Risk factors for hepatitis E virus seropositivity in Dutch blood donors

Hepatitis E has long been considered a disease mainly affecting developing countries and returning travelers [1]. However, over the last decade it became clear that autochthonous infection with hepatitis E virus (HEV) is common in some industrialized countries, and that in Europe the incidence has increased by ten times [2‐4]. The vast majority of HEV-infections in Western Europe are caused by genotype 3, which is known to be a zoonosis with domestic swine, wild boar and deer as the main reservoir [3, 5]. Autochthonous cases may be caused by transmission via food (e.g., consumption of contaminated undercooked meat from pigs, pig-derived blood products used in the food industry, or wild animals) or the environment (e.g., contaminated water or direct contact with infected animals and their feces) [5, 6], but the exact sources and routes of transmission are unknown. Recently Slot et al. demonstrated that the incidence of HEV infection is significantly lower among donors not eating meat compared with meat-eating donors [7]. Moreover, reasons for the current increase of HEV infections in the Netherlands and surrounding countries [4], as observed among young Dutch blood donors [8] and in our national sentinel laboratory surveillance system [9], are currently unclear.

The Netherlands is densely populated by both humans and livestock, including 5.8 million fattening pigs, 1.2 million breeding sows and 5.6 million piglets in 2015 [10]. HEV genotype 3 RNA has been detected in 53% of pig farms, 4% of wild boar feces, and 17% of surface water samples in the Netherlands from 2004 to 2006 [6]. A recent study (2017) showed that residential proximity to pig farms was not associated with increased presence of antibodies against HEV in humans who live in a pig livestock-dense area in the South of the Netherlands [11], which might indicate a primarily food-borne transmission of HEV. Recently in the Netherlands HEV-RNA genotype 3 was detected in a large majority of “liver sausages” (43/55) and “pork liver pâté samples” (12/15) [12], and in 33/36 nonheated liquid pig derived blood products and 7/24 spray dried powder products from pig blood [13]. Other developed countries also reported a broad distribution of HEV-RNA genotype 3 in pork products, among others in liver sausages and dry raw sausages [14‐19]. It is currently unclear whether these pork products contained viable and infectious virus, however the level of HEV-RNA-positive samples among common food products is alarmingly high.

Most human infections by HEV genotype 3 are asymptomatic, or present with mild self-limiting symptoms of hepatitis including jaundice, fever, abdominal pain and fatigue. Most symptomatic cases occur among middle-aged and elderly men [1]. In immunocompromised patients there is a risk of chronic infection which can quickly progress to liver cirrhosis, with substantial morbidity and mortality [1]. Therefore, prevention of HEV infection is especially urgent in these high-risk groups. Increased insight in the epidemiology, risk factors and transmission routes of HEV are essential to be able to formulate and implement food regulatory and public health measures for prevention of hepatitis E infection. In this study, we aimed to identify risk factors for HEV seropositivity.

With this study we identified risk factors for past hepatitis E infection that were primarily related to consumption of (pork) food products and contact with contaminated water. The pork food products “traditional Dutch dry raw sausages (called “cervelaat”, “fijnkost”, “salami” and “salametti”)” and “other dry sausages” were independent risk factors for hepatitis E seropositivity. These dry raw sausages are generally eaten unheated, as they are meant to be consumed sliced on bread. While it is currently unknown whether these products, generally made from raw pork and beef, indeed contain (infectious) HEV, transmission of HEV via raw [20] and processed [21] pork products has been suggested before and is biologically plausible. To completely inactivate HEV it is necessary to heat food to an internal temperature of 70° Celsius for at least 20 min [22]. Production and consumption of raw dry sausages like the Dutch “cervelaat”, “fijnkost”, “salami” and “salametti” mainly takes place in Europe. The sausages can vary slightly between different regions and countries. In general, this type of sausages is produced from uncooked, cured pork belly meat mixed with beef. Regularly, the mix of raw meat is fermented, filled in a porous synthetic casing and is hung out to dry until it is solid enough to cut. Remarkably, the pork food products that we found to be risk factors for anti-HEV IgG seropositivity do not contain pork liver tissue. In our study, pork food products that do contain liver (“liver sausages” and “pork liver pâté”) were only independent risk factors in the univariate model, and were no longer significant risk factors in the multivariate model. However, a correlation between the consumption of pork food products that do contain liver and consumption of traditional Dutch dry raw sausages was found. Usually diaphragm tissue is added to Dutch dry raw sausages, which could lead to the presence of very small amounts of liver in the sausages [23].

Besides the liver, HEV is present as well in blood and muscle tissue of pigs during infection with the virus [24, 25]. Pig meat might contain HEV due to the presence of residual blood (1-2%), however the viral dose is probably low due to the small amount of blood. In addition, pig-derived blood products (e.g. spray-dried plasma, red blood cells, globin and collagen) might be used as additives to enhance color and texture of the sausages [26]. Unfortunately, we could not measure the consumption of pig-derived blood products directly, as it is not possible to query about blood products in the questionnaire. Whether HEV transmission might occur following consumption of products containing pig-derived blood products is still unclear. Recently, HEV RNA was detected in 33/36 porcine blood product samples, which are produced for application in the food industry. According to Boxman et al. the liquid blood products are not heat-treated during production, leaving the HEV detected most likely infectious [13]. A study of Szabo et al. detected HEV RNA in 20% of raw sausages that did not contain liver [19]. Furthermore, it might be possible that traditional Dutch dry raw sausages contain residual infectious virus as they do not have a long period of curing, while pork liver sausages are often briefly cooked during processing. It might be possible that although pork products containing liver contain a high amount of HEV RNA, this is associated with heat-inactivated virus particles that are unable to cause infection. Possibly, the level of HEV RNA in the traditional Dutch dry raw sausages is much lower, but associated with viable virus particles.

