Morbidity and Mortality of Hepatitis E Virus Infections in Belgium

1. Study objectives 1.1. Primary objectives: To examine viral factors, such as gt, subtype and viral load, associated with the morbidity and mortality of HEV infections in Belgium. 1.2. Secondary objectives:

• To describe host factors associated with the morbidity and mortality of HEV infections in Belgium.
• To find independent host and/or viral factors that are associated with the morbidity and mortality of HEV infections in Belgium.

1. Investigational plan 2.1. General study design Retrospective cohort study. 2.2. Study duration, enrollment and number of sites Subjects for which a serum sample is sent to the Belgian National Reference Center for Viral Hepatitis Sciensano to diagnose a HEV infection between January 2010 and June 2018. 2.3. Study population All patients with a confirmed or possible HEV infection at Sciensano between January 2010 up to and included June 2018. 2.3.1. Inclusion criteria:

• HEV RNA positive AND/OR HEV immunoglobulin (Ig)M positive
• Lab confirmation of the HEV infection performed at Sciensano
• All ages will be included
• Referring hospital in Belgium 2.3.2. Exclusion criteria:
• No confirmation of HEV serology AND/OR polymerase chain reaction (PCR) performed at Sciensano
• Cases not diagnosed in Belgium
• Cases without clinical outcome data

1. Study procedures Review of records. Re-evaluation of available liver biopsies by an expert liver pathologist (Prof. Tania Roskams, UZ Leuven). 3.1. Data sources In Belgium, Sciensano acts as a National Reference Center (NRC) for viral hepatitis. Most of the documented HEV infections in Belgium are confirmed in Sciensano and samples on which HEV serology (IgM and IgG), PCR and genotyping have been performed are biobanked. 3.2. Data collection A questionnaire will be sent to all treating physicians of patients with HEV infections in Belgium. If no response is received, a reminder will be sent by regular mail and later by e-mail. Different options will be offered to the treating physicians: to fill in the questionnaire, to complete the questionnaire by telephone with an investigator from Sciensano, to let the data to be collected at the center site by an investigator of Sciensano. Available liver biopsies from confirmed HEV cases will be retrieved from the pathology department of each center and re-evaluated by an expert liver pathologist. 3.3. Variables collected Demographic data: Age and geographical origin. Date of symptom onset, date of diagnosis, date of loss to follow-up / death. Pregnancy at the time of diagnosis. Diabetes mellitus at the time of diagnosis. Immunocompromised status at the time of diagnosis. Known cirrhosis (fibroscan / biopsy / …), hematological or oncological disease. Other comorbidities. Blood transfusion history. Solid organ or stem cell transplantation preceding the HEV infection, type of organ and date. Renal function or renal replacement therapy. Infection in other family members. Use of alcohol and number of units per week. Use of illicit drugs. Travel abroad preceding the HEV infection and name of the country visited. Consumption of pork meat. Treatment data: Treatment with ribavirin. Treatment with other known/putative antiviral therapy. Reduction of immunosuppressive medication because of the HEV infection. Stop of immunosuppressive therapy because of HEV infection. Disease course and outcomes data: Acute hepatitis with complete recovery. Liver failure. Hospitalisation. Duration of hospitalisation in days. Admission to the intensive care unit. Duration of hospitalisation at the intensive care unit in days. Variceal bleeding. Bacterial peritonitis. Hepatic encephalopathy. Extrahepatic manifestations and type. Neurologic disease and type. Chronic HEV infection (as defined by HEV RNA positive for 3 months or more). Progression to cirrhosis as a consequence of HEV infection. Liver transplantation as a consequence of HEV infection and date of transplantation. Death and date of death. Death as a consequence of HEV infection. Total duration of symptoms in weeks. Biochemical data: HEV IgM, HEV IgG, HEV (quantitative) PCR, HEV genotype and subtype. HEV genotyping was performed by Sanger sequencing of a 348 base pair (between 2010 and 2016) or a 493 base pair (from 2017) region of ORF2. The obtained sequences were aligned against HEV reference genomes and submitted to HEVnet for genotype assignment (13). Presenting ALT, peak ALT, presenting AP, peak AP, presenting bilirubin, peak bilirubin, presenting INR, peak INR, presenting albumin, lowest albumin, presenting eGFR (CKD-EPI), lowest eGFR (CKD-EPI). Liver biopsy scoring: Histological grading and staging on H\&E stained liver biopsy slides will be applied in a blinded manner. One expert liver pathologist, who is unaware of the infection status of the patient, will score the liver biopsies. The following parameters will be scored from 0-4 in a semi-quantitative manner: general architecture, lobular and portal inflammation, presence of immune cells and necrotic cells, presence of interfaces. Per patient, a global score will be assigned taken into account the spread, mean and mode of the different scored parameters.
2. Statistical analysis plan

The characteristics of patients will be analysed with the chi-square test for categorical variables and student t-test (Welch test if unequal variances or Wilcoxon Rank sum test when normality is not met) for the continuous variables. Numerical data are presented as the mean and standard deviation or 95% confidence interval, while categorical are expressed as counts and percentages.

The following independent viral variables are defined: HEV gt 3 clade type (abchijklm, efg and ra), viral load.

The following independent host variables are defined, given their known association with disease outcome or severity: alcohol consumption, diabetes, gender, age, cirrhosis and immunosuppression.

Differences between demographic factors, disease severity (expressed by laboratory data and histology) and disease outcomes (as defined by hospitalization status and duration, ICU stay, chronic infection, extrahepatic manifestations and death) will be analysed for the independent viral variables with an univariate analysis. Multivariable linear or logistic regression analysis will be performed with dependent variables that were found to be significantly associated with the independent variables in univariate analysis. Given the retrospective nature of the analysis, an upfront power calculation for the number of included patients for the investigated outcome is not possible. Statistical analyses are performed at a significance level of 5% using the statistical software R or SAS.