Clinical and epidemiological aspects of a hepatitis E outbreak in Bangui, Central African Republic

Hepatitis E virus (HEV), a small, single-stranded, hepatotropic RNA virus, currently classified in the genus Hepevirus [1], is responsible for enteric non-A non-B hepatitis in humans. Infection with HEV, thought to spread via the faecal-oral route, causes outbreaks that have been linked to waterborne sources in developing countries; sporadic cases have also been seen [2]. Although the potential for zoonotic and cross-species transmission has been demonstrated [3, 4], asymptomatic carriers of HEV have been reported, and such cases are potential human reservoirs of the virus [5]. During epidemics, the person-to-person transmission rate appears to be low, although intrafamilial transmission is possible [6]. The course of the disease is generally self-limiting; however, chronic HEV infection has been reported in immunocompromised patients [7‐10]. Reports have indicated an increased risk for disease and as high as 10-20% case mortality from fulminant hepatitis during the third trimester of pregnancy [11].

Outbreaks of hepatitis E frequently occur in tropical Africa during the rainy season due to overflowing drains, short-circuiting of networks of clean water and use of contaminated water from wells [12‐14]. The Central African Republic (CAR), located in tropical Africa, is considered to be an area of high endemicity for the main infectious diseases, including infections with HIV [15], hepatitis B virus and other hepatotropic viruses [16, 17], yellow fever [18], malaria [19], tuberculosis [20] and other infections [21‐23]. No HEV epidemics were documented in the CAR before 2001, although HEV antibodies were detected in 24% of young sexually active adults [24]. The first outbreak of this disease in the CAR was reported in 2001 [25, 26], and a further outbreak occurred in 2004.

In order to improve the diagnosis of hepatitis E in CAR, the aim of this study was to determine the clinical and epidemiological characteristics of HEV infection during the outbreak in Bangui.

Table 1 shows the distribution of HEV serology (IgM and IgG) according to age. The age group 18-34 years were more frequently IgM-HEV positive than the other age groups (1-17 and > 34 years) (p < 0.001).

Amplification by RT-PCR of 127 serum and 25 stool samples positive for IgM-HEV antibody showed that 50 sera and 24 stool samples were positive, indicating that the viral genome was present.

More than 80% of patients with positive IgM anti-HEV antibody had higher ALT (82.1%) and AST (86.8%) values than normal.

The clinical features observed in patients positive for IgM-HEV are listed in Table 2. Although no specific symptoms were reported during the outbreak, jaundice was observed in most patients (93.4%), followed by vomiting (50.7%), hepatalgia (47.4%), asthenia (26.8%) and physical signs like hepatomegaly (30.9%) and distended abdomen (14.5%).

To determine how long the infection persisted in the population, variations in IgM HEV positivity were monitored for 16 months, from June 2004, 1 month after the beginning of the outbreak, to September 2005 (Figure 1). The number of cases with IgM-HEV peaked in July 2004, September 2004 and March 2005. Between October 2004 and February 2005, between the two outbreaks, the level of infection remained constant at the background level.

In order to define the HEV-infected area, we collected blood and stools in different health care centres in the eight districts of Bangui, each of which has at least one public health care centre. As shown in Figure 2, IgM-HEV was detected in patients from almost all districts of Bangui; however, more patients in the fourth and seventh districts presented for medical examination, and higher percentages of HEV infection were found in the seventh and eighth districts.

During the 2004 outbreak of hepatitis E, more than 51% of patients were infected with HEV. The most frequent clinical and biological signs were jaundice, vomiting, hepatalgia, hepatomegaly, asthenia, distended abdomen, fever and high levels of transaminases. A possible limitation of this study is the wide inclusion criteria, as physicians at the health care centres could include all patients who presented with jaundice or uncomplicated malaria. It is possible that this inclusion criterion biased our study, because other patients might not have been included or did not enter the defined categories. Nevertheless, the rate of IgM-HEV obtained was very high and was in accordance with that previously reported; e.g. 42% of patients were infected at the time of an epidemic of hepatitis E in Namibia [14].

