Chapter 60. Viral Hemorrhagic Fevers

• Viral hemorrhagic fevers (VHFs) are seen worldwide as either locally endemic disease or as imported cases in travelers.
• Travelers may be the source of epidemics in countries far from the source.
• Clinical presentation may be a flu-like syndrome, but may also be abrupt, with fever associated with generalized myalgia and headache. Severe cases progress to respiratory and gastrointestinal symptoms followed by systemic “capillary leak” causing pulmonary edema, shock, and bleeding from mucosal surfaces.
• Clinical diagnosis hinges on a history of potential exposure within the preceding few days up to a maximum of 4 weeks. The exposure may be to rodents, ticks, or fresh animal or human blood, or patients, usually in remote rural areas where these diseases are endemic.
• Management is mainly supportive, but effective antiviral therapy is available for some fevers. The critical issues are respiratory support, replacement of blood components as needed, and support of the circulation (using vasopressor drugs if necessary) while avoiding large infusions of intravenous fluids, which worsen pulmonary edema.
• Ribavirin is effective for several VHF infections, and when indicated should be given intravenously and as soon as possible.
• Prevention of transmission of infection is effected by providing routine isolation of suspect cases in a single room, preferably one with negative pressure, and use of standard universal precautions for exposure to contact with blood or body fluids, including gloves, gowns, and other barrier nursing procedures, and careful disinfection of exposed surfaces. With appropriate barrier precautions the risk of nosocomial transmission is minimal.
• With prompt and appropriate supportive therapy, many patients with these infections can make a rapid and complete recovery without significant sequelae.

Viral hemorrhagic fevers (VHFs) are endemic on every continent with the possible exception of Australia. The diseases are characterized by acute onset with high fever, and in some cases, high mortality. The bleeding by which they are known is a complication only of severe disease, but the underlying pathology is of widespread leakage from the capillaries, with prominent pulmonary edema. The clinical syndrome is mainly caused by endothelial cell dysfunction, not by tissue necrosis. Death usually results from hypovolemic shock with or without acute respiratory distress syndrome (ARDS). In survivors, recovery is rapid and usually complete.

The diseases are caused by many different enveloped RNA viruses. Almost all are zoonoses, and infection of humans is accidental, usually the consequence of intrusion of a human into the ecologic niche of the virus. The viruses belong to four major families: bunyaviridae, arenaviridae, filoviridae, and flaviviridae. The bunyaviruses include hantaviruses and Crimean Congo hemorrhagic fever (CCHF), a member of the genus Nairovirus. CCHF is spread by ticks, and occurs widely across Africa, Southeastern Europe, the Middle East, and Asia. Hantaviruses are found throughout the world as natural silent infections of a number of rodents. A New World hantavirus in North and South America is a pathogen of deer mice and related species, and causes the hantavirus pulmonary syndrome (HPS). Arenaviruses are also natural infections of rodents, the most important being Lassa virus, which is confined to West Africa, and the South American hemorrhagic fever viruses belonging to a complex of New World rodent viruses, four of which at present are known to cause VHF. The filoviruses are a unique family of filamentous viruses that includes Ebola and Marburg. African filoviruses have high mortality, but Asian filoviruses seen in monkeys from the Philippines do not appear to be pathogenic for humans. The flaviviridae include yellow fever virus and dengue virus, both spread by mosquitoes. Other viruses may cause hemorrhagic fever, such as Kyasanur Forest disease and Omsk hemorrhagic fever, but these are confined to very local areas and will not be discussed in detail.

Epidemiology

The zoonotic character of these infections means that they are primarily rural diseases in developing communities in areas with poor medical facilities, and they often go undiagnosed. Clinical reports and trials are scant, and data are limited. Essentially, VHFs are diseases of the poor. However, with increasing mobility of populations, and the migration and enrichment of individuals from poor communities who may revisit their villages for family events, infected patients can and do appear almost anywhere in the world. Missionaries and medical staff working in remote areas are also at risk and more likely than locals to reach tertiary care facilities, and thus to be diagnosed. Peace keepers and other aid workers are particularly at risk and several have been infected with Lassa fever in Sierra Leone and a few have died. Epidemics of VHFs such as yellow fever may be devastating, and there can be thousands of cases of this mosquito-borne disease. Some viruses, particularly filoviruses, have only emerged in small epidemics; however their very high mortality has given them prominence disproportionate to the number of patients. Recently Ebola outbreaks in equatorial Africa, specifically Gabon and the Republic of Congo, are associated with a massive epizootic in great apes whose populations are threatened now with extinction due to a combination of Ebola virus, logging, and hunting for bush meat. This epizootic is fueling epidemics in rural communities in the adjoining areas.

History and Clinical Diagnosis

The most critical element in the clinical diagnosis is a thorough history covering the incubation period, a maximum of 4 weeks prior to onset of fever. The element that alerts the physician to a viral hemorrhagic fever (and usually indicates which one it is likely to be) is the contact the patient might have had with known ecologic niches or patients who may have one of these infections (Table 60-1). Questions must include a thorough travel history; information on any possible contact with ticks, fresh animal blood, rodent urine or blood, wild animals, and mosquitoes and other insects; recent camping in exotic areas; possibly entry into bat caves; and attendance at ceremonial funerals. All of these factors can be significant, especially in remote areas. A medical care provider or other worker who might have had contact with blood from a primary case is also cause for alert.

An essential element is a short history of fever, often high and of abrupt onset. However, some patients present with a flu-like syndrome, which is of course difficult to differentiate. Severe body pains and headache are prominent and may be excruciating. Other features may be severe sore throat, nausea and vomiting, petechiae, oozing from the gums, and bradycardia. Proteinuria is frequent. Total peripheral white counts are usually unhelpful, but differential counts usually show lymphopenia and neutrophilia, which may be marked, and may actually mislead one into thinking it is caused by a bacterial infection. Thrombocytopenia is common, and platelet function is impaired. Partial thromboplastin times may be prolonged, but prothrombin times are relatively spared. Disseminated intravascular coagulation (DIC) is not a feature of VHF except as a complication of the terminal phase. As disease progresses, hypovolemic shock, pulmonary edema, and frank bleeding ensue. Serum glutamate oxaloacetate transaminase (SGOT; aspartate aminotransferase) is usually raised, and VHFs distinguish themselves from viral hepatitis because the SGOT is disproportionately high compared with the serum glutamate pyruvate transaminase (SGPT; alanine aminotransferase). Ratios of SGOT:SGPT may be as high as 11:1, and the level of the SGOT also reflects prognosis. Patients are rarely jaundiced, and the bilirubin is usually normal, the exception being yellow fever as its name suggests. The central nervous system (CNS) is relatively spared, but encephalopathy and neurologic sequelae such as ataxia and deafness occur. Isolated ocular sequelae in Ebola including blindness have been reported, as has a syndrome of polyserositis following Lassa fever.1,2

Laboratory Diagnosis

Care must be taken in collection, handling, and transport of specimens, and consultation with the laboratory is essential. Gloves must be worn at all times and the specimens clearly labeled as hazardous. Blood samples should be drawn, preferably into a vacuum tube system. Specimens for transport should be transferred to a leakproof plastic container and double wrapped in further leakproof containers for shipping to a suitable reference laboratory. Sera may be safely handled for immunologic tests by inactivation by heating to 60°C for 30 minutes.

The most rapid and accurate method of diagnosis is by reverse transcriptase polymerase chain reaction (RT-PCR). This technique is reliable and safe to perform directly on serum or tissues. However, availability of the assay specific for a particular virus may not be locally available and specimens need to be directed rapidly to a competent laboratory. The chemical reagents used in RT-PCR will reliably inactivate virus, so the technique also has advantages in safety, provided the initial steps are performed with care. In some circumstances antigen detection techniques in blood and tissues are available using enzyme-linked immunosorbent assay (ELISA) methodology, but once again these are usually only available in reference laboratories. Laboratory diagnosis is also by isolation of virus from serum, but this usually requires a high level of laboratory biocontainment, and takes time, often a week or more. Demonstration of a fourfold rise in antibody titer, or high-titer IgG antibody with virus-specific IgM antibody in association with compatible clinical disease is also diagnostic, but the more rapid VHFs, particularly Ebola and CCHF, evolve so quickly that patients may never develop antibody before death. Antibody detection techniques include immunofluorescence assay (IFA) and ELISA, and are increasingly being produced using recombinant antigens, thus avoiding the use of infectious agents in their preparation.

