Background
Hepatitis E is an acute, self-resolving infectious disease caused by hepatitis E virus (HEV). It is responsible for more than 50 % of adult acute viral hepatitis cases in endemic countries and 1 % in non-endemic countries [
1]. Although most of HEV infections are mild or subclinical, the infection in pregnant women is particularly severe in high endemic countries. It has been reported that a significant proportion of pregnant women with hepatitis E may progress to fulminant hepatitis during epidemics, especially in the third trimester, with a high mortality rate of 15–20 % [
2]. On the other hand, although HEV is mainly transmitted by the fecal-oral route, epidemiological and clinical studies have suggested that vertical transmission of HEV may frequently happen in HEV infected pregnant women and lead to adverse fetal outcomes, such as miscarriage, stillbirth and neonatal death [
3,
4]. Moreover, although scarcely documented, the available data suggest secondary transmission of HEV may occur among family members both in outbreaks and sporadic infections [
5‐
7]. However, fetal outcomes and person-to-person transmission of HEV among household members in China have received limited attention.
Sero-epidemiological studies have showed that the overall seroprevalence of anti-HEV IgG in African and European pregnant women were 11.6–84.3 % and 3.6–29.3 %, respectively [
8‐
10]. In addition, in non-endemic and endemic areas, recent HEV infections could also occur in pregnant women. The positive rate of anti-HEV IgM in pregnant women in Spain and Ghana was 0.67 % [
9] and 18.5 % [
11], respectively. China is an endemic area for HEV. The reported seroprevalence of HEV infection in Chinese pregnant women was 10.24–16.2 % for anti-HEV IgG and was 2.56–3.2 % for anti-HEV IgM [
12,
13]. However, these are cross-sectional studies. In addition, unlike infection with hepatitis A virus (HAV), it is reported that anti-HEV IgG antibodies only persist for months to years after infection [
14] and it is meaningful to study the spontaneous decay of anti-HEV IgG. In the present study, we investigated the prevalence and potential risk factors of HEV infection among pregnant women in Jiangsu Province, and observed anti-HEV evolution in these postpartum women and their children after a 6-year follow-up. In addition, we also observed maternal and fetal outcomes during pregnancy in HEV infected individuals, and explored the possibility of intrafamilial HEV transmission.
Discussion
In the present study, we detected anti-HEV IgM and IgG in a cohort of 497 women at mid-term pregnancy and at 6 years postpartum and also in their 497 children. We found that the positive rate of anti-HEV IgM or the prevalence of anti-HEV IgG was constant in women at the two time points. The unchanged positive rates do not reflect that there was no novel HEV infection occurred during the observation period, but it was resulted from the balance of seronegative conversion due to the natural decay of anti-HEV IgG and the seropositive conversion caused by the novel infections. Additionally, we found that anti-HEV prevalence in Chinese children was extremely low.
Previous studies have indicated that commercially available immunoassays for detection of anti-HEV antibodies differ dramatically in the sensitivity and specificity. The sensitivity could range from 72 to 98 % in detecting the same diagnostic sensitivity panel [
20], and highly discrepant results existed in evaluation of anti-HEV IgG seroprevalence by different assays in the same serum panel, ranging from 4.5 to 29.5 % [
21]. In recent years, antigen derived from ORF2 polypeptide of genotype 4 has been proved to display good diagnostics performance in detecting antibodies against genotypes 1, 3, and 4 of HEV [
22‐
24]. Since HEV genotype 4 is dominant in China [
25], we considered that our ELISA based on the ORF2 of genotype 4 is appropriate to detect anti-HEV in China; the ELISA may largely exclude false-positive results caused by nonspecific binding based on the combination of high reactivity of purified immunodominant ORF2 polypeptide 459–607 and poor reactivity of the truncated polypeptide 472–607 [
16]. Moreover, all anti-HEV positive serum samples in the ELISA were verified by Western blotting, which further reduced the likelihood of false-positive results caused by ELISA. In addition, as an approach for testing the reliability of our assays, all anti-HEV IgM positive samples were also tested to be positive by the highly specific Wantai anti-HEV IgM ELISA kit, indicating good specificity of our two combination assays. Thus, our results would be more reliable to reflect the real situations of HEV infection.
Our results showed that the prevalence of anti-HEV IgG in the second trimester of pregnant women in Jiangsu province was 11.1 %, relatively lower than that (15.5 %) in the pregnant women in Shandong province [
12]. In addition, the positive rates of anti-HEV IgM in the mid-term pregnant women and 6-year follow-up were each 0.6 %, lower than these recently reported in Shandong province (3.2 %) [
12] and Yunnan province (2.56 %) [
13]. The differences in the positive rates of anti-HEV IgG or IgM in different areas of other countries were also documented in the literature. For example, in Japanese blood donors with elevated ALT levels, by detecting anti-HEV IgG with different ELISAs, the positive rates in Honshu and Hokkaido were 1.7 % and 3.2 %, respectively [
26,
27]. One may assume that the variations in the positive rates may reflect the real situations in different areas of the same country. However, except for the factors of different sex and age structure and sampling errors, we considered that the different antigens used in the ELISA assays of detecting anti-HEV IgG may be a more important reason. Fukuda et al. [
26] applied antigen derived from ORF2 protein of HEV genotype 4 and observed a relatively low positive rate, whereas Sakata et al. [
27] applied antigen derived from ORF2 protein of HEV genotype 1 and with a high positive rate. Thus, the relatively lower prevalence of anti-HEV IgG in our study may be attributed to the more specific assay used in the present study. Additionally, unlike previous studies [
12,
28], we did not find significant associations of age and type of residential area (urban/rural) with the prevalence of anti-HEV. It is hard to explain the reason at this time.
