CLINICALLY significant liver fibrosis is a key predictor of adverse outcomes in patients with steatotic liver diseases, such as non-alcoholic fatty liver disease (NAFLD) and alcohol-related liver disease. Accurate and early detection of fibrosis in primary care is essential to ensure timely intervention and reduce long-term complications. However, current diagnostic methods, such as transient elastography, are often limited by factors like accessibility, cost, and the availability of advanced testing. To address these challenges, a software tool called LiverPRO was developed to improve the identification of clinically significant liver fibrosis in primary care.
In this prospective cohort study, LiverPRO was developed and validated using six independent cohorts from Denmark, Germany, and England, which included patients from both primary and secondary care settings with steatotic liver disease linked to alcohol or metabolic dysfunction. The primary outcomes of interest were the presence of clinically significant fibrosis (stage ≥F2) and advanced fibrosis (stage ≥F3). The model development used fractional polynomial regression applied to a set of relevant clinical variables. Diagnostic performance was assessed by predicting elevated liver stiffness via transient elastography (≥8 kPa and ≥12 kPa) and evaluating its prognostic ability over 2- and 5-year periods for liver-related events. Diagnostic accuracy was quantified using the area under the receiver operating characteristic curve (AUC), while clinical performance metrics such as sensitivity, specificity, and Harrell’s C-statistic were used to assess the tool’s prognostic reliability.
In the development cohort (n=462), LiverPRO demonstrated strong performance in detecting clinically significant fibrosis. In a larger cohort (n=6468), LiverPRO maintained good accuracy, with an AUC of 0.80 (95% CI 0.78–0.82), comparable to existing tools such as enhanced liver fibrosis testing (AUC 0.78) and the LiverRisk score (AUC 0.81), but outperforming the Fibrosis-4 index (AUC 0.69) and the NAFLD Fibrosis Score (AUC 0.74). LiverPRO showed the ability to exclude significant fibrosis, demonstrating high accuracy in reducing unnecessary further diagnostic procedures.
The tool’s performance was consistent across multiple validation cohorts, though accuracy varied slightly, especially when dealing with different patient populations and liver blood test analytes. LiverPRO also showed utility using UK Biobank data, predicting liver-related events with a C-statistic of 0.80 (95% CI 0.77–0.84) at 2 years.
While LiverPRO provides a reliable and practical solution for detecting clinically significant liver fibrosis and elevated liver stiffness in primary care settings. Its performance highlights the potential for earlier risk stratification and intervention in patients with steatotic liver disease, particularly where access to more advanced diagnostic tools may be limited.
Katie Wright, EMJ
Reference
Lindvig KP al. Development, validation, and prognostic evaluation of LiverPRO for the prediction of significant liver fibrosis in primary care: a prospective cohort study. Lancet Gastroenterol Hepatol. 2025;10(1):55-67.
