Computerized tomography (CT)-based hepatic radiomics phenotypes associate with response to anti-angiogenics in neuroendocrine tumors (NETs): A subanalysis from the phase II TALENT clinical trial

#3528

Introduction: The phase II TALENT clinical trial provided evidence of the efficacy of Lenvatinib in patients with panNETs and giNETs. However, the role of radiomic predictive biomarkers of response to antiangiogenic drugs remain unknown.

Aim(s): Identify and validate a CT-based radiomic phenotype associated with clinical benefit to antiangiogenic drugs in patients with NETs.

Materials and methods: Radiomic features were extracted from liver metastases at pre-treatment CT scan from 57 patients included in the TALENT clinical trial. The most relevant radiomic features were selected using minimum Redundancy Maximum Relevance (mRMR) and combined in a logistic regression equation for predicting clinical benefit (Progression free survival > 15.7 months). Models were validated internally and tested in an external cohort of 26 patients with NETs treated with Sunitinib. A regression model was used to combine radiomic features with clinical variables (Ki67, tumor origin and liver tumor burden).

Conference:

Presenting Author: Garcia Alvarez A

Authors: Garcia Alvarez A, Hernando J, Ligero M, Delgado E, Ibrahim T,

Keywords: Lenvatinib, Radiomics,