Marburg virus

Marburg virus is a filovirus that causes Marburg virus disease, a severe and frequently fatal haemorrhagic illness closely resembling Ebola virus disease. It was the first filovirus ever recognised, identified in 1967, and it remains one of only two filovirus genera that infect humans, the other being the ebolaviruses. Its genus, Orthomarburgvirus, contains a single species, Orthomarburgvirus marburgense, within which sit two closely related viruses: Marburg virus and the less common Ravn virus, which cause clinically indistinguishable disease.

Like Ebola disease, the illness is dominated by profound fluid loss, shock and multi-organ failure rather than by exsanguinating haemorrhage, and its case fatality has ranged from about 24% to 88% across outbreaks. Two features set Marburg virus apart from its ebolavirus cousins: its reservoir is not merely suspected but proven, and its outbreaks are characteristically seeded by human entry into bat-infested caves and mines.

Marburg disease is a recurring emergency across sub-Saharan Africa, with outbreaks in recent years in Angola, the Democratic Republic of the Congo, Uganda, Equatorial Guinea, Tanzania and Rwanda, and occasional export in returning travellers. Unlike the Zaire species of Ebola virus, it has no licensed vaccine and no licensed specific treatment, so early recognition and supportive care remain the mainstays.

Discovery and historical significance

Marburg virus announced itself in 1967 through a cluster of severe haemorrhagic illness in laboratory and animal-facility workers in the German cities of Marburg and Frankfurt and in Belgrade, then in Yugoslavia. The common exposure was tissue from a batch of African green monkeys imported from Uganda for the production of poliovirus vaccine and research reagents. In all, 31 people were infected and 7 died, and the new agent, a long filamentous virus unlike anything then known, was named after the city of Marburg.

The discovery was historically pivotal: Marburg virus was the first filovirus to be recognised, nine years before Ebola virus, and it established the filamentous morphology that would define the family. A second, genetically distinct member of the species was later identified and named Ravn virus, after a 1987 case in a Danish boy who had visited Kitum Cave in Kenya; it has since caused sporadic cases and small outbreaks alongside Marburg virus, with the same clinical picture.

The 1967 investigation, which identified a filamentous virus by electron microscopy and grew it in cell culture, effectively founded filovirology and prompted the maximum-containment laboratory practices later applied to Ebola virus.

The reservoir was pinned down through outbreaks among people who shared a specific exposure. A protracted outbreak among gold miners at Durba in the Democratic Republic of the Congo between 1998 and 2000, and the large Angolan outbreak of 2004 to 2005, which caused around 250 cases with a case fatality near 90%, pointed repeatedly to caves and mines. Investigation of these sites, and of tourists who fell ill after visiting Ugandan bat caves in 2008, converged on the Egyptian rousette fruit bat, in which infectious virus was eventually found, making Marburg virus the first filovirus with a firmly proven natural host. Outbreaks have continued into the present decade, including in Equatorial Guinea and Tanzania in 2023 and Rwanda in 2024.

Classification, structure, and genome

Classification

Marburg virus belongs to the family Filoviridae, order Mononegavirales, and to the genus Orthomarburgvirus (renamed from Marburgvirus in the 2023 taxonomic revision that introduced binomial species names). The genus contains a single species, Orthomarburgvirus marburgense, which encompasses two distinct viruses, Marburg virus and Ravn virus. The two differ by roughly a fifth of their genome sequence but cause the same disease.

Although Marburg and the ebolaviruses share a family, a genome plan and a clinical syndrome, several biological differences distinguish them, and these have practical consequences for immune evasion, diagnosis and countermeasure design.

