Journal of Medical Virology 34:232-236 (1991)

Outbreak of Acute Hepatitis E Virus Infection Among Military Personnel in Northern Ethiopia ~

Edemariam Tsega, K. Krawczynski, B.-G. Hansson, E. Nordenfelt, Y. Negusse, W. Alemu, and Y. Bahru Department of Internal Medicine, Faculty of Medicine, Addis Ababa University (E.T.,Y.N.) and Department of Epidemiology, Ministry of Health, Addis Ababa (W.A.) and Department of Medicine, Mekane Hiwot Hospital, Asmara (Y.B.),Ethiopia; and Division of Viral Diseases, Centers for Disease Control, Atlanta, Georgia (K.K.); Department of Medical Microbiology, Section of Virology, Malmo General Hospital, University of Lund, Lund, Sweden (B.G.H.. E N . ) An outbreak of acute hepatitis E virus (HEV) infection occurred from October 1988 to March 1989 in military camps in northern Ethiopia. The epidemic was waterborne and entirely confined to military men, of whom 423 hospitalized, icteric patients were studied. The clinical course was mild and short, without any fulminant hepatitis or death. All sera tested for anti-HAV-lgM were negative and among 54 (13%) patients who were positive for HBsAg, 7 (2%) were positive for anti-HBc IgM. On the other hand, 28 of 30 (93%) patients had antibodies against hepatitis E virus (anti-HEV) in contrast to 1 of 29 (3%) asymptomatic controls ( P < .01). The need for an easily available, inexpensive serologic test for HEV infection, protection of water supplies from fecal contamination, adequate chlorination and/or boiling of drinking water, and health education about personal and environmental hygiene, especially in communities at high risk, is emphasized.

ET-NANBH has been noted t o be a unique viral entity and was renamed as hepatitis E virus (HEV) [Reyes et al., 19901. A large outbreak of acute hepatitis E virus infection is reported among military personnel in northern Ethiopia between October 1988 and March 1989.

PATIENTS AND METHODS Background of the Epidemic According to the Regional Health Department (located in Asmara, the capital city of Eritrea), acute viral hepatitis is endemic in the province of Eritrea with epidemic peaks occurring every 2-3 years, the last major epidemic being noted in 1985-1986. The epidemic reported now was first recognized in October 1988 in some military barracks in the Keren district. The outbreak may have started earlier as it was difficult for the Regional Health Department t o differentiate between the last phase of a malaria epidemic and the beginning of this acute viral hepatitis outbreak. A team of health workers, including a physician and a sanitarian, visited the affected areas in January KEY WORDS: hepatitis E, epidemic, Ethiopia 1989, with the aims of investigating the cause(s) of the epidemic of jaundice, of determining the mode of transmission, of collecting information on the number of INTRODUCTION cases affected and of taking appropriate measures for A number of epidemics of enterically transmitted controlling the outbreak. After visiting hospitals, clinnon-A, non-B hepatitis (ET-NANBH) have been re- ics, and different military camps, it was concluded that ported from Southeast Asia [Hillis et al., 1973; Kane et the epidemic was widespread but confined to military al., 1984; Ramalingaswami and Purcell, 19881, the personnel located in Keren, Asmara, Massawa, and Indian subcontinent [Sreenivasan et al., 1978; Wong et Seraye districts (see Fig. 1). There were frequent al., 1980; Khuroo, 19801, Africa [Balabbes et al., 1985; military movements between these places. Field visits Pillot et al., 19871, Mexico [Public Health Service, included inspection of sources of drinking water, food Centers for Disease Control, 1987133,and Central Amer- handling, and sanitary conditions in and around miliica [Villarejos et al., 19751. Except for imported cases tary camps, as well as collection of water samples for [DeCock et al., 19871, ET-NANBH has not been recognized in North America, Australia, and Europe. In 1985-1986, large outbreaks of ET-NANBH have been Accepted for publication March 28, 1991. documented in Ethiopians residing in refugee camps in E. Tsega is now at Montreal General Hospital, Division of eastern Sudan and northwest Somalia [Public Health Gastroenterology, 1650 Cedar Avenue, Montreal, QC H3G 1A4, Service, Centers for Disease Control, 1987al. Recently, Canada. Address reprint requests there. ~

0 1991 WILEY-LISS, INC.

