TYPE Original Research PUBLISHED 20 January 2023 DOI 10.3389/fimmu.2023.1066393 LAG3 genotype of the donor OPEN ACCESS and clinical outcome after EDITED BY José Laurent Cohen, Université Paris-Est Créteil Val de Marne, allogeneic transplantation from France REVIEWED BY HLA-identical sibling donors Paul Joseph Martin, Fred Hutchinson Cancer Research Center, United States David Cruz 1, Rocı́o Rodrı́guez-Romanos 1, Marta González-Bartulos 1, Grégory Ehx, University of Liège, Belgium Irene Garcı́a-Cadenas 2, Rafael de la Cámara 3, Inmaculada Heras 4, *CORRESPONDENCE Ismael Buño 5, Nazly Santos 1, Natàlia Lloveras 1,6, Pilar Velarde 1, David Gallardo Esperanza Tuset 1, Carmen Martı́nez 7, Marcos González 8, dgallardo@iconcologia.net Guillermo F. Sanz 9, Christelle Ferrá 10, Antonia Sampol 11, Rosa Coll 1,6, SPECIALTY SECTION This article was submitted to Jose A. Pérez-Simón 12, Javier López-Jiménez 13, Manuel Jurado 14 Alloimmunity and Transplantation, and David Gallardo 1,6* on behalf of the GvHD/Immunotherapy a section of the journal Frontiers in Immunology Working Party of the Spanish Group of Hematopoietic RECEIVED 10 October 2022 Transplant (GETH) ACCEPTED 04 January 2023 PUBLISHED 20 January 2023 1 Hematology Department, Institut Català d’Oncologia - Hospital Dr. Josep Trueta, Institut d’Investigació Biomèdica de Girona (IDIBGI), Josep Carreras Research Institute, Girona, Spain, 2Hematology Department, CITATION Hospital de la Santa Creu i Sant Pau, Institut d’Investigació Biomèdica Sant Pau, Universitat Autònoma de Cruz D, Rodrı´guez-Romanos R, Barcelona, Barcelona, Spain, 3Hematology Department, Hospital La Princesa, Madrid, Spain, 4Hematology González-Bartulos M, Garcı´a-Cadenas I, Department, Hospital General Universitario Morales Meseguer, Murcia, Spain, 5Hematology Department de la Cámara R, Heras I, Buño I, Santos N, and Genomics Unit, Hospital General Universitario Gregorio Marañón, Gregorio Marañón Health Research Lloveras N, Velarde P, Tuset E, Martı´nez C, Institute (IiSGM), Complutense University, Madrid, Spain, 6Department of Medicine, Universitat de Girona, González M, Sanz GF, Ferrá C, Sampol A, Girona, Spain, 7Hematology Department, Hospital Clínic de Barcelona, Institut d’Investigacions Coll R, Pérez-Simón JA, López-Jiménez J, Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain, 8Hematology Department, Hospital Clínico Jurado M and Gallardo D (2023) LAG3 Universitario, Salamanca, Spain, 9Hematology Department, Hospital Universitari i Politècnic La Fe, genotype of the donor and clinical Valencia, Spain, 10Hematology Department, Institut Català d’Oncologia – Hospital Germans Trias i Pujol, outcome after allogeneic transplantation Josep Carreras Research Institute, Badalona, Spain, 11Hematology Department, Hospital Universitari Son from HLA-identical sibling donors. Espases, IdISBa, Palma de Mallorca, Spain, 12Hematology Department, Hospital Universitario Virgen del Front. Immunol. 14:1066393. Rocío, Instituto de Biomedicina de Sevilla, Universidad de Sevilla, Sevilla, Spain, 13Hematology Department, doi: 10.3389/fimmu.2023.1066393 Hospital Universitario Ramón y Cajal, Madrid, Spain, 14Hematology Department, Hospital Universitario COPYRIGHT Virgen de las Nieves, Granada, Spain © 2023 Cruz, Rodrı´guez-Romanos, González-Bartulos, Garcı´a-Cadenas, de la Cámara, Heras, Buño, Santos, Lloveras, Velarde, Tuset, Martı´nez, González, Sanz, Ferrá, Sampol, Coll, Pérez-Simón, López- Introduction: The association of polymorphisms in molecules involved in the Jiménez, Jurado and Gallardo. This is an immune response (checkpoint inhibitors) with the clinical outcome after open-access article distributed under the allogeneic transplantation (alloHSCT) has been described. Lymphocyte Activation terms of the Creative Commons Attribution License (CC BY). The use, distribution or 3 (LAG3) is a surface protein that plays a regulatory role in immunity as an inhibitory reproduction in other forums is permitted, immune checkpoint molecule. provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is Methods: To determine its role in the alloHSCT setting, we analyzed 797 patients cited, in accordance with accepted transplanted from HLA-identical sibling donors. The LAG3 rs870849 C>T academic practice. No use, distribution or reproduction is permitted which does not polymorphism was genotyped in donors. comply with these terms. Results: We detected a higher incidence of severe acute GVHD in patients transplanted from donors with TT genotype (p: 0.047, HR 1.64; 95% CI 1.01 – 2.67). Overall survival (OS) was worse for patients transplanted from donors with the rs870849 CT/TT genotype (0.020; HR, 1.44; 95% CI 1.06 – 1.96), as well as disease-free survival (DFS) (p: 0.002; HR 1.58, 95%CI: 1.18 – 2.14) and transplant- related mortality (TRM) (p< 0.001; HR: 1.88, 95% CI 1.29 – 2.74). When combining the LAG3 rs870849 and the PDCD1 rs36084323 genotypes of the donor, three Frontiers in Immunology 01 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 genetic groups were well defined, allowing a good stratification of the risk of acute GVHD, TRM, OS and DFS. Discussion: We conclude that the LAG3 genotype of the donor may be considered in donors’ selection. As this selection may be limited in the HLA-identical sibling donor scenario, further studies exploring the impact of LAG3 genotype of the donor in unrelated transplantation are warranted. KEYWORDS LAG3: lymphocyte activation gene 3, graft-versus-host disease, allogeneic transplant, immune response, checkpoint molecules Introduction cleavage by itself seems to be a post-translational regulation of membrane LAG3 expression (18). Genetic polymorphisms in genes coding for molecules responsible LAG-3 has emerged as an important regulatory molecule of the for costimulatory or