CHAPTER 107: EBOLAVIRUS AND MARBURGVIRUS INFECTIONS

Several viruses of the family Filoviridae cause severe and frequently fatal viral hemorrhagic fevers in humans. Introduction of filoviruses into human populations is an extremely rare event that most likely occurs by direct or indirect contact with healthy mammalian filovirus hosts or by contact with infected, sick, or deceased nonhuman primates. Filoviruses are highly infectious but not very contagious. Natural human-to-human transmission takes place through direct person-to-person (usually skin-to-skin) contact or exposure to infected bodily fluids and tissues; there is no evidence of such transmission by aerosol or respiratory droplets. Infections progress rapidly from influenza-like to hemorrhagic manifestations and typically culminate in multiple-organ dysfunction syndrome and shock. Treatment of filovirus infections is of necessity entirely supportive because no specific efficacious antiviral agents or vaccines are yet available.

Filoviruses are categorized as World Health Organization (WHO) Risk Group 4 Pathogens. Consequently, all work with material suspected of containing filoviruses should be conducted only in maximal containment (biosafety level 4) laboratories. Experienced personnel handling these viruses must wear appropriate personal protective gear (see “Prevention,” below) and follow rigorous standard operating procedures. The proper authorities and WHO reference laboratories should be contacted immediately when filovirus infections are suspected.

ETIOLOGY

The family Filoviridae includes three genera: Cuevavirus, Ebolavirus, and Marburgvirus (Table 107-1 and Fig. 107-1). The available data suggest that the only known cuevavirus, Lloviu virus (LLOV), and one ebolavirus, Reston virus (RESTV), are not pathogenic for humans. The remaining four ebolaviruses—Bundibugyo virus (BDBV), Ebola virus (EBOV), Sudan virus (SUDV), and Taï Forest virus (TAFV)—cause Ebola virus disease (EVD; International Classification of Disease, Tenth Revision [ICD-10], code A98.4). The two marburgviruses, Marburg virus (MARV) and Ravn virus (RAVV), are the etiologic agents of Marburg virus disease (MVD; ICD-10 code A98.3).

Filoviruses have linear, nonsegmented, single-stranded, negative-sense RNA genomes that are ~19 kb in length. These genomes contain six or seven genes that encode the following seven structural proteins: nucleoprotein, polymerase cofactor (VP35), matrix protein (VP40), glycoprotein (GP_1,2), transcriptional cofactor (VP30), secondary matrix protein (VP24), and RNA-dependent RNA polymerase (L protein). Cuevaviruses and ebolaviruses, but not marburgviruses, also encode three nonstructural proteins of unknown function (sGP, ssGP, and Δ-peptide). Filovirions are unique among human virus particles in that they are predominantly pleomorphic filaments but also assume torus- or 6-like shapes (width, ~80 nm; average length, ≥790 nm). These enveloped virions contain helical ribonucleocapsids and are covered with GP_1,2 spikes (Fig. 107-2).

EPIDEMIOLOGY

To date (i.e., as of December 3, 2014), a total of 20,012 human filovirus infections and 8058 fatalities have been recorded (Fig. 107-3). These numbers emphasize both the high degree of lethality (number of deaths per number of sick people; 40.3%) and the overall low mortality (impact on healthy population) of filovirus infections. At least for the moment, natural filovirus infections do not pose a global threat. Filoviruses pathogenic for humans appear to be exclusively endemic to Equatorial Africa, although this distribution may change if natural or artificial environmental alterations lead to filovirus host migration and increased contacts between nonhuman hosts and humans (Fig. 107-4). The majority of recorded EVD and MVD outbreaks can be traced back to single index cases who transmitted the infection to others. These chains of contacts suggest that only around 50 natural host-to-human spillover events have occurred since the discovery of filoviruses in 1967. Outbreak frequency, case numbers, and overall lethality probably depend on the particular etiologic agent, the geographic location and socioeconomic conditions of the affected country, and local customs. In particular, the availability of personal protective gear and reusable medical equipment, such as syringes and needles, has affected overall case numbers in the past, and outbreaks have been contained when local burial practices, such as ritual washing, have been either prevented or altered by the use of gloves. The incidence of EVD and MVD may have increased over the past two decades (Figs. 107-3 and 107-4), but researchers debate whether the observed change is due to increased filovirus activity, more frequent contact between filovirus hosts and humans, or continuous improvement in surveillance capabilities.

