Untargeted serum metabolomics reveals metabolic signatures distinguishing basal cell carcinoma risk groups: an exploratory analysis

Abstract

Background and aim. Basal cell carcinoma (BCC) is the most common skin cancer, yet reliable non-invasive markers for distinguishing low-risk (LR) from high-risk (HR) lesions are still lacking. Serum metabolomics provides a promising approach for capturing systemic biochemical changes associated with tumor behavior. In this study, we conducted untargeted serum metabolomic profiling using high-performance liquid chromatography coupled with mass spectrometry in 48 patients with histologically confirmed BCC.

Methods. The cohort included 38 HR and 10 LR lesions. After quality filtering, 99 polar and 54 lipophilic metabolites were retained for analysis. Using a significance threshold of P < 0.01 and absolute fold change > 1.2, 10 metabolites differed between HR and LR BCC.

Results. Principal component analysis showed partial separation between the two groups driven by metabolites including dihydroxybutyric acid, glucosylsphingosine, androsterone, and several lysophosphatidylcholines. A linear discriminant model based on the first 4 principal components achieved an AUC of 0.88, corresponding to a sensitivity of 89% and a specificity of 40% for identifying HR lesions. Enrichment analysis revealed representation of multiple chemical classes, including carboxylic acids, steroids, glycerophospholipids, indoles, diazines, and organooxygen compounds. Several metabolites varied significantly by anatomical location and tumor size, while histologic subtype showed no meaningful influence.

Conclusions. These findings provide initial evidence that serum metabolomics can detect metabolic differences between LR and HR BCC and may serve as a basis for developing non-invasive biomarkers to improve BCC risk stratification