inhibitory effects on T-lymphocytes have been immune response, comparable to CTLA-4 or PD-1, consisting in the reported to be associated with autoimmune diseases and susceptibility inhibition of T-cell activation. As it has been described for PD-1, to tumors (1, 2). It has also been explored the correlation between these constitutive expression of LAG-3 is associated with exhausted T cells polymorphisms and the appearance of GvHD after allogeneic stem cell and Treg. The LAG-3 gene has 8 exons, and several single-nucleotide transplant. In this sense, specific genetic variants in both the cytotoxic T- polymorphisms (SNPs) have been described. The C/T transition encoded lymphocyte antigen 4 (CTLA4) gene, and the programmed cell death 1 by the SNP rs870849 encodes a substitution of the amino acid isoleucine (PD-1) gene of the donor have been associated with differences in the for the amino acid threonine in the transmembrane domain of the LAG- clinical outcome after allogeneic hematopoietic stem cell transplantation 3 protein (Thr455Ile). Zhang et al. described an association between the (alloHSCT) (3–5). LAG-3 rs870849 TT genotype and an increased risk of multiple sclerosis Lymphocyte Activation 3 (LAG-3), also known as CD223, (19), but these results have not been validated by other groups (20). belongs to the immunoglobulin superfamily and contains 4 To date, there are not data exploring whether LAG-3 may be a extracellular immunoglobulin domains and is expressed in T molecule involved in the development of GVHD in patients receiving lymphocytes, B lymphocytes, NK cells, and plasmacytoid dendritic an allogeneic stem cell transplant. This study aims to investigate cells. The LAG3 gene is located next to CD4 in chromosome 12 and whether donor’s LAG-3 genotype influences clinical outcome after contains 8 exons (6). LAG3 transcription is upregulated following alloHSCT from HLA-identical sibling donors. TCR activation and associated cytokine release (IL-2, IL-12, IL-10, IFN gamma) (7–9). Structurally, LAG-3 is a CD4-homologous protein, although it shares Patients and methods only 20% of the aminoacidic sequence. Following lymphocyte activation, LAG3 interacts with molecules of the major histocompatibility complex Patients of class II (MHC-II) by a few residues of the D1 domain, in the most distal part of the cell membrane of the lymphocyte, generating an We retrospectively studied a cohort of 797 patients receiving an inhibitory signal through its intracellular portion and therefore, allogeneic HSCT from an HLA-identical sibling donor between 2000 bringing the lymphocyte to a state of anergy (10–12). In addition to and 2014 in Spanish centers. The conditioning regimen was MHC-II, other LAG3 ligands have been discovered: galectine 3 (Gal-3), myeloablative in 70.1% of cases and peripheral blood was the most fibrinogen-like protein 1 (FLG-1), and its overexpression on cancer cells used stem cells source. The clinical characteristics of the studied have been identified as a mechanism of tumor immune evasion (13, 14). cohort are summarized in Table 1. DNA isolated from peripheral LAG3 is stored in the trans-Golgi vesicles (late endosome) of CD8+ blood from the HSC donors was analyzed for the LAG3 genotype. lymphocytes together with PD-1, and the action of the two molecules is Written informed consent was obtained from patients and donors synergistic by inhibiting lymphocyte activation (15, 16). before DNA storage. DNA samples were obtained from peripheral blood Like CTLA-4, LAG3 has two main isoforms. The transmembrane using the QIAamp DNA Blood Mini Kit (Qiagen, GmbH, Hilden, form, of 70 kDa, is formed by the 4 immunoglobulin domains (D1- Germany), in accordance with the manufacturer’s instructions, and D4) associated with the intracytoplasmic portion (KIELLE) which will stored at -80°C until use. Samples from patients included in this study deliver the inhibitory signal. The soluble form of LAG3 (sLAG3) is were provided by the IDIBGI Biobank (Biobanc IDIBGI, B.0000872), derived from the proteolytic cleavage of surface LAG3 mediated by integrated in the Spanish National Biobanks Network and they were disintegrin and metalloproteinase ADAM10 and ADAM17 at the processed following standard operating procedures with the appropriate connecting peptide between D4 and the transmembrane domain (17). approval of the Ethics and Scientific Committees. The study was This soluble form of LAG3 has no known biological function, and the approved by the Ethics Committee of the Dr. Josep Trueta University Frontiers in Immunology 02 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 TABLE 1 Clinical characteristics of the patients included in the study according to the LAG3 genotype of the donor (n: 797). All cases CC CT TT p Median age (range) 42.0 (2 – 65) 45.5 (2 – 65) 43.0 (4 – 65) 38 (4 – 65) 0.447 Sex (Male/Female) 488 (61.2%) 309 (38.8%) 177 (61%) 113 (39%) 229 (61.1%) 146 (38.9%) 82 (62.1%) 50 (37.9%) 0.974 Male recipient - Female donor 192 (24.1%) 60 (20.7%) 88 (23.5%) 44 (33.3%) 0.018 Diagnosis Acute lymphoblastic leukemia 134 (16.8%) 43 (14.8%) 60 (16%) 31 (23.5%) 0.110 Acute myeloid leukemia 279 (35.0%) 100 (34.5%) 141 (37.6%) 38 (28.8%) Chronic myeloid leukemia 106 (13.3%) 42 (14.5%) 48 (12.8%) 16 (12.1%) Non-Hodgkin’s lymphoma 151 (18.9%) 57 (19.7%) 71 (18.9%) 23 (17.4%) Myelodysplastic syndrome 83 (10.4%) 31 (10.7%) 36 (9.6%) 16 (12.1%) Severe aplastic anemia 34 (4.3%) 10 (3.4%) 17 (4.5%) 7 (5.3%) Other 10 (1.2%) 7 (2.4%) 2 (0.5%) 1 (0.8%) Advanced disease (beyond CR1) 270 (33.9%) 91 (34.5%) 136 (40.2%) 43 (37.4%) 0.349 Source of stem cells (PB) 599 (75.1%) 225 (77.6%) 273 (73%) 100 (75.8%) 0.391 Myeloablative conditioning regimen 558 (70%) 184 (64.6%) 266 (71.9%) 94 (73.4%) 0.074 Total body irradiation 258 (32.4%) 