SEARCH

SEARCH BY CITATION

Keywords:

  • Acute hepatitis;
  • Egypt;
  • genotype;
  • hepatitis E virus;
  • risk factors;
  • sporadic

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

Clin Microbiol Infect 2012; 18: 982–988

Abstract

The aim of the study was to describe the characteristics of acute hepatitis E in Greater Cairo. Patients with acute hepatitis E were identified through a surveillance of acute hepatitis using the following definition: recent (<3 weeks) onset of fever or jaundice, alanine aminotransferase at least three times the upper limit of normal (uln), negative markers for other causes of viral hepatitis and detectable hepatitis E virus (HEV) RNA. Comparison of the liver tests between acute hepatitis E and hepatitis A virus (HAV), case–control analysis (four sex-matched and age-matched (±1 year) HAV controls per case) to explore risk factors and phylogenetic analyses were performed. Of the 17 acute HEV patients identified between 2002 and 2007, 14 were male. Median age was 16 years (interquartile range 13–22). Compared with HAV (n = 68 sex-matched and ±1 year age-matched), HEV patients had higher bilirubin (mean (SD) 10.9 (5.7) uln versus 7.5 (4.4) uln, p 0.05) and aspartate aminotransferase levels (38.6 (27.1) uln versus 18.3 (18.1) uln, p 0.02). Co-infection (hepatitis C virus RNA or hepatitis B surface (HBs) -antigen positive/IgM anti-hepatitis B core (HBc) anitgen negative) was diagnosed in four patients. In univariate matched analysis (17 cases, 68 matched controls), HEV cases were more likely to live in a rural area than HAV controls (matched OR 7.9; 95% CI 2.0–30.4). Of the 16 isolates confirmed as genotype 1, 15 belonged to the same cluster with 94–98.5% identity in the open-reading frame 2 region. Our findings documented the sporadic nature of HEV in Greater Cairo, characterized a large number of Egyptian HEV genotype 1 strains and identified living in a rural area as a potential risk factor for infection.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

Hepatitis E virus (HEV) is transmitted by the faecal–oral route and is responsible for both sporadic infections and outbreaks in developing countries with poor sanitation [1]. In these countries, the prevalence of anti-HEV IgG is high and increases with age, reflecting evidence of HEV infection. However, the overall prevalence remains lower than the prevalence of hepatitis A virus infection (anti-HAV antibody) and is much lower in younger age groups [1].

In rural Egypt the reported prevalence of anti-HEV antibody is among the highest in the world [2,3], starting early in life with >60% of 10-year-old children having detectable antibodies. However, compared with developing countries, where massive outbreaks have been reported such as in Darfur (Sudan) [4] and Uganda [5], outbreaks have not been reported in Egypt. Furthermore, sporadic acute hepatitis caused by HEV is uncommon [6] and none of 34 seroconversions reported in a study conducted in rural Egypt were symptomatic [7]. Isolates of HEV circulating in Egypt belong to genotype 1 and are close to North African isolates but a limited number of isolates have been characterized so far in urban [8] or rural areas [6].

In this context, there is a need to further document the public health burden of hepatitis E in Egypt, in particular, to study the clinical impact, to describe the at-risk population, to explore the modes of transmission, and to see how it may differ from other enteric pathogens such as hepatitis A, as this has implications for the control measures, including future vaccination policies. In this paper, we describe the main clinical and epidemiological characteristics of symptomatic acute hepatitis E in comparison to hepatitis A in Greater Cairo, which comprises both rural and urban areas, and we present the results of the phylogenetic analysis in relation to the epidemiological characteristics.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

Participants’ recruitment and questionnaire

Acute hepatitis patients were identified through a surveillance system of acute hepatitis implemented in April 2002 in two fever hospitals in Cairo, Abassaia and Imbaba Fever Hospitals. These are public hospitals specialized in infectious diseases offering care to the disadvantaged population of Greater Cairo. Until 2008, Greater Cairo included three governorates, Cairo, El Qalyoubia and Giza, the last two being partly rural. Details of the surveillance system have been described elsewhere [9]. Briefly, individuals older than 5 years of age with recent (<3 weeks) fever or jaundice, elevated serum alanine aminotransferase (ALT) at least three times the upper limit of normal were invited to participate. After providing written informed consent (from one of the parents if <18 years of age), they answered orally administered standardized questionnaires covering socio-demographic characteristics, present and past health conditions, and exposure to potential risk factors for viral hepatitis in the 1 to 6 months before onset of illness.

