Hepatitis E seroprevalence and viremia rate in immunocompromised patients: a systematic review and meta‐analysis

Hepatitis E is an infectious disease of the liver caused by the hepatitis E virus (HEV). Immunocompromised patients present a particular risk group, as chronification of hepatitis E leading to life‐threatening cirrhosis occurs when these patients are infected. Therefore, this study aims to estimate and compare the anti‐HEV seroprevalence and the rate of HEV RNA positivity in transplant recipients and patients with human immunodeficiency virus (HIV).


| Introduc tion
Hepatitis E is an infectious disease of the liver caused by the hepatitis E virus (HEV). 1,2 Anti-HEV IgG positivity indicating previous HEV exposure is a common finding, with seroprevalence rates of up to 30% in Germany, France and the Netherlands. 3 Risk of HEV exposure has elevated by almost double in the United States compared to neighbouring countries in Latin America. 4 Although the vast majority of patients undergo an asymptomatic course, a minority develop clinically overt hepatitis. 1,2 Hepatitis E may lead to acute-on-chronic liver failure (ACLF) in patients with underlying chronic liver disease, or it can cause chronic hepatitis in immunocompromised patients. 2,5 Chronic hepatitis E, which has been observed only in genotype (GT) 3 and 4 infections, may proceed to hepatic fibrosis and subsequent cirrhosis. 6,7 It has been assumed that transplant recipients (TR) are at risk of chronification with rates up to 50%. 1,8 In addition to TR, chronic hepatitis E has been described in patients with HIV-1 infection. [9][10][11] Notably, although HEV infection in HIV positive patients is assumed to be strongly associated with a reduced CD4 T cell count, chronic hepatitis E may persist even after the immune system has recovered under antiretroviral therapy. 6,11 In addition to TR and HIV positive patients, chronic hepatitis E has been reported in rheumatological [12][13][14] and haematological [15][16][17][18] patients with chronification rates of up to 33%, in rheumatological patients respectively. 19 In haematological patients ongoing hepatitis E with fatal outcome was observed in 16% of patients, as recently reported. 20 Despite the fact that chronicity of HEV infection is a topic of emerging relevance in transplant recipients and HIV positive patients, there is still no direct comparison of anti-HEV IgG and HEV-RNA positivity rates in these different groups of immunocompromised patients. Thus, we performed a systematic review and meta-analysis to estimate the anti-HEV seroprevalence and the rate of HEV RNA positivity in transplant recipients and HIV positive patients in order to assess the risk of exposure and chronification.

| Search Strategy and Selection Criteria
The literature search was performed using two different databases: Scopus and PubMed. A literature search in both databases was performed by using terms "hepatitis E" or "HEV" in combination with the terms "transplant", "transplantation" or "HIV", "AIDS" or "immuno-

| Study quality
The articles identified were examined for their study quality according to a fixed scheme. The criteria applied were as follows: identification of the employed assay (ELISA or PCR) and confirmation of use according to the manufacturer's instructions; absence of age restriction in the study cohort (e.g. studies reporting on children and/or adolescents < 18 years were excluded) and population-based analysis. Studies that did not meet the study quality criteria were excluded from further analysis. Data were assessed based on their methodological quality according to the Joanna Briggs Institute's well-established critical appraisal tool for prevalence studies. 22,23 Studies were assessed by G.B. and discussed with T.H. accordingly. Any disagreements were resolved by discussing with a third investigator (S.P.).

| Statistical analysis
The prevalence of anti-HEV was estimated by pooling the study data of HEV-RNA positivity rate between HIV and transplant patients, we also used a linear mixed effects regression analysis, but only included the methodological quality score as a moderator. A restricted maximum likelihood estimator (REML) with logit transformation was used for prevalence estimation. Heterogeneity was checked via the quantity I 2 and publication bias was checked via a funnel plot. Analysis was conducted using R (version 3.5.1) and the metafor package.

| Result s
Of 1,138 articles, 120 (10.5%) were included in the final analysis as shown in the study flow chart (Fig. 1). Detailed information on the studies included and their characteristics are provided as supplementary tables (Table S2-S4). anti-HEV seroprevalence, tested with commercially available assays, did not differ significantly between TR and HIV positive patients independently of the assay, methodological quality and study year (Table S1). Using inhouse assays, anti-HEV seroprevalence differed significantly (Table S1)  available in 83 studies. However, we did not observe relevant effects on estimated prevalence rates in an age adjusted subgroup analysis (see "age adjusted anti-HEV IgG Seroprevalence" illustrated in Fig   S1).

