The spatial landscape of the peritumoral interface: ARG1+ PMN-MDSCs reduction and NETosis as indicators of occult metastasis in tongue squamous cell carcinoma.

BACKGROUND: Predicting occult lymph node metastasis in early-stage tongue squamous cell carcinoma remains a clinical challenge because traditional markers, such as depth of invasion, exhibit limited accuracy. This study aimed to identify spatial immunological biomarkers to improve risk stratification and clinical decision-making. METHODS: Olink proteomic profiling was conducted to screen for metastasis-associated biomarkers. Subsequently, multiplex immunofluorescence coupled with artificial intelligence-powered spatial analysis was utilized to evaluate the microenvironment in primary lesions and paired cervical lymph nodes. A novel "Spatial Availability Index" was applied to map cellular distribution, and diagnostic performance was evaluated using receiver operating characteristic curve analysis. RESULTS: Arginase-1 (ARG1) was identified as a top candidate biomarker, exhibiting a unique "tissue-low, serum-high" expression pattern in cases with occult metastasis. Spatial analysis revealed a significant depletion of ARG1+ polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) in both primary tumors and paired lymph nodes during metastatic progression. Specifically, the reduction of these intact cells was most pronounced within a 35-μm proximal segment from the tumor boundary. This localized cellular loss inversely correlated with an accumulation of neutrophil extracellular traps (NETs) at the peritumoral interface, indicating that ARG1+ PMN-MDSC-associated spatial alterations coincide with NET-associated remodeling at the peritumoral interface. Integrating the intrastromal density of ARG1+ PMN-MDSCs with depth of invasion significantly enhanced the diagnostic accuracy for occult metastasis (Area Under the Curve = 0.8318) compared to using depth of invasion alone. CONCLUSIONS: Early metastatic risk is driven by localized alterations within the myeloid compartment, characterized by the transition of intact ARG1+ PMN-MDSCs to NETs within a 35-μm peritumoral hotspot. This spatial biomarker provides a high-resolution tool to optimize patient selection for elective neck dissection.