91 (32%) 127 (34.4%) 40 (31.3%) 0.730 GvHD prophylaxis 0.273 Cyclosporine + Methotrexate 557 (69.9%) 199 (68.6%) 258 (68.8%) 100 (75.8%) Other combinations 240 (30.1%) 91 (31.4%) 117 (31.2%) 32 (24.2%) CR1, first complete remission; PB, peripheral blood. Comparison between genetic groups. Hospital (Girona, Catalonia, Spain) and fulfilled the recommendations of Univariate analyses were performed to evaluate the association the Helsinki declaration. between the rs870849 genotype of the donor and acute GVHD (aGVHD), overall survival (OS), disease-free survival (DFS), Transplant-related mortality (TRM) and relapse. LAG3 genotyping The Kaplan-Meier method was applied for the analysis of OS and DFS. Curves were compared using the log-rank test. Cumulative According to the allele frequency and previous association with incidence estimates were used to explore differences in aGVHD, TRM autoimmune diseases, we decided to explore the rs870849 C>T and relapse. Death without signs of aGVHD was considered as a polymorphism, located in the transmembrane domain. competitive risk for acute GVHD. The competing risk for TRM was The genotype of the donors for this polymorphism was determined death after relapse and the competing risk for relapse was death in via allelic discrimination plots on the QuantStudioTM 6 and 7 Flex complete remission. Patients with non-malignant diseases were excluded Real-Time PCR System (Applied Biosystem, Carlsbad, CA) by using in the relapse analysis. Chronic GVHD incidence was considered as a TaqMan real-time PCR primers and probes obtained as commercially qualitative variable instead a time-dependent one because the onset of available (TaqMan SNPmGenotypingAssay (rs870849); reference this complication was often missing or not clear, comparing incidence number: 4351379 (Applied Biosystem, Carlsbad, CA). The Assay between genetic groups by the chi squared test in univariate analysis and reagents included both the necessary primers and fluorescently logistic regression model in multivariate analysis. labeled (FAM and VIC) TaqMan MGB probes to amplify and detect Multivariate analysis with regression modeling of competing risks each polymorphism. The PCR cycling conditions were as follows: initial (21) was performed to analyze the effects of the LAG-3 genotype of denaturation at 95°C for 10 min, followed by 40 cycles of denaturation the donor and other factors on aGVHD, TRM and relapse. at 95°C for 15 s and annealing/extension at 60°C for 60 s. The presence Multivariate analysis using Cox proportional hazard model was of wild-type and variant alleles was defined by comparing the relative performed for overall survival and DFS. A two-sided p value below end-point fluorescence created by the degradation of each fluorescently 0.05 was considered to be statistically significant. labeled TaqMan probe, according to the manufacturer’s instructions. Results Statistical analysis LAG3 genotype distribution Allele and genotypes frequencies were formulated by direct counting. Homogeneity between genotype groups was performed using the LAG3 rs870849 genotype could be successfully determined in all Chi-square test for qualitative variables and Student’s T-test or the analyzed donors. As previously described in Caucasian ANOVA for continuous variables. population, the C allele was the majoritarian, being detected in Frontiers in Immunology 03 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 83.5% of cases whereas the T allele was identified in 63.7% of the analysis (CC: 20.2%, CT: 27%, TT: 41.3%; p< 0.001) (Figure 1D). studied cases. The genotype distribution showed 290 donors (36.4%) The TRM incidence of patients receiving transplant from donors with homozygous for the C allele, 132 homozygous for the T allele (16.6%) rs870849 TT genotype was higher than those receiving grafts from a and 375 heterozygous CT (47.1%). donor with any other genotype (CC or CT): 41.3% vs. 24.1% in Distribution of relevant clinical variables was comparable multivariate analysis with regression modeling of competing risks; between groups based on the LAG3 rs870849 genotype of the donor (p< 0.001; HR: 1.88, 95% CI 1.29 – 2.74). except a higher proportion of patients with the combination Male Table 2 shows the detailed univariate and multivariate analysis for recipient - Female donor in the group with donors homozygous for severe GVHD incidence, OS, DFS and TRM. the T allele (Table 1). Concerning the relapse incidence at 5 years, we did not find statistically significant differences related to the LAG3 rs870849 genotype of the donor (CC: 26.3%, CT: 23.9%, TT: 33.5%; p: 0.118). LAG3 genotype of the donor and clinical Multivariate analysis with regression modeling of competing risks also outcome after allogeneic hematopoietic failed to detect an association between the LAG3 rs870849 genotype transplant from HLA-identical sibling donors and relapse incidence (p: 0.770; HR 1.07, 95%CI: 0.60 – 1.45). We performed a specific subgroup analysis in patients with acute All patients showed successful engraftment. The cumulative myeloblastic leukemia (AML), and we detected that the LAG3 rs870849 incidence of grades II, III or IV acute GvHD at 180 days was CT/TT genotype of the donor was not associated with relapse (CC: 28% similar in patients receiving grafts from donors that were vs. CT/TT: 50.5%; p: 0.310; HR 1.40, 95%CI: 0.73 – 2.69). homozygous for the LAG3 rs870849 C allele (34.9%), heterozygous Also, we explored whether LAG3 rs870849 genotype of the donor CT (32.5%) or homozygous for the T allele (33.8%) (univariate was associated with relapse in patients with or without acute GVHD analysis showing p: 0.829 and multivariate analysis p: 0.768). but did not find this association neither in patients with aGVHD (p: When considering only severe, grades III or IV acute GVHD at 0.350) nor in those not presenting this complication (p: 0.634). 