Marburg virus compared with the ebolaviruses

Feature	Marburg virus	Ebola viruses
Genus	Orthomarburgvirus	Orthoebolavirus
Species in genus	One	Six
Glycoprotein gene	Single reading frame, no secreted glycoprotein	Editing yields the spike plus a secreted decoy (sGP)
Interferon antagonism	VP40 blocks the JAK-STAT pathway	VP24 blocks STAT1 nuclear import
Natural reservoir	Confirmed: Egyptian rousette bat	Suspected bat host, unconfirmed

Virion structure

Marburg virus has the filovirus form: a filamentous, enveloped virion of uniform 80 nanometre diameter and variable length, often bent into loops or hooks, with a most-infectious length around 790 nanometres. Within the lipid envelope, which carries trimeric glycoprotein spikes, lies a helical nucleocapsid built from nucleoprotein wound around the genome together with the polymerase cofactor VP35, the transcription activator VP30, the polymerase L and the protein VP24. The matrix protein VP40 forms the layer beneath the envelope.

Genome organisation

The genome is a single molecule of negative-sense, non-segmented RNA of about 19 kilobases, one of the largest in the negative-strand RNA viruses, with the seven genes in the fixed order 3’ to 5’ of NP, VP35, VP40, GP, VP30, VP24 and L. The most instructive difference from the ebolaviruses lies in the glycoprotein gene: Marburg virus expresses its glycoprotein from a single reading frame and makes no secreted glycoprotein, whereas the ebolaviruses use transcriptional editing to produce both the surface spike and a secreted decoy. The Marburg surface glycoprotein is cleaved by the cellular protease furin into its two subunits during transport to the cell surface.

Replication cycle

Marburg virus follows the same entry route as the ebolaviruses. Attachment is promiscuous, using host lectins and phosphatidylserine receptors to concentrate the virion on the cell surface rather than binding a single specific surface receptor, and the whole particle is then internalised by macropinocytosis.

Inside the acidifying endosome, host cathepsin proteases trim the glycoprotein to expose its receptor-binding site, which then engages the essential intracellular receptor Niemann-Pick C1 (NPC1), the same cholesterol-transport protein used by the ebolaviruses. Receptor binding triggers fusion of the viral and endosomal membranes and releases the nucleocapsid into the cytoplasm.

Transcription and replication proceed entirely in the cytoplasm. The polymerase L, directed by VP35 and activated by VP30, transcribes the genes into messenger RNAs in a declining gradient set by their order, so nucleoprotein is made in greatest abundance and the polymerase least; as nucleoprotein accumulates, the enzyme switches to replicating full-length genomes. Progeny nucleocapsids are carried to the plasma membrane, where VP40 drives budding of filamentous virions through the host ESCRT machinery, releasing new particles studded with glycoprotein.

Pathogenesis

Marburg disease, like Ebola disease, begins as a subversion of the innate immune system. The first cells infected are macrophages, monocytes and dendritic cells, from which the virus disseminates to the liver, spleen, adrenal glands and endothelium. Infected macrophages release inflammatory cytokines and tissue factor, while infected dendritic cells fail to prime an effective adaptive response.

The virus disarms the interferon system, but by a route that differs from the ebolaviruses and is worth holding onto. VP35 masks double-stranded RNA and blocks the sensor RIG-I, preventing interferon from being made, as in the ebolaviruses. The response step, however, is blocked differently: Marburg virus uses VP40 to disable the JAK-STAT signalling pathway, whereas the ebolaviruses use VP24 to block STAT1 from entering the nucleus. Marburg VP24 instead engages the cell’s antioxidant Keap1-Nrf2 pathway. The net effect is the same, an unchecked early viraemia, but through a distinct molecular mechanism.

The rest of the pathogenesis mirrors Ebola disease. Tissue factor from infected macrophages triggers disseminated intravascular coagulation, consuming clotting factors and platelets, so that bleeding, when it occurs, reflects consumptive coagulopathy rather than a discrete bleeding lesion. Profound lymphopenia develops through bystander apoptosis of uninfected T and natural killer cells, collapsing the adaptive response, and necrosis of the adrenal cortex impairs blood-pressure control and deepens shock. Survival tracks with the ability to mount an antibody response, which fatal cases characteristically fail to do.

