FIbrosis aNd Metabolic Associated Steatotic Liver Disease - HCC risk prediction tool (FIND-HCC)

Research question Can a tool composed of clinical parameters, including liver stiffness measurement (LSM) accurately predict five-year risk of developing hepatocellular carcinoma (HCC) at an individual level in people with metabolic associated steatotic liver disease (MASLD).

Background Current HCC surveillance guidelines may underserve individuals with MASLD as nearly 40% of MASLD-HCC cases occur in patients without cirrhosis. Early cancer detection is vital for curative treatment. MASLD affects 1-in-4 adults therefore targeted selection of individuals at the greatest risk of HCC is required.

LSM obtained at fibroscan is strongly associated with HCC incidence in MASLD.

Aims and objectives Aim: To establish a large dataset of patients with MASLD fibrosis who have undergone fibroscan and use this to develop an HCC prediction tool in surveillance-eligible patients with MASLD. Objectives: 1. Describe the cumulative incidence of HCC for people with MASLD according to LSM using a competing risks model 2.

Describe the natural history of MASLD-HCC using multistate modelling 3. Develop an HCC risk prediction tool for individuals with MASLD 4. Perform an external validation scoping exercise 5.

Develop a health economic analysis protocol (HEAP) for a future cost-effectiveness analysis Methods We will establish a unique dataset of n=5000 people with MASLD fibrosis (LSM≥8kPa) according to fibroscan across seven UK Hospital Trusts. Individuals will be identified from existing datasets stored on fibroscan machines.

We will follow the patient s journey from the point of initial detection of liver fibrosis until their first clinical event (development of HCC, liver decompensation, death), or the end of the study.

Research nurses will search electronic health records to collect covariate data (demographics, clinical data, biochemistry tests) which will be combined with LSM to develop an HCC risk prediction tool. We will aim to identify HCC before liver decompensation as this is the time when HCC treatment will modify survival.

Markov modelling will feed into a HEAP, which will identify the necessary data sources and analytic methods to inform a future cost-effectiveness analysis.

Timelines for delivery Months 1-7: Ethical approval and sponsorship Months 4-15: Development of the database Months 13-24: Data cleaning/analysis, create risk prediction tool Months 24-26: External scoping review Months 21-26: Develop a HEAP Months 27-30: Result dissemination Anticipated impact and dissemination Short term impact: Change in attitudes towards the benefits of HCC surveillance in MASLD supported by an effective predictive tool; increased awareness of MASLD and its complications among the diabetes community.

Longer term impact: A future clinical trial demonstrating superior detection of early HCC in people with MASLD selected for surveillance according to the FIND-HCC risk score compared to standard care has the potential to changes current HCC surveillance practices for patients with MASLD.

Dissemination: (i) Presentation at liver and allied medical conferences, (ii) publication in academic and general readership journals, (iii) liver health workshop (iv) webinar (v) dissemination of results across patient support groups and patient charities, (vi) consultation with the UK National Screening Committee and NHS England, (vii) social media, (viii) manuscript on the impact of patient and public involvement in this study.