Robust, independent and relevant prognostic 18F-fluorodeoxyglucose positron emission tomography radiomics features in non-small cell lung cancer: Are there any?

• DOI: 10.1371/journal.pone.0228793
• Publication Date: 2020-02-25T18:52:12Z
• Authors: Tom Konert; Sarah Everitt; Matthew D. La Fon...; Jeroen B. van de ...; Michael P. MacManus; Wouter V. Vogel; Jason Callahan; Jan-Jakob Sonke
• Keywords: Standardized uptake value; Fluorodeoxyglucose

In locally advanced lung cancer, established baseline clinical variables show limited prognostic accuracy and 18F-fluorodeoxyglucose positron emission tomography (FDG PET) radiomics features may increase for optimal treatment selection. Their robustness added value relative to current factors are unknown. Hence, we identify robust independent PET that have complementary in predicting survival endpoints. A 4D dataset (n = 70) was used assessing the repeatability (Bland-Altman analysis) independence of (Spearman rank: |ρ|<0.5). Two 3D datasets combined 252) were training validation an elastic net regularized generalized logistic regression model (GLM) based on a selection (GLMall). The fitted performance externally validated 40). GLMall (measured with area under receiver operating characteristic curve, AUC) highest 2-year overall (0.66±0.07). No significant improvement observed compared containing only or any endpoint. External led AUC values no higher than 0.55 this study, FDG did not endpoints cancer patients. Improving risk stratification decision making has still been proven difficult