Filoviruses: Ebola and Marburg

The filoviruses are the archetypal agents of viral haemorrhagic fever, and the two that infect people, Ebola virus and Marburg virus, cause the most lethal disease in the group. They take their family name from their appearance, long filamentous particles (Latin filum, thread), and they are maintained in nature by fruit bats, spilling into people through contact with infected wildlife and then spreading from person to person through blood and body fluids. That capacity for human-to-human transmission, combined with case fatality that can exceed half of those infected, is what makes filovirus outbreaks a recurrent international emergency.

A defining feature of the modern response is how species-specific the available countermeasures are. A licensed vaccine and licensed monoclonal antibodies exist, but they protect against Zaire ebolavirus alone; the Sudan, Bundibugyo and Marburg viruses have no licensed vaccine or treatment. The 2026 outbreak of Bundibugyo virus in the Democratic Republic of the Congo and Uganda, declared a public health emergency of international concern, has made that gap concrete, forcing the response to rely on public-health measures and investigational products rather than the tools built for the Zaire species.

Classification and the filovirus family

The family Filoviridae contains two genera of human importance. Orthoebolavirus holds the six ebolaviruses, and Orthomarburgvirus holds Marburg virus and its close relative Ravn virus. The genus names, revised in recent taxonomy, replace the older “Ebolavirus” and “Marburgvirus”.

The six ebolaviruses differ markedly in their threat to people. Four are human pathogens: Zaire, Sudan and Bundibugyo ebolaviruses cause large, lethal outbreaks, while Taï Forest ebolavirus is known from a single non-fatal human case. Reston ebolavirus infects pigs and macaques in Asia but has caused no human disease, and Bombali ebolavirus is known only from bats. Marburg and Ravn viruses are treated together as the cause of Marburg virus disease.

Virus	Genus	Human pathogenicity	Notable outbreaks
Zaire ebolavirus	Orthoebolavirus	High, case fatality up to ~90%	1976 Zaire; 2013 to 2016 West Africa; 2018 to 2020 DR Congo
Sudan ebolavirus	Orthoebolavirus	High	1976 Sudan; 2000 and 2022 Uganda
Bundibugyo ebolavirus	Orthoebolavirus	Moderate to high	2007 Uganda; 2012 DR Congo; 2026 DR Congo and Uganda
Taï Forest ebolavirus	Orthoebolavirus	Low, one non-fatal case	1994 Côte d’Ivoire
Reston ebolavirus	Orthoebolavirus	None in humans	Asia, in pigs and macaques
Bombali ebolavirus	Orthoebolavirus	Unknown	Detected in bats only
Marburg and Ravn	Orthomarburgvirus	High, case fatality ~24 to 88%	1967 Europe; 2004 to 2005 Angola; 2024 Rwanda; 2025 to 2026 Ethiopia

Structure and genome

Filovirus particles are long and filamentous, sometimes branched or looped, with a uniform diameter but a variable length. Each carries a single negative-sense RNA genome of about 19 kilobases wrapped in a helical nucleocapsid within a lipid envelope. The genome encodes seven genes in a fixed order, among them the nucleoprotein, the polymerase, the matrix protein VP40 that drives budding, and the surface glycoprotein (GP) that mediates attachment and entry and is the target of vaccines and antibodies.

The glycoprotein has an unusual twist. In the ebolaviruses the GP gene is transcriptionally edited, so most of what the infected cell makes is a secreted form of the glycoprotein rather than the structural spike, thought to act as a decoy for antibodies; the marburgviruses do not edit their gene and make none. Entry is also distinctive: the particle is taken up by macropinocytosis and only meets its true receptor, the intracellular cholesterol transporter Niemann-Pick C1 (NPC1), deep within the endosome after the glycoprotein has been cleaved.

Interferon antagonism is central to the disease and is shared across the family but differs by genus in its detail. The VP35 protein is the common antagonist, masking viral RNA from the innate sensor RIG-I and blocking interferon induction; the ebolaviruses then add VP24 to block interferon signalling through STAT1, while the marburgviruses use VP40 for the same effect. This blockade lets filoviruses replicate rapidly and to high titre before the immune system can respond.

Reservoir and transmission

The reservoirs of the filoviruses are not equally well established, and the two genera differ sharply. For Marburg virus the reservoir is confirmed as the Egyptian rousette bat, a fruit bat from which the virus has been repeatedly detected and isolated in cave and mine colonies, which is why human spillover has repeatedly followed entry into bat-infested caves and mines. For the ebolaviruses the reservoir remains unconfirmed: fruit bats are the leading suspects, on the strength of antibody and viral-genome-fragment evidence, but infectious Ebola virus has never been isolated from a wild bat, and the definitive host is still unknown. Non-human primates and forest antelope are not reservoirs but amplifying hosts that, like people, sicken and die, and human outbreaks often begin with contact with such infected wildlife.

