth phase of cells was ensured before harvestingcells for engraftment. A total of 0.5 × 106 or 1 × 106 CT26,B16 or MC38 cells were inoculated subcutaneously intothe flank of each mouse. All tumor cells were inoculatedin 100 μl of PBS. Mice bearing CT26 or MC38 tumorsof 40–200 mm3 (depending on model and setting) wererandomized into treatment groups based on tumor size.B16 mice were treated after 4 days. Mice were treatedwith the following antibodies or antibody combina-tions. Anti-mouse PD-1 (29F.1A12, rat IgG2a, RRID:AB_2687796, BioXcell or Assay Genie, RMPI-1 mIgG2a,mIgG1 and mIgG1-N297A or RMPI-1 mIgG2a-AF647or mIgG1-N297A-AF647), anti-mouse CTLA-4 (9H10,hamster IgG, RRID: AB_10949609, BioXcell), anti-FcγRII(AT130-2 mIgG2a and mIgG1-N297A), anti-humanCTLA-4 (ipilimumab, human IgG1, Apoteket), anti-mouse PD-L1 (mIgG1-atezolizumab), anti-mouse TIGIT(10A7 mIgG2b or mIgG2a) and/or control antibod-ies mIgG2a, mIgG1 or hIgG1 isotype. Antibodies wereadministered intraperitoneally in a volume of 200 μl. For tumor growth inhibition studies, mice were treated3 times (2–3 days apart). Animals were continuouslymonitored, and mice were euthanized when any of thefollowing end points were met: study termination, tumorburden ≥ 2,000 mm3, excessive tumor ulceration, or mori-bund appearance. Tumor burden was measured usingcalipers, and tumor volumes were calculated using themodified ellipsoid formula 0.5 × (length × width2). For rechallenge studies, cured mice were inoculatedsubcutaneously into the opposite flank from the curedprimary tumor with 5 × 106 MC38 cells. This was donewithin a minimum of 6 weeks post complete cure. For tumor-infiltrating immune cell analysis (by singlecell RNA sequencing or by flow cytometry) mice wereeuthanized 1 h—24 h after the first or third treatment(depending on experimental set-up) and tumors werecollected. Tumors were diced and incubated with either 1 Wun-sch unit/ml Liberase TL (Roche) in PBS at 37 °C at250 rpm for 20 min or in a solution of 50% Liberase TL(Roche) and 50% DNAse (Roche) for 3 × 5 min with slightvortexing in between at 37 °C, 5% CO2. The digestionreaction was quenched with 5 mL of RPMI (containing10% FCS) and homogenized into a single cell suspen-sion through a 100 μm cell strainer. Lymph nodes andspleen did not undergo enzymatic digestion but werehomogenized into a single cell suspension through a100 μm cell strainer. Red blood cells in the spleen werelysed using Lysis Buffer (Gibco or Bio-Rad). Tumor-infiltrating immune cell analysis using single cellRNA sequencing Cell preparation Tumors harvested as indicated were enzymaticallydigested as described. Viable lymphocytes were iso-lated using gradients (Lympholyte-M Separation Media,Cedarlane). Tumor lymphocytes were stained with viabil-ity dye and CD45 antibody (see antibody Table S1) andviable CD45+ cells were sorted using a FACS Aria Fusion(BD). Cells were counted (NucleoCounter NC-202,ChemoMetec) and using 4% paraformaldehyde (PFA)were fixed for 20 h at 4 °C, followed by long-term stor-age at –80 °C as instructed by ‘10 × Genomics’ protocol.Samples were further processed at the Center for Trans-lational Genomics (CTG) at the medical faculty at LundUniversity, Sweden. In short, prior to library preparation,samples were thawed, washed, and hybridized with bar-coded probe sets according to the Chromium Next-GEMFlex protocol (10 × Genomics). Approximately 104 cellsper sample were targeted for capture, and the pooledsamples were loaded onto a Chromium X instrumentusing a Chip Q. The generated libraries were sequencedon an Illumina NovaSeq 6000, targeting a sequencingdepth of 2 × 104 reads per cell. Single cell RNA analyses The scRNA-seq reads were aligned to the mouse tran-scriptome and a gene expression matrix was generatedusing the Cell Ranger pipeline using default settings.Using the R Seurat package (v4.4.0), the output filesfrom cellranger were converted to a Seurat object fordownstream analysis []. The R package Doubletfinder(v2.0.3) was applied with default parameters to find andremove cell doublets in each sample individually [].The cell cycle phase was controlled with the Seurat Cell-CycleScoring function using cell cycle related genes. Thenumber of detected genes, the number of detected UMIs,house-keeping gene expression and the percentage ofmitochondrial gene expression were used to remove low-quality cells. The SCTransform function was performedto normalize samples and applied separately for eachdataset after finding the top 3000 feature genes usingSelectIntegrationFeatures function. Then we applied thefunctions PrepSCTIntegration, FindIntegrationAnchorsto find integration anchors and IntegrateData (normal-ization.method = “SCT”) to complete data integration. We ran principal component analysis on the gene spaceusing RunPCA and clustering was performed based onthe shared nearest neighbor between cells (FindNeigh-bors) and graph-based clustering (FindClusters). We used ### Page 5 uniform manifold approximation and project (UMAP)with the function RunUMAP with the same number ofprincipal components used for embedding for the visu-alization of clusters. For the UMAP based clustering, weused the top 30 calculated PC dimensions and a resolu-tion of 0.8, eventually identified 25 distinct immune cellsubpopulations. To identify marker genes, we appliedFindAllMarkers function in Seurat (one-sided Wilcoxonrank sum test) with p-value adjusted for multiple test-ing using the Bonferroni correction. Using well curatedcell specific marker genes and based on prior experience,we annotated the immune cell subpopulations. For cell-type specific sub-clustering, we integrated the data acrosssamples and performed SCTransform approach as men-tioned above with parameters ranging from 20–30 PCsand UMAP resolution ranging between 0.2–0.7 withfinal parameters selected to generate consistent visual-izations. AverageExpression function in Seurat was usedto measure the average expression of gene markers andz-scores per genes were calculated for visualization withthe pheatmap package. The R package LSD () was usedto visualize density plots of cells using the UMAP coor-dinates of cells from each condition. All computationalanalyses were performed using publicly available soft-ware packages, as described above (-). No custom code or novel computational algorithmswere developed for this study. Immune cell analysis using flow cytometry Characterization of immune cell contents in tumors acrossdifferent mouse models WT mice were inoculated with either CT26, (BALB/c)or MC38 (C57BL/6J) tumor cells and tumors were moni-tored for growth with calipers. Mice were treated asdescribed for therapy experiments above. When CT26tumors reached approx. 600–700 mm3 or MC38 tumorsreached approx. 