The high prevalence of jaundice among our patients may be explained by the fact that CAR patients are more aware of this sign, because it is usually related to yellow fever and thus requires urgent medical examination. A study conducted in Pakistan [27] also showed that jaundice was present in almost all patients with HEV IgM positive serology. This sign was also found in 10 IgM-HEV-positive patients in Nigeria [28]. Our study shows, however, that the presence of jaundice cannot confirm HEV infection, and vomiting, hepatalgia, asthenia, fever, elevated ALT levels and hepatomegaly, distended abdomen and fever must also be present [29]. As previously reported [30], we also found that adolescents and young adults were preferentially infected by HEV during the epidemic. This observation suggests that active people are more readily in contact with the virus, especially in tropical developing countries. In these countries, many people are farmers, who may be a source of contamination by HEV, because poultry and swine farmers, other professionals with an occupational activity related to farming and veterinarians are occupational risk groups and are more heavily exposed to HEV than other professions [31‐35].

Biochemical determination of hepatic cytolysis by the detection of high levels of transaminases is an important parameter in the control of liver disease. In our study, the level of transaminases was four to five times higher than normal in more than three quarters of the infected patients. This confirms the presence of hepatic cytolysis, which usually accompanies infection with hepatitis viruses [36]. RT-PCR was not positive in all cases with IgM HEV, confirming that detection of the viral genome is generally difficult after the beginning of symptoms during epidemics of hepatitis E [37]. Our results could not formally link HEV infection to transaminases, because of the prevalence of many other infectious diseases, including other types of hepatitis, yellow fever and malaria, which may also increase hepatic enzymes [38].

The main inclusions occurred in July and September 2004 and March 2005, which correspond to the peaks of infection and were linked to positive IgM-HEV serology. In Bangui, the rainy season starts in May and finishes in December, and a recrudescence of cases of acute hepatitis E is observed during that period in tropical countries, due to overflowing drains and short-circuiting of networks of clean water and structures for purifying wastewater. The level of endemicity results in high IgG anti-HEV seroprevalences. As there is generally a single source of contamination, generally from water, epidemics of hepatitis E are characterized by spectacular numbers of infected people [37, 39, 40]. The lack of samples during August was probably related to the absence of health professionals, who take their annual vacation at this time. The outbreak may therefore have peaked during August, and the peak observed in September might be only the downward part of the curve of infection. The period between October 2004 and February 2005 represents the inter-epidemic season. Asymptomatic excretion of HEV particles in the stools and the environmental reservoir of the HEV may explain the background noise observed and circulation of the virus.

Most patients came from the fourth and seventh districts of Bangui, but more cases of HEV infection were detected in patients living in the seventh and eighth districts. As the disease is related to exposure to faeces, hepatitis E is endemo-epidemic in areas where collective hygiene is badly monitored [3]. During epidemics, ingestion of viral particles by consumption of contaminated water is the principal mode of transmission, and more rarely by food soiled by human excreta. The seventh and eighth districts often have problems of overflowing drains and short-circuiting of networks of clean water and of structures to purify wastewater. Latrines are built close to wells, and water is usually used without treatment. All these insalubrities might explain the presence of the virus and the high rate of infection of the population in these two districts. Further, larger studies are necessary to determine the distribution of HEV in the districts of Bangui and to investigate the sources of the frequent outbreaks of HEV in this city since the first outbreak described in the suburbs in 2001 [25, 26].

This study shows that there are no specific clinical or physical symptoms that might allow clinicians to diagnose hepatitis E. The clinical signs, dominated by jaundice, vomiting, hepatalgia and asthenia, are not specific to HEV infection, but could orient the clinician in the event of an association with an evocative epidemiological context. In order to avoid repetitive outbreaks and epidemics of hepatitis E in Bangui, it would be desirable to reinforce the preventive measures of individual and collective hygiene and to set up a monitoring system for this pathology.