Abstract Background PD-1 inhibitors have shown significant efficacy in triple negative breast cancer (BC), however, durable responses are less common in estrogen receptor-positive (ER+) BC. Better markers are therefore needed that will identify likely responders to PD-1 inhibitors among patients with luminal BC. While most efforts have focused on the immune checkpoint protein PD-L1, the alternative PD-1 ligand, PD-L2, has been largely overlooked. We aimed to determine if PD-L2 is associated with unfavorable prognosis in ER+ BC. Methods PD-L2 protein levels in cancer cells and stromal cells were measured retrospectively by quantitative immunofluorescence histocytometry in tissue microarrays of therapy-naïve, localized or locoregional ER+ BC and correlated with progression-free survival (PFS). Evaluable tumor PD-L2 data were derived from a main study cohort A diagnosed between 1988-2005 (n=684) with extensive clinical and outcome data and from an independent validation cohort B diagnosed between 1992-2012 (n=273). Patients received standard-of-care adjuvant therapy without immune checkpoint inhibitors after tumor resection. Results Univariate analysis of the main cohort A revealed that high PD-L2 expression in cancer cells was associated with shorter PFS (HR=1.8; 95%CI:1.3-2.6; p=0.001), an observation that was validated in an independent cohort B (HR=2.3, 95%CI:1.1-4.8; p=0.026). Approximately one third of ER+ BC cases were classified as high PD-L2. After multivariable adjustment for common clinicopathological variables, high cancer cell levels of PD-L2 remained independently predictive of early recurrence (HR=2.0; 95%CI:1.4-2.9; p< 0.001). Sub-analysis of ER+ BC cases treated with adjuvant chemotherapy (n=197) suggested that high PD-L2 levels in cancer cells was associated with particularly increased risk of progression (multivariable HR=3.4; 95%CI:1.9-6.2; p< 0.001). The observed frequent expression of PD-L2 protein in BC provided scientific rationale for the design of our ongoing phase II clinical trial (NCT04243616) of neoadjuvant combined PD-1 inhibitor (cemiplimab; Regeneron Pharmaceuticals Inc) and chemotherapy in patients diagnosed with PD-L1+ and/or PD-L2+ BC. The primary objective of this trial is to assess pathologic responses to neoadjuvant treatment with secondary objective of assessing the correlation between PD-L1/PD-L2 status and tumor responses. Pathologist review of PD-L1 and PD-L2 expression in an initial set of ER+ tumors (n=15) screened for trial eligibility revealed frequent discordance between cancer cell positivity for PD-L1 and PD-L2 protein (PD-L1: median=0%; range 0-2% vs. PD-L2: median=18%; range < 1-60%) as well as immune cell positivity (PD-L1: median=5%; range 0-50% vs. PD-L2: median=0; range 0-50%). Conclusions In treatment-naïve ER+ breast tumors, high cancer cell expression of PD-L2 protein was an independent predictor of poor clinical outcome, with evidence of further elevated risk of progression in patients who received adjuvant chemotherapy. Preliminary analyses of ER+ tumors from our ongoing clinical trial showed frequent discordance between baseline PD-L1 and PD-L2 protein expression in both cancer cells and immune cells. Collectively, our analyses indicate that PD-L2 has prognostic value for ER+ BC, and our progress justify further studies to determine whether PD-L2, alone or in combination with PD-L1, may serve as a predictive marker of response to PD-1 inhibitors. Citation Format: Lubna N. Chaudhary, Inna Chervoneva, Amy R. Peck, Yunguang Sun, Misung Yi, John F. Langenheim, Julie M. Jorns, Sailaja Kamaraju, Yee Chung Cheng, John Burfeind, Christopher R. Chitambar, Jeffrey A. Hooke, Albert J. Kovatich, Craig Shriver, Hai Hu, Juan P. Palazzo, Marluce Bibbo, Terry Hyslop, Richard Pestell, Edith P. Mitchell, Hallgeir Rui. High PD-L2 Protein Expression in Cancer Cells is an Independent Marker of Unfavorable Prognosis in Luminal Breast Tumors [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-11-13.