STARD3NL as a prognostic marker related to an immunosuppressive tumor microenvironment in hepatocellular carcinoma

Abstract

Background Changes in the tumor microenvironment (TME) are closely involved in hepatocellular carcinoma (HCC) development, and lipid metabolic disorders may take part in this process. STARD3NL is known to be related to cholesterol trafficking and lipid homeostasis. However, little is known about whether STARD3NL is linked to immune features of the HCC microenvironment. Methods We first explored STARD3NL expression and its prognostic relevance in HCC using The Cancer Genome Atlas (TCGA). To verify the bioinformatic findings, HCC tissue microarrays (TMAs) were analyzed by immunohistochemistry (IHC) for STARD3NL localization and survival evaluation. Multiplex immunohistochemistry (mIHC), combined with digital image analysis, was further used to characterize immune cell infiltration. Clinicopathological differences between tumors with high and low STARD3NL expression were also examined. In addition, spatial distance analysis was performed to assess how immune cells were arranged in STARD3NL-high regions. Results TCGA-based analysis and mIHC validation both showed higher STARD3NL expression in HCC tissues. Patients with increased STARD3NL expression had poorer survival (p = 0.029), and multivariable analysis supported STARD3NL as an independent risk factor. STARD3NL-high tumors showed increased proportions of CD68+ macrophages and FOXP3+ T cells (p < 0.005). Spatial analysis further indicated that Tregs were more frequently located near macrophages (p = 0.001), with a shorter mean macrophage-Treg distance (p = 0.027). This macrophage-Treg proximity was associated with worse prognosis (p = 0.030). Conclusion High STARD3NL expression is significantly associated with immune-suppressive features and unfavorable outcomes in HCC. These findings indicate that STARD3NL may have value as a prognostic biomarker associated with an immunosuppressive tumor microenvironment.