Hepatitis E Virus as a Causative Agent of Unexplained Liver Enzyme Elevations in HIV-Infected Patients

Abstract

To the Editors: Liver enzyme elevations (LEEs) in HIV-positive patients are mostly due to either viral hepatic coinfections or antiretroviral drug–induced hepatotoxicity, leaving a small number of patients with unexplained LEEs.1 However, it was recently shown that a part of the unexplained LEEs in the general population is attributable to locally acquired hepatitis E virus (HEV) genotype 3 infections—probably as the result of zoonotic transmission.2,3 Recent reports have shown that HIV-infected patients are at risk for persistent HEV carriage due to their immunocompromised state.4–6 Since then, the prevalence of HEV in several cohorts of HIV-infected patients has been described.7–12 Renou et al13 observed a geographical gradient in HEV seroprevalence, being lower in HIV-infected patients in northern France compared with the south. However, data from northern Europe are scarce.14,15 Considering the possibility of zoonotic transmission16,17 and the high HEV seroprevalence among pigs and boars in these northern European countries,18,19 investigating the HEV epidemiology in susceptible populations in these countries is important. Therefore, we conducted a study in the Netherlands evaluating the prevalence of locally acquired HEV infection in a cohort of HIV-infected patients with and without unexplained LEEs under follow-up in our clinic. The laboratory results of all HIV-infected patients under follow-up from January 2007 through February 2011 in the University Medical Center Utrecht (n = 1117) were reviewed for the occurrence of elevations in serum alanine aminotransferase (ALT) above the upper limit of normal (45 U/L in our hospital). Because we focused on locally acquired infections with HEV genotype 3, immigrants from regions where non–genotype 3 HEV is highly prevalent were excluded. A number of selected serum samples of unexplained LEEs from the period—negative for hepatitis A, B, or C infection—were obtained for HEV serological testing. As controls, another group of white patients—matched for age and gender—without LEEs was selected and a random sample obtained. All samples were tested—together, in 3 batches—for the presence of both anti-HEV IgM and IgG antibodies with a commercial enzyme-linked immunosorbent assay kit (recomWell HEV; Mikrogen GmbH, Neuried, Germany); positive and borderline results were tested by line blot assay (recomLine HEV IgG/IgM; Mikrogen GmbH). All tests were performed according to the manufacturer's instructions. Patient characteristics at the time the serum sample was obtained were registered. Continuous data were expressed as median value (with range) and analyzed using the Mann–Whitney U test, whereas categorical data are given as a percentage (with number) and analyzed using the Fisher exact test. Results are 2-sided, and P ≤ 0.05 was considered to be statistically significant. A total of 56 HIV-infected patients with unexplained LEEs were selected and matched for age and gender to a total of 50 HIV-infected patients without LEEs. Except for serum ALT values, there were no significant differences in patient characteristics between both groups (Table 1). In the total population (N = 106), 4 patients (3.8%) tested positive for anti-HEV IgG antibodies. Only 1 patient without LEEs tested HEV IgG positive. Within the group of patients with LEEs, 2 patients were HEV IgG positive, whereas 1 patient tested positive for both anti-HEV IgM and IgG antibodies indicating an acute HEV infection. However, a subsequently performed HEV polymerase chain reaction (PCR) on this serum sample was negative. No PCR was performed in the other patients who tested positive for IgG only because this indicated past exposure to HEV. No IgM was detected in any of the collected samples from the group without LEEs. Regarding the 4 HEV-positive patients with detectable IgG antibodies, 3 (75%) were men, with 2 of them being men who have sex with men. Mean age was 32 years, and the mean baseline CD4+ count was 528 cells per cubic millimeter. None of these characteristics differed significantly from the IgG-negative population. Again, only the baseline mean ALT value was significantly higher in the anti-HEV IgG antibody–positive patients (288 vs. 76 IU/L, P < 0.01).TABLE 1: Patient CharacteristicsThis study shows that locally acquired HEV infection occurs among HIV-positive patients in the Netherlands—probably as the result of zoonotic transmission.20 Furthermore, we observed an anti-HEV IgG seroprevalence rate of 3.8% in our whole cohort of HIV-infected patients, with a higher prevalence in those with LEEs (5.4% vs. 2%, P = 0.62). In the patients without LEEs, we did not detect any acute HEV infections. This observed seroprevalence is similar to the one found in the general Dutch population.2 In contrast, 2 recent studies14,15 reported a slightly higher anti-HEV IgG seroprevalence of 4.9% (German cohort) and 9.4% (British cohort) in a predominantly northern European population. No acute infections were detected in these cohorts either. Differences in patient characteristics and different available assays21,22 may have contributed to the observed differences in epidemiology between the study populations. Because recent observations have suggested a delayed anti-HEV seroconversion in HIV patients with very low CD4 counts,4 questions have been raised about the suitability of these tests for the detection of HEV infections,7 and together with the varying diagnostic value between the available assays, they support the argument that serological screening alone may be insufficient to diagnose HEV and should be complemented with detection of HEV by PCR. The recent developments regarding the epidemiology of HEV and the altered natural course in immunocompromised patients led to the discussion of whether HIV-positive patients with unexplained LEEs should be tested for HEV infection.23 With only 1 immunoblot-confirmed case of locally acquired acute HEV infection in the Dutch HIV-positive population, this might not be a cost-effective strategy. In conclusion, although HEV infection should be considered as a causative agent of unexplained LEEs in HIV-infected patients in northern European countries, the number of acute infections is low.