180 days, the incidence was 13.7% for patients receiving grafts from homozygous rs870849 CC donors, 10.3% for CT genotype, and 18.6% for patients with donors homozygous for the T allele (p: 0.082). Interaction between LAG3 and Patients transplanted from LAG3 rs870849 TT donors had a higher PDCD1 genotypes incidence of grades III or IV aGVHD when compared with any other genotype (considering together CC and CT genotypes): 18.6% vs LAG3 is stored together with PDCD1, and its liberation seems to 11.8% respectively (Figure 1A). Multivariate analysis with regression have synergistic action by inhibiting T-lymphocyte activation. As the modeling of competing risks identified the LAG3 rs870849 TT PDCD1 genotype of the donor has been previously described as genotype of the donor as an independent risk factor for developing having an impact on clinical outcome after HLA-identical sibling severe acute GVHD (p: 0.047; hazard ratio [HR]: 1.64, 95% donor transplant, we explored the additive effect of the LAG3 confidence interval [CI] 1.01 – 2.67). rs870849 C>T and the PDCD1 rs36084323 A>G genotypes of the The incidence of chronic GvHD was similar between the LAG3 donor on the severe acute GVHD incidence. rs870849 genetic groups (CC 36.8%; CT 42.7%; TT: 40.2%: p: 0.402). The genotype at both SNPs was available for 732 donors. The The presence of a LAG3 rs870849 T allele in the donor was LAG3 rs870849 TT genotype was associated to the PDCD1 associated with a significant reduction in overall survival at 5 years. rs36084323 GG (n: 119; 16.3% of the total combinations) or AG (n: This association was dependent of the number of T alleles of the 4; 0.5%) genotypes. The LAG3 rs870849 CC genotype was associated donor (OS at 5 years: CC: 58.7%, CT: 50.2%, TT: 43.1%; p: 0.004) to the PDCD1 rs36084323 GG genotype in 249 cases (34%), AG in 10 (Figure 1B). Patients transplanted from LAG3 rs870849 TT donors cases (1.4%) and AA in 5 cases (0.7%). The LAG3 rs870849 had worse OS than those transplanted from donors with any other heterozygous CT genotype was associated to the PDCD1 genotype (pooling together CC and CT genotypes) (43.1% vs. 53.9%; rs36084323 GG genotype in 328 cases (44.8%), AG in 16 cases p: 0.014). Multivariate analysis using Cox proportional hazard model (2.2%) and AA in 1 case (0.1%). confirmed that the LAG3 rs870849 TT genotype was an independent The presence of the LAG3 rs870849 TT genotype was associated risk factor for worse OS (p: 0.020; HR, 1.44; 95% CI 1.06 – 1.96). with an increased risk of severe acute GVHD independently of the Disease-free survival at five years was also affected by the LAG3 PDCD1 rs36084323 genotype. According to this data and considering rs870849 genotype of the donor (DFS at 5 years: CC: 55.2%, CT: the previously reported results considering the clinical outcome in 43.9%, TT: 38.2%; p: 0.003) (Figure 1C). When comparing DFS basis to the PDCD1 genotype of the donor, we defined three genetic between patients transplanted from homozygous TT donors versus groups in 741 patients: group 1) 132 patients transplanted from any other genotype we detected statistically significant differences in donors homozygous for the LAG3 rs870849 T allele, independently of the Cox proportional hazard model (p: 0.002; HR 1.58, 95%CI: 1.18 – their PDCD1 rs36084323 genotype; group 2) 577 patients 2.14). Evaluation of statistical interaction between the LAG3 TT transplanted from donors genotyped as LAG3 rs870849 CT or CC genotype and the use of female donors for male recipients did not and PDCD1 rs36084323 GG [as the G allele in homozygosis was detect significance in multivariate analysis (p: 0.279; HR 1.38 95%CI: previously suggested to be associated to a higher incidence of acute 1.06 – 1.81). GVHD (5)]; and group 3) 32 patients transplanted from donors This association with OS and DFS was linked with an increased genotyped as LAG3 rs870849 CT or CC and PDCD1 rs36084323 AG incidence of transplant-related mortality at 5 years in univariate or AA. The low number of cases in the last genetic group is related to Frontiers in Immunology 04 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 A B C D FIGURE 1 Clinical outcome according to the LAG3 rs870849 genotype of the donor. (A) Incidence of grades III to IV acute graft-versus-host disease, (B) Overall survival; (C) Disease-free survival and (D) Transplant-related mortality. TABLE 2 Univariate and multivariate analysis for acute GVHD incidence, overall survival, disease-free survival and transplant-related mortality. Grades III to IV acute GVHD Univariate Multivariate Patient’s age p: 0.554; HR: 1.01 (95% CI 0.99 – 1.03) p: 0.220; HR: 1.01 (95% CI 0.99 – 1.03) Diagnosis acute leukemia p: 0.130; HR: 0.72 (95% CI 0.47 – 1.10) p: 0.075; HR: 0.68 (95% CI 0.45 – 1.04) Sex mismatch (male patient/female donor) p: 0.146; HR: 1.41 (95% CI 0.88 – 2.24) p: 0.380; HR: 1.24 (95% CI 0.77 – 2.00) Source of stem cells (peripheral blood) p: 0.319; HR: 1.30 (95% CI 0.77 – 2.19) p: 0.180; HR: 1.51 (95% CI 0.82 – 2.79) Myeloablative conditioning regimen p: 0.893; HR: 1.03 (95% CI 0.65 – 1.64) p: 0.220; HR: 1.39 (95% CI 0.82 – 2.37) T-cell depletion p: 0.187; HR: 0.55 (95% CI 0.22 – 1.35) p: 0.150; HR: 0.48 (95% CI 0.18 – 1.31) Prophylaxis other than cyclosporine + MTX p: 0.120; HR: 0.71 (95% CI 0.46 – 1.10) p: 0.076; HR: 0.67 (95% CI 0.44 – 1.04) Donor with LAG3 rs870849 TT genotype p: 0.039; HR: 1.67 (95% CI 1.02 – 2.73) p: 0.047; HR: 1.64 (95% CI 1.01 – 2.67) Overall survival Patient’s age p: 0.003; HR: 1.02 (95% CI 1.01 – 1.03) p< 0.001, HR 1.03 (95% CI: 1.02 – 1.04) Gender: male p: 0.600; HR: 1.06 (95% CI 0.85 – 1.33) p: 0.611, HR 0.97 (95% CI: 0.72 – 1.29) Diagnosis acute leukemia p: 