Identification of LAG3+ T Cell Populations in the Tumor Microenvironment of Classical Hodgkin Lymphoma and B-Cell Non-Hodgkin Lymphoma
Katsuyoshi Takata,Tomohiro Aoki,Lauren C. Chong,Katy Milne,Tomoko Miyata-Takata,Karanvir Singh,Talia Goodyear,Pedro Farinha,Graham W. Slack,Laurie H. Sehn,Kerry J. Savage,Brad H Nelson,David W. Scott,Christian Steidl
DOI: https://doi.org/10.1182/blood-2020-141462
IF: 20.3
2020-11-05
Blood
Abstract:Background: LAG3 is one of the immune check point receptors that are expressed on activated cytotoxic T-cells and regulatory T cells. Physiologically, T-cell proliferation and memory T-cell differentiation is negatively regulated by LAG3-MHC interaction. In cancer tissues, T-cells that are chronically exposed to tumor antigens might upregulate LAG3 and receive inhibitory stimuli to enter an exhaustion state limiting anti-tumor immune responses. Currently, clinical trials using double blockade of LAG3/PD1 are active in several solid tumours, but there are only a small number of clinical trials using LAG3 monoclonal antibodies in lymphoma. Recently, we published a characteristic LAG3+ T-cell population as a mediator of immune suppression in classical Hodgkin lymphoma (Aoki & Chong et al. Cancer Discovery 2020). However, the abundance and variability of LAG3 positive T-cell populations across a spectrum of B-cell lymphoma has not been well studied and it remains an open question if LAG3 expression is associated with treatment outcome under standard-of-care conditions. Methods: We performed a LAG3 immunohistochemical (IHC) screen in a large cohort of B-cell Non-Hodgkin lymphoma (diffuse large B-cell lymphoma (DLBCL); N=341, follicular lymphoma (FL); N=198 (grade 1-3A), transformed FL to aggressive lymphoma (tFL); N=120, mantle cell lymphoma (MCL); N=179, primary mediastinal large B-cell lymphoma (PMBCL); N=61) and classical Hodgkin lymphoma (HL; N=459) to assess LAG3 expression in the tumor microenvironment (TME). Moreover, we characterized LAG3+ T-cell populations using multi-color immmunohistochemistry (IHC) (LAG3, PD1, CD4, CD8, FOXP3, CD20) in various lymphoma subtypes. Clinical parameters including treatment outcome were correlated with the abundance of LAG3+ T-cell populations in the TME. Results: On average, HL (7%) and PMBCL (6%) showed higher LAG3+ cellular frequency than the other B-cell lymphoma subtypes studied (DLBCL and FL: 2%, MCL: 0.8%). Comparing the frequency of LAG3+ cells according to MHC class I/II status, DLBCL showed a significant correlation with MHC class I status, and LAG3 expression correlated with MHC class II status in HL. Next, we performed multi-color IHC to describe subtype-specific expression patterns of LAG3 in T cell subsets. LAG3+PD1- T-cells were predominantly found in HL and PMBCL with only rare LAG3+PD1+ cells in HL. The majority of LAG3+ T-cells co-expressed CD4 in HL, in contrast to CD8 in PMBCL. DLBCL showed a mixed population pattern with a 1:1 ratio of LAG3+PD1- and LAG3+PD1+ T-cells. In FL, the majority of LAG3+ T-cells were CD4+PD1+, suggesting a more exhausted TME phenotype in FL than in other lymphoma subtypes. Cellular distance analysis showed that LAG3+CD4+ T-cells were in close vicinity to CD20+ lymphoma cells in FL, while in DLBCL and PMBCL, the nearest neighbors of malignant cells were LAG3+CD8+. Triple-positive LAG3+PD1+CD8+ T-cells significantly correlated with high infiltrating M2 macrophage (Pearson's correlation test, P < 0.001) content and the ABC cell-of-origin subtype (Pearson's correlation test, P = 0.002) in DLBCL. The abundance of LAG3+CD8+PD1- cells correlated with a high FLIPI score (Pearson's correlation test, P = 0.033), disease specific survival (HR = 2.8, 95% CI = 1.3-5.9, P = 0.006), time to progression (HR = 2.8, 95% CI = 1.6-5.0, P = 0.001) and transformation (HR = 4.0, 95%CI = 1.7-9.6, P = 0.002) in FL treated with R-CVP (N = 135). Assessing LAG3 expression by single color IHC in FL (cut-off at 5%), patients with LAG3-positive samples showed significantly higher FL transformation rates (P = 0.023) and tFL samples showed higher abundance of LAG3+ cells than the corresponding primary pretreatment FL samples (primary FL: 1.5±1.7% vs. tFL: 4.2±3.8%, t-test, P = 0.01). The increased transformation risk was validated in an independent FL cohort treated with R-CHOP/CVP (N=97, HR = 6.2, 95% CI = 2.8-13.9, P < 0.001). Conclusion: The highest abundance of LAG3+ T-cells in the TME was found in HL and its related entity PMBCL. The differential outcome correlations and co-expression patterns in LAG3+ T cells across B-cell lymphoma subtypes indicate heterogeneity in TME composition and related pathogenic mechanisms. Our results suggest that LAG3 expression patterns will be important in the interpretation of ongoing studies and highlight populations that may benefit from LAG3 checkpoint inhibition. Disclosures Sehn: AstraZeneca: Consultancy, Honoraria; Genentech, Inc.: Consultancy, Honoraria, Research Funding; Amgen: Consultancy, Honoraria; AbbVie: Consultancy, Honoraria; Chugai: Consultancy, Honoraria; TG therapeutics: Consultancy, Honoraria; Verastem Oncology: Consultancy, Honoraria; Teva: Consultancy, Honoraria, Research Funding; Servier: Consultancy, Honoraria; F. Hoffmann-La Roche Ltd: Consultancy, Honoraria, Research Funding; MorphoSys: Consultancy, Honoraria; Takeda: Consultancy, Honoraria; Apobiologix: Consultancy, Honoraria; Seattle Genetics: Consultancy, Honoraria; Gilead: Consultancy, Honoraria; Kite: Consultancy, Honoraria; Merck: Consultancy, Honoraria; Lundbeck: Consultancy, Honoraria; Karyopharm: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Celgene: Consultancy, Honoraria; Acerta: Consultancy, Honoraria. Savage:Merck, BMS, Seattle Genetics, Gilead, AstraZeneca, AbbVie, Servier: Consultancy; BeiGene: Other: Steering Committee; Roche (institutional): Research Funding; Merck, BMS, Seattle Genetics, Gilead, AstraZeneca, AbbVie: Honoraria. Scott:Celgene: Consultancy; Abbvie: Consultancy; AstraZeneca: Consultancy; NIH: Consultancy, Other: Co-inventor on a patent related to the MCL35 assay filed at the National Institutes of Health, United States of America.; Roche/Genentech: Research Funding; NanoString: Patents & Royalties: Named inventor on a patent licensed to NanoString, Research Funding; Janssen: Consultancy, Research Funding. Steidl:Bayer: Consultancy; Juno Therapeutics: Consultancy; Roche: Consultancy; Seattle Genetics: Consultancy; Bristol-Myers Squibb: Research Funding; AbbVie: Consultancy; Curis Inc: Consultancy.
hematology