Outbreak of HEV in Northern Ethiopia

233

I

Keren 0 Seraye

r

Massawa

Asmara

ERITREA .

e .w

Fig. 1. Miniature map of Ethiopia and an enlarged section showing affected regions in the province of Eritrea.

biological and chemical analyses (from Hamelmal, Gizgiza, Felfel, Awgaro, and Semen Keren localities). Routine cultures and chemical analysis of the water samples revealed significant numbers of coliform bacteria, ammonia, and nitrates, all indicative of fecal (human and/or other animal) contamination. Assuming that the hepatitis epidemic was transmitted by the fecal-oral route, appropriate health education was given to those living in the affected area. As a likely result of this intervention, the epidemic was on the decline by the end of February 1989. However, the exact cause of the hepatitis could not be determined and new patients continued to appear.

chronic liver disease, severe infection and/or the use of hepatotoxic drugs, and serum transaminase levels greater than 2.5 times the upper limits of normal. Hepatomegaly was diagnosed when the vertical span of the liver along the right mid-clavicular line (determined by palpation and percussion) was greater than 14cm. The biggest civilian hospital in Asmara, Mekane Hiwot Hospital, was selected for this study because most military patients with jaundice and other medical problems were referred there. After interview and physical examination, blood samples were taken (end of March 1989) from every patient for haemoparasites, serum glutamic oxaloacetic transaminase (SGOT), serum glutamic pyruvate Clinical Investigation of Patients transaminase (SGPT), serum alkaline phosphatase, By the third week of March 1989, a physician (E.T.) serum bilirubin, and hepatitis immunological markers and an epidemiologist (W.A.) were sent to Eritrea by (IgM-specific antibody to hepatitis A virus [anti-HAVthe Ministry of Health, Addis Ababa, to evaluate the IgMl and to hepatitis B core antigen [anti-HBc-IgMl, clinical and epidemiological aspects of the outbreak of hepatitis B surface antigen [HBsAgl, antibody to hepacute viral hepatitis. A protocol dealing with social atitis C virus [anti-HCV], and antibody to hepatitis E data (age, sex, religion, marital status, family size, virus antigen [anti-HEVI). Stool specimens collected income, and place of abode at the time of illness), from patients who had had jaundice for less than a demographic and clinical data (history of previous week and water samples taken from sources suspected jaundice, contacts, pregnancy, injections, blood trans- to be contaminated, both in a phosphate buffer at pH fusion, intravenous drug abuse, vaccination during the 7.2-7.4, could not be examined for viral particles as 6 months prior to illness, dental extraction, cultural facilities were not available. The sera for hepatitis practices such as tonsillectomy, uvulectomy, circumci- virus markers were kept in a deep freezer a t -2O"C, sion, tattooing and ear piercing, nutritional status, and until shipped by air in ice to the Virology Section of the signs and symptoms) was prepared and completed by Department of Medical Microbiology, Malmo General the attending physicians for hospitalized patients ful- Hospital, Lund University, Sweden. filling the following selection criteria: icterus with Blood Smears for Haemoparasites onset within the past 4 weeks, blood films negative for Giemsa and methylene blue stains were used to look haemoparasites (Faleiparurn malaria and Borrelia recurrentis or louse-borne relapsing fever), absence of for malaria parasites and Borrelia species.

Tsega et al.

234

Symptoms and signs

TABLE I. Frequency of Symptoms and Signs No. of patients Frequency (%) of seen and examined symptoms and signs

Symptoms Malaise Anorexia Nausea Vomiting Fever Headache Abdominal pain Pruritis Arthralgia Signs Icterus Abdominal tendernessa Hepatomegaly Splenomegaly Ascites Dermatitis (rash) Urticaria Arthritis Angioneurotic oedema Dark-coloured urine Clay-coloured stool

423 423 420 395 408 403 417 421 421

99.5 99.8 99.5 93.6 97.0 96.3 98.6 13.6 7.4

423 41 1 409 410 409 421 420 422 421 420 412

100.0 82.0 9.7 4.7 0.2 4.6 0.0 0.0 0.0 94.4 36.1

aRight upper quadrant and epigastrium.

Biochemical and Serological Tests Serum alkaline phosphatase, SGOT and SGPT were quoted in terms of mill-international unit or unitsiml (Boehringer-Mannheimer and E. March) with upper limits of normal: 45,40, and 35, respectively. Both total and direct bilirubin were determined by standard methods. Commercial radioimmunoassay (RIA) kits (Abbot Laboratories) were used to test anti-HAV-IgM, HBsAg, and anti-HBc-IgM. Antibodies to HCV were determined by the “Ortho-HCV” enzyme-linked immunosorbent assay (ELISA) (Ortho Diagnostics, Raritan, NJ). Sera from 30 patients selected randomly and 29 asymptomatic controls (who were also military personnel selected randomly from the same areas and of comparable age group) were tested for antibodies to HEV antigen by a fluorescent antibody blocking assay [Krawczynski and Bradley, 1989; Krawczynski et al., 19911.

