1. Introduction
The annual hepatitis E virus (HEV) infection rate is approximately 20 million globally [
1]. The prevalence of HEV infection varies largely across geographical and occupational settings. Previous studies revealed that the HEV antibody seropositivity rate in the general population ranges from 2% to 77.7% in Southeastern Asia and from 2.17% to 52.2% in European countries [
2,
3,
4]. In China, the estimated prevalence of HEV infection in the general population is from 13% to 14.22% in some provinces [
5,
6]. However, a much higher anti-HEV seropositivity rate (38.34%) was reported in the population in Guangdong, China [
7]. Our previous study also demonstrated a relatively high HEV seropositivity rate in both hepatocellular carcinoma (HCC) patients (40.93%) and healthy individuals (41.81%) from Guangdong [
8]. However, there are no comprehensive estimations of HEV seroprevalence in cancer patients or in each common cancer type.
Cancer is a leading cause of death worldwide, and 12–15% of cancer cases are attributable to chronic viral infection [
9,
10]. HEV is both an enveloped and non-enveloped single-stranded RNA virus with a genome size of 7.2 kb [
11]. Up to now, four main HEV genotypes have been identified. HEV genotypes 1 and 2 are transmitted through a fecal–oral route and only infect humans, while HEV genotypes 3 and 4 are able to be spread from animals to human beings [
12]. The main site of HEV replication is the liver as well as extrahepatic organs, including the small intestine, colon, spleen, stomach, and kidney [
11]. Acute or chronic infection can directly damage liver cells, resulting in impaired liver function. Previous studies indicated that HEV replicated in the hepatocyte cytoplasm and, afterwards, was released into the bile and blood [
13]. Moreover, HEV infection may cause liver damage due to immune regulation from natural killer (NK) cells and cytotoxic T cells [
14]. On the other hand, extrahepatic infection can cause neurological, renal, gastrointestinal, and hematological disorders [
15]. Nevertheless, the detailed pathogenesis of hepatitis E remains largely unknown. A previous study revealed that around 26% of cancer patients in eastern China were infected with HEV and that the prevalence of HEV in them was much higher than that in non-cancer patients, suggesting a positive correlation between HEV infection and cancer development, including extrahepatic cancers [
6]. Our previous study also revealed that HEV exposure might abate the detrimental effect of HBV on HCC, indicating that the association between HEV exposure and cancer might be modified by other factors [
8]. Nonetheless, there exists only very limited epidemiological information on the interaction between HEV exposure and other cancer risk factors in cancer populations, including HCC and extrahepatic cancers.
Collectively, we hypothesized that HEV infection was associated with cancer risk in some extrahepatic organs and that the association between HEV and cancer risk can be modified by other significant oncogenic factors. Therefore, we conducted a current, single-center, case-control study to address these important questions.
4. Discussion
In this study, we found a robust association between HEV seroprevalence and gastric cancer risk. We also found that HEV seroprevalence was associated with an elevated risk of cancer development in male patients younger than 45 years old. However, no modification effect between HEV and HBV, HEV and smoking, or HEV and alcohol on cancer development was discovered in our study.
HEV exposure was more commonly observed in the cancer patients as compared to non-cancer patients [
6,
23]. There are a few possible explanations. First, cancer patients are immunocompromised and unable to form a potent immune response against HEV. Second, similar to other viruses, HEV may drive or accelerate carcinogenesis by causing genome instability mediated by persistent inflammation [
24]. Third, aging is regarded as one of the most important cancer risk factors. In our study, we observed that cancer patients were significantly older than non-cancer patients (55.61 vs. 43.15 years) (
Table S1). A U.S. study revealed that increased age was the only factor associated with HEV seropositivity, probably a consequence that resulted from repeated HEV exposure and accumulated antibodies over time [
18]. Collectively, these mechanisms might lead to more HEV exposure prevalence in cancer patients.
In our study, HEV seroprevalence was correlated with a slightly elevated cancer risk in all cancer patients. However, this association was not statistically significant as determined by the sensitivity analyses (
Table S2). Mara et al. demonstrated that HEV was involved in lots of vital cancer pathways related to apoptosis, oxidative stress, proliferation, growth and angiogenesis [
25]. By disrupting the cell signal of tumor suppressor proteins or upregulating the proliferation pathways, HEV promotes cancer development. Moreover, chronic inflammation is a fundamental factor for promoting the oncogenesis process. HEV invasion led to continuous cell death and inflammatory cell infiltration, after which the active immune response was initiated and a series of oncogenic processes were stimulated. Our observation might result from the heterogeneity in cancer types, that is, in other words, HEV plays differential roles in oncogenesis in different cancers. We performed a subgroup analysis to further test whether HEV seropositivity was a risk factor in individual cancer types. Previous research revealed that tissue-specific factors secreted by stromal cells would influence the outcome of immune response to different pathogens in different tissues [
26]. Moreover, age associates differentially with different cancers, which, in turn, are linked to diverse degrees of HEV seropositivity. Therefore, it is highly possible to observe inconsistent degrees of HEV prevalence among different cancer types.
