Measurements of ctDNA that reflect minimal residual disease following chemoradiotherapy (CRT) for non-small cell lung cancer (NSCLC) are highly predictive of ultimate patient outcomes. But biomarkers that offer such predictions during treatment could help clinicians adapt therapy to improve outcomes for their patients. Researchers co-led by Ludwig Stanford’s Ash Alizadeh and Maximilian Diehn reported in an April publication in Cancer Discovery their development and validation of a dynamic risk model for this purpose. This model, termed Continuous Individualized Risk Index, was developed to address chemoradiotherapy (CRT)outcomes for locoregional lung tumors (CIRI-LCRT). Built on the analysis of 418 NSCLC patients undergoing such therapy, their model accurately predicted ultimate progression-free survival outcomes. The researchers showed that mid-CRT concentrations of circulating tumor (ct) DNA in patients strongly predict disease progression. They then integrated additional pre-CRT risk factors, including pre-treatment tumor histology and features from radiomic techniques—the use of image processing and statistical analysis to discern quantitative tumor features like shape and texture—with mid-CRT ctDNA measurements to develop a combined model, CIRI-LCRT, that improves outcome prediction. The researchers argue that tumor features, radiomics, and mid-CRT ctDNA analysis are complementary and, taken together, can accurately identify the risk of progression in patients. They are hopeful that this novel model will enable the fine-tuning or personalization of therapeutic strategies—for instance, changing systemic therapy—to improve treatment outcomes.
Integrating ctDNA Analysis and Radiomics for Dynamic Risk Assessment in Localized Lung Cancer
Cancer Discovery, 2025 April 29