Recently, several studies have surveyed changes in anti-HEV IgG prevalence over time [
29‐
31]. However, all those studies enrolled different populations at various time points, leaving the comparison less meaningful. In our present study, through a cross-sectional and a 6-year interval follow-up study in the same population, we observed that the prevalence of anti-HEV IgG was dynamically constant. This is in agreement the notion that the prevalence of anti-HEV IgG remains essentially stable after people reaching a certain age [
32]. Additionally, during the 6-year postpartum follow-up, we observed that 18 women with positive anti-HEV IgG during pregnancy became seronegative, which is conforming the phenomenon of spontaneous decay of anti-HEV IgG antibodies in the years after infection [
14]. According to this serological feature of HEV, we suppose that the constant prevalence of anti-HEV IgG in adults may be resulted from the balance of natural decay of anti-HEV IgG and seroconversion to anti-HEV IgG due to incidental infection. Moreover, for those who were persistently positive for anti-HEV IgG during the pregnancy and at the follow-up, we could not clarify whether it was due to the long lasting of anti-HEV IgG or due to the re-infection in these women since re-infection of HEV may occur in the presence of anti-HEV [
33].
Our results based on the longitudinal observation in the same cohort of women imply that cross-sectional survey may underestimate the prevalence of anti-HEV IgG. Based on the cross-sectional surveys, the positive rate of anti-HEV IgG either during pregnancy or postpartum happened to be equal to 11.1 %. However, the accumulated prevalence at 6 years postpartum should be at least 14.7 %, since we just tested anti-HEV at an interval of 6 years and could not identify the possible cases of anti-HEV seroconversion which then turned seronegative in the study period. It was also reasonable to speculate that some of the women with negative anti-HEV during the mid-trimester actually had past HEV infections and subsequently became seronegative.
Generally, children are especially vulnerable to infectious diseases transmitted via the fecal-oral route. For example, HAV infection usually occurs before the preschool age, reaching a peak at about 6 years old [
34]. HEV, which is also an enterically transmitted virus, however, the infection rate in children is low when compared with adults. For example, the seroprevalence of anti-HEV IgG increased from 1 % in 0–17 years to 18 % in German adults [
21,
35]. In our study, the prevalence rate of anti-HEV IgG in children was 0.4 %, also significantly lower than that in their mothers. We assume that one possible reason is that, compared with the adults, children have extremely low chances to expose to the virus, although they are susceptible to enterically transmitted hepatitis. On the other hand, the reported prevalence of anti-HEV IgG by Jia et al. [
36] in 1–14 years old Chinese children was 6.77 %, which is much higher than our results. The different laboratory methods, just like mentioned above, may be responsible for the variations. Thus, the prevalence of anti-HEV IgG in children merits further investigation by high quality reagents or by several assays in parallel. In addition, person-to-person transmission of HEV has been reported and secondary attack rate among household transmission of HEV was ranging from 0.7 to 2.0 % [
5‐
7]. In our study, none of the 18 children whose mothers experienced postpartum anti-HEV IgG seroconversion was anti-HEV positive, indicating the low risk of intra-family HEV infection.
In the present study, all anti-HEV IgM positive individuals were in normal ALT levels. However, since we did not examine the pregnant women at regular intervals and ALT is not always elevated in acute HEV infections [
37], we could not figure out whether they had subclinical infections or symptomatic hepatitis E before the normalization of ALT levels. In addition, none of tested anti-HEV IgM positive serum samples had detectable HEV RNA. Excluding the low detection efficiency of RT-nested PCR, the transient existence of viremia [
38] or low viral load [
39] in these patients may be the main reasons. Additionally, studies have demonstrated that HEV infection in pregnant women can result in intrauterine growth restriction, abortions, still births or neonatal death [
3,
4]. In the present study, we did not observe any adverse effects on the pregnant and neonatal outcomes in the three pregnant women with new infections. It may be due to the small number of cases or due to the possibility of less severe conditions of these pregnant women.
One limitation in the present study is that we did not comprehensively evaluate the risk factors for infection of HEV in the study subjects, leaving the possible reasons for the novel infections happened in postpartum period unresolved. Second, one woman who was positive for both anti-HEV IgM and IgG over 6 years had normal ALT levels and had no detectable HEV RNA. Whether she was chronically infected or had unusual profile of anti-HEV response or had the possibility of repeated infection remains to be elusive. Third, the lack of HEV RNA detection makes it impossible to analyze the genotypes of sporadic HEV infections in Jiangsu, China.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
Y-HZ, YH, GG and HH conceived and designed the study. GG, HH, LZ and YB performed the experiments. HH and Y-HZ analyzed the data. GG, HH, LZ, YB, YH and Y-HZ contributed reagents/materials/analysis tools. GG, HH and Y-HZ wrote the paper. All authors read and approved the final manuscript.