Abstract 5666: Octaplex spectral imaging of non-small cell lung cancer (NSCLC): In-depth insights into lung cancer immune micro-environment

Abstract

Abstract Introduction: Lung cancer continues to be the lead cause of death among all cancer types. Checkpoint Inhibitors (CPI) showed great promise in curing or prolonging survival in various cancer types including advanced NSCLC, and others (HNSCC, RCC, Melanoma). The key limiting steps to such success is to find the right target population, minimize toxicities and better understand resistance mechanisms. Despite recent discoveries (1), there continues to be a huge need for development of better predictive biomarkers. Combining immunology and pathology knowledge will help guide the field of immune-oncology. Our group is interested in more in-depth anatomic dissection of immune TME (iTME). Methods: To better understand the iTME in NSCLC and get deeper insight into the interaction of immune infiltrate with tumor cells, we successfully performed 8-color multispectral imaging (MSI) on 7 available cases of early stage NSCLC at Georgia Cancer Center. All 7 cases were successfully stained for CD4, CD8, FOXP3, PDL1, Ki67, CD68, CK, and DAPI. The regions of interest were identified by a research pathologist and images were captured using the Vectra 3 system (Perkinelmer). Results: Octaplex staining was successful in this pilot study. All cases had significant immune infiltrate consistent of T and B lymphocytes as well as Tregs and CD68+ macrophages. The distribution of most infiltrates took a network like picture highlighting the piecemeal approach of immune infiltrate to attack cancer cells. In many instances, the infiltrate traces the tumoral capillary network. None of the cases followed the patterns described previously (2), but ather a mixture of these morphologies. Areas highly expressing PD-L1 tended to be devoid of Tcells but highly expressed on the edges of highly infiltrated areas. Tumor cells cruise through iTME by forming small agglomerates of cells or elongated shaped individual cells. Surprisingly CD4:CD8 ratio was inversely related to tumor differentiation with low ratio in poorly differentiated, solid patterns. In addition, immune infiltrates seem to be locked within areas where PD-L1 is highly expressed by tumor or its vascular network. Conclusion: This pilot study provides insights from guided pathology. Under sampling is a major problem in studies and might lead to false conclusions. New insights into immune resistance through “Egress Lock” is a potential resistance mechanism to immune therapies. Type of infiltrate and its behavior still need further in-depth examination.