A Systematic Review of Hepatitis E Virus Infection in Children

A systematic review was conducted of the epidemiology, clinical and laboratory features, and outcome of hepatitis E virus infection in children.

A systematic review was conducted, seeking all literature relevant to the epidemiology, clinical and laboratory features, and outcome of hepatitis E virus (HEV) infection in children. Transmission is thought to be primarily from fecal–oral transmission, with the role of transmission from animal reservoirs not being clear in children. Worldwide, seroprevalence is <10% up to 10 years of age, with the exception of 1 of 5 studies from India and the sole study from Egypt. Seroprevalence increases with age, but it is not clear if it is increasing over time. The clinical presentation of HEV infection has broad similarities to hepatitis A virus (HAV) infection, with most cases being subclinical. However, HEV differs from HAV in that infectivity is lower, perinatal transmission can result in neonatal morbidity and even mortality, and a chronic carrier state exists, accounting for chronic hepatitis in some pediatric solid organ transplant recipients.

hepatitis E, pediatric

Hepatitis E virus (HEV) is a major cause of epidemic water-borne hepatitis in tropical and subtropical countries in areas with poor sanitary conditions. The infection is endemic to southeast and central Asia, the Middle East, and northern and western Africa [1, 2]. In endemic countries, HEV is a major cause of acute sporadic hepatitis as well as epidemic outbreaks of hepatitis related to fecal contamination of drinking water during heavy rainfall or floods [3, 4]. In nonendemic regions such as North America, Western Europe, and developed countries of the Asia-Pacific, HEV is an infrequent cause of acute hepatitis and is usually thought to be acquired during travel to disease-endemic areas, although autochthonous (locally acquired) HEV infection is now increasingly recognized [2].

Hepatitis E virus is a single-stranded RNA virus that is the sole member of the genus Hepevirus in the family Hepeviridae. Of the 4 genotypes of HEV known to cause human disease, genotypes 1 and 2 are restricted to humans, whereas genotypes 3 and 4 have been found in other mammals as well. Genotype 1 is the major cause of both epidemic and sporadic hepatitis in HEV-endemic countries of Asia and Africa, whereas genotype 2 has been associated with HEV outbreaks in Mexico and Western Africa. Genotype 3 has been documented to cause sporadic locally acquired hepatitis in the United States, Europe, Japan, Australia, and New Zealand and has been sequenced from human samples from Bolivia and Argentina. Genotype 4 is known to cause sporadic cases of human hepatitis in Asian countries including China, Taiwan, Vietnam, and Japan. Only a single known serotype of HEV is recognized [5].

HEV is similar to hepatitis A virus (HAV) in that both are single-stranded RNA viruses with a single serotype, predominantly enteric transmission, and similar incubation periods. However, in endemic countries, HAV is commonly acquired in early childhood whereas HEV has its peak incidence in early adulthood [6]. As compared with HAV, HEV infection in adults has been documented to give rise to more severe disease with protracted coagulopathy and cholestasis, and is associated with a higher mortality rate, especially in pregnant women [7, 8] and in those with underlying liver disease such as cirrhosis [9].

Minimal attention has been paid to HEV infection in children. The objective of this study was to summarize the published literature on the epidemiology, clinical and laboratory features, and outcome of HEV infection during childhood.

METHODS

Search Strategy and Selection Criteria

The following databases were searched: Ovid Medline In-Process & Other Non-Indexed Citations, Ovid Medline Daily, and Ovid Medline 1946 to November 2013. Search terms used were as follows: “hepatitis E” limited to “children 0–18” and “hepatitis E” plus “vertical transmission.” Any article that discussed the pathophysiology, incidence, clinical manifestations, treatment, or prevention of HEV infection during childhood was selected. Excluded articles included those on basic principles already addressed by other included studies, articles with inadequate description of children within their study population, or serological studies without data on ages of children included in the study. The initial search yielded 543 articles, of which 196 articles were selected, yielding 86 articles that fit the inclusion criteria, as well as 10 articles acquired by cross-referencing within the articles that fit the inclusion criteria. Because this is a descriptive systematic review, it was not possible to assess the quality of the included studies. Only English-language studies were included.