1300–1400 mm3 in area, mice were sac-rificed, and samples were processed as described above.Immune cell populations were identified using fluoro-chrome labeled antibodies (see table S1). After cell sur-face staining, cells were permeabilised using the Foxp3/Transcription Factor staining buffer set (Invitrogen)before staining for intracellular proteins. Samples wereacquired on an LSR Fortessa × 20 (Beckton Dickenson)flow cytometer and analysed using FlowJo software(RRID:SCR_000410). PD-1 expression on human tissue Cryopreserved, dissociated tumor cells were purchasedfrom Discovery Life Sciences. Material from 8 differentindications with 1–5 tumors per indication were includedin the analyses and stained with antibodies against CD45,CD3, CD8 and PD-1 (see table S1). αPD-1 (clone 1D6E10;pembrolizumab) was produced in-house and labeledwith Alexa Fluor 647. PD-1 expression on CD8+ T cellswas analysed by flow cytometry, samples were acquiredon a FACSAria Fusion (BD) flow cytometer. Dead cellswere excluded based on positive staining for the fixableviability dye eFluor 780 (Thermo Fisher Scientific). Cytokine analysis in murine tumors and serum For protein analysis, MC38 tumors were harvested day 7post first treatment, weighed and placed into homogeni-zation liquid (Pierce RIPA buffer (Merck) and 1:100 Pro-tease inhibitor cocktail (Merck)), 3 μ1/mg tumor. Tumorswere homogenized (GentleMACS, Miltenyi) and subse-quently centrifuged (10 min at 10,000 × g). Supernatantwas stored at −80 °C degrees until analysis. Blood was drawn from vena saphena (BD VacutainerHeparin tubes) from MC38 tumor-bearing mice treatedin total 3 times with 10 mg/kg of aFcγRIIΔFc, alone or incombination with 1 mg/kg of ipilimumab (day 1, 4 and 8).Blood was drawn from 5 mice at each time-point (24, 72and 168 h). Levels of 5 cytokines (TNF-α, IFN-γ, CCL22, IL-β andIL-6,) were assessed in serum and tumor using a U-plexassay (MSD) and read by a QuickPlex SQ 120 Reader(MSD). Data were analyzed using Discovery Workbenchv.4.0. Generation of human-PD-1 expressing Jurkat cells The full-length human PD-1 gene was cloned intopCDNA3 and nucleofected into Jurkat cells using theAmaxa cell line Nucleofector kit V and program X-001on the nucleofector II device. Positive cells were selectedwith 1 mg/ml geneticin (Life Technologies) for 2 weeksbefore being stained for PD-1 expression and FACSsorted to obtain pure populations of differing expressionlevels (FACSAriaII, BD Biosciences). Cells were main-tained in complete RPMI + 1 mg/ml geneticin. Generation of monocyte derived macrophages (MDM) andin vivo activated T cells Leukocyte cones were obtained from Southampton NHSBlood and Transplant service. Use was approved by theUniversity of Southampton Faculty of Medicine EthicsCommittee and East of Scotland Research Ethics Ser-vice, Research ethical committee reference number 16/ES/004. Peripheral blood mononuclear cells were isolatedby density gradient centrifugation (Lymphoprep, StemCell Technologies). MDM were then isolated via adher-ence or CD14+ isolation. For adherence, 1 × 107 PBMCsin 2 ml RPMI + 1% human AB serum were incubated in6-well plates for 2 h at 37 °C + 5% CO2. Non-adherentcells were collected and used to generate activated Tcells. Plates were washed with PBS and MDM differenti-ated over 7 days in complete RPMI + 100 ng/ml M-CSF ### Page 6 (in-house). Non-adherent PBMCs were resuspendedat 1 × 106 cells/ml in cRPMI with the addition of Dyna-beads human T-activator CD3/CD28 at 0.4 × 106 beads/ml + 30 IU/ml IL-2 (Peprotech). Beads and cytokine werereplaced twice over 8 days with beads removed using amagnet. For phagocytosis with subsequent FcγR staining,CD14+ cells were isolated by positive selection usingCD14+ MicroBeads (Miltenyi Biotec). Cells were dif-ferentiated in a 96 well plate over 7 days in completeRPMI + 100 ng/ml M-CSF. Phagocytosis assay Adherence isolated MDM were harvested and re-platedat 5 × 104 macrophages/well in a 96 well plate in 100 μlcRPMI + 100 ng/ml M-CSF and incubated overnightat 37 °C with 5% CO2. For phagocytosis assays macro-phages were treated with αFcγRIIB at 20 μg/ml (2 × finalconcentration) in 50 μl cRPMI for 30–45 min or 10 μg/ml overnight (FcγR detection). Meanwhile, target cells(Jurkat-hPD1 or activated T cells) were labelled with5 μM CFSE before being resuspended at 5 × 106/ml incRPMI and opsonized with αPD-1 at 2 × final concen-tration for 15 min at room temperature. 50 μl target cellsuspension (5:1 target:effector ratio) was added per wellof MDM and incubated for 1 h at 37 °C. Macrophageswere stained with αCD14 APC/BV421 (Biolegend/BDBiosciences, table S1). For FcγR staining, cells were addi-tionally stained with αFcγR specific antibodies. Phago-cytosis assessed by flow cytometry as the proportion ofCD14+CFSE+ cells as a total of CD14+ macrophages usinga FACS Canto II (BD Biosciences) or Cytoflex (BeckmanCoulter). For imaging, CD14+CFSE+ cells were FACs sorted ontoSuperfrost slides (ThermoFisher) using a FACSMelody(BD Biosciences). Slides were fixed with 4% paraformal-dehyde for 10 min and mounted with VECTASHIELDhard set with DAPI. Slides were imaged with an OlympusIX81 inverted microscope (Xcellence pro software) andimages processed in ImageJ (Fiji). Trogocytosis assay MDM were differentiated over 7 days and plated in 96well plates as described. 2 h after plating 80 μM dynasoredynamin inhibitor (Abcam) was added and MDM incu-bated at 37 °C for 24h. FcγRIIB antibodies were addedat 20 μg/ml (2 × final concentration) and incubated for30–60 min. Meanwhile, human PD1 Jurkat cells wereopsonized with 2 or 20 μg/ml (2 × final concentration)Alexa Fluor488 labelled nivolumab. Opsonised cells wereadded to MDM at a 5:1 target:effector ratio and incu-bated at 37 °C for 1 h. Jurkat cells and CD14+ macro-phages were analysed by flow cytometry (FACS Canto II,BD Biosciences) for loss of labelled nivolumab from theJurkat cells and uptake by MDM. Statistical analysis Statistical significance was performed using GraphPadPrism software (GraphPad, RRID:SCR_002798). Forcomparisons between two groups with normal distribu-tion, Student’s T test was used. If the data points in eachgroup originated from the same donor, paired Student’sT test was used. For comparisons regarding one param-eter using three or more groups with normal distribution,one-way ANOVA test was used. If the data points in eachgroup originated from the same donor, Repeated Mea-sures ANOVA test was used. For comparisons regard-ing one parameter but over several e.g. cell types usingthree or more groups with normal distribution, two-wayANOVA test using Tukey’s adjustment for multiple com-parisons was used. For survival curves, Log-Rank Man-tel Cox test was used. Number of data points, numberof experiments, graphic interpretation (mean and SDetc.), units, statistical test used and p-value limits havebeen stated in figure legends. Individual data points wereshown when possible. ScRNA data processing has beendescribed above, and no statistical analysis was per-formed in this data set. Results Generation of mFcγRII-blocking surrogate antibodies We previously developed human antibodies that effec-tively block the inhibitory hFcγRIIB []. Two antibodyvariants derived from the hFcγRIIB-specific antibody6G11 were generated: a hIgG1 antibody (BI-1206) witha