Hepatitis E in blood donors: investigation of the natural course of asymptomatic infection, Germany, 2011

Asymptomatic hepatitis E virus (HEV) infections have been found in blood donors from various European countries, but the natural course is rarely specified. Here, we compared the progression of HEV viraemia, serostatus and liver-specific enzymes in 10 blood donors with clinically asymptomatic genotype 3 HEV infection, measuring HEV RNA concentrations, plasma concentrations of alanine/aspartate aminotransferase, glutamate dehydrogenase and bilirubin and anti-HEV IgA, IgM and IgG antibodies. RNA concentrations ranged from 77.2 to 2.19×105 IU/mL, with viraemia lasting from less than 10 to 52 days. Donors showed a typical progression of a recent HEV infection but differed in the first detection of anti-HEV IgA, IgM and IgG and seropositivity of the antibody classes. The diagnostic window between HEV RNA detection and first occurrence of anti-HEV antibodies ranged from eight to 48 days, depending on the serological assay used. The progression of laboratory parameters of asymptomatic HEV infection was largely comparable to the progression of symptomatic HEV infection, but only four of 10 donors showed elevated liver-specific parameters. Our results help elucidate the risk of transfusion-associated HEV infection and provide a basis for development of screening strategies. The diagnostic window illustrates that infectious blood donors can be efficiently identified only by RNA screening.


Introduction
The hepatitis E virus is a single-stranded RNA virus; there are currently four human pathogenic genotypes 1 to 4 [1]. Genotypes 1 and 2 are hyperendemic in developing countries, restricted to humans, and transmission occurs by the faecal-oral route [2,3]. In industrialised countries, genotypes 3 and 4 are responsible for sporadic cases of HEV infection. However, the incidence of non-travel-associated HEV infections has increased and hepatitis E is now recognised as an emerging and often undiagnosed disease [1,4,5]. The genetic similarity of strains isolated from humans and other mammalian species suggests zoonotic or foodborne transmission [6,7]. Hepatitis E presents asymptomatically or symptomatically. Symptomatic infection presents as an acute, mostly self-limiting hepatitis with clinical characteristics similar to hepatitis A [2]. Clinical manifestations of HEV infections caused by the different genotypes are indistinguishable. Genotype 3 and 4 patients are usually middle-aged and elderly men, whereas genotypes 1 and 2 also cause acute hepatitis in healthy children and adolescents [8]. The pathogenic impact of genotype 1 and 2 and genotype 3 and 4 differ considerably. HEV genotype 1 and 2 infections lead to a high mortality among pregnant women in developing countries (8-20% [9,10]) while no serious infections among pregnant women with genotypes 3 and 4 were described in industrialised countries. HEV genotype 3 and 4 infection proceed asymptomatically in immunocompetent individuals [8], but severe or fatal HEV infections have been observed in individuals with chronic liver disease [11,12], in transplant patients [13,14] and in immunosuppressed individuals [8]. Asymptomatic HEV infection has often been observed in blood donors [15][16][17], with reported prevalence rates of HEV RNA-positive donors of 1:2,848 (England [18]), 1:1,240 (Germany [17]) and 1:1,761 (the Netherlands [19]).
The progression of viraemia and the serological course of anti-HEV antibodies during clinically apparent HEV infection is well characterised [2,20,21], but so far little is known about the progression of infection in asymptomatic individuals, in whom HEV infection usually remains undetected. Therefore, we conducted a prospective study to characterise the duration of viraemia, the antibody response (IgA, IgM and IgG), and the progression of liver-specific enzymes in 10 HEV genotype 3-infected German blood donors [17].

Specimens
From July to September 2011, a total of 16,125 individual German blood donors were routinely screened for the presence of HEV RNA by the Uni.Blutspendedienst Ostwestfalen-Lippe. Their geographical origins were North Rhine-Westphalia, Lower Saxony and Hesse; 57.5% (n = 9,271) were male, with a median age of 33 years (± 13; range: 18-72), and 42.5% were female (n = 6,867), with a median age of 32 years (± 13; range: 18-71) [17]. The screening recovered 13 HEV RNApositive donors. Retrospectively, residual plasma samples of one donation preceding and several donations following the initial HEV RNA-positive donation, taken within a short time distance from each other, (Table  1) were available for 10 donors (D1 to D10, all male). The day of the detection of HEV RNA by PCR screening was defined as day 0, but HEV infection is most likely to have occurred before the beginning of our study period. This aspect limits the exact calculation of the diagnostic window between the detection of HEV RNA and anti-HEV antibodies. In addition, the period of detectability of antibodies may have started before the first positive sample and lasted beyond the last positive sample. To take this into account, we calculated two intervals of HEV-RNA positivity: Interval 1 started on the day of the first positive and ended on the day of the last positive sample, whereas interval 2 started at half of the interval between the last negative and first positive sample and lasted until half of the interval between the last positive and first negative sample. The duration of anti-HEV seropositivity was calculated according to interval 2.
All HEV-infected donors underwent pre-donation medical examination and negated current diseases or any known risk factors for viral infection. Post-donation questionnaires to elucidate risk factors for HEV infection were returned by six donors. The study protocol followed the ethical guidelines of the Ruhr University, Bochum, and was approved by the institutional review board. All donors provided informed consent.

