Researchers have created an artificial intelligence (AI) algorithm capable of reading CT scans to predict how patients will respond to a specific immunotherapy, according to an August Lancet Oncology study.
Lead author Roger Sun, MD, and colleagues developed a machine learning algorithm capable of teasing out biological and clinical information from CT scans to produce a radiomic signature. The measure can define the level of lymphocyte infiltration of a tumor and produce a predictive immunotherapy effectiveness score.
The radiomic signature was retrospectively validated in four independent cohorts of 500 patients with advanced solid tumors. Sun and colleagues validated their machine learning-based creation genomically, histologically and clinically, making it all the more “robust,” according to a statement.
The signature was also validated using the Cancer Genome Atlas cohort along with patients participating in five phase I trials of anti-PD-1/PD-L1 immunotherapy. Those patients who responded well to treatment at three and six months tended to have higher radiomic scores, as did patients with better overall survival, according to the statement.
“This imaging predictor provided a promising way to predict the immune phenotype of tumors and to infer clinical outcomes for patients with cancer who had been treated with anti-PD-1 and PD-L1,” the authors wrote.
Future studies, according to the statement, will test the signature both retrospectively and prospectively in larger groups of participants separated by specific cancer types.
“Our imaging biomarker could be useful in estimating CD8 cell count and predicting clinical outcomes of patients treated with immunotherapy, when validated by further prospective randomized trials,” Sun and colleagues concluded.
Matt joined Chicago’s TriMed team in 2018 covering all areas of health imaging after two years reporting on the hospital field. He holds a bachelor’s in English from UIC, and enjoys a good cup of coffee and an interesting documentary.