Laboratory testing

A 10-ml venous blood sample was collected. Patients were tested for standard liver functions (ALT, aspartate aminotransferase (AST), bilirubin, alkaline phosphatase) and for the following hepatitis markers: anti-HAV IgM (HAVAB®, M EIA; Abbott Laboratories, Diagnostics Division, Abbott Park, IL, USA), HBs antigen (AUSZYME MONOCLONAL®, third-generation EIA; Abbott Laboratories, Diagnostics Division) and anti-HBc IgM (CORZYME®, M rDNA; Abbott Laboratories, Diagnostics Division). In patients with non-A non-B hepatitis, anti-hepatitis C virus (HCV) antibody and HCV-RNA were assessed serologically (INNOTEST® HCV Ab IV; Innogenetics, Ghent, Belgium) and using PCR (nested reverse transcription PCR by in house assay using 5′-untranslated repeat primers). In patients with positive HCV antibodies and RNA, exacerbation of chronic hepatitis C by other infectious agents was ruled out using serological testing (anti-Epstein–Barr virus (EBV) IgM (ETI-EBV-M reverse P001605; Dia Sorin, Vercelle, Italy), anti-cytomegalovirus (CMV) IgM (AXSYM®_system-CMV-IgM; Abbott Laboratories, Wiesbaden, Delknheim, Germany), and anti-Toxoplasma IgM (AXSYM® system-Toxo-IgM, Abbott Laboratories)). Real-time PCR for HEV-RNA was performed in patients with positive HCV antibodies and RNA as well as those with negative markers for acute hepatitis A, B and C: viral RNA was detected in blood samples by real-time PCR with a hydrolysis probe [10]. Serological analyses were performed with the EIAgen HEV IgG kit and EIAgen HEV IgM kit (Adaltis Ingen, Paris, France) as recommended by the manufacturer.

Definition of acute hepatitis E

Cases were defined as follows: recent (<3 weeks) onset of fever or jaundice, elevated serum ALT at least three times the upper limit of normal, negative IgM anti-HAV, IgM anti-HBc, IgM anti-EBV, CMV IgM, Toxoplasma IgM and detectable serum HEV RNA by reverse transcription PCR.

Phylogenetic analysis

The genotype was determined by sequencing a 189-nucleotide fragment within the open reading frame 2 gene [11]. Phylogenetic trees were created by use of the neighbour-joining method (Kimura two-parameter method). Phylogenetic analyses were performed with genotype information on reference sequences based on the HEV classification proposed by Lu et al. [12].

Statistical analysis

Case–control study  For each case of hepatitis E, four acute hepatitis A patients diagnosed at the same hospitals, in the same year, residing in Greater Cairo and matched on age (±1 year) and sex were randomly selected as ‘control-cases’ (HAV controls).

Laboratory findings (continuous variables) between acute symptomatic HEV and HAV age-matched and sex matched controls were compared using the Wilcoxon signed-rank test. As there were four HAV controls for each HEV case, the mean value of the four HAV controls was used to compute the test.

Matched odds ratios (mORs) and 95% CI were estimated for each potential exposure using conditional logistic regression to account for the matched design. Data were analysed using STATA 10.0 software.