0.001; HR: 1.43 (95% CI 1.15 – 1.80) p: 0.003, HR 1.47 (95% CI: 1.14 – 1.90) Advanced disease (beyond CR1) p<0.001; HR: 2.46 (95% CI 1.94 – 3.12) p< 0.001, HR 2.29 (95% CI: 1.79 – 2.93) Sex mismatch (male patient/female donor) p: 0.083; HR: 1.24 (95% CI 0.97 – 1.59) p: 0.154, HR 1.12 (95% CI: 0.93 – 1.60) Source of stem cells (peripheral blood) p: 0.031; HR: 1.34 (95% CI 1.02 – 1.75) p: 0.671, HR 0.93 (95% CI: 0.68 – 1.25) Myeloablative conditioning regimen p: 0.330; HR: 0.89 (95% CI 0.70 – 1.13) p: 0.047, HR 1.37 (95% CI: 1.00 – 1.87) T-cell depletion p: 0.754; HR: 0.94 (95% CI 0.65 – 1.36) p: 0.881, HR 0.96 (95% CI: 0.62 – 1.48) (Continued) Frontiers in Immunology 05 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 TABLE 2 Continued Grades III to IV acute GVHD Univariate Multivariate Prophylaxis other than cyclosporine + MTX p: 0.011; HR: 1.36 (95% CI 1.07 – 1.72) p: 0.177, HR 1.21 (95% CI: 0.92 – 1.58) Donor with LAG3 rs870849 TT genotype p: 0.014; HR: 1.41 (95% CI 1.07 – 1.85) p: 0.020, HR 1.44 (95% CI: 1.06 – 1.96) Disease-free survival Patient’s age p: 0.018; HR: 1.01 (95% CI 1.01 – 1.02) p< 0.001, HR 1.02 (95% CI: 1.01 – 1.03) Gender: male p: 0.800; HR: 0.97 (95% CI 0.78 – 1.21) p: 0.897, HR 0.94 (95% CI: 0.71 – 1.24) Diagnosis acute leukemia p: 0.003; HR: 1.40 (95% CI 1.12 – 1.74) p: 0.003, HR 1.45 (95% CI: 1.13 – 1.87) Advanced disease (beyond CR1) p<0.001; HR: 2.14 (95% CI 1.69 – 2.70) p< 0.001, HR 2.06 (95% CI: 1.62 – 2.63) Sex mismatch (male patient/female donor) p: 0.546; HR: 1.08 (95% CI 0.84 – 1.39) p: 0.740, HR 1.07 (95% CI: 0.78 – 1.47) Source of stem cells (peripheral blood) p: 0.577; HR: 1.07 (95% CI 0.84 – 1.37) p: 0.107, HR 0.79 (95% CI: 0.60 – 1.05) Myeloablative conditioning regimen p: 0.820; HR: 1.03 (95% CI 0.81 – 1.31) p: 0.024, HR 1.44 (95% CI: 1.05 – 1.98) T-cell depletion p: 0.885; HR: 1.02 (95% CI 0.72 – 1.45) p: 0.649, HR 1.16 (95% CI: 0.79 – 1.70) Prophylaxis other than cyclosporine + MTX p: 0.181; HR: 1.18 (95% CI 0.92 – 1.51) p: 0.772, HR 1.04 (95% CI: 0.77 – 1.40) Donor with LAG3 rs870849 TT genotype p: 0.016; HR: 1.38 (95% CI 1.06 – 1.81) p: 0.002, HR 1.58 (95% CI: 1.18 – 2.14) Transplant-related mortality Patient’s age p: 0.003; HR: 1.04 (95% CI 1.02 – 1.05) p<0.001; HR: 1.04 (95% CI 1.03 – 1.06) Gender: male p: 0.327; HR: 1.17 (95% CI 0.85 – 1.60) p: 0.810; HR: 0.95 (95% CI 0.62 – 1.45) Diagnosis acute leukemia p: 0.941; HR: 0.99 (95% CI 0.73 – 1.34) p: 0.880; HR: 1.03 (95% CI 0.72 – 1.46) Advanced disease (beyond CR1) p<0.001; HR: 2.20 (95% CI 1.58 – 3.04) p<0.001; HR: 1.88 (95% CI 1.34 – 2.62) Sex mismatch (male patient/female donor) p: 0.004; HR: 1.59 (95% CI 1.15 – 2.19) p: 0.013; HR: 1.54 (95% CI 1.09 – 2.17) Source of stem cells (peripheral blood) p: 0.156; HR: 1.30 (95% CI 0.90 – 1.88) p: 0.770; HR: 0.94 (95% CI 0.61 – 1.44) Myeloablative conditioning regimen p: 0.383; HR: 0.86 (95% CI 0.62 – 1.20) p<0.001; HR: 1.88 (95% CI 1.28 – 2.77) T-cell depletion p: 0.194; HR: 0.68 (95% CI 0.38 – 1.22) p: 0.250; HR: 0.69 (95% CI 0.36 – 1.30) Prophylaxis other than cyclosporine + MTX p: 0.097; HR: 1.31 (95% CI 0.95 – 1.82) p: 0.470; HR: 0.86 (95% CI 0.58 – 1.28) Donor with LAG3 rs870849 TT genotype p<0.001; HR: 1.95 (95% CI 1.38 – 2.75) p<0.001; HR: 1.88 (95% CI 1.29 – 2.74) the previously described low frequency of the PDCD1 rs36084323 A TT genotype of the donor was an independent risk factor for a higher allele in Caucasian population. Table S1 (supplementary material) incidence of acute GVHD, independently of the PDCD1 genotype (p: shows the clinical characteristics of each genetic group and the 0.031; HR: 1.76, 95%CI: 1.05 – 2.94). homogeneity comparison between groups. Regarding overall survival, patients receiving grafts from donors with When determining the incidence of grades III to IV acute GVHD at LAG3 rs870849 TT genotype had poorer outcome than those with other 180 days according to these genetic groups we found that patients whose genotypes at this SNP and with PDCD1 rs36084323 GG genotype, donor was homozygous for the LAG3 rs870849 T allele had the higher whereas the minority group with LAG3 rs870849 CT or CC and at incidence of this complication (18.6%), patients with donors with any least an A allele at PDCD1 rs36084323 had an excellent OS (43.1%, 53% other LAG3 genotype but homozygous for the PDCD1 rs36084323 G and 70% respectively; p: 0.022) (Figure 2B). When comparing genetic allele had an intermediate incidence (12%) whereas none of the 30 groups face-to-face we detected that patient with donors homozygous for patients whose donors had LAG3 rs870849 CT or CC and PDCD1 the LAG3 rs870849 T allele had a statistically significant worse OS than rs36084323 AG/AA genotypes developed grades III to IV acute GVHD those of group 2 (p: 0.030) or group 3 (p: 0.025), whereas direct (0%). These differences were globally statistically significant (p: 0.025) comparison between groups 2 and 3 did not show differences (p: (Figure 2A). However, comparison between curves showed statistically 0.132). Multivariate analysis including the PDCD1 genotype identified significant differences only when comparing patients with donors with the donor LAG3 rs870849 TT genotype as a risk factor for worse OS (p: LAG3 rs870849 TT genotype and those transplanted from donors 0.042; HR: 2.17, 95% CI: 1.03 – 4.58). genotyped as LAG3 rs870849 CT or CC and PDCD1 rs36084323 AG As expected, DFS followed a similar pattern (38.2%, 48.8% and 70.2% or AA (p: 0.016), whereas no differences were detected when comparing respectively; p: 0.017) (Figure 2C). Again, multivariate analysis any of these groups with patients transplanted from donors genotyped as considering the PDCD1 genotype detected the LAG3 rs870849 TT LAG3 rs870849 CT or CC and PDCD1 rs36084323 GG (p: 0.075 and p: genotype of the donor