Transcriptomic Characterization of North Queensland Hepatocellular Carcinoma.

INTRODUCTION: Hepatocellular carcinoma (HCC) is a growing burden, particularly in rural, regional, and remote areas, but samples from these communities are underrepresented in public cancer data repositories. It remains unclear whether the findings of large, commonly studied cohorts such as The Cancer Genome Atlas (TCGA) are applicable to these remote communities. METHODS: We profiled paired tumour and adjacent non-tumour liver biopsies from 19 patients admitted to the Townsville University Hospital in rural Australia. We used RNA-seq to characterize transcriptomic and mutational features and compared these with the TCGA Liver Hepatocellular Carcinoma (LIHC) cohort. Furthermore, we used these data to test a transcriptome-only adaptation of our TARGET-SL pipeline for low-cost drug target prediction. RESULTS: Differential expression analysis identified 923 genes altered in our cohort, of which 64% overlapped with TCGA-LIHC, and the cohort-mean gene expression correlated strongly (Spearman rho = 0.96). Somatic variant calling from RNA highlighted mutational heterogeneity, with CTNNB1 (47%) and TP53 (21%) the most frequently mutated genes, consistent with TCGA findings. Copy number inference detected recurrent deletions on 8p, 6q, and 17p, congruous with known HCC patterns. We ran TARGET-SL solely on RNA-seq to identify personalized driver genes in these patients and were able to identify a drug candidate in 63% of patients. CONCLUSION: Our results demonstrate that NQ HCC shares core molecular features with larger TCGA cohorts and that a transcriptome-based approach can feasibly support precision oncology in resource-limited regional settings.