P3.02c-067 Validation of PD-L1 Expression on Circulating Tumor Cells in Lung Cancer: Topic: IT Biomarkers

Abstract

The human immune system recognizes and eliminates certain types of tumor cells, whereas other malignancies are capable of suppressing immune function. For example, a number of cancers cell types express programmed cell death ligand 1 (PD-L1), which binds to its receptor PD-1 on T cells to prevent their activation. High levels of PD-L1 expression are typically associated with poor patient prognosis. Based on these results, researchers have developed immunotherapies (e.g., inhibitors of the PD-1/PD-L1 pathway) to stimulate the immune system, allowing the body's natural defenses to combat the tumor. To determine which patients are suitable candidates for receiving immunotherapy, levels of PD-L1 expression are often determined from tumor biopsies, but tumor heterogeneity can confound these results and obtaining tumor tissue is often not feasible. To enable non-invasive detection and sequential monitoring of tumor-associated PD-L1 expression we have developed a highly sensitive method of detecting PD-L1 levels in circulating tumor cells (CTCs). Here we sought to analytically validate the PD-L1 assay by introducing PD-L1-positive (H358) and PD-L1-negative (BT474) cells into control blood samples, and measuring detection accuracy. PD-L1 expression levels on carcinoma cell lines were identified by flow cytometry. For analytical validation, H727, BT474 H358, HCC78 and H820 cells were spiked into CEE-SureTM blood collection tubes, in replicates and on different days, incubated overnight and thereafter processed. The leukocyte fraction was incubated with our pan-CTC antibody capture cocktail, labeled with biotinylated secondary antibody, followed by enrichment in our streptavidin coated microfluidic channels. Enriched cells were stained for DAPI, cytokeratin, CD45, PD-L1 (clone 28-8) and CEE-Enhanced (pan-CTC stain). After automated fluorescence scanning, 400 spiked tumor cells per microfluidic channel were identified and average PD-L1 intensities were quantified for each cell and cut-off criteria were determined. In our microfluidic PD-L1 assay we demonstrate H727 and BT474 cells to be negative for PD-L1, while H358 cells have low-medium and HCC78 and H820 cells high PD-L1 expression. We determined a cut-off value (average fluorescence intensity value) that yielded 100% concordance between the result of the PD-L1 test and the identity of the introduced cell lines, based on a 95% confidence level and a 3.9% negative cut-off. The Biocept PD-L1 assay can accurately detect added CTCs that express PD-L1 in blood samples. This ability affords a way to identify patients likely to benefit from immune therapy as well as monitor the efficacy of such treatments.