RESULTS The Outbreak Estimation of the number of people affected depended mainly on records of icteric cases admitted to hospitals. Subclinical and asymptomatic cases could not be identified easily by health personnel. Well over 750 patients were seen at Mekane Hiwot Hospital in about 3 months. Considering all the other hospitals in Eritrea, all the possible subclinical and asymptomatic cases, and all the areas affected, the outbreak must have been large and widespread. After exclusion of 12 patients with Falciparum malaria in their blood films, the total number of 423 cases was studied.

Demographic and Clinical Data The mean age of the study group was 25.6 years, with a range from 15 to 56, and 81% of the patients were between 18 and 30 years of age. There was no female patient. The patients were from all parts of Ethiopia. Eighty percent were Christians and the rest Muslims. Sixty-seven percent of the patients were single, 33% were married, the mean family size of each was 5.6 and the mean monthly income was 70.00 Birr (2.05 Birr = U.S.$l;some of the patients, the militia, did not earn a regular salary). A history of jaundice prior to the onset of the current illness was elicited in 84/409 (21%)cases, twice in 12 patients, and thrice in 5. Only 951387 (25%) patients gave a history of contact with jaundiced patients before admission to hospital. Malaria chemoprophylaxis, chloroquine phosphate 500 m g weekly, was distributed to all military personnel, but compliance was not complete. No hepatotoxic drugs, mass vaccination, or immunoglobulin injections had been taken by the patients during the 6 months prior to the onset of the present illness. The symptoms and signs are summarized in Table I. The average hospital stay was about 3 weeks (range 2 - 4 weeks). Neither fulminant hepatitis nor death was observed in the group studied. Long-term follow-up was impossible. Blood Smears and Biochemical Studies Twelve patients had F . malaria and were excluded from the study. The mean serum alkaline phosphatase, SGOT, SGPT, total bilirubin, and direct bilirubin were 98.8 k/ml, 266.2 F/ml, 128.0 p m l , 5.8 mg%, and 3.0 mg%,

Outbreak of HEV in Northern Ethiopia respectively. The highest total and direct serum bilirubin levels were 18 and 12 respectively.

Markers of Hepatitis Virus Of 283 patients whose sera were selected randomly for anti-HAV-IgM testing, none were positive. Fiftyfour of the 423 (13%) patients were positive for HBsAg but only 7 of these were positive for anti-HBc-IgM. Also, 20 of 139 (14%) patients were positive for antiHCV. Twenty-eight of the 30 (93%) randomly collected sera from the patients were positive for anti-HEV as compared to only 1 of the 29 (3%) normal controls (chi-square = 47.6680; P < .01) (Table 2). DISCUSSION The epidemic nature, the clinical features, the contaminated and untreated water sources with crowding around them and the decline of the epidemic after the implementation of appropriate health education are strongly in favour of a waterborne acute viral hepatitis infection with a common aetiologic agent. Previous population surveys have indicated that the prevalence of anti-HAV-IgG was 100% and that HBV markers were noted in about 80% of adult Ethiopians [Tsega et al., 1986, 1987b, 19901, a situation which strongly suggests t ha t any epidemic of viral hepatitis in adults is likely to be due to non-A, non-B virus(es). The fact that all of the patients studied were negative for anti-HAV-IgM corroborates this prediction. Antibodies to cytomegalovirus and Epstein-Barr virus are known to be present in over 80% of adult Ethiopians [Tsega et al., 198713; Mengesha et al., 19881 and, therefore, epidemic jaundice is unlikely to be due to these viruses. Only 7 patients were positive for anti-HBc-IgM, suggestive of acute HBV infection. The 11% HBsAg chronic carrier state is similar to previous population survey results [Tsega et al., 1986, 1987al. Thus, the contribution of acute HBV infection toward this epidemic is negligible. All but two sera from 30 patients were positive for anti-HEV, indicating th at HEV infection was the most likely cause of this and other waterborne epidemic outbreaks i n a population already immune to HAV infection. The single anti-HEV positive case in the control group probably represented past exposure to sporadic AVH [Khuroo et al., 19831. The 14% anti-HCV positivity is indicative of past infection as antibodies to HCV become detectable 4-6 months after the onset of acute infection [Alter and Sampliner, 19891. Over 90% of the patients had malaise, anorexia, nausea, vomiting, fever, headache, abdominal pain, and dark-coloured urine (see Table I). This rate is significantly higher than that documented in similar epidemics [Khuroo, 1980; Chakraporty et al., 19821. The difference may be due to selection of the very sick patients for hospitalization, although exaggerated responses from those brought from the war front cannot be ruled out. The low prevalence of hepatomegaly (10%) is most likely due to the strict definition. I t is difficult