EVD and MVD outbreaks are associated with distinct meteorologic and geographic conditions and are probably associated with distinct hosts or reservoirs. The four ebolaviruses that cause disease in humans are endemic in humid rainforests. EVD outbreaks have often been linked to hunting or contact with bush meat (i.e., meat from apes, other nonhuman primates, duikers, or bush pigs) in forests. Ecologic studies indicate that EBOV may be the etiologic agent of extensive and frequently fatal epizootics among wild chimpanzee and gorilla populations. However, replicating isolates of ebolaviruses from wild nonhuman primates have thus far been obtained only in the case of TAFV, which was isolated from a succumbed western chimpanzee in Côte d’Ivoire in 1994. The marburgviruses MARV and RAVV, on the other hand, seem to infect hosts inhabiting arid woodlands. MVD outbreaks have almost always been epidemiologically linked to visits to or work in natural or artificial caves or mines. A pteropid (fruit) bat, the cave-dwelling Egyptian rousette (Rousettus aegyptiacus), serves as a natural and subclinically infected reservoir for both MARV and RAVV. Although bats are suspected to be the hosts for ebolaviruses as well, definitive proof is still lacking. In fact, thus far, only EBOV and RESTV have been loosely connected to frugivorous and insectivorous bats by means of antibody or genome fragment detection, whereas the hosts of BDBV, SUDV, and TAFV remain unclear.

PATHOGENESIS

Human infections typically occur through direct exposure of skin lesions or mucosal surfaces to contaminated bodily fluids or material or by parenteral inoculation (e.g., via accidental needlesticks or reuse of needles in poorly equipped hospitals). Numerous studies, both in vitro and in vivo (in several animal models of human disease), have shed light on key pathogenetic events that evolve subsequent to filovirion exposure. The GP_1,2 spikes on the surface of filovirions determine their cell and tissue tropism by engaging yet-unidentified cell-surface molecules and the intracellular receptor Niemann-Pick C1.

One of the pathogenetic hallmarks of filovirus infection is a pronounced suppression of the immune system. The first targets of filovirions are local macrophages, monocytes, and dendritic cells. Several structural proteins of filovirions, in particular VP35, VP40, and VP24, then suppress cellular innate immune responses by, for instance, inhibiting the interferon pathway and thereby enabling a productive filovirus infection. The result is the secretion of copious numbers of progeny virions, as evidenced by high titers in the bloodstream (>10⁶ plaque-forming units [pfu]/mL of serum in humans) and the lymphatics, and dissemination to most tissues. Filovirions then infect additional phagocytic cells, such as other macrophages (alveolar, peritoneal, pleural), Kupffer cells in the liver, and microglia, as well as other targets, such as adrenal cortical cells, fibroblasts, hepatocytes, endothelial cells, and a variety of epithelial cells. Infection leads to the secretion of soluble signaling molecules (varying with the cell type) that most likely are crucial factors in immune response modulation and development of multiorgan dysfunction syndrome. For instance, infected macrophages react by secreting proinflammatory cytokines, a response that leads to further recruitment of macrophages to the site of infection. In contrast, infected dendritic cells are not activated to secrete cytokines, and expression of major histocompatibility class II antigens is partially suppressed. Immunosuppression occurs in part by massive lymphoid depletion in lymph nodes, spleen, and thymus in the absence of reactive inflammatory cellular responses. Results from animal studies suggest that depletion is a direct consequence of considerable bystander apoptosis of lymphocytes; this explanation would also account for the severe lymphopenia that develops in patients. The consequence of these events is not only florid filovirus dissemination but also a proclivity of the patient for secondary bacterial and fungal infections.