Ethics  Approval for the study was obtained from the Institutional Review Board of the Egyptian Ministry of Population and Health (MoPH) and the Ethics Committee of the National Hepatology and Tropical Medicine Research Institute (NHTMRI, Egypt).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

Between April 2002 and December 2007, 1950 patients with acute hepatitis were recruited from Abassaia and Imbaba Fever Hospitals. Of these, 858, 771 and 318 were diagnosed with acute hepatitis A, B and C, respectively.

Seventeen patients met the case definition for acute hepatitis E. All were positive for anti-HEV IgM and IgG. Of these, almost two-thirds occurred in the last 4 months of each year. Median age was 16 years (interquartile range (IQR) 13–22) and only three were female, aged 6, 9 and 26 years, respectively. Main clinical characteristics and laboratory findings are presented in Table 1. All but two patients were jaundiced and 12 presented with fever. Median serum bilirubin was 164.2 μmol/L (IQR 121.4–218.4) and median ALT was 1496 IU/L (IQR 848–2359). None of the patients presented with thrombocytopenia.

Table 1.   Demographic, clinical characteristics, laboratory findings. Patients with acute hepatitis E, Greater Cairo, April 2002–December 2007
Variable n = 17
  1. Data are numbers (%) of persons or median (inter-quartile range).

  2. auln, upper limit of normal value.

  3. bOne missing value.

Male sex14 (82.3)
Age16 (13–22)
Rural residency8 (47.0)
Clinical manifestations
 Fever12 (70.6)
 Abdominal pain13 (76.5)
 Dark urine16 (94.1)
 Scleral icterus17 (100)
 Jaundice15 (88.2)
Laboratory findings
 Total bilirubin (μmol/L)164.2 (121.4–218.4)
 Total bilirubin (× ulna)9.7 (7.1–12.8)
 Aspartate transaminase (IU/L)1181 (720–1797)
 Aspartate transaminase (× uln)31.9 (19.4–48.6)
 Alanine transaminase (IU/L)1496 (848–2359)
 Alanine transaminase (× uln)40.4 (22.9–63.8)
 Platelets (× 103)b223.5 (172.5–357.5)

Compared with hepatitis A patients (n = 858), those with hepatitis E were slightly older (median age 12 years, IQR 8–18 and 16 years, IQR 13–22 for hepatitis A and hepatitis E, respectively, p 0.03) and more likely to be male (82.3% versus 61.3%, respectively) although the latter difference was not significant (p 0.07). In terms of laboratory findings, bilirubin and AST were higher in HEV than in sex-matched and age-matched control HAV patients (Table 2).

Table 2.   Comparison of the main clinical and laboratory findings between acute symptomatic hepatitis E patients and sex-matched and age-matched hepatitis A controls,Greater Cairo, April 2002–December 2007
VariableHEV cases (n = 17)HAV controls (n = 68)pa
  1. HAV, hepatitis A virus; HEV, hepatitis E virus.

  2. ap value of the Wilcoxon signed-rank test: 17 HEV cases, 17 HAV controls (mean value of the four HAV controls is used). data are median (inter-quartile range).

  3. bTimes the upper limit of normal value.

  4. cOne missing value in HEV cases, three missing values in HAV controls.

Total bilirubin (× ulnb)9.7 (7.1–12.8)6.9 (4.2–10.3)0.05
Aspartate transaminase (× uln)31.9 (19.5–48.6)13.5 (5.1–24.9)0.02
Alanine transaminase (× uln)40.4 (22.9–63.8)19.2 (14.1–41.6)0.09
Platelets (× 103)c223.5 (172.5–357.5)290 (224–380)0.12

Co-infection with HEV was diagnosed in four patients: one tested positive for HCV RNA, anti-HCV antibody and HBs antigen (IgM anti-HBc negative), one had a diagnosis of acute hepatitis C (HCV RNA positive and anti-HCV antibody negative) and two were HBs antigen positive (IgM anti-HBc negative) (Table 4). None of these four patients presented with clinical features of decompensation such as ascites or hepatic encephalopathy.