Management

VHFs in survivors are self-limiting diseases, so providing the acute crisis can be managed, recovery is rapid and usually complete, though convalescence may be prolonged.3,4 Some viruses are highly treatable using the antiviral agent ribavirin, provided therapy is instituted as early as possible in the disease course (Table 60-2). Therapy with immune plasma has also been advocated, but with the exception of Argentine hemorrhagic fever, it has not been shown to have efficacy.

The main challenge is careful management of fluid balance. Patients will often present with high hematocrit due to capillary leakage aggravated by dehydration. Despite this, pulmonary edema is a real risk, and patients should be infused with caution. Blood and platelet replacement may be necessary. Full intensive care support may be required including mechanical ventilation, monitoring of central venous pressure, and dialysis. Seizures and arrhythmias will need to be controlled. DIC is not generally a major feature, and heparin is contraindicated. Exchange transfusion and use of steroids are controversial and cannot be recommended at this time. Early accurate diagnosis and good intensive care support are the most important underlying principles for the physician.

Patients with VHF do not travel well, since their cardiovascular system is often unstable, and even the moderate trauma of travel can induce bleeding or hypotension. For this reason moving a patient should be avoided whenever possible.5 Moving the patient also increases the risk and the number of people exposed and thus the risk of secondary infections.

Prevention of Spread

The fearsome reputation of some of these viruses, such as Ebola, Lassa, and CCHF, comes from their ability to spread to medical staff and patients in facilities where inadequate barrier nursing and other practices lead to blood-to-blood contact with the virus, such as needlestick injuries, blood spills on unprotected damaged skin, or mouth-to-mouth resuscitation. In some countries, reuse of needles and syringes has produced nosocomial and community outbreaks.6 There are also reports of outbreaks in surgical teams who unwisely performed laparotomy on infected patients.7 In these circumstances mortality has been high.

The key to prevention of transmission both in endemic and nonendemic areas has consistently been good hospital and laboratory practice, with strict isolation of febrile patients at risk of VHF and rigorous use of gloves and disinfection.8 A small number of designated personnel should undertake direct care and be kept fully informed about the nature of the virus and the precautions to be taken.

Intensive care and full life support should not be denied. Essential surgery should be performed with meticulous attention to avoiding accidental inoculation of blood or fluids. Pregnant patients are a major challenge. They often present with absent fetal movements, and maternal survival in Lassa fever and probably other hemorrhagic fevers has been shown to depend on aggressive obstetric intervention to evacuate the uterus.9

Aerosol spread in hospitals has not been documented; indeed there is much published evidence that this is not a major hazard. Past recommendations for isolation of patients in a plastic isolator have long given way to simple strict barrier nursing. Indeed, the risk to personnel is about equal to that of human immunodeficiency virus (HIV) transmission, and the practices now universally adopted for HIV nursing and medical care are quite adequate to prevent transmission of VHFs, provided they are fully implemented and observed. This practice presents no excess risk to hospital personnel and allows substantially better care to be given to the patient.

The major factor in nosocomial transmission is unawareness of the possibility of VHF by a health care provider who is also inattentive to the requirements of effective barrier nursing. Many outbreaks in developing countries have been due to lack of hygiene and to unwise surgery on febrile patients. Once the diagnosis has been entertained and appropriate precautions instituted, the risk of nosocomial transmission is very small.

High risk of infection is associated with direct percutaneous or mucosal contact with blood or body fluids. Postexposure prophylaxis with ribavirin for CCHF and for arenaviruses should be offered to contacts with high-risk exposures. People with medium- or low-risk contact histories (includes most unprotected contact with blood or body fluids, and casual or social contact) may safely be observed for development of persistent high fever for 3 weeks from the last date of contact. The practice of following up airline passengers and other low-risk exposures has been discontinued. Low-risk contacts (entered the room) do not need to be placed under surveillance.

The 1988 Centers for Disease Control and Prevention (CDC) Guidelines for the Management of Patients with VHFs recommends routine patient isolation in a single room, preferably but not necessarily with negative air pressure gradient from the hallway through an anteroom to the patient room.3 Staff education; use of gloves, gowns, and masks; and rigorous disinfection are mandatory. The recommendations issued for patient management and handling of clinical specimens from acquired immunodeficiency virus (AIDS) patients are adequate for containment of VHFs.

Lassa and Ebola viruses in particular are robust and withstand drying. However, all the viruses can be inactivated by heat (60°C), detergents, chlorine, formalin, and ultraviolet (UV) radiation (sunlight). All the hemorrhagic fever viruses have envelopes that are disrupted by detergents and soaps, rendering them inactive. Though detergents cannot be recommended for disinfection, it is useful to remember that their liberal use will greatly reduce the number of infectious particles. Disinfection can be accomplished by washing with 0.5% phenol in detergent (e.g., Lysol), 0.5% hypochlorite solution, formaldehyde, glutaraldehyde, or paracetic acid. Care should be taken to ensure solutions are freshly and correctly made up and time allowed for disinfectant to work on spills.

Finally a problem of VHFs, particularly viruses such as Ebola, is the fear and press attention they receive. A single case can be quite traumatic for an institution unless the situation is carefully handled. A measured and informed approach from a collaborative team of doctors, nurses, administrators, and others is needed. Careful education of all medical staff, emphasizing the real risks and the way to avoid them, allaying unnecessary fears, and avoiding panic will result in appropriate management of the patient and avoid secondary infections. Press attention can be quite disruptive and is best managed by the sharing of accurate and regular information.

Bunyaviruses are a large family of spherical (85 to 100 nm) negative-sense RNA viruses found in animals and insects. Crimean Congo hemorrhagic fever (CCHF) virus is a member of the tick-borne Nairovirus genus of this family, and the hantaviruses, which are rodent-borne, form their own genus.

Crimean Congo Hemorrhagic Fever

CCHF is a tick-borne viral disease initially described in the USSR in the 1930s, and the virus was isolated in the Belgian Congo in the 1950s. The virus is now known to occur from Eastern Europe through Asia, the Middle East, and in all of Africa and the People's Republic of China (PRC), where the disease is known as Xinjiang fever.4,10–13 Humans are infected from ticks or by handling blood or secretions from infected people or domestic animals. Because of the high mortality of infection with CCHF virus, the virus is classified as Biosafety Level 4 (BSL4).

CCHF virus infects at least 24 species of ixodid (hard) ticks, particularly Hyalomma species, which serve as both reservoir and vector of this agent. CCHF infection of humans has emerged as a sporadic but important disease, particularly in arid livestock farming areas. Seasonality depends on local climatic conditions, with peaks corresponding to maximum periods of tick infestation.14,15 A wide range of wild and domestic animals and birds may be infected.

The illness:infection ratio of CCHF virus ranges from 1 : 5 in the southern USSR to more than 50% of infected persons in South Africa, and mortality varies from 5% to 10% in the southern USSR to 35% in South Africa; rates of 35% to 50% or higher are possible in nosocomial outbreaks. Mild or inapparent infection appears to be the rule for most mammalian species, with the exception of humans.

The high risk of nosocomial outbreaks was first recognized in 1976, when a laparotomy was performed on a patient in Pakistan with abdominal pain, hematemesis, and melena. Eleven secondary cases in hospital staff resulted in three deaths, including a surgeon and an operating theater attendant.7 Since then, similar nosocomial outbreaks have been reported in many countries from South Africa through countries of the former Yugoslavia, Iran, Dubai, Afghanistan, China, and Russia.10–13,16–20 In hospital outbreaks heavy blood contamination or needlestick injury is frequently recorded.16

Clinical and Laboratory Features

The incubation period of CCHF is only about 2 to 9 days.21 There is a rapid, dramatic onset with severe headache, high fever, chills, and excruciating myalgia strongly localized to the lower back and joints. There is often epigastric pain. Conjunctivitis and a mild flushing of the face and chest, pharyngeal hyperemia, and petechiae on the palate are frequent. Bradycardia is typical, and diarrhea occasional.

