5 Minutes
Early prediction of severe liver disease with routine blood tests
A new study from Karolinska Institutet shows that a straightforward blood-test algorithm can estimate an individual’s risk of developing severe liver disease up to ten years before clinical onset. A Swedish research team developed a statistical model that uses routine clinical information to stratify long-term risk for outcomes such as cirrhosis, liver cancer and liver transplant. A new study from Karolinska Institutet reveals that a simple blood test may accurately predict a person’s risk of developing severe liver disease years in advance.
How the CORE model works
The prediction tool, called CORE, combines five widely available variables: age, sex and three commonly measured liver enzymes—AST (aspartate aminotransferase), ALT (alanine aminotransferase) and GGT (gamma-glutamyl transferase). These biomarkers are typically reported on routine metabolic or liver function panels performed in primary care and occupational health checks.
Researchers applied advanced statistical modeling to link these variables to long-term liver outcomes. The derivation cohort included more than 480,000 adults from Stockholm examined between 1985 and 1996, followed for up to 30 years. Approximately 1.5% of participants developed severe liver disease during follow-up, allowing the team to calibrate the model against real-world clinical endpoints including cirrhosis, hepatocellular carcinoma (liver cancer) and liver transplantation.
Performance and validation
In the original report in The BMJ, CORE distinguished between people who did and did not develop severe liver disease with 88% accuracy (area under the receiver operating characteristic curve). That performance exceeded the widely used FIB-4 score, a fibrosis marker originally designed for patients with suspected chronic liver disease rather than unselected primary-care populations.
The Swedish team validated CORE in independent cohorts from Finland and the UK, where it again demonstrated strong predictive ability. The investigators caution that further evaluation is needed in subgroups at especially high risk—such as people with type 2 diabetes, obesity or known metabolic dysfunction-associated fatty liver disease (MAFLD).
Integrating screening into primary care
A primary aim of the project was clinical practicality: CORE uses inputs already available in many electronic health records and standard lab panels, which facilitates large-scale screening without new testing infrastructure. The research group has made a web-based calculator available for clinicians at www.core-model.com to estimate individual risk in consultations.
"These are diseases that are growing increasingly common and that have a poor prognosis if detected late," said Rickard Strandberg, affiliated researcher at Karolinska Institutet’s Department of Medicine in Huddinge, one of the developers of the test. “Our method can predict the risk of severe liver disease within 10 years and is based on three simple routine blood tests.”
Hannes Hagström, principal investigator and senior consultant at Karolinska University Hospital, added: "This is an important step towards being able to offer early screening for liver disease in primary care. Drug treatment is now available, soon hopefully also in Sweden, for treating people at a high risk of developing liver diseases such as cirrhosis or liver cancer."
Clinical implications and next steps
If adopted in primary-care workflows, CORE could prioritize patients for additional non-invasive testing (transient elastography, advanced fibrosis panels) or referral to hepatology, enabling earlier intervention. Important next steps include prospective implementation studies, integration into electronic medical records to automate risk flags, and targeted validation in high-risk groups.
Clinicians and health systems should weigh sensitivity, specificity and downstream costs of follow-up testing when considering roll-out. Still, CORE represents a low-cost, scalable approach to identifying individuals at risk for life-limiting liver disease years before symptoms appear.
Conclusion
The CORE model translates routine lab data and basic demographics into a practical long-term risk score for severe liver disease. With additional validation and system-level integration, it has the potential to expand early detection efforts and enable timely preventive or therapeutic interventions in primary care.
Source: scitechdaily
Leave a Comment