Exposures

Seventeen case subjects with acute hepatitis E and 68 individually matched control subjects with acute hepatitis A were included in the case–control study. In univariate matched analysis, rural residency was significantly associated with hepatitis E (mOR 7.9; 95% CI 2.0–30.4). HEV cases were more likely to live in a household with more than seven occupants and to have unsanitary toilets than HAV controls, with mOR 6.0 (95% CI 1.4–25.4) and mOR 10.2 (95% CI 1.8–46.2), respectively. Rural residency and unsanitary toilets were highly correlated variables. Living close to an animal shed (inside or outside the house) had an association with hepatitis E but it was not significant (mOR 2.6; 95% CI 0.6–11.9) (Table 3).

Table 3.   Exposures in the 6 months before onset of symptoms in hepatitis E cases and sex-matched and age-matched hepatitis A controls, Greater Cairo, April 2002–December 2007
 HEV cases, n (%) (n = 17)HAV controls, n (%) (n = 68)Matched OR (95% CI)
  1. HAV, hepatitis A virus; HEV, hepatitis E virus.

  2. aOne missing value among HAV controls.

  3. bContact with a jaundiced person: three missing values in HAV controls; one in HEV cases.

  4. cToilets with no built-in flushing mechanism.

Demographic characteristics
 Rural residency8 (47.1)7 (10.3)7.9 (2.0–30.4)
 Travel outside Egypt10
 Travel in Egypt3 (17.6)10 (14.7)1.3 (0.3–6.8)
 Number of persons in the house >7a5 (29.4)4 (6.0)6.0 (1.4–25.4)
 Ratio of household occupants to sleeping rooms >4a3 (17.7)4 (6.0)3.0 (0.7–13.4)
Contact with a jaundiced personb06 (9.2)
Sanitation
 Water outside the house01
 Unsanitary toiletsc7 (41.2)6 (8.8)10.2 (1.8–46.2)
 Consumption of unwashed vegetables12 (70.6)42 (61.8)1.7 (0.4–6.8)
Contact with animals
 Any contact at home or work4 (23.5)9 (13.2)2.2 (0.5–9.6)
 Animal shed (inside or outside the house versus none)3 (17.6)5 (7.3)2.6 (0.6–11.9)
 Rats a nuisance in the house6 (35.3)19 (28.4)1.3 (0.5–3.9)

Phylogenetic analysis

Hepatitis E virus could be sequenced in 16 of the 17 cases and all were genotype 1. All but one of the 16 isolates belonged to the same cluster, sharing from 94% to 98.5% identity in the open-reading frame 2 region. The isolate that did not belong to the cluster was close to those from the Indian sub-continent and corresponded to a patient who reported travel abroad. The 15 other isolates had 93.8–97.9% identity with the two isolates from rural central Egypt (FJ423078-Egypt) [6] but only 89.2– 91.3% identity with an isolate from urban Cairo (AF051352) [8] (Fig. 1). One group of two and one group of three strains with closed genetic identity were identified. The two individuals from the first group (AF00386; AF00415) were diagnosed in November and December 2002 and lived in the same urban district (Table 4). Of the three individuals from the second group, (AF01642; AF03010; AF03037), the last two were diagnosed in September and October 2005 and lived in the same rural district.

image

Figure 1.  Phylogenetic tree of hepatitis E virus (HEV) isolates recovered in Greater Cairo, April 2002 to December 2007. Phylogenetic tree constructed by neighbour-joining based on the open-reading frame 2 region of the HEV genome, 189 nucleotides (a) or 97 nucleotides (b). The phylogenetic analysis includes HEV sequences recovered from Egyptian patients (sequence names are highlighted in grey for the patients of this study) and GenBank reference sequences. HEV sequence names are labelled as follows: name of the sequence, genotype, country of origin and host.