Within 3 to 5 days after onset, signs of hemorrhagic diathesis may develop with bleeding from the gums, epistaxis, hematuria, and bloody diarrhea. Petechiae and purpura are frequent. Massive hematemesis, melena, extensive bloody effusions, bleeding from virtually every orifice and from venipuncture sites, and hypovolemic shock precede death with development of sometimes spectacular ecchymoses. Despite the high fever, bradycardia is typical, with tachypnea due to pulmonary edema.

Liver function tests may show mild to marked hepatocellular dysfunction, sometimes accompanied by renal failure. SGOT may be markedly raised, but SGPT elevation is usually minimal. There is no evidence of myocardial involvement. There is no objective evidence of direct CNS infection. Intracranial bleeding is a possibility. Changes in affect and mood, including aggressive behavior, are a feature of the convalescent phase.

Severe thrombocytopenia is invariable, frequently with counts below 20,000/μL. Prothrombin times (PTTs) and activated partial thromboplastin times (APTT) may be prolonged but usually not severely so. Platelet counts of less than 20,000/μL, PTT >60 seconds, SGOT >200 IU/L, and SGPT >150 IU/L are associated with a risk of death of more than 90% in untreated patients. Leukopenia may be profound with an early predominance of polymorphonuclear cells, giving way to a marked relative and absolute lymphopenia after 7 to 10 days. Some patients have a relative neutrophilia. During convalescence there may be lymphocytosis with atypical forms.

Diagnosis

Diagnosis is most rapidly made by RT-PCR. Though there is some genetic diversity, sufficient conserved sequences may be targeted for broad specificity. Virus may also be isolated from blood of acutely ill patients, but viremia is early and of short duration, and the virus is labile, so that specimens have to be carefully taken and preserved for transportation. Antigen detection systems are also available using recombinant antigens.11,12 Virus-specific antibodies appear 7 to 10 days after infection and neutralizing antibodies after day 14 to 16, but since the incubation period is very short, often only 3 to 4 days, acute diagnosis by antibody detection is not reliable.22

Specific Treatment

CCHF virus is sensitive to the antiviral ribavirin in vitro and in vivo in concentrations as low as 5 μg/mL, and the CDC recommends the empirical use of the intravenous preparation for treatment of CCHF infections.3 Ribavirin has shown to be effective in vivo.16

Oral ribavirin has been used for postexposure prophylaxis for CCHF infection, but its efficacy has not been formally examined. Though intravenous ribavirin is the treatment of choice, oral therapy may be used if it is the only form available, since ribavirin is well absorbed from the gastrointestinal tract and would be expected to attain adequate levels in the blood. There is presently no role for immune plasma in the treatment of CCHF infection.

Prevention

Avoidance of tick bite using repellents applied to skin or soaked into clothing are methods of choice. Slaughter of potentially viremic animals is hazardous. Care of infected patients should avoid blood contact and high-risk procedures such as mouth-to-mouth resuscitation. The combination of epigastric pain and hematemesis has sometimes led an unwary surgeon to perform laparotomy.16 Generalized oozing results in large blood losses and inevitably difficulty in achieving hemostasis. The patient does not survive, and the surgeon and operating attendants themselves have a high risk of fatal infection.

Hantaviruses

Hemorrhagic fever with renal syndrome (HFRS) loosely describes a complex of diseases in which fever and acute renal failure are associated with bleeding, first reported from Manchuria and from Scandinavia during the 1930s. Until the early 1990s the diseases and the viruses that cause them were thought to belong exclusively to the Old World. In 1993 hantavirus pulmonary syndrome (HPS) was identified in an outbreak in the United States,23 and since then patients and infected rodents have been found from Canada to Argentina. Both disease syndromes are now known to be caused by related viruses collectively called hantaviruses, which are members of the bunyaviridae.24,25

Epidemiology

Hantaviruses are all natural parasites of a wide variety of rodents, in which they persist for life without apparent disease. The rodents are infected at or around birth, and excrete large amounts of virus in urine, probably for life. The virus groups are determined primarily by the rodent host species, which govern geographic distribution of diseases they cause. Thus among the Old World hantaviruses, particularly in Siberia, mainland China, and Korea, the most common rodent host is the Manchurian striped field mouse, Apodemus agrarius. This mouse carries the Hantaan virus strain, which is the cause of the most severe clinical manifestation, HFRS. Throughout Europe, a less pathogenic group of hantaviruses, Puumala virus and related strains, cause nephropathia epidemica (NE). These viruses are endemic in the red bank vole, Clethrionomys glareolus. In the Balkans, Belgrade virus isolated from the yellow-neck mouse, Apodemus flavicollis, causes severe HFRS and is more closely related to Hantaan virus. In urban centers in the Far East and in rural areas in China, yet another strain, Seoul virus, with intermediate pathology, is endemic in house rats. Transmission appears to be from direct contact with rodent urine, but aerosol dispersal in heavily rodent urine–contaminated dust is also reported. Human disease occurs in agricultural areas with high human and rodent population densities, during military campaigns or exercises, in crowded urban housing, or in laboratories housing infected rodents. Person-to-person contact has not been reported, and these viruses do not cause nosocomial outbreaks.26 All ages and both sexes may be affected.

Until the early 1990s, HFRS was thought to be exclusively an Old World disease. In 1993, however, a pulmonary disease emerged in the United States that was found to be associated with hantavirus infection (hantavirus pulmonary syndrome; HPS), without a significant renal component.23 This virus, named the Sin Nombre virus, infects deer mice mainly in the southwestern United States. Risk of infection is related to exposure to infected rodent populations. By 1999, 277 cases had been reported from 31 states.27 Canada had also reported 32 cases, but it was South America that rapidly emerged as bearing the greatest burden, with over 500 reported cases in Argentina, Uruguay, Chile, Paraguay, and Bolivia. Two more related South American viruses were identified, Andes and Laguna Negra viruses, and a number of rodents of the sigmodontine and oligoryzomys species appear to be the natural hosts. The ecology of these viruses and rodents appears to be as complex as in the Old World. On both continents the viruses persist in its chosen rodent species despite robust immune responses, probably because they are able to continuously produce quasispecies, thereby escaping immune surveillance. Such viruses probably coevolved with their rodent hosts over as long as 9 million years.

Clinical and Laboratory Features

Hantaviruses are distinguished from most other hemorrhagic fever viruses by their longer incubation period of 2 to 3 weeks; for the Sin Nombre virus it is 9 to 33 days.28 The underlying pathology of both HPS and HFRS appears to be induction of cytokines in lung or renal tissue, and there is in fact overlap between the two syndromes,27 the difference being the predominant target tissue. The Old World hantaviruses induce a brief period of acute renal impairment, which may be as severe as anuria, requiring dialysis. However, this is usually self-limiting and of short duration. In HRFS vascular impairment and thrombocytopenia may result in bleeding that is usually minor, and the “leaky capillary syndrome,” which manifests variably as shock, renal failure, and in severe cases, ARDS. Renal involvement is seen in some HPS patients, but HPS pathology is centered on the lungs, and the ensuing severe respiratory distress carries a higher mortality than HFRS, earlier thought to be about 50%, but now recorded as being on the order of 27%. Though HPS mortality in South America appears to be higher than in North America, it appears that the illness:infection ratio is lower in South America. Mortality from HRFS is about 5% to 15%, with survival often depending on experience in fluid management, and death in NE patients is rarely recorded. In all these syndromes, raised hematocrit on admission to hospital probably reflects hemoconcentration due to leaking capillaries as in other hemorrhagic fevers, and thrombocytopenia, left shift in white cell counts, and hypoxia are characteristic, with the difference in HPS being that bleeding does not appear to be a problem. Pulmonary edema, which may be aggravated by injudicious IV fluids, can coexist with profound hypovolemic shock. Pathologic lesions are compatible with endothelial dysfunction and leakage, since no evidence of endothelial damage is seen. Viral antigen is seen in lung tissue of patients with HPS, and local production of cytokines is involved in the pathology.24,29 Recovery in survivors is rapid, and sequelae are not reported. Antibody has been found in up to 40% of some populations in South America, demonstrating that a large number of infected people have asymptomatic infection or mild disease, but in the north it is difficult to find antibody-positive people with no history of the disease.27 There is no nonhuman primate model for hantavirus infection in which to study these processes in more detail, but recently infection of cynomolgus macaques with the most virulent strain, Andes virus, has been established.30

In Asia and Europe, HFRS and NE are distinct clinical entities, very familiar to physicians working in endemic areas, who normally have little difficulty in making an accurate clinical diagnosis. HFRS was first described in the literature in Eastern Siberia in 1912, and NE in Scandinavia in the 1930s. During the Second World War Japanese investigators conducted some ethically highly controversial studies in prisoners in China. Throughout the twentieth century the disease was a well recognized hazard of war, particularly with soldiers bivouacking in the open or in trenches, and NE may have been one of the causes of “trench fever” in the First World War. There were probably about 30,000 cases in American troops during the Korean War.

HRFS is classically described to exhibit five stages. First is the febrile stage, during which the patient will have headache, malaise, and muscle pains. Petechiae are found not only on the soft palate and under the axillae, but also on the face, neck, upper hips, and thighs. Flushing and edema are marked around the eyes. The liver may occasionally be palpable, and abdominal and renal tenderness are most common. Platelet counts are reduced and may be very low. White cell counts are variable, but in some severe cases, marked transient neutrophilia can lead to confusion with bacterial infection. Proteinuria is invariable. After 2 to 4 days the patient either recovers spontaneously or progresses to the shock phase. This period of low blood pressure is followed within 24 hours by the oliguric or anuric phase. Within a further 24 hours or so most patients enter the diuretic phase. Finally the convalescent phase leads to full recovery with no well-documented sequelae. Although bleeding is uncommon, when it occurs it is usually in the convalescent phase and hemorrhage may be difficult to control, and death may result from intracranial bleeding. There are no clinical and laboratory data to suggest bleeding is due to hepatic failure and consumption coagulopathy is only recorded in terminal disease.

Laboratory Diagnosis

RT-PCR on blood is the most reliable method of diagnosis for all the hantaviruses. Quantitative RT-PCR is available (qPCR), with the ability to measure as few as 103copies per reaction.27 The viruses are difficult to culture, and it appears that many virions and viral antigens are bound in immune complexes in the blood, increasing the difficulty of isolation. In any event, viremia is an early event, so for success specimens must be taken early. The disease may be immunologically mediated, occurring mostly after disappearance of the virus. Laboratory diagnosis of hantavirus infections can also be made by demonstration of a rise in specific antibodies or demonstration of virus-specific IgM. Over 90% of patients have hantavirus-specific IgM and IgG antibody when first seen. Antibodies persist for up to 3 decades and generally react to highest titer with homologous antigens. Neutralizing antibody can be detected.

Specific Treatment

Intravenous ribavirin may be useful in HFRS if given on or before the fourth day of disease, in doses similar to those used for Lassa fever.31 Early treatment is essential. Ribavirin has similarly been recommended for patients with HPS,32 but some adverse effects were noted (principally anemia) and efficacy was not demonstrated (see Table 60-2). Further studies are needed. For both of these syndromes the focus in supportive management of the patient is on careful fluid balance. Intravenous fluids given during the febrile and shock phases may leak into the extravascular space, aggravating edema, particularly during the period of shock and oliguria. Judicious use of vasopressor drugs for severe hypotension is preferred. When pulmonary edema leads to severe dyspnea and impending acute respiratory failure, intubation and mechanical ventilatory support are mandatory. Diuretics are relatively ineffective in HFRS, but peritoneal dialysis or hemodialysis may be life-saving. With these supportive measures and careful fluid management, most patients have the potential to make a spontaneous and complete recovery.

Prevention

The variety and ubiquity of rodent carriers and their success in many ecologic settings make the reduction of hantaviral infections by rodent control virtually impossible. Improvement of housing, particularly food storage, will help prevent those cases caused by rodent invasion of houses. However, a large number of infections occur in rural settings associated with agriculture and forestry. Education of populations to avoid direct contact with rodents in these circumstances may only have marginal effect because the virus may be made airborne by agricultural practices. Some vaccines have undergone preliminary trials in China and North and South Korea with encouraging results, but none has undergone the safety and clinical trials necessary for licensing in any country.

Rodents experience silent, lifelong infection with arenaviruses, with persistent viruria, and are the primary source of contamination of the environment. Five arenaviruses cause hemorrhagic fevers: Lassa in West Africa, and Junin, Machupo, Guanarito, and Sabia in South America, causing Argentine, Bolivian, Venezuelan, and Brazilian hemorrhagic fevers, respectively. These viruses occupy circumscribed, sometimes remote ecologic niches, intrusion into which determines human infection. The viruses infect primarily through skin cuts and scratches and possibly the mucosae, contaminated with rodent urine, although there is evidence with Argentine hemorrhagic fever that dust generated by agricultural implements and laden with infected rodent urine may infect by aerosol. In endemic foci, the number of infected people may be very high. Human-to-human spread through blood contact is reported for Lassa fever both in community and hospital settings but is apparently rare with the other pathogenic arenaviruses.

Arenaviruses are enveloped, pleomorphic, membrane viruses containing two segments of single-stranded RNA, tightly associated with a nucleocapsid protein. The small RNA strand encodes the glycoprotein precursor and the nucleoprotein. Lassa and the South American hemorrhagic fever viruses are categorized as BSL4 laboratory agents. Disease may be severe and hemorrhagic, but with Lassa fever at least, mild or asymptomatic infection is also common.

Lassa Fever

Epidemiology

Lassa fever is confined to West Africa, where it is responsible for up to 16% of all adult medical admissions and about 30% of adult deaths on medical wards.33 It has been estimated that more than 100,000 infections with Lassa virus may occur each year, with several thousand deaths. All age groups and both sexes are affected. In endemic areas, illness:infection ratios range from 9% to 26%, and the proportion of febrile illness associated with seroconversion is between 5% and 14%. Five to eight percent of infected people may be hospitalized, of whom 17% will die if untreated. However, the fatality among all infections (hospitalized and nonhospitalized) is of the order of 2%.34

Person-to-person spread of Lassa virus occurs within homes as well as in hospitals. Hospital outbreaks are associated with inadequate disinfection and direct contact with infected blood and contaminated needles. Increasing and indiscriminate use of needles for intravenous therapy, or intramuscular injections in West African hospitals along with inadequate needle and syringe sterilization has led to large-scale epidemics. These epidemics can be devastating, resulting in the deaths not only of patients but also medical staff, surgeons, nurses, and other scarce personnel.6,35

Lassa fever is an increasing threat. It now affects communities in West Africa outside of its already broad area of rural endemicity. Indeed, urban Lassa fever in West Africa has been occurring with increasing frequency. In early 2000 hospital epidemics in large towns were again being seen in Nigeria, the most populous country in Africa. Since 1990, severe social disruption from conflicts and terror campaigns in Sierra Leone and Liberia have resulted in displacement of up to two million people—25% of the population of the area—with a substantial increase in the already large number of Lassa fever cases and deaths.36

Lassa fever is the exotic hemorrhagic fever most likely to occur in developed countries due to infection in returning travelers. This is because of the prevalence of disease in the endemic areas, and the relatively long incubation period (up to 3 weeks). In the year 2000 at least four cases were imported into Europe.37 All died, due in great part to delay in diagnosis, and therefore delay in instituting antiviral therapy. Increased cases in non-West Africans in 2000 has been seen as a result of United Nations peacekeeping efforts in Sierra Leone, where the rebels' stronghold is the center of the Lassa fever endemic area. One of the fatal cases in expatriates was an Englishman who had been working to disarm the rebel soldiers in the diamond mining area of Eastern Sierra Leone.38

The only known reservoir is Mastomys natalensis, one of the most commonly occurring rodents in Africa. Direct contact between virus-contaminated articles and surfaces and cuts and scratches on bare hands and feet may be the most important and consistent mode of transmission. The sporadic pattern of human infection in the household does not suggest aerosol transmission. Nosocomial spread in hospitals was and continues to be associated with inadequate disinfection and direct contact with infected blood and contaminated needles. Ill advised surgery performed on infected patients has resulted in infection and death among medical staff. Increasing and indiscriminant use of routine intravenous therapy in West African hospitals along with inadequate needle and syringe care has led to large-scale epidemics. Nevertheless, where simple but rigorous barrier nursing techniques have been applied, Lassa virus infection does not spread.8 In a study in London, none of 173 unprotected hospital contacts of a severely ill Lassa fever patient were infected.39

Clinical and Laboratory Features

Following an incubation period of 7 to 18 days, Lassa fever begins insidiously, with fever, weakness, malaise, severe headache—usually frontal—and a very painful sore throat.15 More than 50% of patients then develop joint and lumbar pain, and 60% or more develop a nonproductive cough. Many also develop severe retrosternal chest pain, and about half will have nausea with vomiting or diarrhea and abdominal pain. On physical examination respiratory rate, temperature, and pulse rate are elevated, and blood pressure may be low. There is no characteristic skin rash in Lassa fever, and petechiae and ecchymoses are not seen. About one third of patients will have conjunctivitis. More than two thirds have pharyngitis, half with exudates, diffusely inflamed and swollen posterior pharynx and tonsils, but few if any ulcers or petechiae. The abdomen is tender in 50% of patients. Neurologic signs in the early stages are limited to a fine tremor, most marked in the lips and tongue.40

Up to one-third of hospitalized Lassa fever patients progress to a prostrating illness 6 to 8 days after onset, usually with persistent vomiting and diarrhea. Patients are often dehydrated with elevated hematocrit. Proteinuria occurs in two thirds of patients. About half of Lassa fever patients will have diffuse abdominal tenderness but no localizing signs or loss of bowel sounds. The severe retrosternal or epigastric pain seen in many patients may result from pleural or pericardial involvement. Bleeding is seen in only 15% to 20% of patients, limited primarily to the mucosal surfaces or occasionally conjunctival hemorrhages or gastrointestinal or vaginal bleeding. Severe pulmonary edema and acute respiratory distress syndrome is common in fatal cases, with gross head and neck edema, pharyngeal stridor, and hypovolemic shock.

Over 70% of patients may have abnormal electrocardiograms including nonspecific ST-segment and T-wave abnormalities, ST-segment elevation, generalized low-voltage complexes, and changes reflecting electrolyte disturbance, but none of these correlate with clinical or other measures of disease severity or outcome, and they are unassociated with clinical manifestations of myocarditis.41 Neurologic signs are infrequent but carry a poor prognosis, progressing from confusion to severe encephalopathy with or without general seizures, but without focal signs. Cerebrospinal fluid is usually normal, but with a few lymphocytes, and low titers of virus relative to serum. Pneumonitis and pleural and pericardial rubs develop in early convalescence in about 20% of hospitalized patients, occasionally in association with congestive cardiac failure.

Although the mean white blood cell count in Lassa fever on admission to hospital is often normal, there may be early lymphopenia and later relative or absolute neutrophilia, as high as 30,000/μL. Thrombocytopenia is only moderate, and petechiae are uncommon. Endothelial and platelet dysfunction (despite adequate numbers of circulating platelets) are characteristic of severe disease,42 and the capillary leak syndrome results in ARDS, as with most hemorrhagic fever viruses.

A serum aspartate aminotransferase (SGOT) level of >150 U/L is associated with a case fatality rate of 50%, and there is a correlation between an increasing level and a higher risk of fatal outcome.34 Alanine aminotransferase (SGPT) is only marginally raised, and the ratio of SGOT to SGPT in natural infections and in experimentally infected primates is as high as 11:1. Prothrombin times and glucose and bilirubin levels are near normal, excluding biochemical hepatic failure, suggesting that some of the SGOT may be nonhepatic in origin.

Nearly 30% of patients with Lassa fever infection suffer an acute loss of hearing in one or both ears, not associated with severity of the disease.43 About half show a near or complete recovery by 3 to 4 months after onset, but the remainder have persistent significant sensorineural deafness, which after about a year will be permanent. Many patients also exhibit cerebellar signs during convalescence, particularly tremors and ataxia, but these usually resolve with time. Infrequent complications are uveitis, pericarditis, orchitis, pleural effusion, ascites, and acute adrenal insufficiency.44 Renal and hepatic failure are not seen.

Lassa fever may be a common cause of maternal mortality in many areas of West Africa, with case fatality about 20%.9 Fetal loss is as much as 87%, and does not seem to vary by trimester. Lassa virus is known to be present in the breast milk of infected mothers, and neonates are therefore at risk of congenital, intrapartum, and puerperal infection with Lassa virus. Lassa fever is common in children, but may be difficult to diagnose because manifestations are so general. In very young babies marked edema has been reported. In older children the disease may manifest as diarrhea or as pneumonia or simply as an unexplained prolonged fever.

Diagnosis

RT-PCR is the most rapid and accurate method for acute diagnosis and has been shown to be more sensitive even than virus isolation.45 Virus may persist in serum into the convalescence phase and coexist with antibody, and virus has been detected in urine as many as 60 days after onset.40 By the sixth day of Lassa illness antibodies are found in about 50% of patients. Virus is easily isolated from serum or tissues in cell culture, but this should be performed in BSL4 laboratory facilities. Virus has also been isolated from breast milk, spinal fluid, pleural and pericardial transudate, placenta, and from autopsy material, and may be recovered intermittently for 1 to 2 months in urine. Neutralizing antibodies to Lassa virus are absent in the serum of patients at the beginning of convalescence, and in most people they are never detectable.

Viral protein may be detected by monoclonal antibodies in tissue imprints (usually liver) on a microscope slide or using ELISA techniques. Efforts to detect antigen in conjunctival scrapings, buffy coat preparations, cells from pharyngeal aspirates, and urinary sediment have not been successful.

Specific Treatment

Ribavirin is effective in treating acute Lassa fever and should be given as early as possible.40,46 A five- to tenfold decrease in the case fatality ratio was demonstrated in patients treated with ribavirin compared with untreated patients when therapy was given within the first 6 days of illness. Treatment later in disease is effective, but less so.

Prevention

Avoidance of contact with rodent urine and with blood and tissues from infected rodents is an obvious precaution in the field. In the hospital, spread of Lassa virus is by blood-to-blood contact and can be prevented by simple barrier precautions.8 The importance of awareness by medical teams of the possibility of Lassa fever in patients in or from endemic areas cannot be overemphasized.47 Complete support should not be denied because of the suspected diagnosis. Carefully conducted intensive care or surgery by informed and trained personnel using maximum precautions (double gloves, educated staff, limited theater personnel) does not carry major risks.

South American Hemorrhagic Fever Viruses

The New World arenaviruses causing human disease are Junin (Argentine hemorrhagic fever; AHF), Machupo (Bolivian hemorrhagic fever, BHF), Guanarito (Venezuelan hemorrhagic fever; VHF), and Sabia (Brazilian hemorrhagic fever). All are endemic in geographically limited areas, but new, related viruses may emerge in other yet unstudied areas. The major rodent hosts are Calomys species, and the viruses are related to numerous other nonpathogenic arenaviruses from South American rodents (“Tacaribe complex”).

Argentine hemorrhagic fever was first recognized in the 1950s in the fertile farmland of northwestern Buenos Aires Province in Argentina, and Junin virus was first isolated in 1958. By 1990 about 21,000 cases had been reported over 30 years, but with the recent introduction of a live attenuated vaccine, this disease has diminished. Before the introduction of the vaccine, the disease was seasonal with peaks each May. The major routes of virus transmission to humans is probably through virus-infected dust and grain products, possibly from mechanical harvesters. There is no recorded person-to-person spread.

