Answers are grounded in this paper and cite their sources.
Hepatic steatosis may coexist with primary biliary cholangitis (PBC), but PBC-related biochemical and lipid abnormalities may complicate its identification using laboratory-based indices. We aimed to develop and internally validate a PBC-specific steatosis index (PSI) using routinely collected variables. This retrospective single-center study included adults with biopsy-confirmed PBC during 2016–2025; hepatic steatosis was defined histologically as involvement of at least 5% of hepatocytes. Missing predictors were multiply imputed within training partitions. A genetic-algorithm formula-search procedure was evaluated by repeated nested five-fold cross-validation against least absolute shrinkage and selection operator (LASSO), a prespecified clinical model, and six conventional indices. Among 681 participants, 98 (14.4%) had histological steatosis. The locked two-variable PSI formula was 5 × triglycerides − 2 × diabetes status (1 = recorded diabetes; 0 = no recorded diabetes). The PSI procedure achieved an out-of-fold AUC of 0.703 (95% CI, 0.646–0.757), Brier score of 0.116, and calibration slope of 1.276. At the exploratory OOF probability cutoff of 0.131, sensitivity was 72.4%, specificity 59.3%, positive predictive value 23.1%, and negative predictive value 92.8%. PSI outperformed HSI and TyG-BMI after multiplicity adjustment but did not demonstrate superiority over TyG or LASSO. In this biopsy-confirmed PBC cohort, the PSI procedure showed moderate internally assessed discrimination for steatosis. The transparent and readily reproducible two-variable formula may help prioritize patients for further steatosis assessment when imaging-based evaluation is not readily available, but requires independent external validation before clinical implementation.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: