Marburg virus

INTRODUCTION — The family Filoviridae consists of three genera: Ebolavirus and Marburgvirus, which are among the most virulent pathogens of humans [1-3], and Cuevavirus, which has only been detected in bats in Spain [3,4]. Marburg virus caused the first recognized epidemics of filovirus disease in humans, when infected nonhuman primates were inadvertently imported from Uganda into Germany and Yugoslavia in 1967 [5,6]. Since then, the virus has caused several outbreaks across Sub-Saharan Africa [7]. A single fatal case of Marburg virus disease was detected in Guinea in August 2021, and two fatal cases were detected in Ghana in July 2022 [8]. In April 2023, outbreaks in Equatorial Guinea and Tanzania were reported [9]. Reliable information on filovirus outbreaks can be obtained from the World Health Organization website.

Similar to Ebola virus, Marburg virus causes a rapidly progressive febrile illness that leads to shock and death in a large proportion of infected individuals. Bleeding is observed in some patients, and the disease has traditionally been labeled "Marburg hemorrhagic fever" [10]. However, few patients develop significant hemorrhage, and it is rarely the cause of death. Thus, the syndrome caused by Marburg virus is now designated "Marburg virus disease," just as the term "Ebola hemorrhagic fever" has been replaced by "Ebola virus disease." Because of the biological differences between filovirus genera, vaccines and specific therapies that have been developed against Ebola virus disease will not be effective against Marburg.

This topic reviews the epidemiology, clinical manifestations, treatment, and other aspects of the disease caused by members of the genus Marburgvirus. Detailed discussions of Ebola virus disease are found elsewhere. (See "Epidemiology and pathogenesis of Ebola virus disease" and "Clinical manifestations and diagnosis of Ebola virus disease" and "Treatment and prevention of Ebola virus disease".)

CLASSIFICATION — The filoviruses are nonsegmented, negative-sense, single-stranded RNA viruses. The family name is derived from the Latin "filum," meaning "thread-like," based upon the filamentous structure of the virion. The filoviruses resemble rhabdoviruses and paramyxoviruses in their genome organization and intracellular replication mechanism. (See "Clinical manifestations and diagnosis of rabies", section on 'Virology' and "Measles: Epidemiology and transmission".)

The genus Marburgvirus contains a single species that consists of two recognized variants, Lake Victoria marburgvirus and Ravn marburgvirus, which show approximately 20 percent overall sequence divergence [10,11]. The Ravn variant was responsible for a single case in South Africa in 1987 and for some of the infections among miners exposed to bats in the eastern Democratic Republic of the Congo; all other human cases, including the case in Guinea that occurred in 2021, have been caused by the Lake Victoria variant [12,13]. (See 'Epidemiology' below.)

Both variants of Marburg virus apparently cause similar diseases in humans and certain laboratory animals. In addition, nonhuman primates that have received an experimental vaccine based on one variant are cross-protected against the other [14,15].

EPIDEMIOLOGY

Endemic areas — The first recognized outbreak of Marburg virus disease occurred at separate sites in Germany and Yugoslavia in 1967; the overall fatality rate was 23 percent [5,6]. Both outbreaks were a result of the inadvertent importation of infected vervet monkeys (Chlorocebus pygerythrus) from Uganda for use in vaccine production. Subsequent field studies of trapped animals in Uganda and adjacent countries failed to identify a viral reservoir host.

Since 1967, all known human infections with Marburg virus have occurred in Africa [12,16,17]. Fatality rates among patients have been as high as 80 to 90 percent [7,10]. Reports of Marburg virus disease include:

●In 1975, a young man who had been hitchhiking in Rhodesia became ill after arriving in South Africa; the infection spread to a travelling companion and to a nurse who cared for them [18]. The source was not identified.

●The next recognized outbreak did not take place until 1998, when cases of severe febrile illness were seen among men working in an abandoned mine in the eastern part of the Democratic Republic of the Congo (DRC), where they were heavily exposed to bats and bat excretions [12]. Genomic analysis of virus isolates obtained over the course of the epidemic revealed that miners had been infected with multiple virus variants.