Clinical and Laboratory Features

The South American hemorrhagic fevers are similar in presentation, although Guanarito may more closely resemble Lassa fever.48After an incubation period of about 12 days there is insidious onset of malaise, high fever, severe myalgia, anorexia, lumbar pain, epigastric pain and abdominal tenderness, conjunctivitis, and retro-orbital pain, often with photophobia and constipation. Nausea and vomiting frequently occur after 2 or 3 days of illness. There is no lymphadenopathy or splenomegaly, sore throat or cough, but there is marked erythema of the face, neck, and thorax and conjunctivitis. Petechiae may be observed in the axillae by the fourth or fifth days of the illness. There may be a pharyngeal enanthem, but pharyngitis is uncommon. Relative bradycardia is often observed.

The second stage of illness begins with epistaxis, hematemesis, or acute neurologic disease. In contrast to the relative infrequency of bleeding in Lassa fever, the South American diseases are associated with hemorrhagic manifestations in nearly half of the patients, manifest as gingival hemorrhages, epistaxis, metrorrhagia, petechiae, ecchymoses, purpura, melena, or hematuria. Hypotensive shock, hypothermia, and pulmonary edema precede death. Renal failure has been reported. There is some electrocardiographic evidence of myocarditis. Fifty percent of AHF and BHF patients also have neurologic symptoms during the second stage of illness, such as tremors of the hands and tongue, progressing in some patients to delirium, oculogyrus, and strabismus. Meningeal signs and cerebrospinal fluid abnormalities are rare.

A low white blood cell count, under 1000/μL, and a platelet count under 10,000/μL are invariable. Bleeding and clot retraction times are concomitantly prolonged, although DIC is apparently not a significant feature. Proteinuria is common, and microscopic hematuria also occurs. Liver and renal function tests are only mildly abnormal. Mortality is about 16% in laboratory-confirmed hospitalized patients with untreated AHF. There are no estimates of overall mortality from population-based surveys.

A late neurologic syndrome has been described in AHF, consisting mainly of cerebellar signs, and associated with high-titer antiserum used in treatment in about 10% of cases.49 The syndrome begins between 4 and 6 weeks after onset of acute illness and lasts less than a week. It is characterized by fever, headache, ataxia, and intention tremors, and a mild cerebrospinal fluid pleocytosis with anti-Junin virus antibody in the CSF. Most patients recover within 3 months. Mild permanent damage to acoustic centers has been detected. AHF is also reported to be severe in pregnancy, but no formal studies are available, and women are less frequently affected.

Despite the different degrees of bleeding, there are sufficient similarities between the course of disease in AHF, BHF, and Lassa fever to speculate that they share similar pathophysiologic pathways. Organ function, other than the endothelial system, appears to remain intact, and the critical period of shock is brief, lasting only 24 to 48 hours. Hepatitis is mild, and renal function is also well maintained. Bleeding is more pronounced with AHF and BHF than Lassa fever, but it is not the cause of shock and death. Capillary leakage is significant, with loss of protein and intravascular volume being much more pronounced than loss of red cells.

In marked contrast to Lassa fever, the antibody response to Junin virus is effective in clearing virus during acute disease and may also be sufficient to protect against infection. Neutralizing antibody may be detectable at the time the patient begins to recover from the acute illness, and the therapeutic efficacy of immune plasma in patients with Junin infection is directly associated with the titer of neutralizing antibody in the plasma given.

Diagnosis

RT-PCR is the most rapid and accurate method of acute diagnosis. The IFA may be positive by the end of the second week of illness. Neutralizing and complement-fixing antibody to Junin are usually detectable 3 to 4 weeks after onset. IgM is more difficult to read by IFA, and an ELISA system may be preferred. Virus may also be cultured from serum, but this should be performed in Biosafety Level 4 conditions.

Specific Treatment

In contrast to Lassa fever, convalescent-phase plasma has been shown to be highly successful in Argentine hemorrhagic fever, reducing the mortality from 16% to 1% in patients treated in the first 8 days of illness.49 Efficacy is directly related to the concentration of neutralizing antibodies. Late initiation of therapy is less successful. Availability of appropriately screened plasma may be a problem. Ribavirin may also be effective in treating South American hemorrhagic fever.

Prevention

The human-rodent encounter resulting in AHF occurs during the crop harvests, and there are no means of controlling feral rodents. A successful live attenuated vaccine, Candid 1, for AHF has now undergone phase III studies, and is in use in the endemic area of Argentina, where it has almost eliminated the disease. The vaccine has proved safe in large-scale trials, and has a protective efficacy of 84%.50

Human infections with filoviruses are exceedingly rare, but their occurrence has invariably been dramatic and mysterious.51 The first appearance was in Marburg in 1967. There were 7 deaths among 32 infected laboratory technicians, medical personnel, animal care personnel, and relatives. Primary cases had been exposed to tissues and blood from African green monkeys imported into Germany and Yugoslavia from Uganda. A unique virus was isolated from these with a strange looped and branched filamentous form, hence named filovirus. In 1976 and 1979, epidemics of a hemorrhagic disease with very high mortality in northern Zaire and in southern Sudan were found to result from two strains of a related yet distinct filovirus, named Ebola virus after a river in Zaire.

At first thought related to rhabdoviruses, it has become clear that these viruses form a family of their own, now designated filoviridae. Nucleotide sequence analyses now place the family in the order mononegavirales, which also includes the paramyxoviridae (e.g., respiratory syncytial virus) and rhabdoviridae (such as rabies). Filoviruses are among the largest known viruses, with highly variable length (up to 14,000 nm), but uniform 80-nm diameter, with a helical nucleocapsid, consisting of a central axis 20 to 30 nm in diameter. The virions contain a single negative-strand RNA genome.

Epidemiology

Since the 1967 Marburg outbreak, there have been three further, isolated, primary human Marburg infections (and only two secondary cases), all in adventurous expatriates in remote parts of Africa. In the late 1990s, epidemics in eastern Congo (formerly Zaire) were traced to miners entering disused, partially flooded gold mines. These outbreaks may be ongoing, but the civil disturbances in the area have made it impossible to get accurate information. Between 1976 and 1979 three separate major outbreaks of Ebola occurred in northern Zaire and the southern Sudan.52,53 In each instance, early index cases were rapidly followed by dissemination, mainly in hospitals and clinics, where reuse of needles for injection and exposure of staff to blood were strongly implicated in transmission. Case mortality ranged from 53% to 88%.

Ebola disappeared after 1979, only to re-emerge in 1994 in Kikwit, Zaire, again amplified in a hospital. The outbreak involved some 315 confirmed cases, among whom 244 died (77% fatality).54,55 The virus from the Kikwit outbreak is apparently identical to the 1976 Zaire strain of Ebola. In 1994 a Swiss animal researcher reported 30 deaths in monkeys due to Ebola in the Ivorial rainforest bordering Liberia, and in 1996, 27 cases were reported from Gabon with 18 deaths (67% fatality) in humans. In this last outbreak, 12 of the fatalities had had direct contact with the blood of a dead chimpanzee.56

Person-to-person spread has been the major mode of transmission in epidemics. Contact with patients ill with Ebola is the most important factor in determining risk of illness. Other risk factors associated with human-to-human transmission are infection from contaminated materials such as needles, contact with blood or secretions, preparation of a body for burial, or occasionally, sexual contact. Close contact with blood or tissues of infected monkeys is also important.56 The virus enters through mucous membranes or skin lesions, and outbreaks have been abruptly terminated when blood transmission was interrupted.57 Relatively frequent epidemics of Ebola Zaire in Gabon and the Congo, Ebola Sudan in Uganda, and Marburg in eastern Congo have been reported, and several outbreaks of Ebola Zaire are ongoing, and apparently closely linked to a major epizootic in great apes. This Ebola epizootic, combined with hunting and loss of habitat, is now thought to threaten chimpanzees and gorillas in the area with extinction.58

Clinical and Laboratory Features

The incubation period is 3 to 12 days, 3 to 7 days for needle transmission, and 6 to 12 days for person-to-person spread. The illness:infection ratio for Marburg and Ebola viruses approaches unity, since few if any asymptomatic infections have ever been observed.

