Background: The Tumor Proportion Score (TPS) and Combined Positive Score (CPS) scoring algorithms are used in conjunction with PD-L1 IHC 22C3 pharmDx for the immunohistochemical evaluation of PD-L1 in certain human cancer tissues; both algorithms include PD-L1 staining tumor cells (TC) and have been correlated with response to pembrolizumab therapy (KEYTRUDA®) at specific expression levels in specific tumor indications. This study aims to evaluate whether PD-L1 expression on TC could be important in the evaluation of TNBC with PD-L1 IHC 22C3 pharmDx and the impact on reaching a diagnostic cut-off, as current PD-L1 testing (SP142) for TNBC relies solely on scoring immune cells (IC). TNBC clinical data from KEYNOTE-119 (ClinicalTrials.gov, NCT02555657) and Agilent’s internal TNBC tumor bank data were evaluated by TPS, CPS, and Quantitative Immune Cell Density (QID) to determine the significance of including PD-L1 expressing TC in the scoring algorithms. PD-L1 IHC 22C3 pharmDx is not currently approved for use with TNBC specimens.Methods: TNBC specimens from clinical trial KEYNOTE-119 and Agilent’s internal tumor bank were stained with PD-L1 IHC 22C3 pharmDx and scored using the TPS, CPS, and QID algorithms. The TPS algorithm is defined as the number of PD-L1 staining tumor cells divided by the total number of viable TC, multiplied by 100. The CPS algorithm includes TC and IC and is defined as the number of PD-L1 staining cells (TC, lymphocytes and macrophages) divided by the total number of viable TC, multiplied by 100. In addition to TPS and CPS, QID was also calculated to quantify the contribution from PD-L1 expressing IC. QID is defined as the CPS minus the TPS (QID = CPS - TPS). The data was analyzed to show the percent of specimens that fell below the diagnostic PD-L1 cut-offs of ≥1 and ≥10 when evaluated using QID as compared to CPS. Both scores measure countable PD-L1 staining cells per 100 viable tumor cells. Results: A total of 120 PD-L1 IHC 22C3 pharmDx stained TNBC specimens from Agilent’s internal tumor bank were analyzed. When evaluated using CPS, 71 specimens were positive for the PD-L1 ≥1 cut-off and 49 were positive for the PD-L1 ≥10 cut-off. When analyzed with QID, 8.5% of specimens fell below the PD-L1 ≥1 cut-off and 26.5% fell below the PD-L1 ≥10 cut-off (Table 1). A total of 964 PD-L1 IHC 22C3 pharmDx stained TNBC specimens from KEYNOTE-119 were analyzed. When evaluated using CPS, 604 specimens were positive for the PD-L1 ≥1 diagnostic cut-off and 284 were positive for the PD-L1 ≥10 diagnostic cut-off. When analyzed with QID, 9.3% of patients fell below the PD-L1 ≥1 diagnostic cut-off and 33.5% fell below the PD-L1 ≥10 diagnostic cut-off (Table 1). Conclusion: PD-L1 IHC 22C3 pharmDx stains both TC and IC in TNBC. Removal of the PD-L1 staining TC from the CPS algorithm (QID score) reduces the number of specimens scored as positive for the PD-L1 ≥1 and PD-L1 ≥10 diagnostic cut-offs. This data highlights that a significant portion of TC express PD-L1 in TNBC and inclusion of TC in the scoring algorithm should be considered when evaluating the diagnostic status of a specimen.

Table 1. Agilent Tumor Bank and KEYNOTE -119: CPS and QID

Diagnostic Cut-OffCPS Positive SpecimensQID (CPS – TPS) Positive SpecimensSpecimens Flipped at Diagnostic Cut-offDifference (%)
Agilent Tumor Bank     
PD-L1 ≥ 1 71 65 8.5 
PD-L1 ≥ 10 49 36 13 26.5 
KEYNOTE-119     
PD-L1 ≥ 1 604 548 56 9.3 
PD-L1 ≥ 10 284 189 95 33.5 
Diagnostic Cut-OffCPS Positive SpecimensQID (CPS – TPS) Positive SpecimensSpecimens Flipped at Diagnostic Cut-offDifference (%)
Agilent Tumor Bank     
PD-L1 ≥ 1 71 65 8.5 
PD-L1 ≥ 10 49 36 13 26.5 
KEYNOTE-119     
PD-L1 ≥ 1 604 548 56 9.3 
PD-L1 ≥ 10 284 189 95 33.5 

Citation Format: Tiffany Evans, Stephanie Hund, Darlene Krohn, Kenneth Emancipator, Jonathon Juco, Bryce Portier, Siena Tabuena-Frolli, Karina Kulangara. Significance of PD-L1 expressing tumor cells in the combined positive score with triple negative breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS4-15.