●The largest epidemic, with more than 250 confirmed cases, occurred in Angola in 2004 and centered on a pediatric ward where infection was apparently spread through the use of contaminated transfusion equipment. In contrast to the outbreak in Uganda described above, genomic analysis found that only a single introduction of virus had occurred [11]. The source of infection was not identified. Two reports provide a detailed description of the response of the international organization Médecins Sans Frontières to the Angola epidemic [19,20].

●Since 2004, Marburg virus disease has been identified primarily in Uganda.

•In 2009, two cases occurred separately in tourists who had entered the same bat-infested cave. One was a Dutch woman who was "bumped" by a bat during a visit to the cave; upon returning home, she developed a rapidly progressive illness, leading to her death [21]. News of the case resulted in the retrospective identification of a nonfatal illness in a woman from Colorado who had visited the same cave in 2008 and subsequently developed a febrile illness with hepatitis, coagulopathy, and encephalopathy [22]. Screening of her serum revealed antibodies to Marburg virus.

•In 2012, cases of Marburg virus disease were identified in four different locations in Uganda [23]. All chains of transmission originated from a single source, with amplification through attendance at a traditional funeral. No history of exposure to bats or other potential reservoir species was obtained. The outbreak was contained within two weeks.

•In October 2017, a small outbreak occurred in Uganda; the index case was attributed to exposure to bats during rock salt mining [24]. Four members of one family became infected, of whom three died. Surveillance of more than 300 potential contacts found no additional cases [25]. The outbreak was declared over in mid-December 2017.

●In August 2021, the first case of Marburg virus disease in West Africa was diagnosed in an individual who died after suffering major hemorrhages in the Guéckédou Prefecture in the Nzérékoré Region of southwestern Guinea [8,26]. The patient was diagnosed with Marburg virus disease post-mortem. Careful monitoring found no secondary cases.

●In July 2022, an outbreak was recognized in Ghana [27]. A 26-year-old man presented with fever, vomiting, diarrhea, and epistaxis, and the diagnosis of Marburg virus disease was made at the Noguchi Memorial Institute for Medical Research. The man's wife and infant child also became infected; only the wife survived.

●In 2023, two outbreaks of Marburg virus disease were reported, one in Equatorial Guinea and one in Tanzania [9]. Sequencing of virus isolates did not indicate any connection between these outbreaks, which are occurred on opposite sides of the continent.

•Equatorial Guinea – An outbreak of Marburg virus disease was confirmed in Equatorial Guinea on February 12, 2023 and was declared over on May 15, 2023 [28-31]. During this outbreak there were 17 laboratory-confirmed cases, of which 12 were fatal [30,32]. Twenty-three probable cases were also reported, and all of those patients died.

•Tanzania – On March 21, 2023 the Ministry of Health of Tanzania announced the first outbreak of Marburg virus disease in the country; the outbreak was declared over on June 2, 2023 [33]. The outbreak occurred in the northwest Kagera region [9]. There were eight laboratory-confirmed cases and one probable case; six were fatal [31].

Viral reservoirs — In 2009, scientists at the United States Centers for Disease Control and Prevention succeeded in recovering infectious Marburg virus from Egyptian fruit bats (Rousettus aegyptiacus) captured in a mine in Uganda where numerous cases had occurred over a 10-year period [12]. The isolates showed considerable genomic variation, suggesting that the virus had been present in the bat population for a long time, permitting significant diversification. Studies in captive R. aegyptiacus have found that chronically infected animals shed virus in oral and vaginal fluids without becoming ill, and that increased shedding occurs when female bats give birth [34-36].

The wide geographic dispersion of cases of Marburg virus disease in South Africa, Uganda, the DRC, Angola, and Guinea, as listed above, suggests that the virus is present among chronically infected bats throughout much of sub-Saharan Africa [37-39]. In 2020, this was confirmed by the recovery of infectious virus from apparently healthy cave-dwelling fruit bats in Sierra Leone [40]. The virus sequence closely resembled that which caused the Angola epidemic in 2004.

