Prior to the 1990s, there had been no recognised cases of Ebola haemorrhagic fever (EHF) in Africa since the 1970s, the decade of the discovery of the disease and the causative Ebola virus (figure). In 1989 however, a new species of Ebola virus, Reston Ebola virus, was recognised in Asian monkeys (1). Fortunately, and unlike the previously recognised African Zaire Ebola virus and Sudan Ebola virus, the Asian virus was not pathogenic in the four human infections that were later diagnosed. Within the past decade, at least six outbreaks have been recognised in Africa. These outbreaks have contributed considerably to increasing the knowledge about this uncommon disease and measures for its control.

1994: first EHF outbreaks in Africa in 15 years

The first appearance of EHF in Africa in the past decade occurred almost simultaneously in late 1994 in Côte d’Ivoire and Gabon, although un-explained deaths in chimpanzees in the Tai Forest in Côte d’Ivoire had been noted in November 1992. Ebola virus was isolated from a woman who performed a necropsy on a chimp that succumbed to the disease in mid-November 1994 (2). From this single human infection, a new species of Ebola virus, Cote d’Ivoire Ebola virus, was isolated and became the fourth recognised species of Ebola virus and the third present on the African continent. This was the first demonstration in Africa of transmission from a non-human primate to man. However, the apes are not the natural reservoir because infection is lethal for these animals, just as it is for humans. The reservoir is presently unknown.

The disease outbreak in Ogooue-Ivindo Province in northeast Gabon was recognised in December 1994 and the last case occurred on 9 February 1995 (3). There were a total of 51 cases and 31 deaths (case fatality rate, CFR = 61%) (4). Deaths of great apes (chimpanzees and gorillas) were reported to have occurred in the same area where the first cases appeared. This was the first time that EHF had been reported from either Côte d’Ivoire or Gabon.

In 1995 in Kikwit

In 1995, a large epidemic of Ebola fever occurred in Kikwit, Democratic Republic of the Congo (formerly Zaire). A total of 315 cases were reported and the overall CFR was 81% (with 250 deaths/310 cases with known outcome) were reported (CFR=79%) (5). Zaire Ebola virus was the etiologic agent. Onset of illness in the first case occurred 6 January but the outbreak only came to the attention of the international community in early May. Nosocomial transmission in two hospitals in April was responsible for amplification of the epidemic. Twenty five per cent of all cases occurred among health care workers, with all but a single case, occurring before personal protective measures were used. Barrier nursing practices (gowns, gloves, masks), active case finding, and social mobilization brought the epidemic to an end by 16 July 1995.

The size of the epidemic and the large international response for investigation and control presented the opportunity to make key observations about this uncommon disease. Epidemiological studies showed that the virus was transmitted by having direct contact with a case or body fluids during the late phase of the disease or after death. Clinical studies described disease manifestations of the eyes, during pregnancy, and the sequelae in survivors. The utility of laboratory tests for antigen detection, including ELISA and immunohistochemistry (skin biopsies) was demonstrated. Also, strategies and protocols for surveillance and early recognition of cases, rational triage, and barrier nursing were developed to appropriately manage suspect cases in the absence of on site diagnostic laboratory services.

Ecological investigations searching for the natural reservoir of the virus were conducted, but there was a 5-month interval from the time the first case was infected and the international recognition of the outbreak. These studies did not yield any conclusive evidence about the reservoir.

Two outbreaks in 1996 in Gabon

The following year (1996), two more outbreaks were reported from Gabon in the same province where EHF appeared in 1994 (3). The first epidemic began in early February when 18 persons became ill after butchering a chimpanzee found dead in the forest. Secondary cases resulted from traditional burial practices where no precautions were taken to prevent virus transmission. There were two other primary cases that were not connected with the chimpanzee episode. In total, there were 31 cases, of which 21 died (CFR=68%). The beginning of the second series of cases was in July and August when unrelated cases occurred in two hunters. In August, it was reported that several chimpanzees died in the same area. An ill Gabonese physician sought medical care in South Africa in late October, where he infected a nurse who died with EHF. The last case was in January 1997. This epidemic included a total of 60 cases and 45 deaths (CFR=75%). Zaire Ebola virus was the cause of all three outbreaks in Gabon. The strains from 1994 and those involved in the two 1996 outbreaks differed by less than 0.1% in the regions of the GP (glycoprotein) and L genes (polymerase) that were sequenced.

Major EHF epidemic in Uganda in 2000

In October 2000, Ebola was reported from Gulu District in northern Uganda approximately 90 km from the Sudanese border. The epidemic had begun several months before and the primary case was never identified. During the epidemic, chains of transmission originating in Gulu also occurred in Mbarara and Masindi Districts. The last case was discharged from Gulu Hospital on 23 January 2001. In preliminary analysis, a total of 425 cases and 224 deaths (CFR=53%) were reported, making this the largest EHF epidemic (6). Amplification of the epidemic resulted from ritual contact with the body of the deceased at funerals, providing care for a case at home, and nosocomial transmission from hospital staff or from other patients. Of concern were the number of infections in health care workers after barrier nursing procedures were put into place, indicating the need for improvements in training and supervision. This was the first appearance of Sudan Ebola virus since 1979, and the first report of EHF in Uganda.