as an independent risk factor for worse DFS (p: 0.057 respectively). Multivariate analysis confirmed that LAG3 rs870849 0.026; HR 2.32, 95%CI: 1.10 – 4.88). Frontiers in Immunology 06 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 A B C D FIGURE 2 Clinical outcome according to the combined LAG3 rs870849 and PDCD1 rs36084323 genotype of the donor. (A) Incidence of grades III to IV acute graft-versus-host disease, (B) Overall survival; (C) Disease-free survival and (D) Transplant-related mortality. When exploring TRM according to these genetic groups, we found GVHD, increased TRM and worse survival. These results agree with that patients transplanted from donors homozygous for the LAG3 the previously described association of the T allele with multiple rs870849 T allele had a 5-years TRM incidence of 38.3%, much higher sclerosis, suggesting that the presence of the T allele impairs the than those with donors genotyped as LAG3 rs870849 CT or CC and inhibitory potential of the LAG-3 molecule in the allogeneic PDCD1 rs36084323 GG (24.3%) or LAG3 rs870849 CT or CC and transplant setting, leading to an increased T-cell response. PDCD1 rs36084323 AG or AA (3.8%). These differences were It has been speculated that the rs870849 Isoleucine/Threonine statistically significant (p<0.001) (Figure 2D). When comparing (Ile/Thr) substitution would alter the conformational and functional between curves, we found statistically significant differences between properties of the protein, reducing the binding affinity of LAG3 and group 1 and 2 (p: 0.001), group 1 and 3 (p: 0.001) and groups 2 and 3 (p: leading to an enhanced expansion of T-cells (19). However, this 0.021). Multivariate analysis considering both genetic polymorphisms hypothesis seems highly unlikely because this change affects the detected the LAG3 rs870849 TT genotype as an independent risk factor alpha-helical transmembrane domain. for increased TRM (p: 0.005; HR: 1.81, 95%CI: 1.20 – 2.73), whereas the The relevance of this polymorphism could be explained by PDCD1 rs36084323 of the donor showed no statistical association (p: looking at the effects driven by the same amino acid polymorphism 0.087; HR: 3.41, 95%CI: 0.84 – 13.88). that has been described within the B cell inhibitory receptor FcgammaRIIB (FCGR2B) gene (rs1050501: FCGR2B c.695T>C) leading to the same amino acid change (Ile/Thr) in the Discussion transmembrane helix (22). In vitro, the FcgammaRIIB 232Thr showed to be significantly less potent than wild type 232Ile in Many studies have explored the correlation between genetic inhibiting B cell receptor signaling. Moreover, FcgammaRIIB variations in genes involved in the inhibition of the T-cell immune 232Thr was shown to be less effectively distributed to detergent- response and the development of autoimmune diseases. Moreover, insoluble lipid rafts, probably due to the substitution of a large genetic polymorphisms in CTLA-4 and PD-1 genes have been also hydrophobic amino acid (isoleucine) by a polar one (threonine) associated with the incidence of graft-versus-host disease after (23). If this model may be extrapolated to the LAG3 gene is alloHSCT (3–5). unknown but is seems plausible to expect that the biological In the present study, we have examined the relationship between consequences of the rs870849 Ile/Thr change at the transmembrane the LAG-3 rs870849 genotype of the donor and the clinical outcome helix of LAG3 could be similar to those observed in the presence of after alloHSCT from HLA-identical sibling donor. Our findings show the FCGR2B rs1050501 Ile/Thr. Although it is well known the critical that patients transplanted from donors homozygous for the T allele at role of ITIM for the inhibitory function of FcgRIIB, it has been the LAG-3 rs870849 SNP have an increased risk of severe acute described that the transmembrane domain is also essential for Frontiers in Immunology 07 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 inhibition by impeding the conformational changes of BCR, co- genotypes with the risk of aGVHD, overall and non-relapse mortality localizing with lipid raft to activate the inhibitory signaling, in the recessive model [TT vs (CC and CT)]. However, they did not blocking the synaptic co-localization of BCR and CD19 show any association of donor LAG3 genotypes with the explored microclusters and conferring fast lateral mobility (24). endpoints with the allelic (additive) model. Although it is possible However, if this hypothesis is true, the rs870849 T allele (coding that the absence of these associations in the Japanese cohort is for Isoleucine) should be associated with a higher inhibitory activity ancestry-specific, they raise some significant doubts about the of LAG3 and therefore with a lower incidence of acute GVHD in the potential clinical application of donor LAG3 genotypes for allogeneic model, which is just opposite to our findings. This prediction of outcomes after unrelated HCT. In this sense, new inconsistence is difficult to explain. One possible explanation would studies looking for the association between LAG3 genotype of the be a different kinetics of proteolytic cleavage of surface LAG3 donor and clinical outcome after unrelated transplantation for mediated by disintegrin and metalloproteinase ADAM10 and Caucasian patients should clarify these doubts. ADAM17. In vitro studies to correlate the rs870849 genotype with These results need to be confirmed by analysis of independent protein expression are currently in development. patient cohorts, but our approach suggests that the routine It has been reported that LAG-3 and PD-1 molecules have a determination of polymorphisms in genes coding for immune synergistic effect on T-cell inhibition (16). We explored the clinical checkpoint inhibitors should