In addition to the recovery of infectious virus, Marburg virus RNA has been detected in rectal swabs of R. aegyptiacus bats in caves in South Africa and in Zambia [41,42]. Laboratory studies of captive fruit bats experimentally infected with Marburg or Ebola virus have found that tolerance of infection results from the upregulation of protective interferon-stimulated antiviral genes, while proinflammatory responses are dampened [43,44].

Further evidence of a role of bat exposure is provided by a serologic survey in Uganda, which found that miners were more than five times more likely to have antibodies against Marburg virus in their serum than a control group in a non-mining area of the country [45].

Transmission — Marburg virus spreads from person to person by the same mechanisms as Ebola virus, through direct contact with the blood or other body fluids of a patient or during preparation of a body for burial. As an example, treatment of the 25 patients during the outbreak in Germany and Yugoslavia resulted in 6 secondary infections among doctors and nurses [5].

Infection during patient care or washing of a cadaver may result from the inadvertent transfer of virus on contaminated hands to the mouth or eyes. When appropriate precautions have been used, secondary spread has been limited [5,18]. In contrast, if clinicians do not suspect the diagnosis and/or are not able to employ adequate infection control measures, the spread of virus may be enhanced. As an example, Marburg virus appears to have been transmitted through improperly sterilized injection equipment during the 2004 outbreak in Angola [46]. Infection prevention and control precautions used to care for patients with Marburg virus should be the same as those employed to care for those with Ebola virus disease. (See "Treatment and prevention of Ebola virus disease", section on 'Infection control precautions during acute illness'.)

Infection may also be transmitted in the semen of men who have recovered from Marburg virus disease. In the 1967 outbreak in Germany, a man who survived infection transmitted the virus to his wife after discharge from the hospital, apparently through sexual intercourse. A laboratory study of male nonhuman primates who survived acute Marburg virus infection revealed persistence of virus in the testes [47]. This is similar to transmission of Ebola virus. (See "Clinical manifestations and diagnosis of Ebola virus disease", section on 'Convalescence'.)

There are no data to suggest that Marburg virus is transmitted among humans by mosquitoes or other biting arthropods. In addition, although laboratory animals have been infected by exposure to aerosolized virus [48], there is no evidence that patients with Marburg virus disease have transmitted the infection to others by the respiratory route.

PATHOGENESIS — The pathogenesis of Marburg virus disease closely resembles that of Ebola virus disease [49-51]. (See "Epidemiology and pathogenesis of Ebola virus disease".)

After the virus enters the body via mucous membrane or skin penetration, macrophages and dendritic cells are the first cells to be infected. Filoviruses replicate readily within these ubiquitous "sentinel" cells, causing their necrosis and releasing large numbers of new viral particles into extracellular fluid (figure 1). The spread of virus via lymphatic channels to regional lymph nodes results in further rounds of replication, followed by bloodstream dissemination to dendritic cells and to fixed and mobile macrophages in the liver, spleen, and other lymphoid tissues. Two virus-encoded proteins, VP24 and VP35, facilitate rapid systemic dissemination by blocking the production of type I interferon by infected cells and inhibiting the response to exogenous interferon [52]. Further spread of virus to hepatocytes, adrenal cortical, and other parenchymal cells results in extensive tissue necrosis [53].

Systemic inflammatory response — As in Ebola virus disease, systemic inflammation plays a critical role in the etiology of vascular dysfunction, shock, and death from Marburg virus disease. Infected macrophages and other target cells produce large quantities of proinflammatory cytokines and chemokines, while the release of nitric oxide, prostacyclin, and other vasoactive substances produces a general increase in vascular permeability [49,53]. In addition to the effects of circulating proinflammatory mediators, the production of tissue factor on the surface of virus-infected cells triggers coagulation cascades, leading to disseminated intravascular coagulation [54,55].

Impairment of adaptive immunity — Patients dying of filovirus disease typically fail to show an antibody response to the virus [56]; this impairment reflects both direct and indirect attacks on immune function. Dendritic cells, which have primary responsibility for the initiation of adaptive immune responses, fail to be activated by filoviral infection and are destroyed through necrosis [57]. Adaptive immunity is also impaired by a massive loss of lymphocytes. These cells are not infected by filoviruses but undergo "bystander" apoptosis, presumably induced by inflammatory mediators and/or the loss of support signals from dendritic cells [58].

