CRISPR screening-informed liver cancer stratification identifies ciclesonide as a novel adjuvant drug to alleviate sorafenib resistance

Sorafenib remains a standard therapy for hepatocellular carcinoma (HCC), but intratumor heterogeneity and drug resistance limit its efficacy. To identify resistance-associated genes, we performed a CRISPR activation screen and integrated results with TCGA and GTEx data. Using univariate Cox and LASSO regression, we constructed a five-gene (AKR1B15, BMF, GBA, GINS2, SMC2) sorafenib sensitivity-related prognostic signature. In the TCGA-LIHC training cohort (n = 371), patients with high-risk scores showed significantly worse overall survival and reduced sorafenib sensitivity, which was validated in the ICGC-US cohort. High risk scores also correlated with adverse clinicopathological features. Drug perturbation Gene Set Enrichment Analysis (dpGSEA) identified ciclesonide as a potential sensitizer to sorafenib. Cell viability assays and a subcutaneous xenograft model confirmed that ciclesonide enhanced the antitumor efficacy of sorafenib. RNA sequencing further suggested that this effect may be mediated through regulation of the cell cycle pathway. Collectively, the sorafenib DS-related prognostic signature may serve as an independent prognostic biomarker and predictor of sorafenib sensitivity in HCC, while the combination of sorafenib and ciclesonide represents a promising strategy to improve targeted therapy. To identify HCC patients who are benefit from sorafenib, we have constructed a sorafenib drug sensitivity – related prognostic signature integrating CRISPR activation screen with TCGA TARGET GTEx dataset. Basing on median risk score calculated by signature, patient with higher risk score indicates worse prognosis and more insensitive to sorafenib in TCGA dataset, which was also validated in ICGC-US. Consistently, we illustrate that our signature was an independent prognostic factors and patients with higher risk score showed poor clinicopathological features. Against patients insensitive to sorafenib, we identify some candidate drugs to enhance the antitumor activity of sorafenib using dpGSEA. In vivo experiments, in vitro experiments and RNA-seq analysis were utilized for validating that ciclesonide might associated with downregulation of cell cycle-related transcriptional programs, which may contribute to its enhanced antitumor activity when combined with sorafenib.