AI, NSCLC & Immunotherapy27 Dec 2024 15:47
This will interest some: 'Deep Learning Model for Predicting Immunotherapy Response in Advanced Non−Small Cell Lung Cancer'
https://jamanetwork.com/journals/jamaoncology/fullarticle/2828578
If you can better identify who will respond to immunotherapy, then the costs involved become more justifiable as outcomes improve; got to be good news for companies like SCLP.
Key Points
Question Can deep learning−based algorithms use histologic images to directly predict response to immune checkpoint inhibitors (ICI) in patients with advanced non−small cell lung cancer (NSCLC)?
Findings This cohort study developed and externally validated a response prediction computational pipeline including 958 patients with NSCLC treated with ICI monotherapy, and demonstrated that deep learning prediction scores were associated with response rate, progression-free survival, and overall survival, with performance comparable to programmed death-ligand 1 (PD-L1). Combined deep learning and PD-L1 scores improved patient stratification.
Meaning These findings indicate that an artificial intelligence pathology model could potentially serve as a new tool for guiding ICI treatment, refining patient selection, and improving clinical outcomes in the treatment of advanced NSCLC.