The goal of this observational study is to create a detailed virtual model to better understand how Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) develops. This model will also help predict heart problem at different stage of the disease.

1. - Glossary CT : Computed Tomography CVD : Cardio-Vascular Disease HCC : Hepato Cellular Carcinoma MASH : Metabolic dysfunction-associated steatohepatitis MASLD : Metabolic dysfunction-Associated Steatotic Liver Disease MRI : Magnetic Resonance Imaging PET : Positron Emission Tomography SLD : Steatotic Liver Disease TACE : Trans-Arterial ChemoEmbolisation TARE : Trans-Arterial RadioEmbolisation TIPS : Transjugular Intrahepatic Portosystemic Shunt US : Ultrasound USE : Ultrasound elastography VCTE : Vibration-Controlled Transient Elastography

2. - Description of the Clininical Study

   ARTEMIs retrospective cohort responds to the definition of a "retrospective collection and analysis of health data obtained from individual patients or healthy persons in order to address scientific questions related to the understanding, prevention, diagnosis, monitoring or treatment of a disease, mental illness, or physical condition" as defined in the work programme of this call. In such, the definition of a clinical study as defined by Regulation 536/2014 (on medicinal products) is not applicable in the framework of our study.

   The cohort will serve the following main objectives:

   • To develop and validate machine-learning or mechanistic models that can predict the evolution of liver diseases, in particular MASLD at various stages (use cases 1 and 2), and the outcome of some intervention or therapies (use cases 3 and 4).
   • To identify new correlations or validate suspected correlations between specific observations, interventions and outcomes, in particular cardiovascular complications.

3. - Study rationale Metabolic dysfunction-associated steatotic liver disease (MASLD) is presently the most common chronic liver disease worldwide, accounting for a global prevalence of 25.24% (2). Its natural history remains unclear, given the multiple pathways through which disease progression takes place (3), as well as to the shortage of population-based studies addressing its long-term prognosis (4). As an attempt to alleviate the paucity of good quality data on MASLD's natural history (5) and to improve patient's care, the ARTEMIS project envisages to constitute a longitudinal cohort comprising patients at various stages of liver diseases, with emphasis on MASLD (use cases 1 and 2).

   Given the remarkable heterogeneity underlying MASLD mechanisms, the deployment of computational models has increased in popularity among the scientific community, as an effective means to unravel this intricate subject (6). In particular, the understanding of the human liver metabolism plays a key role towards a deeper understanding of the main drivers that rule disease progression. In such, mechanistic models play a major role in the representation of the complexity that is inherent to the liver and the gastroenterology system. In a complementary way, machine learning models are expected to respond to more precise questions related to different stages of the disease and related comorbidities, therefore allowing the prediction of diagnosis and prognosis, as well as risk stratification, based upon parameters that are specific to each subpopulation.

   In this light, the ARTEMIS cohort will be used to test new hypotheses, as well as to train, validate and evaluate the performance of computational models - including machine-learning models, mechanistic models and associations thereof - aimed to improve the management of MASLD patients. The ARTEMIs cohort will incorporate retrospective multisource data for MASLD patients along the spectrum of the disease, thus including MASH, cirrhosis and HCC patients. The cohort will include patients from 12 centres in 7 countries. The cohort will also incorporate data related to the most relevant comorbidities associated with these populations, most notably, cardiovascular events.

4. - Extent and evaluation of current knowledge directly linked to the scientific question(s) to be answered by the clinical study

   In addition to the complexities concerning its natural history, MASLD has been associated with an increased risk of developing cardiovascular disease (CVD) and cardiac events, including coronary artery disease, atherosclerosis, heart failure, and arrhythmia. The exact mechanism by which MASLD increases the risk of CVD is not fully understood, but it is thought to be related to the systemic inflammation and metabolic dysfunction associated with the condition.

   Several studies have investigated the relationship between MASLD and cardiac events. A systematic review and meta-analysis published in 2016 (7), analysed 16 prospective and retrospective cohorts with 34,043 adult individuals (36.3% with MASLD) and approximately 2,600 CVD outcomes (>70% CVD deaths) over a median period of 6.9 years. They concluded that MASLD is associated with an increased risk of fatal and non-fatal CVD events, although the design of the observational studies did not allow to draw definitive causal inferences.

   There is a consensus that MASLD patients should be closely monitored for cardiovascular risk factors and managed accordingly to reduce their risk of developing CVD. Nevertheless, given the high current prevalence of the disease and its expected growth, such monitoring may enormously stress the public healthcare systems.

   Solutions that help to stratify those MASLD patients at higher risk of suffering cardiovascular events, are needed. The ARTEMIs cohort is aimed to assist the development of this type of solutions, based on advanced computational models.

5. - Objective(s) of the clinical study

The ARTEMIs project envisages to consolidate a holistic virtual model allowing, on the one hand, a better understanding of the underlying mechanisms involved in MASLD progression, as well as the prediction of cardiovascular events at different stages of the disease. In this light, 4 clinical cases will be considered, wherein theory-based mechanistic and data-driven AI models will be developed and validated, either individually or in association, depending on the clinical questions being raised.

The objective of ARTEMIs cohort is to assess the performance of mechanistic and AI-based models that will be deployed in the different clinical cases, based on their respective sensibility and specificity.