wild-type Fc domain that binds both activating andinhibitory FcγRs (‘αFcγRIIBwt’), and a hIgG1N297Qvariant (BI-1607) with severely impaired Fc-binding toall FcγRs (‘αFcγRIIBΔFc’) [,]. At the time of genera-tion, the αFcγRIIBwt (BI-1206) antibody was intendedto treat cancers in which FcγRIIB is expressed on tumorcells (i.e., B-cell cancers). In preclinical models, it dem-onstrated single-agent activity and enhanced the activ-ity of tumor-directed antibodies such as αCD20 andαCD52 []. In contrast, the αFcγRIIBΔFc antibody (BI-1607) was designed to treat solid cancers, where FcγRIIBis expressed only on immune effector cells, and nottumor cells. Consistent with its highly specific blockadeof FcγRIIB and impaired Fc:FcγR binding, this antibodylacked single-agent activity but enhanced the antitumoractivity of αCD20, αHer2, and other tumor-targetingantibodies [,]. To assess the therapeutic potential of FcγR blockade toenhance the activity of immune-modulatory antibodies,e.g., αPD-1 and αCTLA-4, in solid cancers, we generatedsurrogate blocking antibodies to mFcγRII, the murinehomolog of hFcγRIIB and the sole inhibitory FcγR in ### Page 7 the mouse []. These murine-specific reagents facilitateproof-of-concept studies in immune-competent synge-neic mouse tumor models. Fc:FcγR-proficient and -defi-cient blocking antibodies matching the human BI-1206and BI-1607 clinical candidate antibodies to FcγRIIBwere constructed by fusing Fv-sequences of the mouseFcγRII-specific antibody AT-130-2 to mouse IgG2a andmouse IgG1N297A constant domains, respectively. Allgenerated antibodies showed highly specific, high-affin-ity, Fv-mediated binding to mouse FcγRII, as assessedby surface plasmon resonance (SPR) with recombinantprotein (Fig. S1A). Fc:FcγR proficiency and deficiency ofthe AT-130-2 mIgG2a and mIgG1N297A variants (hereaf-ter referred to as αFcγRIIwt and αFcγRIIΔFc, respectively)were similarly assessed by SPR using recombinant mouseactivating and inhibitory FcγR proteins (Fig. S1A). αFcγRIIwt and αFcγRIIΔFc differently modulate αPD-1 andαCTLA-4 antitumor activity in vivo We first assessed whether αFcγRIIwt and αFcγRIIΔFc couldenhance the therapeutic activity of αCTLA-4 (clone9H10) and αPD-1 (clone 29F.1A12) antibodies in immu-nocompetent mice bearing syngeneic CT26 colorec-tal tumors. This tumor model is infiltrated by immunecells, including CD8+ T cells, Tregs, and macrophages,and partially responds to αCTLA-4 but not, or poorly, toαPD-1 antibody therapy []. The model thus reflects thepartial responsiveness to these agents observed in humancancer, leaving room for improvement in efficacy. Com-bination treatment with αFcγRIIΔFc, but not αFcγRIIwt,significantly enhanced antitumor activity and survival inαCTLA-4-treated animals (Figs. and S1B). Conversely,and of potential clinical significance, combined treatmentwith αPD-1 and αFcγRIIwt, but not αFcγRIIΔFc, increasedthe proportion of mice cured in this otherwise αPD-1-resistant model (Figs. and S1B). Treatment of mice bearing MC38 colon adenocarci-noma tumors confirmed the inverse effects of αFcγRIIantibody variants on αCTLA-4 and αPD-1 therapeuticactivity (Fig.). Importantly, consistent with αFcγRIIantibodies modulating the intrinsic activity of αCTLA-4and αPD-1 and αCTLA-4 antibodies through FcγRblockade rather than direct anti-tumor activity, single-agent treatment with αFcγRII antibody variants did notaffect tumor growth kinetics or survival (Fig. S1C). αFcγRII enhancement of αCTLA-4 requires an FcγR-silenced Fc-domain and correlates with Treg depletion andmacrophage repolarization in vivo Treg cells prevent excessive CD8+ T cell expansion inhealth [], and Treg depletion enhances CD8+ T cellinfiltration into tumors, promoting tumor rejection[,–]. Reported correlations between melanomapatient responses to the αCTLA-4 antibody ipilimumaband FcγR engagement, as well as the enhanced antitumoractivity and Treg depletion of Fc-engaging humanizedαCTLA-4 antibodies [], prompted us to assess whetherαFcγRII improved αCTLA-4 efficacy by promoting Tregdepletion and improving intratumoral CD8:Treg ratios.In line with previous reports, treatment with αCTLA-4 ### Page 8 Fig. 2 (See legend on next page.) ### Page 9 alone improved CD8:Treg ratios, with trends towardreduced intratumoral Tregs and increased CD8+ T cellnumbers (Figs.A and S2A). Consistent with FcγRIIacting as a negative regulator of αCTLA-4-mediatedTreg depletion and with Tregs suppressing CD8+ T cellfunction [,], combined treatment with αFcγRIIΔFcand αCTLA-4 mAb further improved CD8:Treg ratiosand induced CD8+ T cell proliferation and activation(Fig.A). Next, we extended our αFcγRIIΔFc-mediated αCTLA-4mAb enhancement analyses to the clinical mAb ipi-limumab. Human CTLA-4 transgenic (hCTLA-4)mice were transplanted with MC38 tumors and treatedwith ipilimumab, αFcγRIIΔFc, or a combination. Tumorgrowth and animal survival were assessed. Consistentwith our murine αCTLA-4 surrogate antibody obser-vations, αFcγRIIΔFc enhanced ipilimumab in vivo effi-cacy, as shown by delayed tumor growth (Fig. S2B) andimproved animal survival (p < 0.02, Log-rank Mantel-Cox test) (Fig.B). Animals cured by the αFcγRIIΔFc/ipi-limumab combination rejected MC38 tumor cells uponrechallenge, demonstrating that immunological mem-ory was induced (Fig. S2C). Consistent with our flowcytometry data from murine CTLA-4 mice, combinedtreatment with αFcγRIIΔFc and ipilimumab activatedeffector T cells, as evidenced by induction of IFN-γ intumors of hCTLA-4 transgenic mice with trends towardenhanced TNF-α and IFN-γ compared with αCTLA-4single treatment (Figs.C and S2D). Moreover, con-sistent with enhancing effector CD8+ T cell responses,αFcγRIIΔFc combined with ipilimumab induced the IFN-γ-responsive chemokine CXCL10 involved in recruit-ing activated T cells from lymph nodes to the tumorbed (Fig.C). We then performed single-cell RNA-seqto agnostically assess the mechanisms underlying theenhancement of αFcγRIIΔFc in combination with ipilim-umab. Unsupervised clustering and UMAP visualizationof tumor-infiltrating leukocytes (TILs) harvested fromantibody-treated MC38 tumor-bearing hCTLA-4 micerevealed the expected presence of multiple immune cellsubsets, including B, T, NK cells, monocytes, macro-phages, and neutrophils (Fig. S2E). Consistent with ourobservations in the murine CTLA-4/αCTLA-4 modelsystem, αFcγRIIΔFc-enhanced ipilimumab efficacy cor-related with more substantial Treg depletion, decreasedexhausted CD8+ T cells, and increased activated CD8+ Tcells (IFN-γ+ TNF-α+) in hCTLA-4 mice (Fig. S2F). Addi-tionally, combined treatment with αFcγRIIΔFc and ipilim-umab increased activated Nos2hi M1 macrophages (Figs.D- G and S2G) and induced DC3 and M1 macrophagegene signatures (Fig.G). In contrast, M2-related geneswere downregulated (Fig.G). Combined αFcγRIIΔFc andipilimumab treatment increased the number of intratu-moral ILC1 cells (Fig. S2H) and fourfold increased totalneutrophils (PMN), with a preferential expansion ofproinflammatory N1 PMNs, including those positive