Distinct tumor-immune ecologies in NSCLC patients predict progression and define a clinical biomarker of therapy response

Abstract

We investigated multiplexed histological images from nine patients (both pre- and on-treatment) with immunotherapy-refractory Non-Small Cell Lung Cancer (NSCLC) treated with an oral HDAC inhibitor (vorinostat) combined with a PD-1 inhibitor (pembrolizumab). Patient responses comprised of either stable disease (SD) or progressive disease (PD). We built an extensive multiplexed-image analysis pipeline involving both cell segmentation and quadrats, coupled with spatial statistics, machine learning, and deep learning to analyze the spatial and temporal features that predict disease progression and identify potential clinical biomarkers. We found that distinct spatial immune ecologies exist between SD and PD patients. We also demonstrate that tumors from PD patients are already characterized by an immune-suppressive environment prior to treatment. Finally, we show that the learned spatial ecologies can predict disease progression better than PD-L1 status alone, suggesting these ecologies can be used as potential companion biomarkers with PD-L1 in NSCLC. These findings will be investigated in a larger-cohort study generated from an ongoing clinical trial (NCT02638090) that includes a wider range of responses including complete and partial responders. Additionally, the computational infrastructure developed in this study can be generalized to any cancer type.