Once a person is infected, the virus spreads through direct contact with blood and body fluids, and the range of infectious fluids is wide, including blood, vomit, faeces, saliva, breast milk and semen. Two settings amplify transmission characteristically. Healthcare without adequate infection control is a major driver, and health workers are often among the earliest casualties, and traditional funeral practices that involve washing and touching the body have seeded large outbreaks. A further feature emerged from the West African epidemic: filovirus persists in immune-privileged sites, with viral RNA detected in semen for up to around two years after recovery, allowing occasional sexual transmission that can spark a fresh chain of infection long after an outbreak appears to be over.

Pathogenesis

Filoviruses first infect monocytes, macrophages and dendritic cells, then spread to a very broad range of tissues, including endothelium, hepatocytes and adrenal cortical cells, producing the widespread tissue damage that distinguishes filovirus disease from milder haemorrhagic fevers. The interferon blockade allows unchecked early replication, while infected macrophages pour out pro-inflammatory cytokines. A striking feature is a profound lymphopenia even though the lymphocytes themselves are not productively infected; they die instead by bystander apoptosis, which helps explain the collapse of adaptive immunity in fatal cases.

The result resembles septic shock. Cytokine-driven increases in vascular permeability, not direct viral destruction of vessels, drive the fall in blood pressure, and the viral glycoprotein can disrupt endothelium directly even without replication. Coagulation is activated throughout the circulation, driven largely by tissue factor released from infected macrophages, giving a disseminated intravascular coagulation picture that consumes platelets and clotting factors, while adrenal infection impairs the steroid synthesis needed to maintain blood pressure. In fatal cases both antibody and T-cell responses fail, whereas survivors mount an early, vigorous adaptive response.

Clinical course and complications

After an incubation of about 2 to 21 days, illness begins abruptly with fever, severe malaise, headache and myalgia, a phase sometimes called the dry phase because the prominent later features are not yet present. Within days it progresses to a wet phase of profuse vomiting and watery diarrhoea that causes large fluid and electrolyte losses, the main cause of the shock that kills, so aggressive fluid resuscitation is central to survival. A maculopapular rash may appear, and conjunctival injection and facial flushing are common.

Frank haemorrhage occurs in a minority and is usually a late feature, showing as oozing venepuncture sites, gastrointestinal bleeding and bruising rather than the torrential blood loss of popular imagination. Case fatality is high but variable, ranging across outbreaks from around a quarter to as much as 90% depending on the virus, the strain and the quality of supportive care. Survivors face a recognised post-viral syndrome of arthralgia, fatigue and, importantly, ocular disease, with uveitis from virus persisting in the eye and, rarely, a late relapse as meningoencephalitis when virus persists in the central nervous system, alongside the persistence in the male genital tract already noted. Marburg virus disease follows a broadly similar course to Ebola disease.

Epidemiology and outbreaks

Marburg virus announced the family in 1967, when laboratory workers in Marburg and Belgrade were infected by imported African monkeys. Ebola virus followed in 1976 with simultaneous outbreaks of the Zaire and Sudan species in central Africa. For decades the outbreaks were episodic and rural, until the 2013 to 2016 West African epidemic of Zaire ebolavirus, by far the largest ever recorded, spread through Guinea, Liberia and Sierra Leone, caused more than eleven thousand deaths, and transformed both the science and the politics of filovirus disease. A large outbreak in the eastern Democratic Republic of the Congo followed from 2018 to 2020, and Sudan virus returned to Uganda in 2022.

Marburg virus has become more active in recent years, with outbreaks in Rwanda in 2024 and Ethiopia across 2025 and 2026, the latter declared over in January 2026. The dominant event at present is the 2026 outbreak of Bundibugyo virus, which began in the Democratic Republic of the Congo, spread to Uganda, and was declared a public health emergency of international concern on 17 May 2026. By late June it had exceeded 1,100 confirmed cases with a case fatality of around 26%, with a single exported case reaching France in a returning health worker. Its significance is sharpened by the fact that the licensed Zaire-specific vaccine and antibodies do not apply to it.