forCXCL10+, compared with the monotherapy group (Figs.H- I and S2I). Furthermore, consistent with FcγRIIblockade enhancing ipilimumab through reprogram-ming myeloid cells, MIP-2, IL-1β, and IL-6 were inducedor increased in tumors from animals treated withαFcγRIIΔFc and ipilimumab, as quantified by multiplexedcytokine assay (Fig.J). αFcγRIIΔFc enables low-dose αCTLA-4-based dual immunecheckpoint blockade for rejection of poorly T cell-infiltrated tumors Dual treatment with αCTLA-4 and αPD-1 mAb increasesICB efficacy []. It is required for activity in particu-larly resistant cancer patients, e.g., those with mismatch-repair proficient colorectal cancer [] and those whohave relapsed on prior ICB therapy []. However, dose-dependent toxicity limits the use of available αCTLA-4mAb-based regimens [–]. Given αFcγRIIΔFc’senhancing effects on αCTLA-4 efficacy and Treg deple-tion, we tested whether it could increase the efficacy ofsafe, low doses of αCTLA-4. αCTLA-4 dose-dependentantitumor activity was confirmed by treating MC38tumor-bearing mice with 10, 2, or 0.4 mg/kg of anti-body (Fig.A and B). As expected, maximal efficacy wasobserved with 10 mg/kg αCTLA-4, equivalent to thehighest tested, maximally efficacious, but toxic dose inhuman cancer patients. Dosing at 2 mg/kg, intermediatebetween the two clinically approved αCTLA-4 doses of (See figure on previous page.) Fig. 2 αFcγRIIΔFc enhances αCTLA-4-induced TIL modulation. A Antibody-mediated TIL modulation in MC38 tumor-bearing C57BL/6 mice which receivedthree treatments (with 10 mg/kg per antibody and injection) of αCTLA-4 alone or in combination with αFcγRIIwt or αFcγRIIΔFc when tumors had reachedca. 150 mm3. Tumors were taken 24 h after the last treatment and assessed by flow cytometry. B Survival of MC38 tumor-bearing hCTLA4-transgenicC57BL/6 mice. When tumors were ca. 50 mm3, mice received 10 mg/kg of αFcγRIIΔFc, alone or in combination with 1 mg/kg of ipilimumab, comparedto ipilimumab (1 mg/kg) alone or control IgG (10 mg/kg). **P < 0.01, *P < 0.05 by log-rank test. C Cytokine analysis in tumor from MC38 tumor-bearinghCTLA4-transgenic C57BL/6 mice treated as in B). Tumors were harvested 168 h after first treatment. Results are pooled from three different experimentswith each symbol representing an individual mouse and lines representing mean. **** P < 0.0001, *** P < 0.001, **P < 0.01, *P < 0.05 by one-way ANOVA.D-I Single-cell RNA sequencing data from CD45+ sorted cells isolated from five MC38 tumors per group treated as in B). Tumors were harvested on day7 after the first treatment. D Combined UMAP of myeloid subclusters color-coded by cell subsets. The identity of each cluster is based on DEGs shownin G) and Fig. S2G. E Proportions of each cluster across treatment groups, color-coded as in D) F Contour UMAP for each treatment group (G) Heatmapshowing selected DEGs per treatment group (H) Bar-plots showing neutrophil numbers and I neutrophil subsets. J Cytokine analysis in tumor from MC38tumor-bearing hCTLA4-transgenic C57BL/6 mice as in C) ### Page 10 1 and 3 mg/kg, was also efficacious, whereas 0.4 mg/kghad little or no antitumor activity. Strikingly, coadminis-tration of αFcγRIIΔFc with the clinically relevant 2 mg/kgdose of αCTLA-4 enhanced efficacy to that of the maxi-mally efficacious but toxic 10 mg/kg dose. We next assessed whether adding αFcγRIIΔFc to a com-bination of low-dose αCTLA-4 mAb and maximallyefficacious αPD-1 mAb could enable rejection of poorlyCD8+ T cell-infiltrated B16 tumors transplanted intosyngeneic C57BL/6 mice. The highly resistant nature ofthis cancer model was confirmed by the absence of sur-viving mice after treatment with a control antibody,low-dose αCTLA-4, αPD-1, or the combination of low-dose αCTLA-4 and αPD-1. Again, adding αFcγRIIΔFc tolow-dose αCTLA-4/αPD-1 converted the regimen intoan effective one, as shown by delayed tumor growth(Fig.C) and an increased proportion of surviving mice(Fig.D). Selective FcγRII blockade enhances αTIGIT Like αCTLA-4, αTIGIT antibodies were recently shownto confer FcγR-dependent antitumor activity, therebyimproving the efficacy of αPD-L1 immune checkpointblockade []. αTIGIT efficacy and tumor microenvi-ronment remodeling were greater with high A:I FcγRengagement ratio antibody isotypes, suggesting thatactivating FcγRs augment antitumor activity and thatFcγRIIB reduces it. We therefore tested whether selectiveFcγRIIB blockade with αFcγRIIΔFc could enhance αTIGITduring αPD-L1-based immune checkpoint blockade,using an analogous experimental setup []. Tumor-bear-ing CT26 mice were treated with αTIGIT and αPD-L1,with or without αFcγRIIΔFc. For the TIGIT antibody, wechose the mIgG2b isotype to mimic the balanced FcγRA:I engagement ratio of clinically relevant FcγR-engagingαTIGIT antibodies, e.g., tiragolumab. For the αPD-L1antibody, the weakly FcγR-engaging mIgG1 isotype was ### Page 11 Fig. 4 (See legend on next page.) ### Page 12 used. As observed with αCTLA-4, FcγRII blockade withαFcγRIIΔFc enhanced antitumor activity and improvedsurvival in αTIGIT + αPD-L1-treated animals (Fig.Eand F). Consistent with a high A:I FcγR engagement ratiodetermining αTIGIT efficacy, combined treatment withαTIGIT mIgG2b antibody and FcγRII blockade similarlyenhanced αPD-L1 compared to a mIgG2a αTIGIT anti-body with a high FcγR A:I engagement ratio (Fig.E andF). αFcγRII enhancement of αPD-1 mAb requires an FcγR-binding Fc-domain αPD-1/PD-L1 antibodies, owing to their broad efficacyand favorable tolerability, constitute the backbone of ICBand are approved for more than 40 cancer types []. Mostapproved αPD-(L)1 regimens do not include αCTLA-4or other immune checkpoint inhibitors. Consequently,drugs that enhance αPD-1 and overcome αPD-1 resis-tance address a significant unmet medical need. Ourdata indicated that αFcγRIIΔFc, while potently enhancingαCTLA-4 monotherapy and CTLA-4/PD-1 combinationICB, did not modulate or enhance αPD-1 (Fig.). Instead,only the αFcγRIIwt antibody variant, equipped with anFcγR-binding Fc domain, improved αPD-1 efficacy. Theseobservations prompted us to analyze the mechanismsunderlying αFcγRIIwt’s αPD-1-enhancing effects. We first confirmed previous observations that FcγR-engagement reduces αPD-1 efficacy, using αPD-1 vari-ant antibodies that preferentially engage activating FcγRs(mIgG2a, high FcγR A: I engagement ratio), the inhibitoryFcγRII (mIgG1, low FcγR A: I engagement ratio), or noFcγRs (mIgG1N297A). Accordingly, in vivo αPD-1 efficacyranked mIgG1N297A > mIgG1 > > mIgG2a, i.e., in orderof decreasing FcγR engagement, as demonstrated byreduced tumor growth and improved survival in MC38tumor-bearing animals (Fig.A). Next, we assessed theprotective effects of αFcγRIIwt and αFcγRIIΔFc on αPD-1in vivo antitumor activity using weakly (mIgG1) orstrongly (mIgG2a) activating FcγR-engaging αPD-1 vari-ants. As observed with the hIgG4-mimicking, moder-ately FcγR-engaging, rIgG2a αPD-1 antibody, combinedtreatment with αFcγRIIwt, but not αFcγRIIΔFc, enhancedthe efficacy of the weakly FcγR-engaging αPD-1 