Fatty liver disease is a public problem affecting many people around the World. For some people a fatty liver causes no problem. But for others it develops into non-alcoholic steatohepatitis, called NASH. NASH is a serious type of fatty liver disease. If NASH gets worse, it can lead to liver complications, such as liver failure and liver cancer.
There is no medicine to treat NASH, but a medicine called GLP-1 used for type 2 diabetes might work. Type 2 diabetes is a common disease where the blood sugar is too high.
With this study we will explore if type 2 diabetes treatment with GLP-1 leads to fewer patients having liver complications. To do this we will use data from nearly 1,500,000 patients with type 2 diabetes examined by their general practitioner in the United Kingdom (CPRD database). We will follow these patients from the first time they get treatment with GLP-1 and calculate how many develop liver failure or die. We will get the information on liver disease and death from hospital and death records.
We hope this study could help us understand if patients treated with GLP-1 are less likely to develop liver complications. If so, this large group of patients with NASH, who right now do not have any treatment options, will be able to get help and thereby avoid liver complications.
To date, there is no approved pharmacotherapy for either Non-Alcoholic Fatty Liver Disease (NAFLD) or Non-Alcoholic Steatohepatitis (NASH). Previous research has established an association between type 2 diabetes (T2D) and NAFLD. The widely used glucose-lowering drugs glucagon-like peptide-1 (GLP-1s) agonists have shown promising results by reducing the progression of NASH in patients with T2D. However, there is limited real-world evidence regarding long-term effect of GLP-1 agonists on liver-related outcomes. We propose to study the real-world comparative effectiveness of GLP-1 agonists vs. dipeptidyl peptidase 4 (DPP4) inhibitors on clinical liver events.
Specific aim is to:
Assess the long term, comparative real-world effectiveness on clinical liver events of GLP-1 agonists vs. DPP4 inhibitors in patients with T2D living in England.
For this study we use a longitudinal, new user, active comparator cohort design, where individuals newly prescribed GLP-1 agonists (index drug) and individuals newly prescribed DPP-4 inhibitors (comparator drug), are followed for the health outcome(s) of interest. Adults (≥18 years) with T2D, and a first ever prescription (index date) of GLP-1 agonists or DPP4 inhibitors registered (in CPRD) from 01 January 2007 (the year the first incretin-based drugs entered the UK market) are included in the study population. Patients with registration (in the Hospital Episodes Statistics; HES) of alcohol-related disorders, chronic liver diseases other than NAFLD and/or prescriptions of drugs inducing liver disease (in CPRD) are excluded. Index date equals date of first prescription of GLP-1 agonists or DPP-4 inhibitors with no prescription of either prior to index date.
Individuals are followed from date of inclusion until time of first event recorded in HES or Office for National Statistics Death Registration (ONS), migration from the databases or end of follow-up, whichever comes first. Sensitivity analyses addressing T2D, alcohol, obesity, age, calendar period, deprivation, event type and reverse causality will be conducted.
The primary outcome of the study is a composite endpoint for which the components are listed below. Events in the outcome are defined as ICD10 or OPCS4 codes from Hospital Episodes Statistics (HES) Outpatient data, HES Admitted Patient Care Data or Office for National Statistics (ONS) Death Registration Data.
PRIMARY ENDPOINT
Clinical liver events (composite endpoint) - using the time to first report of any of the below events captured in HES or ONS:
• Liver mortality
• Hospital contact for liver cirrhosis
• Hospital contact for hepatocellular carcinoma (HCC)
• Hospital contact for chronic liver failure
• Hospital contact for liver transplantation
•
• Hospital contact for portal hypertension
• Hospital contact for gastro-oesophageal varices (of any grade) or portal-hypertensive gastropathy without bleeding
• Hospital contact for hepatic decompensation events
o Ascites
o Placement of Trans jugular intrahepatic portal shunt
o Hepatorenal syndrome
o Hepatic encephalopathy
o Jaundice (related to liver-cirrhosis)
o Gastro-oesophageal varices (or portal-hypertensive gastropathy) with bleeding
POTENTIAL SECONDARY ENDPOINTS
When analysing a composite endpoint, there is a risk of one (of the less severe) component driving the entire association. There is also a risk that the composite endpoint masks different types of effect of the intervention. To explore if this is the case, several secondary endpoints will be considered, including specific liver events and a modified version of the composite endpoint.
To get a better understanding of how the choice of outcome affect the study results one or more of the following modified endpoints may be analysed:
1) Hospital contact for hepatic decompensation events
o Ascites
o Trans jugular intrahepatic portal shunt
o Hepatorenal syndrome
o Hepatic encephalopathy
o gastro-oesophageal varices with bleeding
2) Hospital contact for hepatocellular carcinoma (HCC)
3) Composite endpoint (without the most frequent, first occurring event)
The primary composite endpoint will be reanalysed after removing the most frequently occurring component. This is expected to be ascites but will be determined from the actual data.
In addition, it will be considered to analyse:
4) All-cause mortality in a subgroup of patients with likely NASH
In this subgroup all-cause mortality is more likely to be affected by potential liver effects of the exposure than would be the case for the entire study population of patients with diabetes.
Katrine Grau - Chief Investigator - Novo Nordisk A/S
Pernille Cromhout - Corresponding Applicant - Novo Nordisk A/S
Jens Tarp - Collaborator - Novo Nordisk A/S
Kamal Kant Mangla - Collaborator - Novo Nordisk A/S
Kamlesh Khunti - Collaborator - University of Leicester
Kevin Tan - Collaborator - Novo Nordisk A/S
Mads Jeppe Tarp-Johansen - Collaborator - Novo Nordisk A/S
Mette Skalshøi Kjær - Collaborator - Novo Nordisk A/S
Quentin Anstee - Collaborator - Newcastle University
Rikke Baastrup Nordsborg - Collaborator - Novo Nordisk A/S
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;ONS Death Registration Data;Patient Level Index of Multiple Deprivation