The surface glycoprotein contributes to disease in its own right. Its heavily glycosylated surface shields the virion from neutralising antibody, and glycoprotein shed from infected cells into the circulation may disturb endothelial function and add to the vascular leak. The endothelium is infected but, as in Ebola disease, is not extensively destroyed, so the increased vascular permeability that drives shock is largely a functional consequence of inflammation and coagulopathy rather than direct viral lysis of the vessel wall, and survivors recover without lasting vascular damage. In sites of local immune privilege the virus can survive after it has been cleared from the blood, the biological basis of the late relapses and the sexual transmission seen in a minority of survivors.

Epidemiology

Marburg virus is a zoonosis of sub-Saharan Africa whose defining epidemiological feature is its confirmed reservoir, the Egyptian rousette fruit bat (Rousettus aegyptiacus). This is the sharpest contrast with the ebolaviruses, whose reservoir remains unproven. The bats roost in large colonies in caves and disused mines, and infectious virus, viral RNA and antibodies have all been recovered from them; juvenile bats shed virus in seasonal pulses that coincide with periods of heightened spillover risk. The species is widely distributed across sub-Saharan Africa, which places the potential range of spillover well beyond the sites of the recorded outbreaks.

Human infection is correspondingly tied to these habitats. Most recorded outbreaks trace to entry into caves or mines harbouring rousette bats, the classic exposures being gold and other miners and, in wealthier travellers, tourists who visit bat caves. From an index spillover the virus spreads person to person by direct contact with blood and body fluids, amplified, as with Ebola disease, by health-care transmission and by traditional funerals; it is not spread by the airborne route in ordinary conditions.

The recorded outbreaks illustrate this point-source pattern. The protracted outbreak among gold miners at Durba and Watsa in the Democratic Republic of the Congo, between 1998 and 2000, arose from multiple independent introductions from the Goroumbwa mine and caused around 150 cases. The Angolan outbreak of 2004 to 2005 was the largest and most lethal, with about 252 cases and a case fatality near 90%, amplified by extensive nosocomial and household transmission. Uganda has seen repeated smaller outbreaks, and in 2008 two tourists from Europe and North America fell ill after visiting the bat-infested Python Cave, one fatally, giving decisive evidence linking the rousette bat to human disease. More recent events include outbreaks in Equatorial Guinea and Tanzania in 2023 and a first-ever Rwandan outbreak in 2024 of around 66 cases, during which investigational vaccines and therapeutics were deployed.

The basic reproduction number is modest, so, as with Ebola disease, outbreaks can be halted by interrupting body-fluid contact through isolation, safe burial and contact tracing. Case fatality has ranged widely, from about 24% in the original 1967 laboratory outbreak to close to 90% in Angola, shaped by the standard of supportive care as much as by the virus; by absolute numbers the marburgviruses are at least as lethal as the ebolaviruses.

Natural history

After an incubation period of 2 to 21 days, most often 5 to 10 days, the illness begins abruptly and follows the same broad arc as Ebola disease. A nonspecific febrile prodrome of high fever, severe prostration, headache and muscle pain lasts a few days and cannot be distinguished clinically from malaria or other tropical fevers.

Around the fifth day a gastrointestinal phase supervenes, with vomiting and watery diarrhoea that drive rapid volume depletion, frequently accompanied by a characteristic rash. In patients who deteriorate, the second week brings hypovolaemic and distributive shock, metabolic derangement and multi-organ failure, and death, when it occurs, is usually between about days 8 and 16. Survivors improve slowly over weeks as an antibody response takes hold, and, because the virus can persist in immune-privileged sites after clearance from the blood, a minority experience late complications or become a source of renewed transmission months later.

Clinical presentations and complications

The acute illness

The prodrome is abrupt and severe: high fever, profound prostration, headache, and muscle and joint pain, often with conjunctival injection. As in Ebola disease, this early phase is indistinguishable from common febrile illnesses, so exposure history, above all contact with a cave or mine, a sick or deceased person, or a known transmission chain, is central to recognition. Around the fifth day of illness, a non-itchy maculopapular rash frequently appears, a feature well described in the original 1967 cases, alongside a gastrointestinal phase of vomiting and profuse watery diarrhoea. The resulting fluid and electrolyte losses drive patients toward hypovolaemic shock.

