Radiomics signature based on CECT for non- invasive prediction of response to anti-PD-1 therapy in patients with hepatocellular carcinoma

PURPOSE: This study aimed to develop a radiomics signature (RS) based on contrast -enhanced computed tomography (CECT) and evaluate its potential predictive value in hepa-tocellular carcinoma (HCC) patients receiving anti-PD-1 therapy. METHOD: CECT scans of 76 HCC patients who received anti-PD-1 therapy were obtained in this study (training group = 53 and validation group = 23). The least absolute shrinkage and selection operator (LASSO) regression was applied to select radiomics features of primary and metastatic lesions and establish a RS to predict lesion-level response. Then, a nomogram com-bined the mean RS (MRS) and clinical variables with patient-level response as the end point. RESULTS: In the lesion-level analysis, the area under the curves (AUCs) of RS in the training and validation groups were 0.751 (95% CI, 0.668-0.835) and 0.734 (95% CI, 0.604-0.864), respectively. In the patient-level analysis, the AUCs of the nomogram in the training and validation groups were 0.897 (95% CI, 0.798-0.996) and 0.889 (95% CI, 0.748-1.00 0), respectively. The nomogram stratified patients into low-and high-risk groups, which showed a significant difference in progression-free survival (PFS) (p<0.05). CONCLUSIONS: The RS is a noninvasive biomarker for predicting anti-PD-1 therapy response in patients with HCC. The nomogram may be of clinical use for identifying high-risk patients and formulating individualised treatments. (c) 2022 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.