We found that eating bovine steak and smoked beef were positively associated with seropositivity. However, participants who reported consumption of bovine steak were more likely to report consumption of traditional Dutch dry raw sausages (61.8% versus 50.4%; OR 1.6 95% CI 1.3-2.0), and consumption of “other dry sausages” (10.2% versus 5.2%; OR 2.1 95% CI 1.4-3.1). Also, participants who reported consumption of smoked beef were more likely to report consumption of traditional Dutch dry raw sausages (65.7% versus 53.3%; OR 1.7 95% CI 1.3-2.1).

Participants who reported to have had contact with contaminated water (i.e., removed water from a stable, worked with a septic tank or had contact with sewage) showed relatively high seroprevalence (46%). This is not an unexpected finding, as it is known that in sewage water, surface water and waste water, HEV RNA sequences can be detected that cluster with sequences found in autochthonous cases, swine and wildlife from the same geographical area [5]. Surprisingly, however, we found an independent negative association between anti-HEV IgG-seropositivity and contact with wild animals and dogs, which is in contrast with previous studies [27]. Participants that reported to go out for dinner more than once per month, or to have eaten raspberries in the last 2 months, were less likely to have had past hepatitis E infection, even after adjusting for age and gender. The level of education was significantly higher among participants who reported raspberry consumption and going out for dinner more than monthly, which were negatively associated with anti-HEV IgG-seropositivity. These negative associations with anti-HEV IgG-seropositivity may possibly be due to a higher level of education of these participants and may indicate a healthy lifestyle in general with less exposure to HEV due to for example less meat consumption and less occupations that involve contact with contaminated water. Indeed, a lower level of education was observed among participants who reported contact with contaminated water, which we identified as a risk factor for anti-HEV IgG-seropositivity.

As most infections go unnoticed, serology studies can provide valuable information on prevalence and incidence of HEV infection over different geographical regions and periods in time. In a recent review among European countries the prevalence of anti-HEV IgG antibodies in blood donors ranged from 1.3% (DIAPRO assay, Italy) to 52% (Wantai assay, France) [3]. Among Dutch blood donors seroprevalence was 27% among donations collected in 2011/2012 (Wantai assay) [28]. Our results of the anti-HEV IgG-seropositivity among blood donors in the Netherlands were very similar to those observed 5 years ago [8], i.e. high (31%), unrelated to gender and increasing with age. In this study, we tried to identify risk factors for anti-HEV IgG seropositivity instead of risk factors for acute infection, to identify risk factors associated with past infection and (eating) habits. Moreover, only 0.076% of 59,474 blood donations collected in 2013/2014 in the Netherlands was PCR positive for HEV-RNA [29]. However, in a group of PCR positive donors we would expect recent consumption of a food product with a particularly high risk for HEV, besides the regular (pork) eating habits. For example, in England and Wales consumption of pork pie and consumption of ham and sausages purchased from a major UK supermarket chain was found to be associated with indigenous infection [21].

Strengths of this study include the relatively large number of participants and the use of a detailed questionnaire specifically designed to assess risk factors for hepatitis E infection. A limitation is the discrepancy in timing between the moment of infection and recall period of the questionnaire. The recall period used to assess risk factors in the questionnaire was set on 2 months, thereby assuming that food consumption and other habits are relatively constant throughout large periods in life. The persistence of anti-HEV IgG antibodies after infection still needs further elucidation, as wide population variation in the extent and timing of anti-HEV IgG antibody decay is reported in literature, in which variation between antibody assays probably plays a large role [30]. In addition, the study population consisted of healthy individuals with relatively few health complaints or underlying diseases, and might therefore be not completely representative for the entire population.

In conclusion, this study identified specific raw pork food products and contact with contaminated water as independent risk factors for past hepatitis E infection. This confirms the general working hypothesis that consumption of pork products is the primary route of transmission in the Netherlands. Raw pork food products that do not contain liver might contain a low amount of infectious HEV, while pork food products that do contain liver might contain high levels of inactivated HEV. There could be some additional transmission of HEV via the environment but our study could show this only in case of intense exposure. Contact with animals seems not to be an important transmission route of HEV for the general population. Further studies are needed to elucidate the role of pig-derived blood products used as additives in food in HEV transmission and it should be determined whether the food products we found as a risk factor indeed contain infectious viral particles. As long as the reasons for the current increase in autochthonous hepatitis E cases in the Netherlands are not understood, it is important for people at risk of chronic hepatitis E infection (e.g., immunocompromised and transplant patients) to avoid eating raw or processed pork products and products that might contain pig-derived blood products.