Long-term liver disease can lead to the formation of scars in the liver. When the liver is too scarred to function properly, we speak of liver cirrhosis. Liver cirrhosis may lead to liver failure and/or liver cancer. Several conditions can cause cirrhosis, two of which are chronic infection with hepatitis B virus, and the building-up of fat in the liver, also called non-alcoholic fatty liver disease (NAFLD). NAFLD is often associated with other metabolic conditions such as diabetes and obesity.
The aim of this study is to understand if infection with hepatitis B virus causes patients with type 2 diabetes and NAFLD to develop cirrhosis more often or faster than those who are not infected with this virus.
Liver cirrhosis is a very serious condition that can require liver transplantation or lead to premature death. Preventing or slowing its development can be beneficial to patients at high risk. The results of this study will help evaluate if infection with hepatitis B virus increases the risk, or speed of liver disease progression, and therefore support prevention of Hepatitis B infection in a population already at increased risk.
Both Hepatitis B Virus (HBV) infection and non-alcoholic fatty liver disease (NAFLD) can cause liver damage. There is a lack of data on the role of HBV infection in the progression of liver disease in diabetic patients with NAFLD.
The main research question is: does HBV-infection increase the risk of development of, or accelerate the progression to, NAFLD-related stages in patients with type 2 diabetes mellitus (T2DM)?
We propose a cohort study based on health-records of the CPRD population eligible for linkage with HES Admitted Patient Care, from 2000 to the latest CPRD release available at the time of the analyses. Three cohorts will be compared: subjects with T2DM and HBV-infection, subjects with T2DM and without HBV-infection, and T2DM-free subjects with HBV-infection. Due to underdiagnoses of NAFLD, and the high proportion of NAFLD in subjects with T2DM, T2DM will be used as a proxy for NAFLD exposure. Subjects with a history of liver disease unrelated to HBV infection or NAFLD will be excluded. We will look at descriptive data, calculate incidence rate ratios (IRRs) using a Poisson regression model, and explore diseases progression in a time-to-event analysis. The latter models will be adjusted for covariates and sensitivity analyses will be performed to assess and minimise the impact of potential confounders.
NAFLD-related cirrhosis The primary outcome NAFLD-related cirrhosis is defined, for the purpose of this study, as cirrhosis for which no other cause was identified. The definition is therefore implemented by exclusion, namely as cirrhosis of other aetiology than those defined as study exclusion criteria (alcoholic liver disease, alcoholism, autoimmune liver disease, any viral hepatitis other than HBV, liver transplantation, liver cancer, hemochromatosis, liver failure). The date of onset will be the date of earliest evidence of any cirrhosis. Preliminary code lists for cirrhosis, including those part of the exclusion criteria, are available in the Appendix 2. Liver fibrosis, non-alcoholic hepatocellular carcinoma For exploratory aim 5 and 6 non-alcoholic liver fibrosis and non-alcoholic hepatocellular carcinoma are defined as fibrosis or hepatocellular carcinoma of other aetiology than those defined as study exclusion criteria (alcoholic liver disease, alcoholism, autoimmune liver disease, any viral hepatitis other than HBV, liver transplantation, hemochromatosis, liver failure). The date of onset will be the date of earliest evidence of any fibrosis or hepatocellular carcinoma. Preliminary code lists for non-alcoholic fibrosis and non-alcoholic hepatocellular carcinoma are available in the Appendix 2. Outcome definition algorithm performance A subset of maximum 100 NAFLD-related cirrhosis cases identified in Cohorts 1 to 3 will be selected at random and ascertained through subject profile review, by two independent reviewers, including all available clinical, therapy, and laboratory test recorded in CPRD, to insure the performance of the NAFLD-related cirrhosis classification algorithm. The positive predictive value of the NAFLD-related cirrhosis classification algorithm and the inter-reviewer agreement will be estimated. Based on the positive predictive value, if lower than 90%, case definitions may be refined and additional sensitivity analyses will be performed. Two P95 clinical reviewers (Anke Stuurman, Germano Ferreira) will do the case profile review and Thomas Verstraeten will resolve any disagreement; Prof Yves Horsmans will be consulted in case of the need for further clinical expert adjudication.
Germano Ferreira - Chief Investigator - Not from an Organisation
Germano Ferreira - Corresponding Applicant - Not from an Organisation
Adrienne Guignard - Collaborator - GSK
Anke Stuurman - Collaborator - P95
Kaatje Bollaerts - Collaborator - P95
Maria Alexandridou - Collaborator - P95
Tom Cattaert - Collaborator - P95
Veronique Bianco - Collaborator - GSK
Yves Horsmans - Collaborator - Catholic University of Louvain
HES Admitted Patient Care