Other pathogenetic hallmarks of filovirus infections are a severe disturbance of the clotting system and the impairment of vascular integrity. Disseminated intravascular coagulation is the cause of the severe imbalance in the clotting system of filovirus-infected patients. Thrombocytopenia, increased concentrations of tissue factor, consumption of clotting factors, increased concentrations of fibrin degradation products (d-dimers), and declining concentrations of protein C are typical features of infection. Consequently, the occlusion of small vessels by widely distributed microthrombi leads to extensive necroses/hypoxic infarcts in target tissues (particularly the gonads, kidneys, liver, and spleen) in the absence of marked inflammatory responses. In addition, petechiae, ecchymoses, extensive visceral effusions, and other hemorrhagic signs are observed in internal organs, mucous membranes, and skin. Actual severe blood loss, however, is a rare event. Aberrance in cytokines or other factors such as nitric oxide and direct infection and activation of endothelial cells most likely are responsible for upregulated permeability of the endothelia of blood vessels. This upregulation leads to fluid redistribution (third spacing); interstitial and myocardial edema and hypovolemic shock are common developments. Clinical improvement is rare and is usually characterized by falling viral titers during the development of a virus-specific immune response.

CLINICAL MANIFESTATIONS

MVD and EVD cannot be differentiated by mere observation of clinical manifestations. The incidence of clinical signs does not differ significantly among infections caused by disparate filoviruses (Table 107-2). The incubation period ranges from 3 to 25 days, after which infected people develop a biphasic syndrome with a 1- to 2-day relative remission separating the two phases. The first phase (disease onset until around day 5–7) resembles influenza and is characterized by sudden onset of fever and chills, severe headaches, cough, myalgia, pharyngitis, arthralgia of the larger joints, development of a maculopapular rash, and other signs/symptoms (Table 107-2). The second phase (approximately 5–7 days after disease onset and thereafter) involves the gastrointestinal tract (abdominal pain with vomiting and/or diarrhea), respiratory tract (chest pain, cough), vascular system (postural hypotension, edema), and central nervous system (confusion, coma, headache). Hemorrhagic manifestations such as subconjunctival injection, nosebleeds, hematemesis, hematuria, and melena are typical (Table 107-2).

Typical laboratory findings are leukopenia (with cell counts as low as 1000/μL) with a left shift prior to leukocytosis, thrombocytopenia (with counts as low as 50,000/μL), increased concentrations of liver and pancreatic enzymes (aspartate aminotransferase > alanine aminotransferase, γ-glutamyltransferase, serum amylase), hypokalemia, hypoproteinemia, increased creatinine and urea concentrations with proteinuria, and prolonged prothrombin and partial thromboplastin times.

Patients usually succumb to disease 4–14 days after infection. Patients who survive experience prolonged and sometimes incapacitating sequelae such as arthralgia, asthenia, iridocyclitis, hearing loss, myalgia, orchitis, parotitis, psychosis, recurrent hepatitis, transverse myelitis, or uveitis. Temporary hair loss and desquamation of skin areas previously affected by a typical maculopapular rash are visible consequences of the disease. Rarely, filoviruses can persist in the liver, eyes, or testicles of survivors and may cause recurrent disease months after convalescence.

DIAGNOSIS

Filovirus infections cannot be diagnosed on the basis of clinical presentation alone. Numerous diseases typical for Equatorial Africa need to be considered in the differential diagnosis of a febrile patient. Almost all of these diseases occur at a much higher incidence than filovirus infections and are therefore the more likely candidates during differential diagnostic deliberations. The most important of the infectious diseases that closely mimic EVD and MVD are falciparum malaria and typhoid fever; also important are enterohemorrhagic Escherichia coli enteritis, gram-negative septicemia (including shigellosis), meningococcal septicemia, rickettsial infections, fulminant viral hepatitis, leptospirosis, measles, and all other viral hemorrhagic fevers (in particular, yellow fever). Other ailments, such as venomous snakebites, warfarin intoxication, and the many transient or inherited platelet and vascular disorders, also must be considered. Visits to caves or mines and direct contact with bats, nonhuman primates (especially apes), or bush meat should raise suspicion of filovirus infection, as should admission to or treatment in rural hospitals or direct contact with severely ill local residents.

If EVD or MVD is suspected on the basis of epidemiologic history, exposure history, and/or clinical manifestations, infectious disease specialists and the proper public health authorities, including the WHO, should be notified immediately. Laboratory diagnosis of EVD and MVD is relatively straightforward but requires maximal containment (biosafety level 4), which usually is not available in filovirus-endemic countries, or the involvement of on-site personnel trained in the use of diagnostic assays adapted for field use. Consequently, diagnostic samples should be collected with great caution and with use of proper personal protective equipment and strict barrier nursing techniques. With adherence to established biosafety precautionary measures, samples should be sent in suitable transport media to national or international WHO reference laboratories. Acute-phase blood/serum is the preferred diagnostic specimen because it usually contains high titers of filovirions and filovirion-specific antibodies.