be considered when evaluating the outcome after alloHSCT according to the combined LAG-3 rs870849 predicted clinical outcome associated for each available donor. The and PD1.1 rs36084323 genotype of the donor, and we detected that limitations of the present study are that we did not account for the LAG-3 rs870849 TT genotype is associated with the highest severe multiple comparisons and the associations have not been validated by acute GVHD incidence and TRM, leading to a worse OS and DFS, testing in an independent cohort. independently of the PD-1 rs36084323 genotype. These results The current results could not be applied clinically in donor suggest that the inhibitory effect of LAG-3 may be stronger than selection for related alloHSCT, due to the limited availability of the negative signal driven by PD-1. HLA-identical siblings, but they potentially could be used in donor This observation may be relevant, as cancer immunotherapy selection for unrelated HCT, if these results are reproducible in a based on immune checkpoint inhibitor blockades is currently one similar study with unrelated donors for Caucasian patients. of the main therapeutical approaches for several neoplasms. It is well known that PD-1/LAG-3 expression is a marker of exhausted T-cells infiltrating tumors (25), and high expression of these molecules in T- Data availability statement lymphocytes is associated with poor survival (26). Moreover, it has been suggested that tumors with an immune evasion predominantly The original contributions presented in the study are included in mediated by LAG-3 would be less sensitive to immunotherapy with the article/Supplementary Materials, further inquiries can be directed PD-1 inhibitors (nivolumab, pembrolizumab). In this sense, the to the corresponding author. usefulness of LAG-3 antagonists (Relatlimab) is currently being explored in clinical trials combining both the anti-PD-1 and anti- LAG-3 agents, with promising results (27). Ethics statement This is the first study correlating the LAG-3 genotype of the donor with the occurrence of acute GvHD, transplant-related mortality and The study was approved by the Ethics Committee of the Dr. Josep overall survival after allogeneic transplant in a large cohort of Trueta University Hospital (Girona, Catalonia, Spain). Written Caucasian patients. informed consent to participate in this study was provided by the In contrast, a recent Japanese study explored the relationship participants’ legal guardian/next of kin. between transplant outcome and polymorphisms in the immune checkpoint genes, including the LAG3 rs870849 SNP in patients receiving an unrelated alloHSCT, and the authors failed to find any Author contributions correlation (28). There are some aspects that need to be considered to explain this discrepancy: Takahasi et al. explored Japanese patients DC and DG designed the research, analyzed data and wrote the transplanted from unrelated donors whereas our study is performed article; DC, RR-R, MG-B, NS, NL, PV, ET, and RC performed in Caucasian patients transplanted from HLA-identical sibling research, IG-C, RdC, IH, IB, CM, MG, GS, CF, AS, JP, JL-J, and MJ donors, avoiding hidden or even known HLA mismatches and with participated in the collection and analysis of data. All authors less genetic differences potentially influencing the development of contributed to the article and approved the submitted version. GVHD. Moreover, we must consider the differences in the genetic background between European and Asiatic population: the described genetic distribution of the LAG-3 rs870849 T allele is 38.7% in Funding European whereas this allele represents only 17% in East Asia and 14.9% of donors in the Takahasi study, making difficult to detect This work has been financed by grants PI17/00815 and PI20/ correlations with the clinical outcome. In this study the number of TT 01353, funded by Instituto de Salud Carlos III (ISCIII) and co-funded homozygous donors was too small to test the association of LAG3 by the European Union. Frontiers in Immunology 08 frontiersin.org Cruz et al. 10.3389/fimmu.2023.1066393 Conflict of interest organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or The authors declare that the research was conducted in the claim that may be made by its manufacturer, is not guaranteed or absence of any commercial or financial relationships that could be endorsed by the publisher. construed as a potential conflict of interest. Supplementary material Publisher’s note The Supplementary Material for this article can be found online at: All claims expressed in this article are solely those of the https://www.frontiersin.org/articles/10.3389/fimmu.2023.1066393/ authors and do not necessarily represent those of their affiliated full#supplementary-material References 1. Ueda H, Howson JM, Esposito L, Heward J, Snook H, Chamberlain G, et al. 15. Woo SR, Turnis ME, Goldberg MV, Bankoti J, Selby M, Nirschl CJ, et al. Immune Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote disease. Nature (2003) 423(6939):506–11. doi: 10.1038/nature01621 tumoral immune escape. Cancer Res (2012) 72(4):917–27. doi: 10.1158/0008-5472.CAN- 2. Prokunina L, Castillejo-Lopez C, Oberg F, Gunnarsson I, Berg L, Magnusson V, 11-1620 et al. A regulatory polymorphism in PDCD1 is associated with susceptibility to systemic 16. Okazaki T, Okazaki IM, Wang J, Sugiura D, Nakaki F, Yoshida T, et al. PD-1 and lupus erythematosus in humans. Nat Genet (2002) 32(4):666–9. doi: 10.1038/ng1020 LAG-3 inhibitory co-receptors act synergistically to prevent autoimmunity in mice. J Exp 3. Perez-Garcia A, de la Camara R, Roman-Gomez J, Jimé nez-Velasco A, Encuentra Med (2011) 208(2):395–407. doi: 10.1084/jem.20100466 M, Nieto JB, et al. CTLA-4 polymorphisms and clinical outcome after