Severe disease progresses to a sepsis-like state with metabolic acidosis, electrolyte disturbance and multi-organ dysfunction: acute kidney injury, liver injury and central-nervous-system features from confusion to seizures. Overt haemorrhage occurs in a minority and is usually a late sign, appearing as oozing from puncture sites, petechiae, or bleeding into the gut, and reflects the underlying coagulopathy rather than a primary bleeding lesion. Marked prostration out of proportion to the fever, noted since the earliest cases, and early neurological or bleeding signs point to a severe course. As with Ebola disease, a high viral load at presentation is the most consistent predictor of death.

Special populations and sequelae

Pregnancy carries a grave prognosis, with high maternal mortality and fetal loss, and the products of conception remain highly infectious. Children fare no better than adults. Survivors often suffer a prolonged convalescent syndrome of fatigue, joint and muscle pain, and ocular and psychological sequelae, and the virus can persist in immune-privileged sites: orchitis and uveitis are recognised late complications, virus can persist in semen and enable sexual transmission, and, rarely, virus persisting in the central nervous system causes a late relapse with meningoencephalitis. Secondary infections may complicate recovery, as in the fatal case of Pseudomonas superinfection recorded in a patient with Ravn virus disease.

Diagnosis

Because early Marburg disease cannot be told apart clinically from common febrile illnesses, diagnosis depends on laboratory confirmation guided by an epidemiological case definition. Reverse transcription polymerase chain reaction (RT-PCR) for viral RNA in blood is the diagnostic mainstay, becoming positive once the patient is symptomatic and rising with the viraemia. A negative test very early in a symptomatic contact may need repeating after 48 to 72 hours, because viraemia can lag the first symptoms.

Antigen-detection assays support rapid triage, and serology has a secondary role: IgM antibodies appear in the second week and IgG somewhat later, useful for retrospective diagnosis and survivor studies rather than for the acute decision to isolate and treat. Because filovirus antigens cross-react, a reactive screening result is confirmed with a virus-specific assay, and the choice of specimen and test is coordinated with the reference laboratory to protect staff. Virus isolation is definitive but confined to maximum-containment laboratories. The biosafety burden governs the whole process: live Marburg virus is a biosafety level 4 (BSL-4) agent, so isolation and any aerosol-generating procedure are restricted to maximum-containment facilities, and even routine clinical-pathology testing on a suspected case demands strict precautions.

Management

There is no licensed specific treatment for Marburg virus disease, so aggressive supportive care is the foundation of management. Early and generous replacement of fluid and electrolytes to match gastrointestinal losses, correction of metabolic and electrolyte disturbance, and support of failing organs reduce mortality substantially on their own, and are the core of care in a treatment unit. As in Ebola disease, careful correction of potassium, magnesium, glucose and acid-base status, guided by the scale of gastrointestinal loss, is the substance of effective care. Attention to co-infections such as malaria and to secondary bacterial infection is part of that care.

Specific therapy remains investigational. Unlike Zaire ebolavirus, for which monoclonal antibodies are licensed, Marburg virus has no licensed antiviral, antibody or vaccine. Candidate monoclonal antibodies against the Marburg glycoprotein, the broad-spectrum antiviral remdesivir and the nucleoside analogue favipiravir have shown activity in animal models, and investigational antibodies and antivirals have been offered under trial and compassionate-use protocols during recent outbreaks, including in Rwanda in 2024. Their use remains within research frameworks rather than established practice.

Prevention and public health

Vaccination

There is no licensed Marburg vaccine, but the field has advanced considerably. The leading candidates include a chimpanzee-adenovirus-vectored vaccine (cAd3-Marburg) and recombinant vesicular stomatitis virus and other adenovirus-based designs, built on the same platforms as the licensed Ebola (Zaire) vaccines; several have progressed through early-phase trials and have been offered in rapid-response studies, including during the 2024 Rwandan outbreak. A central obstacle to licensure is that Marburg outbreaks are sporadic and small, which makes conventional efficacy trials difficult, so much of the evidence rests on immunogenicity and on protection in non-human primates. Until a vaccine is licensed, prevention rests on interrupting exposure and transmission rather than on immunisation.