Download figure to PowerPoint

Table 4.   Characteristics and selected exposures of the 16 patients with acute hepatitis E included in the phylogenetic analysis, Greater Cairo, April 2002-December 2007
Sequence id Year Month Age years Sex Co-infection ALT rural Contact animals Travel outside Egypt Other & comments
  1. ALT, aminotransferase (IU/L); EIA, enzyme immunoassay; HBs, hepatitis B surface; HCV, hepatitis C virus.

AF002682002 Aug16MNo2646YesYesNo 
AF002052002 Aug14MNo848YesYesNo 
AF002712002 Sep16MNo2513YesNoNo 
AF003862002 Nov 422MHBs antigen positive1496NoYesNoNatural herb consumption for medical purpose Same district as AF00415
AF004152002 Dec 820MNo1596NoNoNoSame district as AF00386
AF009342003 June24MHBs antigen positive HCV EIA, &PCR positive5300NoNoNoTooth extraction +anesthesia Same governorate as AF01320 and districts next to each other
AF011022003 Sep9FNo1777NoNoNo 
AF012982003 Sep9MNo903NoNoNo 
AF013202003 Oct6MNo625YesYesNoSame governorate as AF00934 and districts next to each other
AF016422004 Jan26FHCV PCR positive, HCV EIA negative500NoNoNoCesarean section. IV injections
AF017382004 Feb21MHBs antigen positive2359YesYesNo 
AF023832004 Sep14MNo2356NoNoNo 
AF025102005 Feb35MNo750NoNoNo 
AF030102005 Sept 1813MNo3199YesNoNoSame city/district as AF03037
AF030372005 Oct 3019MNo442YesNoNoSame city/district as AF03010
AF50172007 Nov22MNo1390NoNoYes 

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

The surveillance system for symptomatic acute viral hepatitis implemented in two fever hospitals in Greater Cairo since mid-2002 using a systematic and well-defined algorithm for the diagnosis of viral hepatitis yielded only 17 cases of acute symptomatic HEV infection over 5 years. This system was not meant to be exhaustive though, because it was limited to two Cairo hospitals, restricted to symptomatic forms of hepatitis, and used an 80% sensitive method (serum detection of HEV RNA) [13]. Its merits have been to confirm the sporadic nature of acute hepatitis E in Greater Cairo during the study period, and would be to detect an outbreak if it were to happen in the large catchment area of the two hospitals. This study brings to light some important epidemiological and clinical characteristics of hepatitis E in Greater Cairo. Most patients were young, one-third of them being <14 years of age, despite the fact that children are less likely than young adults to present with symptoms, as documented in HAV and hepatitis B virus (HBV) infections [14,15]. Acute HEV infection is mostly asymptomatic. In Nepal, a country where HEV is considered to be endemic, the proportion of infections without symptoms estimated in the context of an outbreak [16] or during a cohort survey [17] was ∼70% in young male adults and adolescents. This finding of a young age at infection in Greater Cairo matches earlier observations in rural areas of Egypt where >60% of children aged over 10 years already had anti-HEV IgG [2]. Less than 1% of acute viral hepatitis in Greater Cairo was caused by HEV. These data confirm that most HEV infections take a clinically silent course. The majority of the patients were male, which may reflect either a higher exposure, a higher susceptibility to infection, or a higher risk of developing symptoms once infected. In outbreaks that occurred in developing countries, the male to female ratio is somewhat contrasted, males largely outnumbering females in some outbreaks [18] but not in others [4,19]. Hence, a difference in exposure is a possible hypothesis for the higher proportion of males in our study of sporadic cases.

Regarding the biological characteristics, when compared with HAV-infected patients, and taking into account age (and gender), hepatitis E patients had higher values of AST and bilirubin. These differences were not observed between sporadic HEV and HAV infections in Europe [20,21]. However, these HEV infections concerned much older patients.