235

TABLE 11. Frequency of Hepatitis Virus Markers Markers Anti-HAV-IgM HBsAg Anti-HBc-IgM Anti-HEV a) Patients b) Controls Anti-HCV

Total tested

Positive

283 423 423

0 (0) 54 (13%)a 7 (2%)

30 29 139

28 (93%) 1 (3%)’ 20 (14%)

a47 (11.1%)are chronic camers and the remaining 7 (1.7%)repre sent acute infection.

to ascribe the splenomegaly to viral hepatitis in such an area of endemic malaria. The presence of icterus, pruritis (14%), and clay-coloured stools (36%) suggests that a significant number of patients had cholestasis. Also, the presence of skin rash and arthralgia is worth noting a s a possible immune-complex manifestation; however, a n immunological study for confirmation is necessary. Although long-term follow-up was not possible, the early recovery, and the absence of fulminant hepatitis and death during hospitalization are indicative of mild disease. This was also observed by others who noted this disease to be even milder than acute HAV infection [Khuroo, 1980; Chakraporty et al., 1982; Molinie et al., 19881. However, it should be emphasized that HEV infection is associated with high mortality in pregnant women and infants during the perinatal period [Khuroo et al., 1981; Kane et al., 19841. There are indications that ET-NANBH epidemics occur frequently in Ethiopia where crowding and poor hygiene prevail [De Cock e t al., 19871. A clinical study of viral hepatitis in pregnant women [Tsega, 19761 and the aetiology of sporadic acute viral hepatitis suggest that HEV infection may account for high morbidity and mortality among pregnant women and infants during the perinatal period. While there is a n urgent need for a n easily available and inexpensive serological test for the diagnosis of HEV infection, it is more important to protect water supplies from fecal contamination, to chlorinate adequately or boil drinking water, and to repeatedly give health education about personal and environmental hygiene, especially to communities a t high risk, e.g., people in refugee camps, military camps, etc. Although immune globulin prepared from local donors is said to be protective [Joshi et al., 19851, this approach appears impractical. However, as the virus is now identified and cloned, mass active immunization, preferably as part of a polyvalent vaccine, which may include HAV, HBV, and HCV vaccines, may be the ideal approach in developing countries where these viruses are hyperendemic.

ACKNOWLEDGMENTS This study was supported financially by a grant from the Swedish Agency for Research Cooperation with Developing Countries (SAREC) through the Science and Technology Commission of Ethiopia and the Re-

236

Tsega et al.

