Published online Apr 14, 2022. doi: 10.3748/wjg.v28.i14.1479
Peer-review started: November 9, 2021
First decision: January 9, 2022
Revised: January 22, 2022
Accepted: March 6, 2022
Article in press: March 6, 2022
Published online: April 14, 2022
Sorafenib is regarded as a first-line systematic treatment option for patients with advanced hepatocellular carcinoma (HCC), but its efficacy is largely influenced by raising resistance. The phosphorylation status of β-arrestin1 influences its function as a signal strongly related to sorafenib resistance.
Identifying patients who are more likely to benefit from sorafenib treatment and discovering related biomarkers associated with sorafenib treatment response can guide personal management.
The purpose of this study was to develop and validate radiomics-based models for predicting β-arrestin1 phosphorylation in HCC with contrast-enhanced computed tomography (CT).
We included ninety-nine HCC patients (training cohort: n = 69; validation cohort: n = 30) who received systemic sorafenib treatment after surgery. Radiomics features were generated and selected to build a radiomics score and then combined with clinical and imaging features to establish clinico-radiological (CR) and clinico-radiological-radiomics (CRR) models. The performance and clinical usefulness of the models were measured by receiver operating characteristic and decision curves. Their association with prognosis was also evaluated using the Kaplan-Meier method.
Our study found that the ALT level, tumor size and tumor margin were significant independent factors for predicting β-arrestin1 phosphorylation. The CRR model showed better discriminative performance than the radiomic score or the CR model. The β-arrestin1 phosphorylation status predicted by the CRR model was shown to be significantly associated with overall survival.
The radiomics signature is a reliable tool for evaluating β-arrestin1 phosphorylation, and may help to better identify patients who would benefit from sorafenib treatment.
The results of this study suggests that CT-based radiomics may provide promising and noninvasive biomarkers for the evaluation of β-arrestin1 phosphorylation and may help to identify the subset of HCC patients who are more sensitive to sorafenib treatment, thus potentially guiding personalized treatment strategies.