| Time trends of articles related to HEV and immunocompromised patients
The articles included were sorted by date of publication (Fig.S3).

| Discussion
Chronic HEV infection in immunocompromised patients including transplant recipients, HIV positive patients or patients with haematological malignancy has been considered increasingly relevant in recent years (Fig.S3). 24 Regarding chronic HEV infection, we have yet to thoroughly understand its potential risk factors, long-term sequalae and management. Rapid disease progression to cirrhosis, a feared and potential life threatening complication, has frequently been observed in immunocompromised patients with chronic HEV infection. 7 Even though relevance of HEV infection in the transplant setting has been acknowledged by the international scientific community, there are still conflicting standpoints on the overall anti-HEV seroprevalence and rate of chronic HEV infection in immunocompromised patients. 8 To address this, we performed this systematic review and meta-analysis to compare anti-HEV seroprevalence and HEV viremia in The present study confirms high rates of anti-HEV seroprevalence up to 30% in the special cohorts of transplant recipients and HIV positive individuals (Fig. 2). These rates are far higher than previously described anti-HEV seroprevalence rates in the general population (2% to 17%), as reported in two recent systematic reviews and meta-analyses on anti-HEV seroprevalence in Europe and the Americas. 3,4 The highest anti-HEV seroprevalence rates were observed in heart transplant recipients. This is in line with Pischke et al., who found the highest anti-HEV seroprevalence in heart transplant recipients compared to non-transplant patients undergoing cardiac surgery and healthy controls. Thus, heart transplant recipients, frequently receiving blood products, seem to present a particular risk group for HEV exposure. This is remarkable, as heart transplant recipients are strongly immunosuppressed and therefore, a lack of sensitivity concerning serological testing is anticipated. However, anti-HEV IgG rates observed in this single centre study were much lower in comparison to the present study. 25 A known difficulty of seroprevalence studies is the large variety of commercially available assays and in particular, their different levels of sensitivity and specificity. 3 Unlike what was initially hypothesized, anti-HEV seroprevalence in transplant recipients was not significantly higher than in HIV positive patients tested using commercially available assays. Overall, these findings support the assumption that in immunocompromised patients, prevalence could be underestimated because of undetectable anti-HEV antibodies.
Interestingly, significantly lower anti-HEV seropevalence in transplant recipients was observed in the subgroup tested with non-stan-  30 However, affection of the immune system such as a reduction of CD4 count has been associated with increased anti-HEV seroprevalence rates and potential risk for chronification. 26 In this study, we observed comparable HEV exposure rates in HIV positive patients and transplant recipients.
In addition, we aimed to assess rates of HEV RNA positivity in two cohorts of immunocompromised patients. Importantly, detecting HEV RNA via PCR is recommended in immunocompromised patients. 31,32 In immunocompetent patients detectable viremia is assumed to last up to six weeks, whereas immunocompromised patients are at risk of developing chronic hepatitis E, which is defined as HEV RNA presence for at least three months. 33 In the present study, we found a high rate of HEV RNA positivity The present study provides novel insights on the nature of HEV infection in immunocompromised patients. However, our investigation has some limitations. First, findings depend on the quality of the studies included for review. However, in order to prevent potential bias, data were assessed by experienced scientists according to the Joanna Briggs Institutes' critical appraisal tool, well-established for prevalence studies. 22 The statistical analyses and the interpretation of the data were performed by a well-experienced experts Second, the studies did not provide clinical data such as serum total IgG levels to interpret the immunological conditions. In particular, the HIV positive patient cohort is presumably undergoing standard ART treatment and thus potentially does not have a substantially TA B L E 1 Subgroup analyses of transplant recipients: Anti-HEV seroprevalence rates and HEV RNA are shown in percent with 95% confidence interval (CI) In conclusion, in this systematic review and meta-analysis we

ACK N OWLED G EM ENTS
The authors thank Elaine Hussey for proofreading and linguistic correction of the manuscript. Open access funding enabled and organized by Projekt DEAL. Open access funding enabled and organized by ProjektDEAL.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflicts of interest regarding this study.