Krawczynski K, Bradley D, Ajdukiewicz A, et al. (1991): Virus associated antigen and antibody of epidemic non-A, non-B hepatitis: serology of outbreaks and sporadic cases. In Shikata T. Purcell RH, Uchida T (eds): “Viral hepatitis C, D, and E.” Amsterdam: Elsevier Science Publishers, pp 229-236. Mengesha B, Tsega E, Nordenfelt E, Hansson B-G, Lindberg J (19881: Sero-epidemiologicalsurvey of cytomegalovirus infection in Ethiopia. Ethiopian Medical Journal 26:179-184. Molinie C, Saliou P, Roue R, Denee JM, Farret 0,Vergeau B, Vindrios J (1988): Acute epidemic non-A, non-B hepatitis: A clinical study of 38 cases in Chad. In “Viral Hepatitis and Liver Disease.” New York: Alan R. Liss, Inc., pp 154-157. Pillot J, Sharma MD, Lazizi Y, et al. (1987): Immunological characterization of a viral agent involved in epidemic and sporadic non-A, non-B hepatitis. Annals of the Institute Pasteur Virology 138:145158. Public Health Service Centers for Disease Control (1987a): Enterically transmitted non-A, non-B hepatitis-East Africa (19871. Morbidity Mortality Weekly Report 36:241-244. Public Health Service, Centers for Disease Control (1987b): EnteriREFERENCES cally transmitted non-A, non-B hepatitis-Mexico (1987).Morbidity Mortality Weekly Report 36:597-602. Alter MJ, Sampliner RE (1989): Hepatitis C-And miles to go before we sleep. New England Journal of Medicine 321:1538, 1539. Ramalingaswami V, Purcell R (1988): Waterborne non-A, non-B hepatitis. Lancet 1:571-573. Balabbes H, Bongeumouth A, Benatallah A, Illoul G (1985):Epidemic non-A, non-B viral hepatitis in Algeria: strong evidence for its Reyes GR, Purdy MA, Kim JP, Luk K-C, Young LM, Fry KE, Bradley spreading by water. Journal of Medical Virology 16:257-263. DW (1990): Isolation of a cDNA from the virus responsible for Chakraporty S, Datta M, Pasha ST,Kumar S (1982): Non-A, non-B enterically transmitted non-A, non-B hepatitis. Science 247:1335viral hepatitis: A common-source outbreak traced to sewage 1339. contamination of drinking water. Journal of Communicable Dis- Sreenivasan MA, Banergee K, Pandya PG, e t al. (1978): Epidemioeases 14:41-46. logical investigations of a n outbreak of infectious hepatitis in Ahmedabad city during 1975-76. Indian Journal of Medical ReDeCock KM, Bradley DW, Sandford NL, Govindarajan S, Maynard J E , Redeker AG (1987): Epidemic non-A, non-B hepatitis in search 67:197-206. patients from Pakistan. Annals of Internal Medicine 106:227-230. Tsega E (1976): Viral hepatitis during pregnancy in Ethiopia. East Hillis A, Shrestha SM, Saha NK (1973): An epidemic of infectious African Medical Journal 53:268-277. hepatitis in the Kathmandu valley. Journal of the Nepal Medical Tsega E, Mengesha B, Nordenfelt E, Hansson B-G, Lindberg J Association 11:145-151. (1987a): Prevalence of hepatitis B virus markers among Ethiopian Joshi YK, Babu S, Sarin S, Tandon BN, Gandhi BM, Chaturvedi VC blood donors: Is HBsAg screening necessary? Tropical and Geo(1985): Immunoprophylaxis of epidemic non-A, non-B hepatitis. graphic Medicine 39:336-340. Indian Journal of Medical Research 81:18, 19. Tsega E, Mengesha B, Nordenfelt E, Hansson B-G, Lindberg J Kane MA, Bradley DW, Shrestha SM, et al. (1984): Epidemic non-A, (198713): Seroepidemiological and demographic survey of Epsteinnon-B hepatitis in Nepal. Recovery of a possible etiologic agent and Barr virus infection in Ethiopia. Transactions of the Royal Societv transmission studies in marmosets. Journal of the American of Topical Medicine and Hygiene 81:677-680. Medical Association 252:31403145. Tsega E, Mengesha B, Hansson B-G, Lindberg J , Nordenfelt E (19861: Khuroo MS (1980): Study of a n epidemic of non-A, non-B hepatitis: Hepatitis A, B, and Delta infection in Ethiopia: A serologic survey possibility of another human hepatitis virus distinct from postwith demographic data. American Journal of Epidemiology 123:344-351. transfusion non-A, non-B type. American Journal of Medicine 68:818-824. Tsega E, Nordenfelt E, Mengesha B, Hansson B-G, Tsega M, Lindberg J (1990): Age-specific prevalence of hepatitis A virus antibody in Khuroo MS, Duermeyer W, Zargar SA, Ahanger MA, She MA (1983): Ethiopian children. Scandinavian Journal of Infectious Diseases Acute sporadic non-A, non-B hepatitis in India. American Journal 22:145-148. of Epidemiology 118:360-364. Khuroo MS, Teli MR, Skidmore S, Sofi Ma, Khuroo MI (1981): Villarejos VM, Visona KA, Eduarte ACA, Provost PJ, Hilleman MR (1975): Evidence for viral hepatitis other than type A or type B Incidence and severity of viral hepatitis in pregnancy. American Journal of Medicine 70:252-255. among persons in Costa Rica. New England Journal of Medicine 293:1350-1352. Krawczynski K, Bradley DW (1989): Enterically transmitted non-A, non-B hepatitis: Identification of virus-associated antigen in ex- Wong DC, Purcell RH, Sreenivasan MA, Prasad SR, Pavri KM (1980): perimentally infected cynomolgus macaques, Journal of Infectious Epidemic and endemic hepatitis in India: evidence for a non-A, non-B hepatitis virus aetiology. Lancet 2:876-879. Diseases 159:1042-1049.

search and Publication Office of Addis Ababa University, as well as by the Ministry of Health, Addis Ababa, Ethiopia. We are grateful t o Dr. Gertachew Tadesse, Vice-Minister of Health, for his encouragement and support. The Regional Health Department of Eritrea contributed to the preliminary field report and the Medical Director of Mekane Hiwot Hospital, Asmara, facilitated our work. In addition, the authors gratefully acknowledge the significant contributions of the physicians in the Department of Medicine, Mekane Hiwot Hospital, Asmara, the Residents of the Department of Medicine, Faculty of Medicine, Addis Ababa University, Addis Ababa, and the senior technicians of the Mekane Hiwot Hospital, Asmara.