Externally validated non-invasive diagnosis of obese MASH: CTL/S-serum biomarker integrated nomogram

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Context To develop and validate a non-invasive nomogram for predicting metabolic dysfunction-associated steatohepatitis (MASH) in obese patients, reducing reliance on liver biopsy. Methods This retrospective multicenter study included 387 severely obese patients undergoing laparoscopic sleeve gastrectomy with intraoperative liver biopsy at Tianjin Medical University General Hospital (training cohort: n = 272; internal validation: n = 115). Baseline demographics, laboratory tests including alanine aminotransferase (ALT) and high-density lipoprotein cholesterol (HDL-C), anthropometric measurements, and non-contrast CT liver-to-spleen attenuation ratio (CT L/S ) were collected. After univariate analysis and multicollinearity assessment, multivariable logistic regression identified independent predictors. A nomogram was constructed and validated internally (bootstrap, 1000 resamples) and externally (n = 141, China-Japan Friendship Hospital). Performance was evaluated by Area Under the Curve (AUC), calibration, and decision curve analysis (DCA). Results Among 387 severely obese patients, 223 were diagnosed with MASH (overall prevalence 57.62%). In the external validation cohort ( n = 141), the MASH prevalence was 60.28% (85/141). ALT (OR: 1.207, p = 0.007), HDL-C (OR: 0.705, p = 0.002) and CT L/S (OR: 0.139, p = 0.001) were independent predictors. The nomogram demonstrated good discrimination: training cohort AUC = 0.768 (95% CI: 0.712–0.824), internal validation AUC = 0.789 (95% CI: 0.705–0.872), and external validation AUC = 0.816 (95% CI: 0.747–0.885). Calibration was good (Hosmer–Lemeshow p > 0.05), and DCA showed clinical utility. Conclusion A nomogram integrating ALT, HDL-C, and CT L/S provides a reliable, non-invasive tool for predicting MASH risk in obese metabolic surgery candidates, validated internally and externally. This tool enables preoperative risk stratification, facilitates shared decision-making between clinicians and patients, and has the potential to reduce unnecessary liver biopsies in low-risk individuals. However, prospective validation in broader clinical settings is warranted before routine implementation.