RESULTS

Transmission and Pathogenesis

Most HEV infections in endemic areas are due to enteral transmission, usually due to fecal contamination of drinking water. Parenteral transmission of HEV through blood transfusion [10] has also been documented, as has HEV transmission from mother to infant [7].

HEV has been found in domestic and wild pigs worldwide. Elevated alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels [11] and increased anti-HEV seropositivity in pig handlers and swine veterinarians compared with controls [12], as well as high sequence homology between swine and human HEV on RNA testing [13], suggest that pigs can act as reservoirs of HEV in nonendemic countries. Food-borne zoonotic transmission of genotype 3 or 4 HEV has been linked to consumption of undercooked pork or wild game [14–16]. Food-borne or contact zoonotic infection from pigs may represent a significant mechanism of acquisition of sporadic HEV infection in countries that are not endemic for HEV.

Serological evidence of HEV has been found in cattle, sheep, rabbits, mongooses, chickens, goats, cats, dogs, and rhesus monkeys, although the virus has not been isolated or sequenced from these animals [12]. Evidence for cross-species transmission of HEV from these species to humans is lacking, with the exception of a single case report from Japan where the pet cat of a 47-year-old with locally acquired HEV infection and no other known source of infection was found to be HEV immunoglobulin G (IgG) positive [17]. In contrast, genotypes 1 and 2 have neither been isolated from animals nor transmitted to animals in experimental studies, suggesting that these genotypes are incapable of crossing species barriers [5].

Seroprevalence

Epidemiological studies from most countries show that HEV seroprevalence (as determined by HEV IgG) tends to be low in early childhood (Table 1) [6, 18–52], increasing with age from <10% in children aged <10 years (with a few exceptions) to up to 76% in adolescents in Egypt [37]. Within countries, HEV seroprevalence varies by region studied, as well as by residence and socioeconomic status. Two studies from India have documented that HEV seroprevalence tends to be lower in rural areas compared with urban areas [19, 20]. One of these studies showed that HEV seroprevalence in urban children varied with socioeconomic groups, being 1.4% in the higher socioeconomic group compared with 6.4% in the lower socioeconomic group in children aged 6–10 years, and 3% in the higher socioeconomic group vs 10.3% in the lower socioeconomic group in children aged 11–15 years. These differences in socioeconomic groups were not seen after the age of 25 years where the highest seropositivity was documented [19].

Table 1.

Seroprevalence of Hepatitis E Virus in Children <20 Years of Age in Selected Studies