WHO and over 20 institutions in the Global Outbreak Alert and Response Network assisted the Ugandan government in bringing the epidemic to an end. A field laboratory for Ebola diagnosis was set up for the first time during an epidemic. This laboratory, established by the US Centers for Disease Control and Prevention, made possible real time decisions on management of suspect cases and contact tracing. Having an on site laboratory also facilitated sample collection and storage, and provided clinical chemistry services. This information should result in a better understanding of the pathophysiology of EHF. Meaningful ecological studies could not be performed because it was not possible to identify the primary case.

2001: EHF reported in Gabon and Republic of Congo

The next appearance of Ebola fever was only one year later in Gabon. In late November 2001, health authorities in Zadie District, Ogooue-Invindo Province (same province as the 1994-96 EHF cases) were notified of deaths in a family living 30 km from the border with the Republic of the Congo. The onset of illness in the primary case was 25 October and the description of the disease was compatible with a viral haemorrhagic fever. By the end of November, a nurse in Mekambo hospital was infected. She sought care in the provincial hospital in Makokou where she died. Gabonese patients went to the Republic of Congo to be treated by traditional healers where they infected others. Blood samples were collected from suspect cases in Mekambo and Makokou and tested at the Centre International de Recherche Médicale de Franceville (CIRMF) in Gabon. Gabonese officials notified WHO in early December about the outbreak. CIRMF confirmed EHF and reported that the virus was Zaire Ebola virus, and different from the 1994 and 1996 strains (7).

An international team of WHO staff and partners1 in the Global Outbreak Alert and Response Network quickly arrived in Gabon and assisted the Ministry of Health in controlling the outbreak. International staff also worked with Congolese Health officials to establish surveillance and control activities in Mbomo District just across the border, and later in Kelle District when a separate chain of transmission began in late December.

The epidemic ended on 18 and 19 March 2002 in Congo and Gabon respectively. A preliminary analysis of available data indicates that there were a total of 123 cases and 97 deaths (CFR=79%). Sixty five cases and 53 deaths were reported from Gabon and 58 cases and 44 deaths from the Republic of the Congo. Twenty five of the Congo cases (23 deaths) were from remote villages northeast of Kelle and do not appear to be linked to the other cases in Congo or Gabon. This was the first time that EHF has been reported in the Republic of the Congo.

Lessons learned

There are several remarkable aspects of this epidemic. There is both epidemiological and virological evidence for at least six separate introductions (primary infections) in this epidemic. This differs from other EHF outbreaks, except possibly the Gabon epidemic in the autumn of 1996, where there was a single primary case that then infects others via person-to-person transmission. As with previous outbreaks in Gabon, deaths in great apes were reported and several of the transmission chains began after reported contact with these animals. For example, the last chain of transmission in Gabon began in February 2002 after several hunters had contact with a gorilla. CIRMF detected Ebola virus in the carcass when it was found several weeks later. This confirmed that Ebola virus had a role in the epizootic that was first reported in November. The apparent wide distribution of the virus in the forest presented an excellent opportunity to conduct ecological studies for the natural reservoir. Collections were made in February 2002 and laboratory investigations are presently underway. Genetic sequencing and comparisons of viruses detected from cases in the various chains of transmission are also being performed.

Control efforts in both Gabon and in the Republic of Congo were hampered by the remoteness of the affected areas, making it difficult to establish efficient communication, as well as transport for personnel and materials for barrier nursing. Problems in getting the full cooperation of the communities in identifying and hospitalising cases, and reporting contacts, emphasised the need for more effective social mobilisation strategies and activities at the onset of Ebola outbreaks.

Future challenges

The occurrence of recognised epidemics of EHF has clustered temporally in 1976-79, 1994-96 and 2000-02. The explanation may be related to climatic conditions. Wilson et al. (8) have recently reported that ‰ anomalies of low rainfall, as measured by ground measurements and remote sensing satellites, appear to ‘trigger’ the emergence of EHF during the vegetation recovery period after the rains begin. Monitoring climatic conditions in areas such as Gabon where EHF has appeared at least four times in the past seven years should be part of the programme to improve surveillance and outbreak preparedness. Within the past decade, EHF has been recognized for the first time in four countries.

The case fatality rates for Zaire Ebola virus (70-90%) and Sudan Ebola virus (50-70%) infections in the past decade were, with the exception of the Gabon 1994 outbreak, unchanged from that observed in the 1970s although patient care improved during recent outbreaks. Effective anti-viral therapies are needed to reduce mortality. As mentioned earlier, improvements are also needed in the areas of social mobilisation and more effective approaches for insuring the safety of healthcare workers. Immunization has been shown to protect non-human primates against Ebola virus infection, but several issues, eg. current lengthy immunization schedule and the context in which a vaccine would be used, will have to be addressed before it will be of practical use in epidemic situations. Finally, in the search for the natural reservoir, the recent ecological studies in the hot forest areas in Gabon from which multiple transmission chains emerged will hopefully yield clues about this unsolved virological mystery.

Acknowledgement
I thank Pierre Formenty, Cathy Roth, Roberta Andraghetti and Guenael Rodier for reviewing the manuscript and providing helpful comments.