Selection and prediction of targeted serum protein biomarkers for early diagnosis of esophageal squamous cell carcinoma

Abstract

Achieving early diagnosis of esophageal squamous cell carcinoma (ESCC) is crucial for reducing the disease burden. This study aims to identify potential serum protein biomarkers associated with the early diagnosis of ESCC in two independent populations and further develop diagnostic models. This case-control study identified and verified ESCC-shared biomarkers in two independent populations from Taixing and Jinan (China). Utilize differential analysis and functional enrichment analysis, combined with bulk RNA-seq and single-cell RNA-seq data, were used to validate the diagnostic proteins at the tissue and cellular levels. Multiple machine learning algorithms were employed to screen core diagnostic proteins and develop models for ESCC diagnosis. Indicators such as area under the curve (AUC) were used to evaluate the models’ predictive capabilities for ESCC diagnosis. Lastly, survival data were employed to further expand the ability of genes encoding core diagnostic proteins to predict the prognosis of ESCC patients. Differential analysis validated across two independent populations revealed 13 proteins with diagnostic value for ESCC, with most proteins showing consistent expression patterns in mRNA and single-cell RNA compared to serum samples. The diagnostic model, developed using the gradient-boosted decision tree algorithm based on the variables retained after Lasso regression (CDKN1A, ADAMTS15, KLK13, RSPO3, ANXA1, MetAP2), demonstrated best discriminative effects, calibration, and clinical utility in both the derivation dataset [AUC = 0.932, 95% confidence interval (CI): 0.894–0.970] and the validation dataset (AUC = 0.968, 95% CI: 0.944–0.992). Risk scores constructed from the transcriptome gene set of core diagnostic proteins also demonstrate certain predictive capability for risk stratification. The diagnostic proteins identified demonstrate promising performance in both the diagnosis and prognostic assessment of ESCC. This achievement refined and enhanced the diagnostic capabilities for ESCC at its earliest stages, potentially opening new avenues for timely intervention to improve patient outcomes.