Proteomic signature predicts the metastatic risk of primary cutaneous squamous cell carcinomas.

Abstract

Cutaneous squamous cell carcinoma (cSCC) is a heterogeneous skin malignancy worldwide. While most tumours are effectively treated by surgical excision, between 5% and 37% of biologically aggressive cases metastasise to regional lymph nodes or distant organs. Identifying tumours at risk of metastasis remains challenging, particularly for the 16-30% of cases lacking clear clinical or histopathological high-risk features. In this study, we applied a mass spectrometry-based proteomic approach to profile archival primary cSCC samples with and without confirmed metastasis. A total of 4,819 protein groups were identified, of which 284 were differentially abundant between the groups. The differential abundance of a subset of proteins was further validated in silico using independent transcriptomic datasets. These proteins were enriched in pathways associated with metastatic hallmarks, including reduced apoptosis, decreased cell adhesion and differentiation, and increased angiogenesis and keratinocyte migration. Classification analysis using support vector machine models achieved 88.66% accuracy in predicting metastatic potential. Collectively, these findings demonstrate the potential of proteomics to improve metastatic risk stratification and guide clinical management of cSCC, ultimately supporting better patient outcomes.