A groundbreaking study uses machine learning to analyze tumor-infiltrating lymphocyte levels in non-small cell lung cancer patients, providing a robust and cost-effective method to predict the success of immune checkpoint inhibitor treatments.
In a game-changing study reported in JAMA Oncology, researchers led by Dr. David J. Kwiatkowski discovered that machine learning can provide essential insights into the success of immune checkpoint inhibitor therapy for non-small cell lung cancer (NSCLC) patients. By analyzing tumor-infiltrating lymphocyte (TIL) levels from standard histologic images, the team was able to predict patient response to treatment with impressive accuracy.
The multicenter study included 685 patients treated between February 2014 and September 2021. The team developed a machine-learning automated method to count tumor, stroma, and TIL cells in whole-slide hematoxylin-eosin-stained images of NSCLC tumors. In multivariate analysis, high TIL levels were independently associated with improved progression-free survival and overall survival. The study also found that TIL levels had greater classification accuracy for immune checkpoint inhibitor response vs nonresponse compared to tumor mutational burden alone.
The researchers concluded that TIL assessment, easily incorporated into pathology laboratory workflows at minimal additional cost, may enhance precision therapy. This groundbreaking study marks a significant leap forward in utilizing machine learning to predict the success of cancer treatments, ultimately leading to improved patient outcomes and potentially saving lives.