mIgG1and, remarkably, overcame resistance to the stronglyFcγR-engaging αPD-1 mIgG2a antibody, as assessed byreduced tumor growth and improved animal survival(Fig.B). Our finding that αFcγRII antibodies required an FcγRbinding Fc domain to enhance αPD-1 demonstratedthat αFcγRII-blockade alone was insufficient to enhanceαPD-1. Previous reports indicated roles for both acti-vating and inhibitory FcγRs, limiting αPD-1 therapeu-tic efficacy through macrophage-mediated depletion(ADCP) of αPD-1-coated CD8+ T cells and trogocytosis(transfer) of T cell-bound αPD-1 antibodies from CD8+T cells, respectively. We assessed the relevance of thesemechanisms to αFcγRII modulation of αPD-1 in vivo.Combined treatment with αFcγRIIwt, but not αFcγRIIΔFc,and αPD-1 increased intratumoral activated CD8+ Tcells (Figs.C and S3A). In contrast, both αFcγRIIwt andαFcγRIIΔFc similarly prevented in vivo macrophage-medi-ated trogocytosis of αPD-1 from CD8+ T cells (Fig.D). Human αFcγRIIBwt blocks macrophage phagocytosisof nivolumab and pembrolizumab-treated T cells withclinically relevant PD-1 expression Our results demonstrated that αFcγRIIB antibodiesrequired an FcγR-binding Fc domain to enhance αPD-1in vivo and suggested that αFcγRIIBwt achieves this bysparing αPD-1-coated CD8+ T cells from macrophage-mediated depletion rather than by reducing the transfer(trogocytosis) of αPD-1 antibodies from T cells to mac-rophages. We assessed the human relevance of thesemechanisms for αFcγRIIB modulation of αPD-1 using aco-culture system in which Jurkat cells expressing intra-tumorally relevant levels of PD-1 (Fig.A and B) wereco-cultured with FcγR-expressing monocyte-derivedmacrophages (MDM) (Fig.C). Jurkat-PD-1/macrophageco-cultures were treated with human αPD-1 antibodiesnivolumab or pembrolizumab in the presence or absenceof clinically relevant human αFcγRIIBwt (BI-1206) orαFcγRIIBΔFc (BI-1607) mAb, and macrophage phagocy-tosis of Jurkat-PD-1 cells (Fig.D-F) or macrophage tro-gocytosis of αPD-1 from Jurkat-PD-1+ T cells (Fig.G)was assessed. The human αPD-1 antibodies mediatedmacrophage phagocytosis of Jurkat-PD-1 cells. Strikingly,αFcγRIIBwt (BI-1206) but not αFcγRIIBΔFc (BI-1607)completely blocked αPD-1-mediated phagocytosis (Fig.D-E). Moreover, a similar result was observed using invitro-activated primary human T cells as targets, whereonly αFcγRIIBwt inhibited αPD-1-mediated phagocytosis (See figure on previous page.) Fig. 4 αFcγRIIwt enhances αPD-1 treatment for tumor rejection. A Tumor growth in MC38 tumor-bearing mice treated three times with αPD-1 (10 mg/kg)in different isotype formats compared to negative control IgG, alone or B in combination with 10 mg/kg of αFcγRIIwt or αFcγRIIΔFc. Tumors were ca. 100mm3 at start of treatment. Numbers in the graph indicate totally cured mice/total number of mice in the group. C MC38 tumor-bearing mice were treatedwith 25 mg/kg of αFcγRIIΔFc or αFcγRIIwt in combination with αPD-1 in comparison to αPD-1, αFcγRIIwt or control IgG alone on day 0, 3 and 7. Tumors wereanalyzed by flow cytometry on day 4 and 8 after treatment start. Each dot represents one mouse. **P < 0.01, *P < 0.05 using ordinary one-way ANOVA. D Invivo trogocytosis of αPD-1 from CD8+ T cells. MC38 tumor-bearing mice were treated (i.v) with αPD-1 or αPD-1-N297A labeled with AF647. Two groupswere pretreated with αFcγRIIΔFc or αFcγRIIwt. The CD8+ cells in the tumors were analyzed for αPD-1 staining 1 h post injection. Each dot represents onemouse. ****P < 0.0001, ***P < 0.001 by ordinary one-way ANOVA ### Page 13 of target cells (Fig.F). In contrast, and as observedin vivo, both αFcγRIIBwt (BI-1206) and αFcγRIIBΔFc(BI-1607) reduced αPD-1 trogocytosis, as shown bydecreased αPD-1 mAb transfer to the macrophage sur-face (Fig.G). Immune checkpoint subcellular expression determinesFcγR blockade requirements Our results demonstrated that tailored FcγR blockadeeffectively enhances the in vivo therapeutic activity ofimmune checkpoint-blocking antibodies. We and oth-ers have previously shown that intratumoral Tregsexpress higher levels of CTLA-4 than CD8+ T cells [,,]. Accordingly, Tregs are more sensitive to deple-tion by an αCTLA-4 antibody. Conversely, intratumoralCD8+ T cells express higher levels of PD-1 than Tregsand are therefore more sensitive to depletion by anαPD-1 antibody [] (Figs.C and S3B-C). Accordingly,Fc:FcγR-silenced αFcγRII enhanced αCTLA-4, whereasFc:FcγR-proficient αFcγRII boosted αPD-1, correlating ### Page 14 with enhanced and blunted depletion of CTLA-4HighTreg cells and PD-1High effector CD8+ T cells [,,], respectively. Neither FcγRII-blocking antibody vari-ant showed single-agent antitumor activity or modu-lated TIL numbers, suggesting that the observed effectsdid not result from depletion of FcγRII-expressing cellsthemselves but rather from modulation of αCTLA-4and αPD-1 antibody interactions with FcγRs. We there-fore hypothesized that the differential boosting effects ofFcγR-silenced and FcγR-proficient αFcγRII in each caserelated to their ability to either bind and block FcγRIIselectively through Fv-mediated interactions (Fc-silencedαFcγRII) or bind and block all FcγR through a combi-nation of Fv- and Fc-mediated interactions (Fc:FcγR-proficient αFcγRII) once bound to the target cell surfacethrough FcγRII. We tested this in vivo and in vitro, assessing the avail-ability of free, unbound, activating Fc gamma receptorsand FcγRIIB, and the resulting A:I ratio on macrophagesfollowing treatment with αFcγRII(B)wt or αFcγRII(B)ΔFcmAb. Consistent with our hypothesis, macrophagesisolated from animals treated with αFcγRIIΔFc showedimproved A:I ratios, whereas macrophages treated withFcγRIIwt showed blunted FcγR availability (Fig.A). Theanalogous findings were made with the human-relevanttherapeutic antibodies in the T cell/macrophage co-culture system in vitro. While the Fc:FcγR-impairedαFcγRIIB antibody BI-1607 improved A:I ratios byselectively blocking the inhibitory FcγRIIB, the Fc:FcγR-proficient αFcγRIIB antibody BI-1206 reduced bothinhibitory and activating macrophage FcγRs (Fig.B). Discussion Several activating and a single inhibitory FcγR regulatethe efficacy of tumor-targeting and immune-modulatoryantibodies by promoting and reducing, respectively, Fc-dependent activation of FcγR-expressing immune effec-tor cells [,]. Accordingly, preclinical and clinical datademonstrate that αCTLA-4, αTIGIT, and other antibod-ies depend on activating FcγRs to trigger Treg deple-tion [], myeloid reprogramming, and robust CD8+ Tcell antitumor immunity [], mechanisms that underlieimproved survival in ICB-responding cancers such aslung cancer []. Critically, FcγR-dependent mechanismsmay extend the clinical benefit of ICB to otherwise resis-tant cancers, e.g., microsatellite stable colorectal cancer[], and to patients who have relapsed on αPD-1 [].Accordingly, Fc-engineered FcγR-enhanced αCTLA-4mAbs are being developed to treat poorly immunogenicand ICB-refractory cancers []. We are exploring tailored FcγR blockade as an alter-native to Fc