The disease caused by the African filoviruses is dramatic.52,53,59,60 Onset is abrupt with fever, severe headache (usually periorbital and frontal), myalgia, arthralgia, conjunctivitis, and extreme malaise. Sore throat is a common symptom, often associated with severe swelling and dysphagia, but no exudative pharyngitis. A papular, eventually desquamating rash may occur in some patients, especially on the trunk and back; a morbilliform rash has been observed on white skin. Gastrointestinal symptoms develop in most patients on the second or third day of illness with abdominal pain, cramping, diarrhea, and vomiting. Jaundice is not a feature of Marburg or Ebola disease. Persistence of vomiting and the onset of any signs of mucosal bleeding carry a high risk of fatal outcome. Bleeding begins about the fifth day of illness and is most commonly from the mucous membranes, gastrointestinal tract, gingiva, nasopharynx, and vagina. The most profound physiologic alteration is shock (manifested by hypotension, effusions, and facial edema), which is invariably fatal. Severe, acute fluid loss often with frank bleeding into the tissue and into the gut results in dehydration, and electrolyte and acid-base imbalance. Infection in pregnancy results in high maternal fatality and virtually 100% fetal death. CNS involvement has led to hemiplegia and disorientation and sometimes frank psychosis. Even in convalescence patients show prolonged weakness and severe weight loss, and in a few survivors serious but reversible personality changes are recorded, namely confusion, anxiety, and aggressive behavior.

Thrombocytopenia is invariable, but bleeding is not usually of sufficient volume to account for the shock, nor is it associated with solid evidence of DIC in the small number of animals or humans studied so far. Platelet dysfunction has also been described in experimentally infected nonhuman primates. Profound lymphopenia early in disease is accompanied by marked neutrophilia. Laboratory evidence of moderate DIC appears only in the terminal stages. Liver enzymes (SGOT and SGPT) are raised, but the rise in SGOT is disproportionately higher than SGPT, as in Marburg disease.

At autopsy there is widespread hemorrhagic diathesis into body cavities, membranes, and soft tissue, with focal necrosis in liver, lymph nodes, ovaries, and testis. Most prominent are eosinophilic inclusion bodies in hepatocytes (Councilman-like), without significant inflammatory response.

Very little is understood about the immunology of Ebola virus infections except for the observation, made many years ago, that neutralizing antibodies are difficult to demonstrate in both humans and primates, and that like many zoonotic viruses, notably Lassa virus and now the severe acute respiratory syndrome (SARS) human coronavirus, Ebola virus appears to be able to circulate in humans in the presence of detectable antibody and to show varying ability to persist at least for short periods following acute infection. Nevertheless protection must be achieved, since reactivation disease should have been observed by now, were it occurring. Studies using specimens from patients in Gabon have shown that the innate immune system plays a very important role in disease.61 Evidence from these studies and from animal studies suggest that the proinflammatory response is a central figure both in pathogenesis of severe disease and in protection from disease, in that an early, orderly innate immune response was observed in infected individuals who never developed disease.62 Conversely primate studies have shown that infection of monoculear phagocytes is critical, and that these trigger a cascade of cytokines/chemokines and oxygen free radicals, and that it is this process, not direct viral replication destroying critical cells, which lead to the manifestations of disease. These manifestations are associated with massive intravascular apoptosis.61,63

Diagnosis

RT-PCR is the method of choice for acute diagnosis of Ebola hemorrhagic fever.64 Extreme care should be taken in both drawing and handling blood specimens, since virus titer may be extremely high, and the virus is stable for long periods even at room temperature. High- or rising-titer filovirus-specific IgG is diagnostic, as is the presence of IgM by ELISA or IFA. Patients may die, however, before developing detectable antibodies. Virus may be isolated and identified within 2 to 3 days if suitable facilities are available. An antigen detection ELISA system is also used, but is less sensitive than RT-PCR.64 Antigen may also be detected in biopsies, including skin.65 Unexplained, nonspecific reactions in antibody assays (sometimes up to 15% of sera), particularly with IFA, have plagued serologic studies.

Specific Treatment

Patients will require full intensive care support, including mechanical ventilation, along with blood, plasma, or platelet replacement. Provided strict barrier precautions are observed, intensive care should not be denied and may be life saving. Every effort is justified, since the crisis is short-lived, and complete recovery can be expected in most survivors.

No antiviral therapy (including convalescent plasma, ribavirin, or related compounds) has been shown to be effective against either Marburg or Ebola virus infection in patients or in experimentally infected nonhuman primates. Suitably screened and stored human plasma is in any event unavailable. Human interferon is ineffective in vitro.

Flaviviruses are positive-strand RNA viruses, spread by mosquitoes. Yellow fever and dengue virus cause hemorrhagic disease. Dengue is the only exclusively human VHF.66

Yellow Fever

Epidemiology

Since it was first described in 1498, yellow fever has been a recognized scourge of West Africa, and subsequent to colonization and the slave trade, the West Indies, and Central and South America. Outbreaks in the Americas in the nineteenth century were reported as far north as Hartford, Connecticut, and frequent devastating epidemics were familiar to the southern United States. Mosquito eradication programs successfully controlled epidemics from the beginning of the twentieth century, and with the introduction of an effective vaccine in the 1930s, yellow fever became a preventable disease. Nevertheless it continues to circulate in West and Central Africa, and in the interior of South America. Large epidemics are regularly seen in Africa, particularly Nigeria and the Cameroon, in areas where for political or economic reasons vaccine coverage is low.67

Yellow fever virus is transmitted to humans by mosquito bite; the principal vector is Aedes aegypti, but other Aedes species are also capable of transmission. Two epidemiologic forms are recognized, sylvatic and urban. In the sylvatic cycle the virus circulates from monkey to monkey via mosquitoes, with humans occasionally being infected peripherally. Several species of monkey are involved, but in general, South American nonhuman primates are much more susceptible to disease than Old World primates, and they mostly die from the infection. In the urban form of the disease, the virus is maintained by transmission from humans to humans by mosquito. Direct person-to-person spread has not been reported. Distribution in Africa and South America generally follows that of the equatorial rain forests. For reasons that are unclear, Asia has been spared this virus. Nevertheless, globally the disease affects more than 200,000 persons a year in endemic areas and is a significant risk to unvaccinated travelers.68

Clinical and Laboratory Features

Though the disease has been well described in the past, there are no recent data concerning either clinical disease or pathogenesis,66,68 relegating this fulminating disease to the neglected diseases of the poor of the developing world. The incubation period is usually 3 to 6 days. Clinical presentation varies from a nonspecific low-grade febrile disease to a fulminant hepatitis with renal failure and hemorrhage, which is usually fatal. Onset is abrupt with fever and chills, intense headache, generalized myalgia, nausea, and vomiting. High fever continues for about 3 days, with conjunctival injection, edema and hyperemia of the face and neck, and occasionally a macular or scarlatiniform rash. Then epistaxis and gum bleeding may develop, and there will be proteinuria. Some patients then deteriorate to the “yellow” phase, with jaundice, hemorrhages, violent epigastric pains, and vomiting that may become black from altered blood, the “vomito negro” from which they rarely recover. The renal syndrome progresses from oliguria to anuria. Hypotension and shock appear late. Electrocardiographic changes consist of increase in PR and QT intervals, with bradycardia accompanying high fever. Terminal phases are characterized by agitation, delirium, and convulsions, leading to coma and death. Mortality as high as 50% has been recorded in urban outbreaks in Nigeria.67 When recovery occurs, it is complete, without sequelae.

