Fractal-Based Radiomic Approach to Tailor the Chemotherapy Treatment in Rectal Cancer: A Generating Hypothesis Study

• DOI: 10.3389/fonc.2021.774413
• Publication Date: 2021-12-09T06:05:07Z
• Authors: Carmela Di Dio; Giuditta Chiloiro; Davide Cusumano; Francesco Catucci; Luca Boldrini; Angela Romano; Elisa Meldolesi; Fabio Marazzi; Barbara Corvari; Brunella Barbaro; Riccardo Manfredi; Vincenzo Valentini; Maria Antonietta ...
• Keywords: 03 medical and health sciences; 0302 clinical medicine; Oncology; radiomics; oxaliplatin; Neoplasms. Tumors. Oncology. Including cancer and carcinogens; rectal cancer; predictive modeling; RC254-282; MRI; 3. Good health

Introduction The aim of this study was to create a radiomic model able calculate the probability 5-year disease-free survival (5yDFS) when oxaliplatin (OXA) is or not administered in patients with locally advanced rectal cancer (LARC) and treated neoadjuvant chemoradiotherapy (nCRT), allowing physicians choose best chemotherapy (CT) regimen. Methods LARC cT3–4 cN0 cT1–4 cN1–2 were according an nCRT protocol that included concomitant CT schedules without OXA radiotherapy dose 55 Gy 25 fractions. Radiomic analysis performed on T2-weighted (T2-w) MR images acquired during initial tumor staging. Statistical separately for cohort OXA. ability every single feature predicting 5yDFS as univariate assessed using Wilcoxon–Mann–Whitney (WMW) test t-test. Two logistic models (one each cohort) calculated, their performance area under receiver operating characteristic (ROC) curve (AUC). Results A total 176 image features belonging four families (morphological, statistical, textural, fractal) calculated patient. At analysis, only showing significance maximum fractal dimension subpopulation identified considering 30% 50% threshold levels (maxFD 30–50 ). Once developed feature, AUC 0.67 (0.57–0.77) 0.75 (0.56–0.95) obtained OXA, respectively. maxFD >1.6 correlated higher Conclusion This suggests T2-w can be used define optimal regimen stage III patients. In particular, by providing indication gross volume (GTV) spatial heterogeneity at staging, seems predict 5yDFS. New studies including larger external validation sets are recommended verify results hypothesis-generating study.