Integrated Bulk and Single-Cell Transcriptomics Reveal Hepatocyte-Like Reprogramming in Metastatic Small Intestinal Neuroendocrine Tumors

Liver metastasis is the primary determinant of poor prognosis in small intestinal neuroendocrine tumors (siNETs), yet the molecular mechanisms enabling metastatic adaptation remain largely unknown. Here, we integrate bulk and single-cell transcriptomic analyses to dissect tumor-intrinsic programs underlying this process. Bulk RNA-seq analysis of 44 primary and 37 metastatic siNET patient samples revealed extensive transcriptional remodeling during liver metastasis. Differential expression and pathway enrichment analyses demonstrated upregulation of metabolic and various liver-related pathways, alongside downregulation of intestine-specific programs in metastatic tumors. These results suggest a tumor-intrinsic phenotypic shift toward hepatic functional states. Further, application of the ESTIMATE algorithm, which infers tumor purity from the relative abundance of stromal and immune transcriptional components, demonstrated consistently high purity in metastatic samples, indicating that the observed transcriptional programs reflect tumor-intrinsic alterations rather than contamination from normal liver tissue. To validate these observations at single-cell resolution, we analyzed five matched primary and liver metastatic siNET patient samples using in-house single-cell RNA-seq data. Integrative clustering revealed distinct neuroendocrine (NE) and hepatocyte-like tumor populations, reflecting the cellular heterogeneity within the metastatic niche. Pseudotime and trajectory analyses highlighted a continuous progression from primary tumor-enriched NE clusters toward metastasis-enriched hepatocyte-like clusters, suggesting a gradual reprogramming of tumor cells along a liver-adaptive axis. These hepatocyte-like states exhibited activation of biological processes like hepatic functions and xenobiotic metabolisms. Together, these indicate a selective transcriptional plasticity that enables metastatic tumor cells to acquire liver-specific functional traits, thereby facilitating their survival, metabolic integration, and persistence within the liver microenvironment. In conclusion, by integrating bulk and single-cell transcriptomic data, this study reveals that siNET liver metastasis arises through a tumor cell-state transition toward a hepatocyte-like identity, representing a novel paradigm of liver- specific adaptation in neuroendocrine tumor metastasis and highlights liver-adaptive transcriptional programs that may be therapeutically targeted to disrupt siNET metastatic persistence.