Continent and Country	Population Screened	Year of Screening	HEV Seroprevalence by Age Group	Reference
Asia
Taiwan	600 children aged 6–20 (stored sera from vaccine trial, Taipei)	1981–1992	0.3% in 6–10 y, 0.3% in 11–20 y	[18]
Taiwan	2538 children aged 3–6 from kindergartens in Central Taiwan	2004	3.3% in 3–4 y, 3.4% in 5–6 y	[19]
India	201 schoolchildren aged 4–15, Pune	1982	3.5% in 4–10 y, 16% in 11–15 y	[6]
India	246 schoolchildren aged 4–15, Pune	1992	5.6% in 4–10 y, 12% in 11–15 y	[6]
India	2651 urban and rural schoolchildren aged 6–15 y, Pune	1998	3.8% in 6–10 y, 6.3% in 11–15 y	[20]
India	1020 healthy children, population sample, Vellore	1999–2000	1.6% in 1–5 y, 6.0% in 6–15 y	[21]
India	456 asymptomatic children, Mumbai	2003–2004	17.75% in 0–10 y, 24.7% in 11–20 y	[22]
Nepal	64 children aged 12–19	1992–1993	16% in 12–19 y	[23]
Vietnam	270 children aged 0–19 y	1994	3% in 0–4 y, 5% in 5–9 y, 4% in 10–19 y	[24]
Tibet	227 children aged 10–19 y	2001	23.8% in 10–19 y	[25]
Bangladesh	402 healthy children	2003–2004	0.67% in 1–5 y, 7.76% in 6–10 y 11.68% in 11–15 y	[26]
Iran	439 children aged 2–19 y, Sari District	2003	0 in 2–4 y, 1.2% in 5–9 y, 3.8% in 10–19 y	[27]
Iran	235 children aged 6–19 y, Isfahan Province	2005	0.9% in 6–9 y, 2.4% in 10–19 y	[28]
Yemen	220 children (clinic sample)	2005	7.7% in 0–18 y	[29]
Mongolia	520 schoolchildren	2004	0.6% in 7–12 y	[30]
Mongolia	717 children aged 0–20 y from Ulaanbaatar	2005–2006	0 in age 0–6 y, 1.6% in age 7–9 y, 0 in 10–20 y	[31]
Indonesia	453 healthy children aged 1–20	2005	1.3% in 1–10 y, 1.3% in 11–20 y	[32]
Korea	196 children aged 10–14	2005	1% in 10–14 y. 0 in 15–19 y	[33]
China	1566 stored sera from children aged 0–19, eastern China	2006	7.9% in 0–9 y, 10.76% in 10–19 y	[34]
China	>3000 healthy children aged 0–20 from Beijing and 5 provinces	2007	5.38% in 0–5 y, 4.2% in 6–10 y, 9.75% in 11–20 y	[35]
Africa
Ghana	803 children aged 6–18 y	1998	4.4% in 6–18 y	[36]
Sierra Leone	120 children aged 6–12 y	1997	8% in 6–12 y	[37]
Egypt	5112 healthy children aged 0–19 y	2000	36.2% in 0–4 y, 64.7% in 5–9 y, 75.6% in 10–14 y, 75.5% in 15–19	[38]
Europe
United Kingdom	2731 sera from ages 1–85 y collected in 1991 and 2004	1991 & 2004	<5% in ages 1–9 and 10–19	[39]
Greenland	48 Inuit children aged 7–19 y	1994	0%	[40]
Spain	321 children, stored sera, Granada	1995	1.8% in 5–13 y	[41]
Spain	1249 schoolchildren aged 6–15, Catalonia	2001	4.6% in 6–15 y	[42]
Turkey	909 children aged 6 mo to 16 y, Istanbul	1997–1998	4.8% in 6 mo–2 y, 3.1% in 2–5 y, 2.1% in 5–10 y, 0.3% in 10–16 y	[43]
Turkey	210 asymptomatic children aged 0–18 y, Konya	2001–2002	0 in 1–6 y, 6.8% in 7–12 y, 8.9% in 13–18 y	[44]
Turkey	515 school children aged 6–13 y in 2003, Ankara	2003	1.7% in 6–13 y	[45]
Turkey	Same children as above in 2005	2005	2.1% in 6–13 y	[45]
Americas
Mexico	1848 children 1–19 y	1987–1988	1.1% in 1–4 y, 4.4% in 5–14 y, 9.6% in 15–19 y	[46]
United States	18 695 sera from ages ≥6 y, NHANES survey	1988–1994	≤5% in ages 6–11 and 12–19	[47]
Venezuela	99 rural children	1992	1% in 4–15 y	[48]
Venezuela	120 Amerindian children	1992	3.5% in 4–9 y, 8.8% in 10–15 y	[48]
Argentina	1304 sera from children, mean age 6.4 y	1994–1995	0.15% in children of mean age 6.4	[49]
Bolivia	246 children aged 1–20 y, Chaco region	1997	3.9% in 1–5 y, 3.8% in 6–10 y, 5.5% in 11–20 y	[50]
Bolivia	103 children aged 1–20 y, Bartolo and Casas Viejas, southeastern Bolivia	2006	8.7% in 1–5 y, 0% in 6–10 y, 2.4% in 11–20 y	[51]
Cuba	117 healthy children aged 5–20 y	2003	4.2% in 5–11 y, 4.2% in 12–20 y	[52]

Abbreviations: HEV, hepatitis E virus; NHANES, National Health and Nutrition Examination Survey.