engineering to improve FcγR engagement.Here, we showed that FcγR blockade with Fc-mutant andwild-type αFcγRIIB antibodies enhanced αCTLA-4 andαPD-1 efficacy, respectively. The ability of αFcγRIIBΔFcto enhance αCTLA-4 mAb efficacy matched αCTLA-4’s known reliance on activating FcγRs, correlating withdeeper Treg depletion, myeloid reprogramming, andCD8+ T cell activation. Interestingly, αFcγRIIBΔFc alsoinduced intratumoral CXCL10, part of the CXCL9/10/11-CXCR3 axis responsible for recruiting CD8+ and CD4+ Tcells from lymph nodes to the tumor bed [], predict-ing response to ICB [], and extending ICB activity tootherwise resistant, poorly CD8+ T cell-infiltrated “cold”tumors []. Accordingly, αFcγRIIBΔFc, besides enhanc-ing low-dose αCTLA-4 efficacy to fully therapeutic lev-els in the ICB partially responsive MC38 tumor model,extended its activity to the poorly CD8+ T cell-infiltratedICB-resistant B16 tumor model. Without FcγRIIB block-ade, such efficacy required an αCTLA-4 dose of 10 mg/kg, which is a clinically prohibitive dose owing to itssevere toxicity []. Whether CTLA-4 mAb efficacy canbe separated from its toxicity remains a topic of debate[–]. While only human clinical trials can answer thisquestion, given the lack of tox-predictive mouse models,independent observations support the clinical relevance ### Page 15 of low-dose αFcγRIIBΔFc-enhanced αCTLA-4. Nota-bly, hIgG1 ipilimumab and hIgG2 tremelimumab showsimilar toxicities at comparable exposure levels [–],suggesting that systemic blockade of the CTLA-4:B7immune checkpoint, rather than FcγR engagement, is theprimary cause of toxicity. In contrast, as discussed above, preclinical and clini-cal data demonstrate that tumor-associated FcγR-engagement enhances αCTLA-4 efficacy and is requiredfor activity in poorly CD8+ T cell-infiltrated and ICB-refractory cancers [,]. Notably, FcγRIIB is upregu-lated in the tumor microenvironment but shows limitedexpression on blood myeloid cells and is not expressed onNK cell effectors [,,,]. Relative to Fc-enhancedαCTLA-4 mAbs, which are engineered to bind stronglyto the activating FcγRIIA and FcγRIIIA ubiquitouslyexpressed in blood, targeting FcγRIIB may offer a moretumor-selective approach to enhance antibody immu-notherapy. We recently demonstrated the safety of thisapproach in the context of tumor-targeting antibodyimmunotherapy by co-administering the FcγRIIB-blocking antibody BI-1607 with trastuzumab in breastcancer patients []. While combined ICB with FcγR-engaging mAb, e.g.,αCTLA-4, is required for efficacy in poorly CD8+ T cell-infiltrated tumors, αPD-(L)1 single ICB, owing to itssuperior tolerability, is the most widely approved ICBand the preferred treatment in T cell-inflamed cancers.Enhancing efficacy and overcoming resistance to αPD-1ICB has tremendous medical potential. In this setting,we find that αFcγRIIBwt enhances therapeutic efficacyby blunting αPD-1 engagement of both inhibitory andactivating FcγRs. This was demonstrated by αFcγRIIBwtexclusively restoring and enhancing FcγR-engagingαPD-1 antibodies in vivo and by sparing activated PD-1high-expressing T cells treated with pembrolizumabor nivolumab from macrophage phagocytosis in vitroand in vivo. These observations were consistent withthe differential effects of αFcγRIIΔFc and αFcγRIIwt onαCTLA-4, with αFcγRIIΔFc promoting, and αFcγRIIwtreducing αCTLA-4 antibody-mediated ADCP, reflectedby decreased Treg and increased CD8+ T cells follow-ing combination therapy in vivo. In contrast, both anti-bodies reduced macrophage-mediated αPD-1 antibodytrogocytosis. Together, these observations suggest thatαFcγRIIwt primarily enhances αPD-1 efficacy by reduc-ing macrophage-mediated ADCP of PD-1+ CD8+ Tcells, rather than by preventing αPD-1 trogocytosis.This notion is consistent with earlier observations thatboth activating and inhibitory FcγRs mediate trogo-cytosis [], and align with previous observations thatmacrophage-dependent phagocytosis controls antibodyefficacy and target cell depletion in vivo [,–]. Adetrimental role for FcγRs in αPD-1-based therapy hasbeen previously suggested, including with human-rele-vant αPD-1 antibodies of the hIgG4 isotype [,,].Consistent with the hIgG4 isotype being a relatively poorengager of FcγRs compared with hIgG1, we find thatmacrophage phagocytosis of human T cells coated withnivolumab or pembrolizumab required robust yet clini-cally relevant PD-1 expression. In contrast to CTLA-4,which is most highly expressed on intratumoral Tregcells, PD-1 is progressively increased with activation oneffector CD8+ T cells and in exhausted CD8+ T cell statessensitive to αPD-1 ICB. Sparing and maximizing theactivity of these critical effector cells, e.g., by minimizingFcγR-dependent depletion, is highly warranted. Interest-ingly, while intratumoral anti-PD-1 engagement is asso-ciated with deleterious effects that compromise efficacy,recent data suggest that low anti-PD-1 engagement ofFcγRs in tumor-draining lymph nodes (TDLN) may bebeneficial []. FcγR engagement was shown to mobilizeimmune cells, including tumor-specific CD8 + T cells,in the TDLN. Increased CD8 + T cell mobilization wasassociated with improved therapeutic responses. Our tai-lored approach to FcγR blockade provides an attractivestrategy for spatiotemporally tuning FcγR engagement.It would be interesting to assess whether staggered treat-ment with a single-agent Fc-engaging αPD-1, followed bycombined therapy with FcγRIIBwt, could enhance antitu-mor CD8 + T cell expansion and αPD-1 efficacy. Taken together, our results show that FcγRIIB-selectiveblockade promotes ADCP, whereas pan-FcγR blockadeprevents it. Accordingly, our findings suggest that thedifferential abilities of the antibodies to enhance anti-CTLA-4 and anti-PD-1 efficacy are due to CTLA-4 andPD-1 being more highly expressed on Tregs and CD8+T cells, and Tregs and CD8+ T cells being sensitive toαCTLA-4- and αPD-1-mediated ADCP, respectively [,,]. Conclusions In conclusion, we provide preclinical proof of con-cept that tailored FcγRIIB blockade with Fc-mutated orwild-type antibodies achieves highly specific FcγRIIBor pan-FcγR blockade, thereby enhancing the efficacyof clinically relevant αCTLA-4 and αPD-1 ICB throughdistinct mechanisms that respectively enhance or blockantibody-mediated ADCP of Treg and CD8+ T cells. Human Fc-tailored αFcγRIIB antibodies BI-1607(NCT06784648) and BI-1206 (NCT04219254,NCT03571568), which have the potential to enhance ICBefficacy and extend its use to currently untreatable, resis-tant cancers, are in clinical development. Abbreviations FcγR Fc gamma receptor ICB Immune checkpoint blockade mAb Monoclonal antibody ### Page 16 Supplementary Information The online version contains supplementary material available at. Additional file 1. Supplemental figures and tables. Acknowledgements We thank members of the Antibody and Vaccine group and BioInventInternational for useful discussions and the preclinical unit staff for animalhusbandry. Authors’ contributions Conceptualization: MSC and BF. Methodology: RJO, LM, MS, NY, PH, LVH,MCT, KLC, MY, JFB, AR, IK. Investigation: RJO, LM, MS, NY, PH, LVH, MCT, KLC,MY, JFB, AR, IK. Visualization: RJO, LM, MS, NY, PH, LVH, MCT, KLC, MY, JFB, AR,IK. Funding acquisition: SAB, and MSC. Project administration: MS, MSC, andBF. Supervision: IT, SAB, MSC, and BF. Writing – original draft: RJO, MS, and BF.Writing – review & editing: SAB, IT, RJO, PH, IK, MS, MSC, and BF. Funding This work was supported by Blood Cancer UK (Award number: 12050 and14043), CRUK (Award numbers: A24721 and DRCRPG-May23/100001), andBioInvent International. Data availability The processed data and raw data (single cell RNA sequencing) have beenuploaded to the NCBI Gene Expression Omnibus (GEO) database () with accession number GSE335831. Declarations Ethics approval and consent to participate For in vivo experiments performed at BioInvent, all experimental animalstudies were conducted under the approval of the Swedish Ethical Committee(approval numbers 5.8.18–17196/2018 and 5.8.18–19686/2022) and carriedout in accordance with the Federation of Laboratory Animal ScienceAssociations (FELASA) guidelines. For in vivo experiments performed atUniversity of Southampton, all mice were housed in IVC units and maintainedin accordance with the Animal (Scientific Procedures) Act 1986 andexperiments were performed under approved project licenses (P4D9C89EAand PP5396109), and ethical review by local University of Southamptonanimal welfare committee (approval numbers 69840, 65400). Consent for publication Not applicable. Competing interests BioInvent International may financially benefit from this publication. RO, LM,MS, NY, PH, LVG, MCT, MY, JFB, AR, IK, IT, and BF are employed or partly researchfunded by BioInvent International. SAB reports grants and personal fees fromBioInvent International during the conduct of the study; personal fees fromLTZ Therapeutics, Epsilogen, ImCheck Therapeutics, and F-Star Therapeutics,and grants from Boehringer Ingelheim outside the submitted work. MSCreports grants from BioInvent International and personal fees from BioInventInternational during the conduct of the study; personal fees from variousBiotechs and Pharma outside the submitted work. The remaining authorsdeclare no conflicts of interest. Author details 1Antibody and Vaccine Group, School of Cancer Sciences, University ofSouthampton, Southampton, UK 2BioInvent, Lund, Sweden Received: 30 March 2026 / Accepted: 10 July 2026 References 1. Paul J, Mitchell AP, Kesselheim AS, Rome BN. Overlapping and non-overlap-ping indications for checkpoint inhibitors in the US. J Clin Oncol. 2004;42(16_suppl June 2024):11057. 2. Sharma P, Hu-Lieskovan S, Wargo JA, Ribas A. Primary, adaptive, and acquiredresistance to cancer immunotherapy. Cell. 2017;168(4):707–23. 3. VanderWalde A, Bellasea SL, Kendra KL, Khushalani NI, Campbell KM, ScumpiaPO, et al. Ipilimumab with or without nivolumab in PD-1 or PD-L1 blockaderefractory metastatic melanoma: a randomized phase 2 trial. Nat Med.2023;29(9):2278–85. 4. Clynes RA, Towers TL, Presta LG, Ravetch JV. Inhibitory Fc receptors modulatein vivo cytotoxicity against tumor targets. Nat Med. 2000;6(4):443–6. 5. Nimmerjahn F, Ravetch JV. Fcgamma receptors: old friends and new familymembers. Immunity. 2006;24(1):19–28. 6. Nimmerjahn F, Vidarsson G, Cragg MS. Effect of posttranslational modifica-tions and subclass on IgG activity: from immunity to immunotherapy. NatImmunol. 2023;24(8):1244–55. 7. Nimmerjahn F, Ravetch JV. Divergent immunoglobulin g subclass activitythrough selective Fc receptor binding. Science. 2005;310(5753):1510–2. 8. Roghanian A, Teige I, Martensson L, Cox KL, Kovacek M, Ljungars A, et al.Antagonistic human FcgammaRIIB (CD32B) antibodies have anti-tumoractivity and overcome resistance to antibody therapy in vivo. Cancer Cell.2015;27(4):473–88. 9. Teige I, Martensson L, Frendeus BL. Targeting the antibody checkpoints toenhance cancer immunotherapy-focus on FcgammaRIIB. Front Immunol.2019;10:481. 10. Galvez-Cancino F, Simpson AP, Costoya C, Matos I, Qian D, Peggs KS, et al.Fcgamma receptors and immunomodulatory antibodies in cancer. Nat RevCancer. 2024;24(1):51–71. 11. Arce Vargas F, Furness AJS, Litchfield K, Joshi K, Rosenthal R, Ghorani E, etal. Fc effector function contributes to the activity of human anti-CTLA-4antibodies. Cancer Cell. 2018;33(4):649-63 e4. 12. Dahan R, Sega E, Engelhardt J, Selby M, Korman AJ, Ravetch JV. FcgammaRsmodulate the anti-tumor activity of antibodies targeting the PD-1/PD-L1 axis.Cancer Cell. 2015;28(3):285–95. 13. Simpson TR, Li F, Montalvo-Ortiz W, Sepulveda MA, Bergerhoff K, Arce F,et al. Fc-dependent depletion of tumor-infiltrating regulatory T cells co-defines the efficacy of anti-CTLA-4 therapy against melanoma. J Exp Med.2013;210(9):1695–710. 14. Waight JD, Chand D, Dietrich S, Gombos R, Horn T, Gonzalez AM, et al. Selec-tive FcgammaR co-engagement on APCs modulates the activity of therapeu-tic antibodies targeting T cell antigens. Cancer Cell. 2018;33(6):1033-47 e5. 15. Bullock AJ, Schlechter BL, Fakih MG, Tsimberidou AM, Grossman JE, GordonMS, et al. Botensilimab plus balstilimab in relapsed/refractory microsatellitestable metastatic colorectal cancer: a phase 1 trial. Nat Med. 2024.. 16. Yofe I, Landsberger T, Yalin A, Solomon I, Costoya C, Demane DF, et al.Anti-CTLA-4 antibodies drive myeloid activation and reprogram the tumormicroenvironment through FcgammaR engagement and type I interferonsignaling. Nat Cancer. 2022;3(11):1336–50. 17. Guan X, Hu R, Choi Y, Srivats S, Nabet BY, Silva J, et al. Anti-TIGIT antibodyimproves PD-L1 blockade through myeloid and T(reg) cells. Nature.2024;627(8004):646–55. 18. Arlauckas SP, Garris CS, Kohler RH, Kitaoka M, Cuccarese MF, Yang KS, et al. Invivo imaging reveals a tumor-associated macrophage-mediated resistancepathway in anti-PD-1 therapy. Sci Transl Med. 2017.. 19. Bennion KB, Tariq M, Wyatt MM, Duneton C, Baecher KM, Paulos CM, et al.FcgammaRIIB expressed on CD8(+) T cells limits responsiveness to PD-1checkpoint inhibition in cancer. Sci Transl Med. 2023;15(710):eadd1868. 20. Moreno-Vicente J, Willoughby JE, Taylor MC, Booth SG, English VL, WilliamsEL, et al. Fc-null anti-PD-1 monoclonal antibodies deliver optimal checkpointblockade in diverse immune environments. J Immunother Cancer. 2022.. 21. Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM 3rd, et al.Comprehensive integration of single-cell data. Cell. 2019;177(7):1888-902e21. 22. McGinnis CS, Murrow LM, Gartner ZJ. DoubletFinder: doublet detection insingle-cell RNA sequencing data using artificial nearest neighbors. Cell Syst.2019;8(4):329-37 e4. 23. Cortes J, Priego A, Garralda E, Rojas K, Lord SR, Goetze TO, et al. A first-in-classmonoclonal antibody (BI-1607) targeting FcgammaRIIB: preclinical data and ### Page 17 first-in-human studies in patients with HER2-positive advanced solid tumors.Clin Cancer Res. 2025;31(23):4953–63. 24. Williams EL, Tutt AL, French RR, Chan HT, Lau B, Penfold CA, et al. Develop-ment and characterisation of monoclonal antibodies specific for the murineinhibitory FcgammaRIIB (CD32B). Eur J Immunol. 