Title: MA 06.03 Programmed Death-Ligand 1 (PD-L1) Expression in Clinical Practice: Comparison of Temporally or Spatially Separated Test Results
Abstract: Advances in understanding of immune checkpoint inhibitors, have resulted in FDA approvals of anti-PD-1/PD-L1 inhibitor therapies for clinical use in nonsmall cell lung cancer (NSCLC). Detecting PD-L1 expression, as a predictive biomarker using companion diagnostic test (PD-L1 IHC 22C3 pharmDx), helps us identify NSCLC patients eligible for anti-PD-1 therapy (Pembrolizumab). Tumor Proportion Score (TPS) ≥50% and TPS ≥1% qualitatively estimated, by PD-L1 IHC 22C3 pharmDx test, are cut-offs which indicate use of Pembrolizumab as monotherapy in first line (TPS ≥50%) or second line (TPS ≥1%) settings for NSCLC. Intratumoral heterogeneity of PD-L1 expression in NSCLC is known. Approximately 60% of NSCLC present with advanced stage of disease. Tissue sampling of metastatic sites for initial diagnosis using core needle biopsy or fine needle aspiration techniques is common clinical practice. Significant body of literature is not available to address the issue of PD-L1 expression at metastatic sites and its concordance/discordance with primary lung tumor. We decided to look at cases with repeat request for PD-L1 testing at alternate sites or on subsequent tumor resections. Our departmental anatomic pathology database was queried to search for NSCLC cases wherein PD-L1 immunohistochemistry was performed in our laboratory using companion diagnostic test (PD-L1 IHC 22C3 pharmDx) on AutoLink 48 autostainer as per protocol, and reported by one of our pathologists. Analysis was performed to determine additional PD-L1 IHC test requests for same patient and subsequent subgroup analysis to determine test results and other parameters such as type of specimens, tumor sites, and concordant/discordant results. PD-L1 IHC 22C3 pharmDx test request was received on 460 NSCLC patient specimens in last six months. Of these, in twenty-five patients testing was attempted/performed on two tissue specimens, with final results reported in eighteen patients. Discordant results are noted in four patients (22.22%). In an additional patient, reported level of PD-L1 expression (low) was concordant; however reported TPS (5% & 45%) was different. Currently, in routine clinical practice, PD-L1 IHC test results are usually reported on a single tissue specimen. However, when tested on separate site/s or specimen type/s, our results suggest, that one can observe discordant results. At the lower end of results (PD-L1 negative or low expression), this can impact therapeutic decisions. Though a larger study is necessary to address this issue, one can suggest, that PD-L1 IHC testing should be performed on multiple site specimens, especially when temporally separated, in best interests of patient care.