Diagnosis

Diagnosis rests on detection of viral RNA by reverse transcription polymerase chain reaction (RT-PCR) on blood, which is sensitive during the high viraemia of acute illness; oral swabs can be used where a blood draw is unsafe, and antigen-detection rapid tests support field triage. Because the viruses are biosafety level 4 pathogens, confirmatory testing is centralised in maximum-containment reference laboratories, though molecular assays run on inactivated samples allow safe testing at lower containment. As with every viral haemorrhagic fever, malaria must be excluded in parallel, since it is far more common in the same patients and settings.

Vaccines and therapeutics

The filovirus countermeasure landscape is defined by species specificity. For Zaire ebolavirus there is a licensed single-dose vaccine, Ervebo (rVSV-ZEBOV), a live recombinant vesicular stomatitis virus carrying the Zaire glycoprotein, used in outbreaks through ring vaccination, in which the contacts of a case and their contacts are vaccinated to build a protective ring. A separate two-dose regimen, Zabdeno followed by Mvabea, provides prophylactic protection where the dosing interval can be met. For treatment, two monoclonal antibody products against the Zaire glycoprotein, Inmazeb (a three-antibody cocktail) and Ebanga (ansuvimab), were shown in the 2018 to 2019 PALM trial to reduce mortality and are now standard of care for Zaire ebolavirus disease.

Countermeasure	Type	Target	Status
Ervebo (rVSV-ZEBOV)	Live vaccine, single dose	Zaire ebolavirus	Licensed 2019; ring vaccination
Zabdeno and Mvabea	Two-dose prime-boost vaccine	Zaire ebolavirus	Approved 2020; prophylactic
Inmazeb	Three-antibody cocktail	Zaire ebolavirus	Licensed 2020
Ebanga (ansuvimab)	Single monoclonal antibody	Zaire ebolavirus	Licensed 2020
Investigational vaccines	Vaccine	Sudan, Marburg	In clinical trials, none licensed
MBP134 and maftivimab	Monoclonal antibody, investigational	Prioritised in the 2026 Bundibugyo outbreak	Under evaluation

The gap is that none of the licensed tools is proven against Sudan, Bundibugyo or Marburg virus, because the Zaire glycoprotein they target differs enough that cross-protection cannot be assumed. This is why, in the 2026 Bundibugyo outbreak, the World Health Organization advised against programmatic use of Ervebo and instead prioritised investigational products for evaluation: the antiviral remdesivir and the monoclonal antibodies MBP134 and maftivimab. Investigational vaccines against Sudan virus and against Marburg virus are in clinical trials, but none is yet licensed. Across all filovirus disease, meticulous supportive care remains the foundation of treatment, since fluid, electrolyte and organ support save lives regardless of the species.

South African context

No filovirus is endemic to South Africa, but the region has both a history of importation and a standing preparedness role. The first Marburg cases recognised outside the 1967 laboratory outbreak occurred in 1975, when a traveller fell ill in Johannesburg after journeying through the region and infected a companion and a nurse. The current concern is imported Ebola disease: the 2026 Bundibugyo outbreak in central Africa prompted national preparedness guidance, and any suspected case is managed through the same viral haemorrhagic fever pathway of risk assessment, isolation, safe specimen handling and reference-laboratory confirmation, coordinated through the National Institute for Communicable Diseases. Filoviruses are handled at maximum containment, and confirmatory testing is centralised at the national reference laboratory.

References and recommended reading

• Kuhn JH, Amarasinghe GK, Perry DL. Filoviridae. In: Fields Virology, 7th edition, Chapter 11. Wolters Kluwer; 2023. The mechanistic reference for filovirus structure, entry through Niemann-Pick C1, interferon antagonism and reservoir biology.
• Carson G, Bray M, Roth C. Viral Haemorrhagic Fevers. In: Clinical Virology, 4th edition, Chapter 9. ASM Press; 2016. The comparative account of filovirus pathogenesis, transmission and clinical disease.
• Flórez-Álvarez L, de Souza EE, Botosso VF, et al. Hemorrhagic fever viruses: pathogenesis, therapeutics, and emerging and re-emerging potential. Frontiers in Microbiology. 2022;13:1040093. The current account of interferon antagonism and the Ebola vaccine and monoclonal-antibody landscape.
• Hewson R. Understanding Viral Haemorrhagic Fevers: Virus Diversity, Vector Ecology, and Public Health Strategies. Pathogens. 2024;13(10):909. Current taxonomy, reservoir ecology and the vaccine-development landscape.
• World Health Organization. Ebola disease caused by Bundibugyo virus, Democratic Republic of the Congo and Uganda: disease outbreak news and emergency guidance. WHO; 2026. The current outbreak status and the guidance on Ervebo and investigational therapeutics.