Infection with HEV has been recognized as a possible cause of hepatic decompensation in patients with liver cirrhosis in endemic regions [22]. This was not observed in this study, despite four patients having co-infection with HBV or HCV. However, these patients were too young to have already developed a severe liver disease as the result of HCV or HBV infection and one of them had an acute HCV infection. It is of interest that the prevalence of chronic HCV infection (1/17) among hepatitis E patients was not different from that expected among young adults of Greater Cairo [23]. It would suggest that chronic HCV infection does not greatly predispose to symptomatic forms of acute hepatitis E. This may not be the case for chronic HBV infection, whose prevalence (3/17) in this group exceeded the 5% expected among young adults of Greater Cairo. One woman, aged 26 years, had evidence of acute HCV infection (positive HCV RNA without HCV antibodies) following a caesarean section without any blood transfusion reported. The simultaneous occurrence of acute HCV and HEV infections, despite different modes of transmission (although cases of hepatitis E after blood transfusions have been reported [24]), is surprising.

Living in a rural area was associated with a higher risk of hepatitis E. This, together with the absence of waterborne outbreaks reported in Egypt, would indicate contamination from domestic animals raised in close proximity to or within the household. However, contact with animals was not significantly associated with hepatitis E. Stoszek et al. [7] found positive anti-HEV but no HEV genome in domestic animals and rodents in the two villages (Nile delta and upper Egypt) where they conducted a sero-incidence survey among the residents, and HEV genotype 1 has been detected in work horses in old Cairo [25]. However, a zoonotic reservoir and evidence for possible transmission have been documented only for HEV genotype 3 in Europe and genotype 4 in Japan [26]. In developing countries where outbreaks occurred, they were related to consumption of faecally contaminated drinking water. All the 17 patients in this study had access to water inside their houses. According to the 2008 Egyptian Demographic Health Survey, 90.0%/97.4% of the households in rural/urban areas in lower Egypt have access to water piped into the residence [23]. However, access to water from the tap in the dwelling may not prevent episodes of contamination of the water in cases of leaky water pipes passing through soil that is contaminated with sewage [27]. Although episodes of water contamination can provoke outbreaks, they may not be a common cause for sporadic cases. Groups of strains with closed genetic identity clustered in time and place (same district) observed in this study may support common exposures, among which sporadic contamination of water is a possibility.

The proportion of individuals living in a household with a high number of occupants was significantly higher in patients infected with HEV than in HAV controls. This may reflect poor living conditions and the potential for person-to-person transmission. However, in contrast to HAV [14], the role of person-to-person transmission of HEV, either in epidemic or in sporadic context, is still debated [28–30].

This study has several limitations. As a result of the limited sample size for the case–control analysis, we could not perform a multivariable analysis. Because of the nature of the comparison group (HAV cases), associations between exposures and acute hepatitis E should be interpreted with caution. Environmental exposures were not fully explored through the questionnaire.

The present work has characterized a large number of Egyptian HEV strains. Our data indicate that the majority belong to an unclassified subtype, different from the one described by Tsarev et al. [8] and classified as subtype 1e by Lu et al. [12]. Until recently, only genotype 1 strains were found in Egypt in humans [6,8] or in animals [25]. Surprisingly, a genotype 3 virus was recently identified in the stools of a child with acute hepatitis [31].

In conclusion, this study documented the sporadic nature of acute symptomatic hepatitis E in greater Cairo and the early exposure to HEV in life. Serious clinical disease seems to be rare. Our findings suggest the potential role of living in rural areas as a risk factor for HEV infection which deserves to be further explored in studies including an in-depth environmental investigation and combined virological investigations to identify the source of the different HEV strains circulating in Egypt.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

We are indebted to the clinicians and the non-medical staff at Abassaia and Imbaba hospitals for help with patient recruitment.

Transparency Declarations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References

Potential conflict of interest: none. This study was funded by Agence Nationale de Recherche sur le SIDA et les hepatites virales, France (ANRS Grant numbers 1203 and 12122. Symptomatic acute hepatitis cohort in Egypt).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Transparency Declarations
  9. References