Moreover, the interactions of HEV exposure and other common oncogenic risk factors in cancer were explored. We discovered that age and gender, but not HBV infection, smoking, and alcohol intake, would be the modification factors on the association between HEV and cancer risk. We then stratified patients by age and gender to test their specific effects. Remarkably, we found that HEV seroprevalence was indicated as a significant cancer risk factor in young male patients but not in elderly male patients and all female patients. We speculated that the change of hormonal levels due to aging may contribute to some extent to this gender- and age-related discrepancy. Higher progesterone and estradiol levels were indicated to predispose individuals to HEV exposure according to previous studies [
27]. In addition, estrogens largely enhance immune functions while androgens mainly suppress immune effects [
28]. Thus, young female patients were protected by the hormone when they were infected with HEV. Therefore, we observed an increased risk between HEV seropositivity and cancer incidence in young male but not in young female groups. The endocrine and immune systems change with advancing age, especially among females. A protective effect of estrogen in menopausal women was reduced; a similar trend but not a significant risk effect was observed between HEV exposure and cancer risk among male and female old patients.
Aging imposes a complicated impact on the immune response against virus infection with respect to gender. It is well documented that somatic mutations increase with aging, their patterns vary between the two genders, and the effect on the immune system exerted by aging is divergent between sexes. Meanwhile, growing evidence has suggested that the tumor immune microenvironment has a remarkable impact on the differences in disease development and prognosis between males and females [
29]. These studies directly and indirectly provide theoretic support for the summative effects on the different associations between HEV and cancer risks among different gender and age subgroups. Nevertheless, the potential mechanisms are needed to be further addressed.
Interestingly, HEV infection was not linked to HCC development based on the current study. The liver is the vital organ for HEV residence and replication. Acute and chronic HEV infection directly damages liver cells; therefore, HEV infection has been suspected to be a risk factor of liver cancer. The postulation is supported neither by the current study nor our previous study [
8]. More interestingly, the group analyses found that HEV was statistically significantly associated with an elevated cancer risk in one type of extrahepatic malignancy, gastric cancer. To further evaluate the probability of HEV as a cancer risk factor, we adjusted the analyses by excluding the patients with different diseases successively in sensitivity analyses to eliminate the confounding effects induced by the co-existing comorbidities. Additionally, the analyses revealed that the association between HEV seroprevalence and gastric cancer risk was robust and not modified by other comorbidities.
Although the detection of HEV in the stomach has been reported [
11], the mechanisms of gastric damage caused by HEV infection were largely unknown. Previous animal studies revealed that HEV RNA was present in the small intestine and stomach in rabbits, implying that the fecal–oral transmission of HEV might replicate with a high possibility in the gastro-intestinal tract and trigger a wide range of oncogenesis processes [
30]. In addition, previous investigations indicated that certain host factors such as HSPG, GRP78, and ATP5B were involved in the cell entry of HEV [
31]. Moreover, GRP78 and ATP5B were expressed in normal gastric tissue and were good prognostic factors of gastric cancer according to the HPA (Human Protein Atlas) database and previous reports [
32,
33]. Thus, we have rationale to speculate that HEV infects gastric tissue by binding to receptors such as GRP78 and ATP5B and contributes to gastric cancer development. Nevertheless, further studies are warranted to explore the mechanisms underlying the increased risk of gastric cancer development in individuals infected with HEV.
Finally, we found a relatively weak association between HEV seropositivity and hematological malignancies in comparison with other cancers. Only limited studies have demonstrated that HEV may elevate the mortality and liver-related morbidity in patients with hematological malignancies [
34]. As blood-borne transmission is recognized as a main HEV exposure source in hematological malignancy, we collected the transfusion information of hematological patients. Our results showed that there was no significant difference in the transfusion frequencies between malignant cases and non-malignant cases. In our study, a relatively high positivity rate of anti-HEV antibodies was also observed in the cases with non-malignant hematological disorders. Even though a high HEV seropositivity was reported in blood donors, HEV was not routinely detected currently in China before blood transfusion [
5]. Therefore, blood recipients, especially when they have hematological diseases, are at a high risk of HEV exposure through the route of blood transfusion.
However, our study has several limitations. First, this is a single-center study. The patients included might not be representative of all cancer and non-cancer patients in China since they might experience unique environmental influences and lifestyles. Therefore, future multi-center studies should be performed to validate our findings. Second, for some specific cancers, certain risk factors were not well documented, such as Helicobacter pylori in gastric cancer. Unfortunately, a Helicobacter pylori test was not regularly conducted in hospitalized patients in China due to the health economic concern, and we cannot include the results of Helicobacter pylori in this study. Wang et al. illustrated that the infection of Helicobacter pylori was related to chronic hepatitis B [
35]. However, the association between Helicobacter pylori and HEV was not addressed and should be addressed in the future. Third, the HEV seropositivity rates of non-cancer hospital controls in this study were higher than those in other areas of China [
6]. However, since HEV antibody tests were seldomly conducted in healthy individuals, we could not include enough control samples of healthy populations to further compare the seroprevalence between healthy individuals and patients with non-cancerous diseases. Epidemiology studies should be designed to address HEV-exposed distribution issues based on various geographic areas and population groups in the future. Moreover, similar to other retrospective studies, a temporal sequence of HEV exposure and cancer development could not be confirmed. Therefore, we cannot claim a conclusive cause–effect relationship between HEV exposure and cancer risk. Future prospective studies on this subject may be useful but certainly time consuming, while new epidemiology methods, such as Mendelian randomization, could be alternatives to address this important question.