Integrated Genetic and Topological Analysis Reveals a Hodgkin-like Mechanism of Immune Escape in T-Cell/Histiocyte-Rich Large B-Cell Lymphoma
Gabriel K. Griffin,Margaretha G. M. Roemer,Mikel Lipschitz,Jason Weirather,Christine J. Pak,Allysa Kelly,Daniel Gusenleitner,Evisa Gjini,Bjoern Chapuy,Jie Xu,Benjamin Chen,Aliyah R. Sohani,Scott B. Lovitch,Christopher Fletcher,Gordon J. Freeman,Stephen Hodi,Philippe Armand,Azra H. Ligon,Margaret A. Shipp,Scott J. Rodig
DOI: https://doi.org/10.1182/blood-2018-99-118610
IF: 20.3
2018-01-01
Blood
Abstract:Abstract Introduction: T-cell/histiocyte-rich large B-cell lymphoma (TCRLBCL) is an aggressive large B cell lymphoma that typically presents with disseminated disease. In contrast to diffuse large B-cell lymphoma, not otherwise specified (DLBCL), TCRLBCL is characterized histologically by rare malignant B-cells within a robust but ineffective inflammatory background composed of numerous T cells and macrophages. TCRLBCL shows a "tolerogenic" immune signature by gene expression profiling, as well as frequent upregulation of PD-L1 (Van Loo et al. PMID: 19797726; Chen et al. PMID: 23674495). Although these features suggest that active immune evasion is central to TCRLBCL pathogenesis, its mechanistic basis is poorly understood. Accordingly, we performed an integrated analysis of tumor genetics and cell-cell interactions within the tumor microenvironment to comprehensively study PD-1:PD-L1 interactions in a multi-institutional cohort of TCRLBCL. Methods: 34 cases of TCRLBCL were identified from the pathology archives of four academic medical centers. Control cohorts containing 21 cases of DLBCL and 106 cases of classic Hodgkin Lymphoma (CHL) were used as comparators. An established fluorescence in situ hybridization (FISH) assay was used to identify copy number changes and structural rearrangements of CD274 (PD-L1) and PDCD1LG2 (PD-L2) on chromosome 9p24.1, which represents the primary genetic mechanism of PD-L1/L2 expression in CHL (Roemer et al. PMID: 27069084). Tumor-specific expression of PD-L1 and PD-L2 protein was assessed by immunohistochemistry (IHC) and scoring by two pathologists using a modified H-score (percentage of positive tumor cells [0-100%] multiplied by the mean staining intensity [0-3+]). The topology of PD-L1/PD-1 expression and cell-cell interactions in the tumor microenvironment was determined by multispectral immunofluorescence (IF) microscopy and spatial image analysis, as previously performed for CHL (Carey et al. PMID: 28893733). Results: By FISH, copy gain or amplification of PD-L1 and PD-L2 was identified in 22/34 (64.7%) cases of TCRLBCL (Figure 1A) and was associated with a 4.9-fold increase in tumor PD-L1 expression relative to cases with disomy or polysomy (mean PD-L1 H-score 72 vs 14.7, p = 0.02). A rearrangement of PD-L2 was identified in one case and associated with diffuse expression of PD-L2. These findings contrasted with those observed in the DLBCL cohort, which showed a low overall frequency of 9p24.1 copy gain/amplification (5/21 cases, 23.8%) and only minimal tumor PD-L1 expression (mean PD-L1 H-score 15.6), and were intermediate to those observed in CHL, which shows near universal copy gain/amplification of 9p24.1 (98/106 cases, 92%) and extensive tumor PD-L1 expression (mean PD-L1 H-score 143.7; Figure 1B). By multispectral IF, TCRLBCL showed prominent infiltration by PD-L1+ tumor-associated macrophages (TAM) (Figure 1C), which were 5.5-fold increased relative to DLBCL and 6.6-fold increased relative to CHL (p < 0.001). TCRLBCL also showed marked infiltration by PD-1+ T cells, which were 12.3-fold increased relative to DLBCL and 3.4-fold increased relative to CHL (p < 0.001). By spatial analysis, PD-L1+ TAMs in TCRLBCL were located in closer proximity to tumor cells than PD-L1- TAMs (p < 0.001, Figure 1D-E) and also showed frequent direct interactions with PD-1+ T cells. These findings contrasted with those in DLBCL, where no local enrichment of PD-L1+ TAMs or PD-1+ T cells was identified, and were similar but more prominent than those observed in CHL. Conclusion: TCRLBCL is characterized by recurrent gains of PD-L1 and PD-L2 on chromosome 9p24.1 in association with tumor-specific expression of PD-1 ligands, as well as prominent infiltration by PD-L1+ TAMs and PD-1+ T cells. PD-L1+ TAMs in TCRLBCL are enriched around individual tumors cells and also show frequent direct interactions with PD-1+ T cells, consistent with the establishment of an immunoevasive-niche. These findings contrast with those observed in DLBCL and are most similar to those identified in CHL. Relative to CHL, however, TCRLBCL shows less frequent gains of 9p24.1 and tumor cell expression of PD-L1, and a greater degree of infiltration by PD-L1+ TAMs and PD-1+ T cells. These findings suggest that the PD-1:PD-L1 pathway is central to immune evasion in TCRLBCL and highlight the need to test the clinical efficacy of PD-1 blockade in this patient population. Disclosures Griffin: Moderna Therapeutics: Consultancy. Freeman:Novartis: Patents & Royalties; AstraZeneca: Patents & Royalties; Dako: Patents & Royalties; Roche: Membership on an entity's Board of Directors or advisory committees; Bristol-Myers-Squibb: Patents & Royalties; Merck: Patents & Royalties; EMD-Serono: Patents & Royalties; Roche: Patents & Royalties; Xios: Membership on an entity's Board of Directors or advisory committees; Boehringer-Ingelheim: Patents & Royalties; Bristol-Myers-Squibb: Membership on an entity's Board of Directors or advisory committees; Origimed: Membership on an entity's Board of Directors or advisory committees. Hodi:Merck: Consultancy. Shipp:Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Merck: Research Funding; Bayer: Research Funding; AstraZeneca: Honoraria. Rodig:KITE: Research Funding; Affimed: Research Funding; Merck: Research Funding; Bristol Myers Squibb: Research Funding.