RNA extraction and real-time RT-PCR
Total RNA from individual samples was extracted from 500 µl plasma using the NucliSens easyMAG (bioMerieux, Nürtingen, Germany) automated RNA/DNA extraction system. Amplification using the RealStar HEV RT-PCR Kit (Altona Diagnostic Technologies (ADT), Hamburg, Germany) was performed on the Rotor-Gene 3000 system (Corbett Life Sciences, Sydney, Australia). HEV virus titre in positive plasma was quantified using the first World Health Organization (WHO) international standard for hepatitis E virus RNA for NAT-based assays (Paul-Ehrlich institute, Langen, Germany) [22].
Comparative testing was performed using the Wantai HEV IgM and IgG ELISA (Sanbio B.V., Uden, the Netherlands), and results were classified into three categories: (i) no antibodies detectable (cut-off < 0.9: negative), (ii) evidence for the presence of antibodies (cut-off 0.9-1.1: borderline) and (iii) antibodies detectable (cut-off > 1.1: positive). Confirmatory testing with an immunoblot assay was performed on 22 samples using the recomLine HEV-IgM/IgG immunoassay according to the manufacturer's instructions (Mikrogen GmbH, Neuried, Germany). Sample selection included those samples taken at the first positive detection of anti-HEV antibodies and up to two consecutive samples.
Concentrations of glutamate dehydrogenase (GLDH), aspartate aminotransferase (AST), alanine aminotransferase (ALT) and total bilirubin were measured in plasma samples using the respective enzymatic assays (Abbott Diagnostics Europe, Wiesbaden, Germany) on the Architect ci8200 system (Abbott Diagnostics Europe).

Progression of hepatitis E virus RNA and antihepatitis E virus antibodies
The progression of RNA concentration in follow-up samples from infected patients is shown in Figure panel A, and the key observations of HEV RNA progression are summarised in Table 1. HEV viraemia persisted up to 52 days (D2 and D8, interval 1) with considerably different RNA concentrations in individual donors, ranging from 1.86×10 1 to 2.19×10 5 IU/mL. High RNA concentrations were observed in the window period ranging from 1.36×10 3 to 2.19×10 5 IU/mL. Taking the second interval into account, the duration of viraemia was as long as 20 to 80 days. The maximum viraemia was observed after 20 days, with a mean duration of 29 days for interval 1 and 49 days for interval 2 (Table 1).

Table 2
Occurrence of anti-hepatitis E virus antibody classes (IgA, IgM and IgG) and first detection and duration of IgM and IgG seropositivity in two serological assays, blood donors, Germany, 2011 (n = 10)

Comparison of the diagnostic window using different serological assays
We further compared the timing of the first detection of different antibody classes during the window period when only HEV RNA was detectable and the duration of seropositivity of HEV-specific IgM and IgG antibodies using two different serological assays (Table 2). For IgM and IgG antibodies, the diagnostic window differed depending on the assay used (