allogeneic stem cell 17. Li N, Wang Y, Forbes K, Vignali KM, Heale BS, Saftig P, et al. Metalloproteases transplantation from HLA-identical sibling donors. Blood (2007) 110(1):461–7. doi: regulate T-cell proliferation and effector function via LAG-3. EMBO J (2007) 26(2):494– 10.1182/blood-2007-01-069781 504. doi: 10.1038/sj.emboj.7601520 4. Bosch-Vizcaya A, Pé rez-Garcı́a A, Brunet S, Solano C, Buño I, Guillem V, et al. 18. Graydon CG, Mohideen S, Fowke KR. LAG3’s enigmatic mechanism of action. Donor CTLA-4 genotype influences clinical outcome after T cell-depleted allogeneic Front Immunol (2021) 11:615317. doi: 10.3389/fimmu.2020.615317 hematopoietic stem cell transplantation from HLA-identical sibling donors. Biol Blood 19. Zhang Z, Duvefelt K, Svensson F, Masterman T, Jonasdottir G, Salter H, et al. Two Marrow Transpl (2012) 18(1):100–5. doi: 10.1016/j.bbmt.2011.05.021 genes encoding immune-regulatory molecules (LAG3 and IL7R) confer susceptibility to 5. Santos N, Rodrı́guez-Romanos R, de la Cá mara R, Brunet S, Nieto JB, Buño I, et al. multiple sclerosis. Genes Immun (2005) 6(2):145–52. doi: 10.1038/sj.gene.6364171 PD-1 genotype of the donor is associated with acute graft-versus-host disease after HLA- 20. Garcı́a-Martı́n E, Agú ndez JAG, Gó mez-Tabales J, Benito-Leó n J, Millá n-Pascual J, identical sibling donor stem cell transplantation. Ann Hematol (2018) 97(11):2217–24. doi: Dı́az-Sá nchez M, et al. Association between LAG3/CD4 genes variants and risk for 10.1007/s00277-018-3438-y multiple sclerosis. Int J Mol Sci (2022) 23(23):15244. doi: 10.3390/ijms232315244 6. Triebel F, Jitsukawa S, Baixeras E, Roman-Roman S, Genevee C, Viegas-Pequignot 21. Scrucca L, Santucci A, Aversa F. Regression modeling of competing risk using r: an E, et al. LAG-3, a novel lymphocyte activation gene closely related to CD4. J Exp Med in depth guide for clinicians. Bone Marrow Transpl (2010) 45:1388–95. doi: 10.1038/ (1990) 171:1393–405. doi: 10.1084/jem.171.5.1393 bmt.2009.359 7. Bae J, Lee SJ, Park C-G, Lee YS, Chun T. Trafficking of LAG-3 to the surface on 22. Li X, Wu J, Carter RH, Edberg JC, Su K, Cooper GS, et al. A novel polymorphism in activated T cells via its cytoplasmic domain and protein kinase c signaling. J Immunol the fcgamma receptor IIB (CD32B) transmembrane region alters receptor signaling. (2014) 193:3101–12. doi: 10.4049/jimmunol.1401025 Arthritis Rheumatol (2003) 48(11):3242–52. doi: 10.1002/art.11313 8. Annunziato F, Manetti R, Tomasé vic I, Guidizi MG, Biagiotti R, Giannò V, et al. 23. Kono H, Kyogoku C, Suzuki T, Tsuchiya N, Honda H, Yamamoto K, et al. Expression and release of LAG-3-encoded protein by human CD4 + T cells are associated FcgammaRIIB Ile232Thr transmembrane polymorphism associated with human systemic with IFN - gamma production. FASEB J (1996) 10:769–76. doi: 10.1096/ lupus erythematosus decreases affinity to lipid rafts and attenuates inhibitory effects on b fasebj.10.7.8635694 cell receptor signaling. Hum Mol Genet (2005) 14(19):2881–92. doi: 10.1093/hmg/ddi320 9. Bruniquel D, Borie N, Hannier S, Triebel F. Regulation of expression of the human 24. Wang J, Li Z, Xu L, Yang H, Liu W. Transmembrane domain dependent inhibitory lymphocyte activation gene-3 (LAG-3) molecule, a ligand for MHC class II. function of FcgRIIB. protein cell. 2018 Dec;9(12):1004-1012.Grosso JF, Goldberg MV, Immunogenetics (1998) 48:116–24. doi: 10.1007/s002510050411 getnet d et al. functionally distinct LAG-3 and PD-1 subsets on activated and chronically 10. Maçon-Lemaı̂tre L, Triebel F. The negative regulatory function of the lymphocyte- stimulated CD8 T cells. J Immunol (2009) 182:6659–69. doi: 10.1007/s13238-018-0509-8 activation gene-3 co-receptor (CD223) on human T cells. Immunology (2005) 115:170–8. 25. Datar I, Sanmamed MF, Wang J, Henick BS, Choi J, Badri T, et al. Expression doi: 10.1111/j.1365-2567.2005.02145.x analysis and significance of PD-1, LAG-3, and TIM-3 in human non–small cell lung 11. Huard B, Tournier M, Hercend T, Triebel F, Faure F. Lymphocyte-activation gene 3/ cancer using spatially resolved and multiparametric single-cell analysis. Clin Cancer Res Major histocompatibility complex class II interaction modulates the antigenic response of CD4 (2019) 25:4663–73. doi: 10.1158/1078-0432.CCR-18-4142 + T lymphocytes. Eur J Immunol (1994) 24:3216–21. doi: 10.1002/eji.1830241246 26. Zhang X, Zhao H, Shi X, Jia X, Yang Y. Identification and validation of an 12. Huard B, Mastrangeli R, Prigent P, Bruniquel D, Donini S, El-Tayar N, et al. immune-related gene signature predictive of overall survival in colon cancer. Aging (2020) Characterization of the major histocompatibility complex class II binding site on LAG-3 12:26095–120. doi: 10.18632/aging.202317 protein. Proc Natl Acad Sci USA (1997) 94:5744–9. doi: 10.1073/pnas.94.11.5744 27. Tawbi HA, Schadendorf D, Lipson EJ, Ascierto PA, Matamala L, Castillo Gutié rrez 13. Chung LY, Tang SJ, Wu YC, Sun GH, Liu HY, Sun KH. Galectin-3 augments E, et al. Relatlimab and nivolumab versus nivolumab in untreated advanced melanoma. N tumor initiating property and tumorigenicity of lung cancer through interaction with Engl J Med (2022) 386(1):24–34. doi: 10.1056/NEJMoa2109970 Beta-catenin. Oncotarget (2015) 6:4936–52. doi: 10.18632/oncotarget.3210 28. Takahashi H, Okayama N, Yamaguchi N, Nomura M, Miyahara Y, Mahbub MH, 14. Wang J, Sanmamed MF, Datar I, Su TT, Ji L, Sun J, et al. Fibrinogen-like protein 1 et al. Analysis of relationships between immune checkpoint and methylase gene is a major immune inhibitory ligand of LAG-3. Cell (2019) 176:334–47. doi: 10.1016/ polymorphisms and outcomes after unrelated bone marrow transplantation. Cancers j.cell.2018.11.010 (Basel) (2021) 13(11):2752. doi: 10.3390/cancers13112752 Frontiers in Immunology 09 frontiersin.org