CLINICAL MANIFESTATIONS — Marburg virus causes a rapidly progressive febrile illness that leads to shock and death in a large proportion of infected individuals [7,59]. The clinical manifestations and laboratory findings seen in patients with Marburg virus disease are similar to those produced by Ebola virus (see "Clinical manifestations and diagnosis of Ebola virus disease", section on 'Clinical manifestations'). These include:

●Following an incubation period averaging one week, Marburg virus disease typically begins with the abrupt onset of fever, chills, and general malaise [10,60]. Other signs and symptoms may include weakness, anorexia, severe headache, and pain in the muscles of the trunk and lower back. Vomiting and diarrhea may lead to acute, severe fluid loss.

●Bleeding is rare in the early phase of illness but may become manifest later in the form of petechial hemorrhages of mucous membranes, ecchymoses, and continued oozing from venipuncture sites. Significant bleeding from the gastrointestinal tract or other sites is usually observed only during the terminal phase of fatal infections.

●Additional clinical findings may include a diffuse maculopapular rash, hiccups, chest pain, shortness of breath, headache, confusion, seizures, obtundation, and coma.

Abnormalities in clinical laboratory tests include leukopenia, thrombocytopenia, elevations in serum transaminase (aspartate aminotransferase [AST] and alanine aminotransferase [ALT]) levels, increases in serum blood urea nitrogen (BUN) and creatinine, and abnormal coagulation indices [10]. As an example, in the 1967 Marburg outbreak, many patients had striking leukopenia, with total white blood cell counts as low as 1000/microL, and immature forms at the time of clinical presentation [5]. Serum aminotransferase levels rose rapidly on days 6 to 8 of illness and became highest in patients who died from infection. (See "Clinical manifestations and diagnosis of Ebola virus disease", section on 'Laboratory findings'.)

DIAGNOSIS — The diagnosis of Marburg virus disease is based upon the detection of specific RNA sequences by reverse-transcription polymerase chain reaction (RT-PCR) or viral antigens by enzyme-linked immunosorbent assay (ELISA). However, in contrast to Ebola virus disease, for which rapid diagnostic tests underwent field testing during the West African epidemic [61], rapid tests for Marburg virus are still confined to the research laboratory. A discussion of the tests used to diagnose Ebola virus disease is presented elsewhere. (See "Clinical manifestations and diagnosis of Ebola virus disease", section on 'Diagnostic tests'.)

MANAGEMENT — Management of Marburg virus disease relies on aggressive supportive care to prevent the development of shock while the patient's immune system mobilizes the responses needed to eliminate the virus. The approach to supportive care as well as infection prevention and control precautions used when caring for patients with Marburg virus are similar to those employed to care for patients with Ebola virus disease. (See "Treatment and prevention of Ebola virus disease", section on 'Supportive care' and "Treatment and prevention of Ebola virus disease", section on 'Infection control precautions during acute illness'.)

There are no approved specific treatments for Marburg virus disease. Although antiviral therapies have been approved for treatment of Ebola virus disease based upon findings from randomized trials [62], they consist of monoclonal antibodies against the surface glycoprotein of the Zaire ebolavirus and will not be effective against Marburg virus. (See "Treatment and prevention of Ebola virus disease", section on 'Ebola-specific therapies'.)

In the event of a new, large Marburg epidemic in Africa, it can be expected that there will be clinical trials of therapies for Marburg virus disease similar to the PALM trial for Ebola virus, which was conducted during the 2018-2019 Ebola epidemic in the Democratic Republic of the Congo [62].

Possible experimental agents may include.

●Monoclonal antibodies – In 2015, a coalition of researchers isolated B cells from a Marburg survivor and created a panel of monoclonal antibodies (mabs), which they tested in laboratory animals [63]. A single dose of a mab designated MR191-N given four days postchallenge was able to prevent the death of guinea pigs infected with either the Ravn or Angola virus [64]. The same mab protected rhesus macaques when given in two doses on days 5 and 8 after an otherwise lethal challenge with either virus. Further development of these mabs produced MR186-YTE, a version of MR-191 modified to have an extended serum half-life [65].

