Poster

Machine learning based identification of predictive features of the tumor microenvironment and vasculature in NSCLC patients using the IMpower150 study

ASCO 2020

Study Background

IMpower150 is a phase 3 study measuring the effect of carboplatin and paclitaxel (CP) combined with atezolizumab (A) and or bevacizumab (B) in patients with advanced non-squamous NSCLC.

To better understand pathologic signatures of benefit from IO and VEGF inhibition, we apply a machine-learning based approach to quantify cell composition and tissue architecture of the tumor micro-environment (TME) and vasculature. After model training and evaluation we apply model predictions to identify novel TME features predictive of clinical outcome in IMpower150.
View Poster

Authors

Taylor-Weiner et al.