There is generalized leukopenia and thrombocytopenia. Liver function tests show high SGOT and SGPT, but as in other VHFs, SGOT is disproportionately high. Unlike other VHFs, however, bilirubin rises rapidly and continues to be high in convalescence. Elevated bilirubin and serum aminotransferase levels reflect the severity of illness, and as in Lassa fever, are of prognostic value. In fatal cases hepatic failure contributes to the hypoglycemia and metabolic encephalopathy. Virus may be recovered from the liver, and it is likely that direct viral injury accounts for most of the observations. There is marked albuminuria and oliguria, and in severe cases renal failure may be observed. Metabolic acidosis is an important component of the terminal stages of this disease.

Bleeding is common, often with upper gastrointestinal hemorrhage. Thrombocytopenia is prominent, and prolongation of the prothrombin time, APTT, and clotting time are reported along with reduced levels of clotting factors, particularly those derived from the liver. Disseminated intravascular coagulation may be an important component of the terminal stages and may be associated with the severe hepatic injury. Unlike some other flaviviruses (e.g., West Nile virus) direct involvement of the central nervous system by this virus probably does not occur; however, metabolic encephalopathy results in agitation, mania, delirium, convulsions, and coma. There are no focal neurologic signs.

Yellow fever is by far the most hepatotropic of the VHF viruses. Histology shows the characteristic Councilman bodies, consisting of eosinophilic degeneration in mesolobular or midzonal regions. Generally the liver shows coagulative necrosis with some sparing about the centrilobular veins and portal vessels.

Diagnosis

RT-PCR is the method of choice. Diagnosis is made by isolation of the virus from blood, but the virus is labile and culture may be difficult, requiring suckling mouse inoculation. Though person-to-person spread is not a feature of this disease, care should nevertheless be taken that specimens are handled by persons vaccinated against yellow fever. Serologic data are more difficult to interpret than in other VHFs because of the notorious cross-reactivity of flaviviruses, particularly in persons from endemic areas who may have encountered many of the myriad of related viruses that abound. “Original antigenic sin” is a major feature of all flavivirus infections, since infected people tend to have highest antibody titers to whichever virus was their first exposure, regardless of the virus causing their current or most recent infection. Specific IgM assays using modern ELISA techniques may be helpful, but again in patients from endemic areas pre-existing flavivirus antibody may yield confusing results. For all these reasons diagnosis of yellow fever outside the setting of a known epidemic may be difficult, and RT-PCR performed on specimens (blood or tissue) early in infection is the surest means to identify the virus. Antigen detection techniques using ELISA have also been developed, but await full evaluation.

Specific Treatment

Management is supportive. Yellow fever virus is not susceptible to ribavirin or to any other known antiviral agent.

Prevention

Prevention is by vaccination with the 17D strain of yellow fever virus, one of the oldest and most successful vaccines. This is a live attenuated vaccine. A single shot apparently provides lifelong protection, although WHO recommends validity for 10 years. The current preparation has been one of the safest vaccines in existence, with millions of doses having been given over more than 50 years. Recently there have been isolated reports of actual yellow fever–like disease following vaccination. Nevertheless, the long history of safety of the 17D strain now makes it an important candidate for insertion of other genes, particularly other flavivirus genes such as dengue and hepatitis C. Mosquito control is the ultimate control method, resulting in near complete eradication of disease, as was demonstrated in Brazil and the southern United States. However, the recent resurgence of yellow fever in Brazil that accompanied the reintroduction of Aedes aegypti to many areas of that country is a warning that eradication is a mirage given a zoonotic focus of the virus, and that constant surveillance is essential.

Yellow fever is one of four notifiable diseases in the international health regulations of WHO. Notification of suspected as well as confirmed cases is mandatory.

Dengue Hemorrhagic Fever

Dengue virus infection is frequently asymptomatic or a mild febrile illness. However, dengue hemorrhagic fever (DHF) usually occurs in children less than 15 years of age and in some circumstances in adults. There are four dengue virus serotypes, and DHF is apparently due to infection with a second serotype, though in outbreaks due to newly introduced virus in nonimmune populations, fulminant hemorrhagic disease may be seen in young adults. It is hypothesized that the replication of the second infecting serotype is enhanced by pre-existing antibody at suboptimal levels, but this remains controversial.66

The epidemiology of dengue is recent, dating back to the Second World War. Whether it was the disruption of this war which allowed the virus to spread, or more likely the rapid and explosive population increases and development of large urban areas and megacities is unclear. However, the end result is clear: dengue is endemic in the cities and the countryside of Southeast Asia and has spread widely east and south to Africa and even threatens Australia.69 Deterioration of urban environments with crowding and mosquito infestation also favors the virus. Reintroduction of dengue has now been documented in Central and South America including Brazil, Ecuador, Bolivia, Peru, Venezuela, and many Caribbean islands. It has also been reintroduced into the Pacific. Epidemics are reported each year from Central and South America, and cases are now seen in south Texas. The potential for further spread is great, particularly in the southern United States and more widely in South America. The disease must be considered in febrile travelers from all the affected areas, which now includes most of the tropical regions of the world and many subtropical regions. Unlike yellow fever, the natural history is a cycle involving only mosquito and human, the main vector for its spread is humans, but still it has proved difficult to control given the ubiquity of both mosquito and human hosts.

Large epidemics occur, often annually in all endemic areas. In areas where the virus has been recently reintroduced the most severe disease may be in adults, but in Southeast Asia, DHF is a disease of young children, thousands of whom are affected annually. Epidemics tend to occur in the hot rainy season. The vector is Aedes aegypti, but Aedes albopictus has also been implicated. Aedes prefer to breed in clean cool water, and in Southeast Asia are abundant in drinking water pots kept in the shade. Intercontinental trade in used tires containing moisture was thought to be responsible for the recent introduction of Aedes albopictus in the United States, but no disease or indeed presence of virus has been demonstrated in association with these tires.

Clinical and Laboratory Features

The incubation period is 2 to 10 days. Onset is abrupt, with fever, severe headache, and a combination of myalgia, backache, photophobia, and retro-orbital pain, accompanied by anorexia and vomiting. Most patients then recover uneventfully, but some may progress to dengue hemorrhagic fever or dengue shock syndrome (DSS). These conditions are characterized by low platelet counts, petechiae, and hypovolemic shock. This is normally of short duration and self-limiting, but a few patients may deteriorate and die with pulmonary edema or intracranial bleeding. Patients may also have marginally raised liver function tests, with SGOT levels higher than SGPT levels.

The cardinal signs of dengue hemorrhagic fever are shock, with narrow pulse pressure and petechiae. Hepatic involvement is variable. There is often severe thrombocytopenia and marked intravascular volume depletion with edema as in HFRS. During the acute phase there is lymphopenia. Serous effusions are common. These are transudates with high albumin content. Endothelial cell biopsies show swelling of endothelial cells, with enlarged endothelial gaps in some sections, but no necrosis of cells themselves. Indeed since most patients recover promptly, responding to careful fluid management, it must be assumed that the increased permeability is the result of some acute metabolic or biochemical dysfunction.

The pathogenesis of DHF has been the subject of much debate. Various studies have reported complement consumption, activation of the plasma kinin system, circulating immune complexes, leukopenia, depression of T-cell function, increased natural killer cell activity, or interferon-γ production by dengue-stimulated T lymphoctyes. It has been suggested that an initial infection with one type of dengue virus, followed by a second infection with another serotype, induces an immune response resulting in enhancement of virus entry into monocytic cells where it replicates. This observation, which has been under consideration for decades, still remains controversial.

Diagnosis

Serologic diagnosis is difficult to interpret, again because of cross-reactions with pre-existing flavivirus antibodies, particularly in endemic areas. The RT-PCR detecting dengue virus RNA in serum or white cells is likely to be the most practical approach where it is available and affordable. IgM assays may be useful in identifying acute infection. Identification of the specific virus type is difficult by antibody or antigen detection methods, and may require virus isolation. However, this information is principally of importance for epidemiologic evaluation of epidemic and endemic behavior of the viruses.

Specific Treatment

Management of DHF and DSS is supportive. Dengue virus is not susceptible to ribavirin or to any other antiviral agent.