●Nucleoside/nucleotide analogs – Several nucleos(t)ide analogs have demonstrated activity against Marburg virus:

•The nucleoside analog prodrug favipiravir (T-705) was protective in mice and nonhuman primates infected with an otherwise lethal challenge dose of Marburg virus [66,67]. Favipiravir was assessed in patients with Ebola virus disease in West Africa but did not meet efficacy targets, and its use was discontinued.

•The nucleotide analog remdesivir prevented the death of five of six cynomolgus macaques when given once daily for 12 days following an otherwise lethal challenge with the Angola strain of Marburg virus [68]. In addition, data suggest that there may be a role for combination therapy with remdesivir and mabs for the treatment of Marburg virus disease. Treatment of rhesus macaques infected with the Angola strain of Marburg virus with a combination of remdesivir and MR186-YTE gave better results than either treatment alone; all animals survived when therapy was initiated six days after virus challenge [65]. Remdesivir was evaluated for treatment of Ebola virus disease in the PALM trial in the DRC but was found to be less effective than mab therapies [62].

•The nucleoside analog BCX4430 (galidesivir), a potential broad-spectrum drug against RNA viruses [69], has been evaluated in animal studies for treatment of Marburg virus. It was efficacious in guinea pigs challenged with either the Ravn variant or the Musoke strain of the Lake Victoria variant of Marburg virus and also prevented the death of cynomolgus macaques when treatment began two days after challenge with the Musoke strain [70].

VACCINE DEVELOPMENT — There are no approved vaccines for the prevention of Marburg virus disease and those used to prevent Ebola virus disease will provide no benefit against Marburg. Until a vaccine is available, a strong public health response is key to controlling outbreaks of Marburg virus disease [17,23,37]. (See "Treatment and prevention of Ebola virus disease", section on 'Public health response'.)

Efforts to develop Marburg vaccines are ongoing under the guidance of a World Health Organization-sponsored consortium (MARVAC) [71]. Candidate vaccines include:

●VSV vaccines – In 2005, a single inoculation of a recombinant vesicular stomatitis virus (VSV) vaccine encoding the viral surface glycoprotein was shown to completely protect cynomolgus macaques against otherwise lethal challenge with the Musoke strain of Lake Victoria Marburg virus [72]. The following year, the same vaccine was found to provide postexposure protection when administered 24 or 48 hours after virus challenge [73].

Similar efficacy has been demonstrated against the Angola variant, using a vaccine encoding the Angola GP [74-77]. In one report, this vaccine resulted in complete protection of nonhuman primates challenged with the Angola strain [78]. A different study found that equal protection against the Angola strain could be achieved using a 1000-fold lower vaccine dose [79].

In 2020, a collaborative group subsequently developed a quadrivalent vaccine containing recombinant VSV encoding the surface glycoproteins of Marburg Angola, the Zaire and Sudan species of Ebola virus, and Lassa virus, and the vaccine protected cynomolgus macaques against challenge by each pathogen [80]. In early 2023, there was a new report of an experimental quadrivalent VSV vaccine containing an equal mixture of vaccines against the Zaire, Sudan, and Bundibugyo species of Ebola virus and the Angola variant of Marburg virus [81]. Cynomolgus macaques were given the quadrivalent vaccine and challenged seven days later with one of the four viruses. All animals except one challenged with the Zaire virus survived; by contrast, all control animals died.

●Adenovirus/modified Vaccinia virus Ankara (MVA) vaccines – A multivalent antifilovirus vaccine encoding glycoproteins from Zaire Ebola virus, Sudan virus, Taï Forest virus, and Marburg virus has been developed. These glycoproteins are expressed by adenovirus serotypes 26 and 35 and MVA vectors.

Cynomolgus macaques receiving the combination vaccine were protected against subsequent challenge by any of the four viruses [82]. In addition, phase I studies of the Ad26/MVA combination against Ebola and Marburg viruses in both adults and children have been conducted in Sierra Leone, and no safety concerns were identified [83,84].

More detailed information on adenovirus/MVA Ebola vaccines are presented separately. (See "Treatment and prevention of Ebola virus disease", section on 'Ebola vaccines'.)

