Novel Tissue-Imaging Technology Used for Predicting Recurrence of Lung Cancer
Posted: Thursday, April 5, 2018
Recently developed computerized tissue-imaging technology may help to determine whether patients with non–small cell lung cancer (NSCLC) face an earlier recurrence of the disease, according to an article published in Scientific Reports. The new project, led by Anant Madabhushi, PhD, of Case Western Reserve University, and Vamsidhar Velcheti, MD, of the Cleveland Clinic, may prove to be a predictive support tool for deciding who may benefit from the use of adjuvant treatment in early-stage lung cancer and who may not. The National Cancer Institute recently awarded the researchers a $3.16 million grant to advance the project.
“If this can tell us that only 10% of patients have a chance of the cancer returning, that means the other 90% won’t have to go through chemotherapy,” said Dr. Velcheti in a Case Western press release.
The computational histomorphometric image classifier developed by the researchers uses nuclear orientation, texture, shape, and tumor architecture to predict disease recurrence in early-stage NSCLC. Scientists used a cohort of 70 patients with early-stage NSCLC to construct a supervised classification model involving the most predictive features of disease recurrence and then validated the model on 2 other cohorts of 119 and 116 people, respectively.
The model produced an 81% accuracy in predicting recurrence in the first cohort, 82% in the second cohort, and 75% in the third cohort. A multivariable Cox proportion hazard model of the second cohort, incorporating gender and traditional prognostic variables, demonstrated that the computer extracted histomorphometric score was an independent prognostic factor (hazard ratio of 20.81).