The current methods of choice for the diagnosis of filovirus infection are reverse-transcription polymerase chain reaction (detection limit, 1000–2000 virus genome copies per milliliter of serum) and antigen capture enzyme-linked immunosorbent assay (ELISA) for the detection of filovirus genomes and filovirion components, respectively. Direct IgM and IgG or IgM capture ELISA is used for the detection of filovirion-targeting antibodies from patients in later stages of disease—i.e., those who have been able to mount a detectable immune response, including survivors. All these assays can be conducted on samples treated with guanidinium isothiocyanate (for polymerase chain reaction) or cobalt-60 irradiation (for ELISA) or subjected to other effective measures that render filoviruses noninfectious. Virus isolation in cell culture and plaque assays for quantification or diagnostic confirmation is relatively easy but must be performed in maximal-containment laboratories. If available, electron microscopic examination of properly inactivated samples or cultures can confirm the diagnosis because filovirions have unique filamentous shapes (Fig. 107-2). Formalin-fixed skin biopsies can be useful for safe postmortem diagnoses.

TREATMENT

COMPLICATIONS

Given the severe immunosuppression induced by filovirus infection, secondary infections should be kept in mind and appropriately treated as early as possible. Pregnancy and labor cause severe and frequently fatal complications in filovirus infections due to clotting factor consumption, fetal loss, and/or severe blood loss during birth.

PROGNOSIS

The prognosis of filovirus infections is generally poor, although outcome probably depends somewhat on which particular virus causes the infection (Fig. 107-3). Convalescence may take months, with skin peeling, alopecia, prostration, weight loss, orchitis, amnesia, confusion, and anxiety as typical sequelae. Rarely, filoviruses persist in apparently healthy survivors and are either reactivated by unknown means at a later point or transmitted sexually. Condom use or abstinence from sexual activity for at least 3 months after disappearance of clinical signs is therefore recommended for survivors.

CONTROL AND PREVENTION

Currently, filovirus vaccines are not available. Prevention of filovirus infection in nature is difficult because the ecology of the viruses is not completely understood. As stated above, frugivorous cave-dwelling pteropid bats (Egyptian rousettes) have been identified as healthy carriers of MARV and RAVV. Avoidance of direct or indirect contact with these bats is therefore useful advice to people entering or living in areas where the animals can be found. Prevention seems to be more difficult in the case of ebolaviruses, for which definite reservoirs have not yet been pinpointed. EVD outbreaks have been associated not with bats but rather with hunting or consumption of nonhuman primates. The mechanism of introduction of ebolaviruses into nonhuman primate populations is unclear. Therefore, the best advice to locals and travelers is to avoid contact with bush meat, nonhuman primates, and bats.

Relatively simple barrier nursing techniques, vigilant use of proper personal protective equipment, and quarantine measures usually suffice to terminate or at least contain filovirus disease outbreaks. Isolation of filovirus-infected people and avoidance of direct person-to-person contact without proper personal protective equipment usually suffice to prevent further spread as the pathogens are not transmitted through droplets or aerosols under natural conditions. Typical protective gear sufficient to prevent filovirus infections consists of disposable gloves, gowns, and shoe covers and a face shield and/or goggles. If available, N-95/N-100 respirators may be used to further limit infection risk. Positive air pressure respirators should be considered for high-risk medical procedures such as intubation or suctioning. Medical equipment used in the care of a filovirus-infected patient, such as gloves or syringes, should never be reused unless safety-tested sterilization or disinfection methods are properly applied. Because filovirions are enveloped, disinfection with detergents, such as 1% sodium deoxycholate, diethyl ether, or phenolic compounds, is relatively straightforward. Bleach solutions of 1:100 and 1:10 are recommended for surface disinfection and application to excreta/corpses, respectively. Whenever possible, potentially contaminated materials should be autoclaved, irradiated, or destroyed.