●Recombinant chimpanzee adenovirus vaccine – A single injection of a replication-deficient recombinant chimpanzee adenovirus type 3 (chAd3)-vectored vaccine encoding a wild-type Marburg virus Angola glycoprotein (chAd3-Marburg) was completely protective in nonhuman primates [85]. The same chAd3 vaccine was found to be safe and immunogenic in 40 healthy adults [86]. After a single vaccination, a glycoprotein-specific antibody response was detected in 95 and 70 percent of participants at 4 and 48 weeks, respectively. No serious adverse events related to vaccination occurred.

Future development might also include mRNA vaccines, given their success in the COVID-19 pandemic [87]. (See "Treatment and prevention of Ebola virus disease", section on 'Ebola vaccines' and "COVID-19: Vaccines".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Ebola virus".)

SUMMARY AND RECOMMENDATIONS

●The filoviruses are nonsegmented, negative-sense, single-stranded RNA viruses. The genus Marburgvirus contains a single species, which consists of two variants, Lake Victoria marburgvirus and Ravn marburgvirus. (See 'Classification' above.)

●The first recognized outbreak of Marburg virus disease occurred in Germany and Yugoslavia in 1967 as the result of the inadvertent importation of infected vervet monkeys from Uganda. Since then, all human infections have occurred in Africa, and fatality rates have been as high as 80 to 90 percent. Τhe largest epidemic, with some 250 confirmed cases, occurred in Angola in 2005. In 2023, outbreaks occurred in Equatorial Guinea and Tanzania. (See 'Endemic areas' above.)

●Infectious Marburg virus has been recovered from cave-dwelling fruit bats in Uganda and in Sierra Leone, and viral RNA has been identified in bats in other parts of sub-Saharan Africa. Chronically infected animals shed virus in bodily fluids without becoming ill. (See 'Viral reservoirs' above.)

●Initial cases have resulted from contact with bats or from other unidentified exposures. Marburg virus disease then spreads from person to person through direct contact with the blood or other body fluids of a patient with Marburg virus disease or during preparation of a body for burial. (See 'Transmission' above.)

●After the virus enters the body via mucous membrane or skin penetration, macrophages and dendritic cells are the first cells to be infected. Filoviruses replicate readily within these cells, causing their necrosis and releasing large numbers of new viral particles into extracellular fluid (figure 1). (See 'Pathogenesis' above.)

●Marburg virus causes a rapidly progressive febrile illness that can lead to shock and death. Abnormalities in clinical laboratory tests include leukopenia, thrombocytopenia, elevations in serum transaminase (aspartate aminotransferase [AST] and alanine aminotransferase [ALT]) levels, increases in serum blood urea nitrogen (BUN) and creatinine, and abnormal coagulation indices. (See 'Clinical manifestations' above.)

●The diagnosis of Marburg virus disease is based upon the detection of specific RNA sequences by reverse-transcription polymerase chain reaction (RT-PCR) or of viral antigens by enzyme-linked immunosorbent assay (ELISA). (See 'Diagnosis' above.)

●Patient management is based primarily upon supportive care to limit tissue damage and prevent shock while the patient’s immune system mobilizes the responses needed to eliminate the infection. There are no approved therapies for Marburg virus disease. The monoclonal antibodies that are now in use to prevent or treat Ebola virus disease will provide no benefit against Marburg. (See 'Management' above.)

●Infection prevention and control precautions used to care for patients with Marburg virus disease should be similar to those employed for patients with Ebola virus disease. These are discussed in a separate topic review. (See "Treatment and prevention of Ebola virus disease", section on 'Infection control precautions during acute illness'.)

●There are no approved vaccines for Marburg virus disease, and those used to prevent Ebola virus disease will provide no benefit against Marburg. Efforts to develop Marburg vaccines are ongoing; candidate vaccines include a recombinant vesicular stomatitis virus vaccine, a prime-boost strategy with a recombinant adenovirus followed by a recombinant modified vaccinia Ankara vaccine, and a recombinant chimpanzee adenovirus vaccine. (See 'Vaccine development' above.)