Discussion
HEV viraemia in symptomatic cases usually lasts from four to six weeks but can remain more than 100 days in some cases [23]. Liver enzyme values reach a peak about six weeks post exposure before decreasing towards normal levels by week 10 [20]. The typical serological course of an HEV infection shows an initial rise in short-lived anti-HEV IgM after three to four weeks that decline to baseline levels within three to six months, followed by an increase of IgG which remains detectable for up to 15 years [2,20,21]. However, the knowledge about the natural course of HEV infection in asymptomatic HEV-infected individuals is limited.
The clinically asymptomatic cases analysed in this study represent the preselection of apparently healthy individuals voluntarily donating blood and lacking physically detectable symptoms of infection. The retrospective character of this study limited the availability of consecutive samples from the same donor taken less than 30 days apart and the accuracy of the calculated durations (viraemia, seropositivity). The observed differences in the sensitivity of the serological assays further influenced the calculation of the diagnostic window. For example, it has been shown that the performance of anti-HEV IgG assays strongly influences the estimation of hepatitis E seroprevalence [24]. The progression of HEV RNA in a Japanese cohort of 15 patients with acute symptomatic hepatitis E was largely comparable with what we observed in our study [25]. In contrast to our results, anti-HEV IgA and IgM (first detection: day 8-42) and IgG antibodies (first detection: day 13-59) in the Japanese cohort were detectable in symptomatic cases in parallel to the presence of HEV RNA at first sampling [25], pointing towards an earlier onset of viraemia in the patients without symptoms. Accordingly, anti-HEV IgA and IgM remained detectable until the end of the observation period in symptomatic cases in the Japanese cohort while two different progressions were observed in the asymptomatic cases in our study. Antibodies in some asymptomatic cases showed the same persistence as in symptomatic cases, whereas antibody levels in other asymptomatic cases continuously decreased and reached undetectable levels. Furthermore, we observed IgM positivity for a significantly longer period compared with the Japanese cohort with seropositivity (longer than 100 days in D3, D4 and D6). However, these differences between symptomatic and asymptomatic cases could be related to the performance of the ELISAs used. There is no consensus on whether immunoblot assays (rather than ELISAs) are needed in order to detect anti-HEV antibodies accurately. The immunoblot results in our study did not add informative value; the immunoblot provided negative results for samples with divergent results in the two different ELISAs, most probably because of inferior sensitivity.
Unexpectedly, anti-HEV IgG antibodies declined under detectable levels in samples from donor D3. Previous studies have shown that the period when anti-HEV IgG remains detectable can vary individually from six months to 14 years, but HEV IgG antibodies have also been shown to disappear [26][27][28]. Remarkably, a rise in liver-specific enzymes was observed only in four of 10 asymptomatic individuals, although high viral loads were detected in plasma. The elevation of ALT may have been missed in donors D9 and D10 because of the long delay of 42 and 38 days between two samples, respectively, but for the other eight donors, samples within the first 50 days after detection of HEV viraemia were taken at average intervals of less than 10 days.
There is an ongoing debate about HEV genotype 3 and 4 infection and blood safety. Published reports of HEV infections transmitted by contaminated blood products [29,30] and of the detection of HEV genotypes 3 and 4 in plasma fractionation pools [31] and blood donors [15][16][17] suggest that transfusion transmission of HEV is probably not uncommon, with many undiagnosed subclinical infections [15,16]. In a recent study by Hewitt et al., transmission of HEV genotype 3 via contaminated  blood was demonstrated in 42% of transfusion recipients [18]. The clinical course (asymptomatic, mild hepatitis or acute liver failure) and severity of HEV infection in transfusion recipients are variable, most probably depending on predisposition or immune status. The vast majority of HEV genotype 3 and 4 infections are most likely to result in an asymptomatic course [32] but, for instance, chronic manifestations of HEV genotype 3 infection in immunosuppressed persons can become important in industrialised countries [33]. Feray et al. concluded that transfusion of blood products not screened for HEV RNA is associated with the risk of chronic infection in immunocompromised patients [34]. Nevertheless, the clinical relevance of transfusion-associated HEV infection is insufficiently understood and more data are needed regarding the duration of viraemia, the infective dose, the role of anti-HEV in the recipient and the frequency of clinically apparent transfusion-transmitted HEV infection [35].
Our results on the progression of HEV viraemia illuminate at least one of these questions. To our knowledge, neither the length of HEV window periods nor the course of HEV viraemia during window periods in blood donors have been studied so far. The observed high level viraemia during window period infection could represent an underestimated risk of HEV transmission.
Post-donation questionnaires returned by six donors did not reveal a potential source of HEV infection. None of the infected donors had travelled within two months before the HEV-positive donation. The consumption of pork meat was described by five of the six donors. The number of returned questionnaires in our study is too small for a statistically significant analysis. We currently perform routine HEV blood donor screening and ask those with positive results to answer a questionnaire.

Conclusion
We observed a diagnostic gap between the detection of high viral loads and the detection of anti-HEV antibodies, independently of the antibody class (IgA, IgM or IgG), in our cohort of clinically asymptomatic HEVinfected blood donors. The progression of viraemia and anti-HEV immunoglobulins was comparable to symptomatic cases, but a rise in liver-specific enzymes was infrequent in our blood donor cohort. Asymptomatic HEV infection make NAT screening methods necessary to detect infection and avoid transfusion of contaminated blood donations. However, the majority of infections are transmitted via the zoonotic or food-borne route. It is therefore important to focus public health measures both on blood safety and also on other infection routes for patients at risk, including immunosuppressed patients.