Only a few studies have addressed the issue of changing seroprevalence over time in the same population. Two of these, from Pune, India [6], and Ankara, Turkey [44], showed minor or no significant increase in rates of HEV seropositivity over time. A third study, however, from a tribal population of the Andaman and Nicobar Islands, India showed a significant rise in HEV seropositivity, from 13% in 1989 to 40% in 1999 in children <15 years of age. The authors were unable to identify specific factors responsible for this change over time [53].

Egypt, which, along with the Indian subcontinent and parts of China, is considered hyperendemic for HEV, has one of the highest HEV seroprevalences in young children, suggesting that most HEV in Egypt is acquired early in life [37]. HEV is a significant cause of acute sporadic hepatitis in young Egyptian children, causing 12% of acute hepatitis in children aged 1–13 years in one study [54] and 22% of acute hepatitis in children <10 years of age in another [55]. In contrast, a study from India reported that only 0.7% of sporadic hepatitis in children under 10 years of age was due to HEV [21]. It is not clear to what degree widespread asymptomatic HEV infections earlier in life leading to immunity contribute to the fact that large epidemics of enterally transmitted hepatitis have not been demonstrated in Egypt [7, 37].

Infectivity

A Cuban study reported 8% HEV immunoglobulin M (IgM) positivity, 66% HAV IgM positivity, and 26% dual positivity in outbreaks of viral hepatitis in institutionalized children <15 years of age, and postulated that a shorter period of excretion accounts for fewer HEV outbreaks [56]. Secondary attack rates among household contacts of HEV infection appear to be lower than with other fecal–orally transmitted infections, ranging from 0.7% to 2.0% [57, 58].

Clinical Manifestations

Seroprevalence studies indicate that most HEV infections in childhood are asymptomatic, with one study reporting a history of jaundice in only 5.2% of HEV-seropositive individuals of all ages, compared with 10% of those seropositive for HAV [6].

Signs and symptoms of HEV infection do not differ significantly from those seen in other viral hepatitides. The earliest descriptions of clinical manifestations of hepatitis E come from 2 epidemics of enterally transmitted non-A, non-B hepatitis in India in 1956 and 1978 [59, 60]. After an incubation period ranging from 15 to 60 days, the illness presents with a prodromal preicteric phase with symptoms of fever, anorexia, vomiting, abdominal pain, diarrhea, or constipation lasting from 2 to 3 days. This is followed by an icteric phase lasting 10–14 days during which fever subsides and appetite returns to normal. Mild hepatomegaly is often present, as well as elevations of liver enzymes that persist for up to 6 weeks. Cholestatic symptoms with pale stools and itching have been documented in up to 20% of cases during this phase [59]. During these epidemics, the preicteric and icteric stages of HEV infection have also been documented in children.

Nonhepatic manifestations of hepatitis E documented in children include acute pancreatitis [61], immune hemolysis and thrombocytopenia [62], and Henoch-Schonlein purpura [63].

The majority of patients recover without sequelae or chronic hepatitis, although fulminant hepatic failure has been documented, albeit rarely among children <15 years of age [64].

In one small case series of 36 children from India with acute decompensation of chronic liver disease where underlying liver disease was often first diagnosed after an acute hepatic insult, HEV was the cause of hepatitis in 75% of children [65].

Histology

In a case series of 44 children with acute liver failure, HEV infection was diagnosed either singly or with other hepatotropic viruses in 18 children. Nine of these children died, with postmortem liver biopsies showing acute massive hepatic necrosis [66]. Other histopathological findings found on liver biopsies in 2 small case series of adults and children with fulminant and subacute hepatic failure include varying degrees of mixed portal and lobular inflammation, spotty or confluent necrosis, ballooning degeneration of hepatocytes with apoptotic bodies, pseudo-rosette formation and prominent Kupffer cells, bile ductular proliferation, steatosis, and intracytoplasmic and intracanalicular cholestasis [67, 68]. Although liver biopsies are less commonly performed with uncomplicated hepatitis E, biopsies from one small series of adults with acute HEV infection showed features of acute hepatitis with portal and acinar mixed inflammatory cell infiltrate, apoptotic hepatocytes, focal hepatocyte necrosis, and cholangiolitis [69].

