167 Increased Programmed Cell Death 1 Ligand 2 (PD-L2) Protein and Transcript Levels in Triple Negative Breast Cancer Patients Improve Prognostic Value Through Immune Pathway Activation

Background Recent advances in immunotherapy have shown promising results in triple-negative breast cancer (TNBC). Studies regarding PDL1/PD1 co-inhibitory pathway demonstrate that PDL1 positive subgroup of TNBCs had better prognosis and increased clinical benefit from immunotherapy.1 Even then, other predictive biomarkers beyond PDL1 to better stratify immunotherapy responsive patients is necessary, where tumor infiltrating lymphocytes (TILs) may help predict prognosis when used in conjunction with PDL1 expression.2 Thus, we explored the prognostic significance of another known ligand of PD1,3 programmed cell death ligand 2 (PDL2), in TILs of TNBC patients. Methods 389 TNBC cases diagnosed between 2003–2014 in Singapore General Hospital (SGH) were used. Tissue microarray blocks was stained with anti-PDL2 antibody (D7UHC). Immunostaining was scored based on TILs expression and positivity was defined as ≥ 2 TILs count. PDL2 mRNA ( PDCD1LG2 ) and their association with survival outcomes were assessed in four cohorts (SGH/METABRIC/TCGA/FUSCC). Differentially expressed genes (DEGs) between PDL2 and PDCD1LG2 sample groups were identified using Welch’s t tests with multiple testing corrections. Protein-protein interaction (PPI) of identified DEGs was constructed to characterize hub-genes. Prognostic models were compared using ANOVA to evaluate the delta in log-likelihood of the models (ΔLRx 2). Results 22.9% (71/309) TNBCs expressed higher densities of PDL2 TILs, where PDL2+ patients demonstrated significantly better disease-free survival (DFS, P = 0.044) and a trend towards improved overall survival (OS, P = 0.058) ( figure 1A,B ). Similarly, higher PDCD1LG2 transcript levels significantly improved survival rates in multiple cohorts ( figure 1C,D ). Additionally, the inclusion of PDL2 immune infiltrates and PDCD1LG2 expression, individually or combined, added significant prognostic value for DFS (PDL2: ΔLRx 2 = 0.71, PDCD1LG2 : ΔLRx 2 = 3.3, PDL2 + PDCD1LG2 : ΔLRx 2 = 2.36) and OS (PDL2: ΔLRx 2 = 0.61, PDCD1LG2 : ΔLRx 2 = 3.47, PDL2 + PDCD1LG2 : ΔLRx 2 = 2.79) beyond that provided by classical clinicopathological variables ( table 1–2 ). Hub genes were identified between DEGs of PDL2 TILs (SGH: PTPRC, CD4, CD8A, IL10RA, SELL, ITGB2, CTLA4, LCP2, IFNG, STAT1 ) and PDCD1LG2 (All cohorts: PTPRC, CD4, CD86, IL10RA, IL1B, ITGB2, CD80, IL10, IFNG, CCL5 ) ( figure 2 ). All hub genes were remarkably downregulated in negative PDL2 TILs or low PDCD1LG2 TNBCs with poorer survival rates ( table 3 ). Conclusions Overall, PDL2 expression in TILs could represent as an effective prognostic predictor through its involvement in antitumor immunity within the TNBC tumor microenvironment. These findings further corroborate PDL2 as a potential therapeutic target for TNBC, highlighting the importance of considering immune-related factors in TNBC prognosis and treatment. Acknowledgements We like to thank the team at the Cancer Epigenome research laboratory at National Cancer Centre for the processing and running of the NanoString experiments in our study. References 1. Schmid P, Adams S, Rugo HS, Schneeweiss A, Barrios CH, Iwata H, et al . Atezolizumab and Nab-Paclitaxel in Advanced Triple-Negative Breast Cancer. N Engl J Med . 2018; 379 (22):2108–21. Epub 2018/10/23. doi: 10.1056/NEJMoa1809615. PubMed PMID: 30345906. 2. Marra A, Viale G, Curigliano G. Recent advances in triple negative breast cancer: the immunotherapy era. BMC Med . 2019; 17 (1):90. Epub 2019/05/10. doi: 10.1186/s12916–019-1326–5. PubMed PMID: 31068190; PubMed Central PMCID: PMCPMC6507064. 3. Baptista MZ, Sarian LO, Derchain SF, Pinto GA, Vassallo J. Prognostic significance of PD-L1 and PD-L2 in breast cancer. Hum Pathol . 2016; 47 (1):78–84. Epub 2015/11/07. doi: 10.1016/j.humpath.2015.09.006. PubMed PMID: 26541326. Ethics Approval Ethics approval was obtained from The Centralized Institutional Review Board of Singhealth, Singapore. Consent Consent was waived for this study from The Centralized Institutional Review Board of Singhealth, Singapore. ![Abstract 167 Figure 1][1] Abstract 167 Figure 1 ![Abstract 167 Figure 2][1] Abstract 167 Figure 2 View this table: Abstract 167 Table 1 View this table: Abstract 167 Table 2 View this table: Abstract 167 Table 3 [1]: pending:yes