INCLUSION CRITERIA:

1. - Clinical Use Case 1: Liver disease staging in MASLD patients - Prediction model of fibrosis changes (progression and regression), with ability to distinguish between fast and non-fast fibrosis progression among MASLD patients.

   • Age ≥18 years
   • Diagnosis of MASLD confirmed by radiological imaging (any type: MR, CT, PET, VCTE, US, USE...) or histology (gold standard, following MASH SAF score)
   • With at least one follow-up of minimum 1 year after diagnosis of MASLD, with radiological imaging or histology

2. - Clinical Use case 2: MASLD and progression of cardiovascular diseases

   • Age ≥18 years MASLD patients regardless of disease stage of severity (from simple steatosis to cirrhosis)
   • Patients without known heart disease
   • Cardiovascular assessment available

3.1- Clinical Use case 3-TIPS: Patients with cirrhosis and portal hypertension who receive TIPS placement.

• Age ≥18 years
• TIPS indication (Baveno VII), except pre-emptive and salvage TIPS.
• Recurrent variceal bleeding after failure of the usual pharmacological and endoscopic methods
• Refractory or recurrent ascites or difficult to treat
• Refractory Hydrothorax
• Patients with diagnosis of liver cirrhosis (based on laboratory parameters, clinical, endoscopic, radiological or histological findings), of any aetiology.

3.2.- Clinical Use Case 3-LT: Patients with cirrhosis and portal hypertension who received liver transplantation.

• Age ≥18 years

• All patients with cirrhosis (all aetiologies) who were transplanted

  4.- Clinical Use Case 4: Prediction of cardiac complications due to HCC treatments* (*Note: includes surgical interventions, ablation, TACE, TARE, SIRT and immunotherapies)

• Age ≥18 years

• Diagnosis of HCC (any aetiology)

• Cross sectional imaging follow-up (any modality) of liver diseases 6 months after treatment

• Non-cirrhotic or no more than Child-Pugh B cirrhosis.

• Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1

• Patients without history of prior HCC

• Patients with a history of hypertension should be well controlled (< 140/90 mmHg) on a regimen of antihypertensive therapy.

• With a minimum follow-up of two years or until death, after diagnosis of HCC

  5.- Other populations (participation in control arms)

• Age ≥18 years

• Subjects presenting cardiac fibrosis, without a known MASLD diagnosis (as controls for use case 2)

EXCLUSION CRITERIA:

1. - Clinical Use Case 1: Liver disease staging in MASLD patients - Prediction model of fibrosis changes (progression and regression), with ability to distinguish between fast and non-fast fibrosis progression among MASLD patients.

   • Missing data on blood glucose, BMI and metabolic status.
   • Patients who have received systemic chemotherapy
   • Patients with hepatitis B (HBV) and hepatitis C (HCV), alcoholic liver disease (more than 5 years of drinking history, equivalent to alcohol volume ≥ 30g / D in male and ≥ 20g / D in female), drug-induced liver disease or autoimmune hepatitis.
   • Subjects having a significant risk of bleeding (platelet < 50x109 / L, prothrombin activity < 50%)
   • Presence of any other form of chronic liver, at the time of MASLD diagnosis.

2. - Clinical Use case 2: MASLD and progression of cardiovascular diseases

   • Association with another cause of liver disease
   • History of hepatitis B or C
   • Already known coronary artery disease
   • History of cardiovascular events

3.1- Clinical Use case 3-TIPS: Patients with cirrhosis and portal hypertension who receive TIPS placement.

• Non-cirrhosis TIPS
• Portosinusoidal vascular disease
• Complete portal vein thrombosis
• Patients with surgical porto-caval shunts.
• Patients with evidence of current locally advanced or metastatic malignancy
• Patients with acute or chronic heart failure (New York Heart Association [NYHA]).
• Patients with chronic obstructive pulmonary disease GOLD grade III/IV
• Patients with chronic kidney disease requiring renal replacement therapy
• Patients with a known infection with human immunodeficiency virus (HIV) or have clinical signs and symptoms consistent with current HIV infection
• Patients with previous liver transplantation
• Patients lost to follow-up and therefore have an incomplete 1-year follow-up

3.2.- Clinical Use Case 3-LT: Patients with cirrhosis and portal hypertension who received liver transplantation.

• Patients who were transplanted due to acute liver failure.

• Patients who were already transplanted before (retransplant)

• Patients who are lost to follow-up in the first 5 years after liver transplant.

  4.- Clinical Use Case 4: Prediction of cardiac complications due to HCC treatments* (*Note: includes surgical interventions, ablation, TACE, TARE, SIRT and immunotherapies)

• Mixed-tumor HCC based on radiological and/or pathological examination

• Uncontrolled inter-current illness or psychiatric illness or social situations that would limit compliance with study requirements.

• Subjects with history of another primary cancer

• Fully recovered from any prior surgery and/or radiation and none within 2 weeks of initiating treatment.

• Subjects with active hepatitis B or C on antiviral compounds may remain on such treatment, except for interferon.

• Subjects with diagnosis of tumor of mixed origin, either from radiological or biopsy report.

  5.- Other populations (participation in control arms)

• Patients with diagnosis of MASLD

The study population description involve a retrospective collection and analysis of health data obtained from individual patients or healthy persons. Patients along the MASLD spectrum will be recruited in 12 participant sites. A researcher will select cases fulfilling the inclusion and exclusion criteria and process them according to the Project work plan.