Identification of secretory pathway‐related genes based on Random Forest algorithm to predict the prognosis and immunotherapy response of hepatocellular carcinoma

Abstract

BackgroundThe dysfunction of secretory pathways may represent biomarkers or therapeutic targets of cancer. The hepatocellular carcinoma (HCC) phenotype was studied in relation to the genes in the secretory pathway and to screen for a combination of genes that may be a viable therapeutic target for HCC and connected to the pathophysiological features of the tumor.MethodsUsing the HCC information from The Cancer Genome Atlas, somatic mutation and prognostic association analysis were performed on the secretory pathway genes. Based on prognostic genes in the secretory pathway, the samples were consensus clustered, and a Random Forest model was built. The clinical characteristics, tumor mutation burden, functional status and potential responses to immunotherapy and tumor suppressor medications of various subtypes and risk groups were discussed.ResultsOf the 84 genes for secretory pathway, 32 were prognostic genes related to HCC, which divided HCC into two categories: C1 and C2. By comparing the two types of HCC samples, it was found that the survival outcome of C1 was inferior, with stronger adaptive and innate immunity, but less sensitive to immunotherapy than C2. The constructed prognostic signature included seven of the 32 prognostic genes in the secretory pathway, which showed significant correlation with the prognosis, somatic mutation, biological pathway status, potential response to immunotherapy and sensitivity of 72 tumor suppressor drugs from different HCC cohorts, and had a feasible prognostic effect for 31 types of cancer and immunotherapy cohorts.ConclusionsIn this study, HCC was divided into two molecular subtypes according to prognostic genes in the secretory pathway, and seven of them were combined into one signature, which produced significant results in evaluating the prognosis of different HCC cohorts, pan‐cancer cohorts and immunotherapy cohorts, and had potential guiding significance for prophylactic immunotherapy in patients with HCC.

Development and evaluation of a risk score model based on a WNT score gene-associated signature for predicting the clinical outcome and the tumour microenvironment of hepatocellular carcinoma

“…The clusters also markedly differed in terms of their immune microenvironments and chemotherapy sensitivity. 39 divided HCC into two clusters based on their secretory pathway-related genes and compared their prognoses, adaptive and innate immunity, and immunotherapy sensitivity. Li et al (2023) 40 classified patients with HCC into two clusters based on their intratumoral hypoxia gene transcriptomes and found that the they differed in terms of their tumour-infiltrating immune cells and genetic mutations.…”