Infection prevention and control

Because transmission requires contact with infected fluids, barrier precautions are highly effective when properly applied. Care of a suspected or confirmed case demands full personal protective equipment, single-room isolation, and rigorous decontamination, and safe, dignified burial removes the highly infectious corpse from traditional funeral practices. A distinctive primary-prevention measure follows from the reservoir: avoiding entry to caves and mines that harbour fruit-bat colonies, or using respiratory and skin protection when entry is unavoidable, reduces the risk of spillover. Laboratory work on live virus is confined to biosafety level 4.

Post-exposure prophylaxis

No post-exposure prophylaxis of proven efficacy exists for Marburg virus. Monitoring of exposed contacts through the maximum incubation period, with prompt isolation and testing at the first sign of illness, is the standard approach, and investigational antibodies or antivirals may be considered for high-risk exposures under expert guidance.

Surveillance and notification

Marburg disease is a high-consequence pathogen governed by the International Health Regulations, and a significant outbreak can be declared a public health emergency of international concern. Rapid case detection, laboratory confirmation, and contact tracing with follow-up through the incubation period are the core public-health tools, supported where possible by surveillance of the bat reservoir and of people with cave or mine exposure.

Outbreak response

Control of an outbreak rests on the same measures proven against Ebola disease: case finding and isolation in dedicated treatment units, exhaustive contact tracing, safe and dignified burials, community engagement, and, increasingly, the deployment of investigational vaccines and therapeutics through embedded trials. The recent Rwandan and East African outbreaks have shown that early, coordinated responses can contain Marburg disease within a limited number of transmission chains.

South African context

South Africa has never had an indigenous Marburg outbreak, but it holds a notable place in the virus’s history. In 1975 the first recognised Marburg cases outside the original 1967 European cluster occurred in Johannesburg, when a young traveller who had journeyed through what is now Zimbabwe fell ill and died, and his travelling companion and an attending nurse were infected and survived. The source was never identified, but the episode remains the country’s defining filovirus event and a lasting lesson in nosocomial transmission from an unrecognised case.

The realistic risk today is importation: a traveller or health worker returning from an affected country, or a patient evacuated for care. Marburg disease, as a viral haemorrhagic fever, is a Category 1 Notifiable Medical Condition requiring immediate reporting on clinical suspicion, before laboratory confirmation. Definitive testing is centralised at the National Institute for Communicable Diseases, which houses the country’s maximum-containment diagnostic capability, and suspected cases are managed along a defined pathway of isolation, senior consultation and specialised referral rather than routine testing, as set out in the national viral haemorrhagic fever guidance.

References and recommended reading

• Kuhn JH, Amarasinghe GK, Perry DL. Filoviridae. In: Fields Virology, 7th edition. Philadelphia: Wolters Kluwer; 2023. The primary reference for filovirus classification, structure, replication and pathogenesis, including the Marburg-specific interferon-antagonism mechanism.
• Bray M, Chertow DS. Viral Hemorrhagic Fevers. In: Richman DD, Whitley RJ, Hayden FG (eds.), Clinical Virology, 4th edition. Washington: ASM Press; 2016. Source for the clinical course, the primacy of fluid loss, and the cave and mine epidemiology.
• Burrell CJ, Howard CR, Murphy FA. Fenner and White’s Medical Virology, 5th edition. London: Academic Press; 2017. Concise foundational account of Marburg virology, the rousette-bat reservoir and outbreak history.
• Florez-Alvarez L, de Souza EE, Botosso VF, et al. Hemorrhagic fever viruses: pathogenesis, therapeutics, and emerging and re-emerging potential. Front Microbiol. 2022;13:1040093. Review of filovirus pathophysiology and the developing vaccine and antibody landscape.
• Hewson R. Understanding Viral Haemorrhagic Fevers: virus diversity, vector ecology, and public health strategies. Pathogens 2024;13(10):909. Source for current case fatality figures and countermeasure availability across the VHF families.