Chronic Infection

Recently, chronic infection with persistently elevated aminotransferases and chronic viremia has been documented in small case series of immunosuppressed children who have received solid organ transplants [70, 71]. Chronic HEV infection has also been documented in adults and in case reports of children with conditions associated with immunosuppression such as HIV infection and hematological malignancies [72, 73], although less commonly than in those with solid organ transplants. In one series of 124 children with solid organ transplants, 4 (3.2%) children (2 post–liver transplant and 2 post–renal transplant) were found to be anti-HEV IgG positive. Three of the 4 had no clinical or biochemical evidence of hepatitis. The fourth (post–renal transplant) with evidence of viral shedding of HEV in stool developed chronic hepatitis with persistently elevated liver enzymes and continued to have chronic HEV fecal shedding 24 months after diagnosis [70]. In another series, stored serum samples from children who underwent liver transplant during 1992–2010 at the University of Montreal were tested by HEV serology and RNA by reverse transcriptase polymerase chain reaction (PCR). Among 66 children without chronic hepatitis, 10 (15%) had HEV IgG but none had HEV IgM on follow-up sera. Of 14 children with abnormal transaminases and evidence of chronic hepatitis, 6 of 7 who had been documented as HEV negative before transplant became HEV IgG positive after transplant; 4 of these were also HEV IgM positive on repeated testing after transplant. Results of donor serology were not reported; it remains unclear if HEV was acquired from the donor liver or from exposure after transplant. Hepatitis E virus RNA was detected from only 1 of the 14 children but was found in 5 annual samples starting 10 years after transplant (2 years after the child developed IgM and IgG to HEV) with concomitant development of cirrhosis during that period. Sequencing revealed 2 different strains of HEV, both similar to swine strains of genotype 3 [71].

Chronic HEV infection has also been documented in an adolescent following bone marrow transplant who later went on to develop cirrhosis and portal hypertension. Phylogenetic analysis revealed genotype 3 HEV with high similarity to swine sequences as well as with HEV strains isolated from strawberries in his province of residence (Quebec, Canada). In this child with no history of travel to HEV-endemic areas, the authors postulate that HEV infection may have been acquired by direct or indirect contamination of food through contaminated irrigation systems or fields fertilized with infected swine manure [73].

Chronic HEV infection needs to be considered as one of the differential diagnoses in transplant recipients with chronic hepatitis.

Vertical Transmission and Neonatal Infection

Several studies from developing countries have documented that HEV infection in women carries an increased risk of mortality of up to 20%. Elevated HEV viral load and prolonged HEV viremia have been documented in pregnancy [74, 75], contributing to severe liver injury in the mother as well as transmission of virus from mother to infant [75–80]. Mother-to-child transmission of HEV infection has been documented when mothers have evidence of active HEV infection in the third trimester of pregnancy. The rates of transmission documented range from 30% [76, 77], through 50% [78], 70% [79], and 79% [75] to up to 100% [80] in 6 different small case series. Vertical transmission has been associated with neonatal HEV infection that can present with jaundice at birth and can cause death within the first 48 hours, usually due to severe hypothermia, hypoglycemia, and fulminant hepatic failure (2 of 10 babies and 6 of 15 babies in 2 studies) [75, 80]. Vertical transmission of HEV can also lead to mild hepatitis (icteric or anicteric) in the neonatal period that usually resolves within 8 weeks. No chronic carrier state has been documented in children following vertical transmission of HEV [80].

Diagnosis

Within the first 3 days of clinical illness with jaundice and elevated aminotransferases (ALT), HEV viremia is detectable in 71% of sera and both IgM and IgG to HEV in about 60% ((Figure 1). By the end of the first week when ALT level is maximum, HIV viremia and HEV IgM and IgG are found in up to 90% of sera. Fecal shedding of HEV virus is present up to 4 weeks after onset of illness, being present in half of patients within the first 3 days of illness and in 70% between 8 and 11 days. By 6 weeks from the onset of acute hepatitis E, ALT levels have normalized and HEV viremia has cleared, HEV IgM is found in only a third of patients, and IgG to HEV is present in nearly all patients [81, 82]. IgM to HEV becomes undetectable in most children and adults between 3 and 8 months [83–85], whereas IgG to HEV persists for years [85, 86].