2012;42(8):2109–20. 25. Mosely SI, Prime JE, Sainson RC, Koopmann JO, Wang DY, Greenawalt DM, etal. Rational selection of syngeneic preclinical tumor models for immuno-therapeutic drug discovery. Cancer Immunol Res. 2017;5(1):29–41. 26. Kim JM, Rasmussen JP, Rudensky AY. Regulatory T cells prevent cata-strophic autoimmunity throughout the lifespan of mice. Nat Immunol.2007;8(2):191–7. 27. Sutmuller RP, van Duivenvoorde LM, van Elsas A, Schumacher TN, WildenbergME, Allison JP, et al. Synergism of cytotoxic T lymphocyte-associated antigen4 blockade and depletion of CD25(+) regulatory T cells in antitumor therapyreveals alternative pathways for suppression of autoreactive cytotoxic Tlymphocyte responses. J Exp Med. 2001;194(6):823–32. 28. Quezada SA, Peggs KS, Curran MA, Allison JP. CTLA4 blockade and GM-CSFcombination immunotherapy alters the intratumor balance of effector andregulatory T cells. J Clin Invest. 2006;116(7):1935–45. 29. Klages K, Mayer CT, Lahl K, Loddenkemper C, Teng MW, Ngiow SF, et al. Selec-tive depletion of Foxp3+ regulatory T cells improves effective therapeuticvaccination against established melanoma. Cancer Res. 2010;70(20):7788–99. 30. Dittmer U, He H, Messer RJ, Schimmer S, Olbrich AR, Ohlen C, et al. Functionalimpairment of CD8(+) T cells by regulatory T cells during persistent retroviralinfection. Immunity. 2004;20(3):293–303. 31. Noyes D, Bag A, Oseni S, Semidey-Hurtado J, Cen L, Sarnaik AA, et al.Tumor-associated Tregs obstruct antitumor immunity by promoting T celldysfunction and restricting clonal diversity in tumor-infiltrating CD8+ T cells.J Immunother Cancer. 2022.. 32. Wolchok JD, Chiarion-Sileni V, Rutkowski P, Cowey CL, Schadendorf D,Wagstaff J, et al. Final, 10-year outcomes with nivolumab plus ipilimumab inadvanced melanoma. N Engl J Med. 2024.. 33. Chalabi M, Verschoor YL, Tan PB, Balduzzi S, Van Lent AU, Grootscholten C, etal. Neoadjuvant immunotherapy in locally advanced mismatch repair-defi-cient colon cancer. N Engl J Med. 2024;390(21):1949–58. 34. Feng Y, Roy A, Masson E, Chen TT, Humphrey R, Weber JS. Exposure-responserelationships of the efficacy and safety of ipilimumab in patients withadvanced melanoma. Clin Cancer Res. 2013;19(14):3977–86. 35. Ascierto PA, Del Vecchio M, Mackiewicz A, Robert C, Chiarion-Sileni V, AranceA, et al. Overall survival at 5 years of follow-up in a phase III trial comparingipilimumab 10 mg/kg with 3 mg/kg in patients with advanced melanoma. JImmunother Cancer. 2020.. 36. Shulgin B, Kosinsky Y, Omelchenko A, Chu L, Mugundu G, Aksenov S, et al.Dose dependence of treatment-related adverse events for immune check-point inhibitor therapies: a model-based meta-analysis. Oncoimmunology.2020;9(1):1748982. 37. Semmrich M, Marchand JB, Fend L, Rehn M, Remy C, Holmkvist P, et al.Vectorized Treg-depleting alpha CTLA-4 elicits antigen cross-presentationand CD8(+) T cell immunity to reject ‘cold’ tumors. J Immunother Cancer.2022;10(1):1–14. 38. Chand D, Savitsky DA, Krishnan S, Mednick G, Delepine C, Garcia-Broncano P,et al. Botensilimab, an Fc-enhanced anti-CTLA-4 antibody, is effective againsttumors poorly responsive to conventional immunotherapy. Cancer Discov.2024;14(12):2407–29. 39. Chow MT, Ozga AJ, Servis RL, Frederick DT, Lo JA, Fisher DE, et al. Intratumoralactivity of the CXCR3 chemokine system is required for the efficacy of anti-PD-1 therapy. Immunity. 2019;50(6):1498-512 e5. 40. House IG, Savas P, Lai J, Chen AXY, Oliver AJ, Teo ZL, et al. Macrophage-derived CXCL9 and CXCL10 are required for antitumor immune responsesfollowing immune checkpoint blockade. Clin Cancer Res. 2020;26(2):487–504. 41. Yilmaz A, Haerri L, Granda ME, Coquoz O, Lan Q, Ruegg C. The CXCL10/CXCR3axis is essential for sustaining immunological dormancy in triple-negativebreast cancer. NPJ Breast Cancer. 2026.. 42. Ascierto PA, Del Vecchio M, Robert C, Mackiewicz A, Chiarion-Sileni V, AranceA, et al. Ipilimumab 10 mg/kg versus ipilimumab 3 mg/kg in patients withunresectable or metastatic melanoma: a randomised, double-blind, multi-centre, phase 3 trial. Lancet Oncol. 2017;18(5):611–22. 43. Boutros C, Tarhini A, Routier E, Lambotte O, Ladurie FL, Carbonnel F, et al.Safety profiles of anti-CTLA-4 and anti-PD-1 antibodies alone and in combi-nation. Nat Rev Clin Oncol. 2016;13(8):473–86. 44. Wei SC, Duffy CR, Allison JP. Fundamental mechanisms of immune check-point blockade therapy. Cancer Discov. 2018;8(9):1069–86. 45. Postow MA, Sidlow R, Hellmann MD. Immune-related adverse events associ-ated with immune checkpoint blockade. N Engl J Med. 2018;378(2):158–68. 46. Arce Vargas F, Furness AJS, Solomon I, Joshi K, Mekkaoui L, Lesko MH, et al.Fc-optimized anti-CD25 depletes tumor-infiltrating regulatory T cells andsynergizes with PD-1 blockade to eradicate established tumors. Immunity.2017;46(4):577–86. 47. Hussain K, Liu R, Smith RCG, Muller KTJ, Ghorbani M, Macari S, et al. HIFactivation enhances FcgammaRIIb expression on mononuclear phagocytesimpeding tumor targeting antibody immunotherapy. J Exp Clin Cancer Res.2022;41(1):131. 48. Boross P, Jansen JH, Pastula A, van der Poel CE, Leusen JH. Both activating andinhibitory Fc gamma receptors mediate rituximab-induced trogocytosis ofCD20 in mice. Immunol Lett. 2012;143(1):44–52. 49. Uchida J, Hamaguchi Y, Oliver JA, Ravetch JV, Poe JC, Haas KM, et al. Theinnate mononuclear phagocyte network depletes B lymphocytes throughFc receptor-dependent mechanisms during anti-CD20 antibody immuno-therapy. J Exp Med. 2004;199(12):1659–69. 50. Biburger M, Aschermann S, Schwab I, Lux A, Albert H, Danzer H, et al.Monocyte subsets responsible for immunoglobulin G-dependent effectorfunctions in vivo. Immunity. 2011;35(6):932–44. 51. Montalvao F, Garcia Z, Celli S, Breart B, Deguine J, Van Rooijen N, et al. Themechanism of anti-CD20-mediated B cell depletion revealed by intravitalimaging. J Clin Invest. 2013;123(12):5098–103. 52. Gul N, Babes L, Siegmund K, Korthouwer R, Bogels M, Braster R, et al. Macro-phages eliminate circulating tumor cells after monoclonal antibody therapy.J Clin Invest. 2014;124(2):812–23. 53. Biburger M, Lux A, Nimmerjahn F. How immunoglobulin G antibodies killtarget cells: revisiting an old paradigm. Adv Immunol. 2014;124:67–94. 54. Wohner M, Brechtelsbauer S, Friedrich N, Vorsatz C, Bulang J, Liang C, etal. Tissue niche occupancy determines the contribution of fetal- versusbone-marrow-derived macrophages to IgG effector functions. Cell Rep.2024;43(2):113757. 55. Hashimoto M, Nasti TH, Jin HT, Bu M, Araki K, Lee J, et al. Selective deple-tion of virus-specific CD8 T cells from the liver after PD-1 therapy withFc-intact antibody during chronic infection. Proc Natl Acad Sci U S A.2026;123(15):e2427192123. 56. Guerin MV, Ruggiu M, Feldmann LC, Corre B, Ombredanne P, Iannascoli B, etal. Lymph node fine-tuning FcgammaR signaling boosts anti-PD-1 therapy. JImmunother Cancer. 2026;14(6):1–16. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.