Figure 1.

Changes in markers of acute hepatitis E infection at various time points after onset of symptoms (data derived from references [81–83].) Abbreviations: ALT, alanine aminotransferase; HEV, hepatitis E virus; IgG, immunoglobulin G; IgM, immunoglobulin M.

Although detection of HEV RNA by PCR in blood is more sensitive and specific than serology and confirms the diagnosis of HEV infection, the shorter period of HIV viremia necessitates a combined approach using both IgM and PCR to improve diagnostic sensitivity. Acute HEV infection may be diagnosed in the presence of elevated aminotransferases and either a positive HEV IgM or a positive HEV PCR. A positive isolated HEV IgG in the absence of the other markers would indicate remote infection.

Although several HEV serological assays are available commercially and through research laboratories, wide variability in sensitivity and specificity of both IgM [86–88] and IgG [86, 87, 89, 90] assays, as well as poor concordance between assays testing for different antigenic epitopes [86, 90], means that none are formally approved by national or health authorities such as the US Food and Drug Administration for diagnosis of HEV infection, nor validated for estimation of seroprevalence in the case of IgG assays. Until specific and broadly immunoreactive assays become available, clinicians will have to depend on assays that are available locally or from the Centers for Disease Control and Prevention to diagnose HEV infection using a diagnostic algorithm as mentioned above.

Treatment

Treatment of HEV infection has been mainly supportive up until now, with measures for relief of symptoms and management of liver cell failure when present. However, the recognition of chronic HEV infection in transplant recipients has led to attempts at treatment with pegylated interferon (peginterferon) and ribavirin [91, 92]. In a recent study, 3 patients who developed chronic HEV infection after liver transplant were given peginterferon alfa-2a for 3 months after reduction of dosage of immunosuppressive medication failed to eradicate the virus. In 2 of the 3 patients, a sustained virological response as demonstrated by negative HEV RNA in serum and stool samples was documented 5 and 6 months after cessation of therapy, along with normalization of liver enzymes. These patients continue to be followed up. The third patient relapsed at week 2 after completion of peginterferon therapy [91]. In another study by the same group, 6 patients with renal transplants and chronic HEV infection received ribavirin for 3 months. Four patients had a sustained virological response; 2 patients relapsed at 1 and 2 months, respectively, after ribavirin therapy ended [92]. Neither of these studies included children.

Prevention

Epidemiological evidence that people previously infected with HEV are protected from infection during outbreaks, as well as experimental evidence of protection following passive immunoprophylaxis and challenge in macaques has given rise to the hope of protection through vaccination against hepatitis E [93]. Two genotype 1 candidate vaccines, baculovirus-expressed 56 kDa protein and Escherichia coli–expressed HEV 239 protein have been clinically evaluated. The first vaccine was studied in a randomized placebo-controlled phase 2 trial in 2000 Nepalese Army personnel, and was shown to have a protective efficacy against hepatitis E of 95.5% after 3 doses given at 0, 1, and 6 months [94]. The second vaccine, also given in a 0-, 1-, and 6-month schedule, was studied in a randomized placebo-controlled phase 3 trial of safety and efficacy among 112 604 healthy men and women aged 16–65 years in China. This vaccine was well tolerated, showed a 4-fold or greater increase in antibody response in 98.7% of vaccinees 1 month after the third dose, and was protective against both genotype 1 and 4 HEV infection, with a vaccine efficacy of 95.5% after the first dose and 100% after the second or third dose on follow-up for 1 year compared with the placebo group [95]. This vaccine was approved for use in China, and the first batches of vaccine rolled out in October 2012 [96]. Data on further duration of protection as well as efficacy in younger age groups can be expected in the future.

In conclusion, it seems likely that HEV infection in children is underrecognized. Further characterization of neonatal infection and chronic infection and improved, readily available diagnostic tests will be required to advance the field.

Notes

Author contributions. V. P. V. helped to design the protocol, completed the systematic review, and wrote the manuscript. J. L. R. conceived the study, helped to design the protocol, and reviewed all drafts of the manuscript.

Potential conflicts of interest. Both authors: No reported conflicts.

Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.