eMedicine - Dermatologic Aspects of Bioterrorism Agents : Article by Chad M Hivnor

You are in: eMedicine Specialties > Dermatology > ENVIRONMENTAL

Dermatologic Aspects of Bioterrorism Agents

Article Last Updated: Jun 19, 2008

In light of recent events in the United States, Americans find themselves at the beginning of the 21st century living with the specter of biologic terrorism delivered to their nation's doorstep. As demonstrated by the recent anthrax threat in the United States, dermatologists have a unique opportunity to be the first to recognize a bioterrorism attack.

Of the top 12 category-A, high-priority biologic agents that pose a risk to national security, only botulism toxin and the Lassa fever virus do not elicit any cutaneous manifestations.¹ The US Centers for Disease Control and Prevention (CDC) has listed the following agents as those of greatest concern: Bacillus anthracis (anthrax), Yersinia pestis (plague), variola (smallpox), Clostridium botulinum toxin (botulism), Francisella tularensis (tularemia), filoviruses (Ebola and Marburg hemorrhagic fevers), and arenaviruses (Lassa fever and Junin or Argentine hemorrhagic fever).^{2, 3} See Dermatologic Aspects of Bioterrorism Agents, Anthrax for a detailed discussion of anthrax infection.

By 1991, the Iraqi regime had weaponized anthrax, botulinum toxin, and aflatoxin. At least 17 countries, not including terrorist organizations, possess biologic weapons. Therefore, dermatologists must become familiar with the cutaneous manifestations of these potential threats.

The Medscape Disaster Preparedness and Aftermath Resource Center may be of interest.

Yersinia pestis

Y pestis is the cause of plague and is responsible for as many as 200 million deaths throughout recorded history. The plague mainly occurred in 3 pandemics during the 6th, 14th, and 19th centuries.⁴ In 2006, 13 cases of human plague were reported in the United States, occurring in New Mexico, California, Colorado, and Texas. This is the largest number of human plague cases in a single year in the

United States

since 1994.⁵ Because of the ease with which it has caused death throughout history, Y pestis has been an attractive agent in offensive biologic warfare.

Y pestis is a gram-negative, nonmotile, non–spore-forming coccobacillus that exhibits bipolar, safety pin–type staining with Giemsa, Wright, or Wayson stains. Y pestis possesses a number of virulence factors that make it an especially dangerous pathogen.

The F1 capsular antigen is particularly important for pathogenesis and infectivity.^{4, 6} A carbohydrate protein envelope forms around the bacilli when it grows at temperatures higher than 33°C.⁴ At lower temperatures, Y pestis does not express the F1 capsular antigen; therefore, neutrophils and monocytes can easily engulf the bacilli. Neutrophils can destroy bacilli, but monocytes cannot. In monocytes, Y pestis multiplies and develops its antiphagocytic capsule (ie, F1 capsular antigen) that prevents even neutrophils from digesting it. The bacilli then multiply in the lymph nodes and blood, and they may then travel throughout the body.⁴ Ten to 50-fold fewer bacteria can cause death in mammals at 37°C compared with the amount needed at 25°C.⁴

At least 31 types of fleas and more than 200 species of mammals in 73 genera serve as reservoirs.⁴ The Oriental rat flea Xenopsylla cheopis is the classic vector, although any flea could probably transmit plague under the right conditions, and the human flea Pulex irritans may have been the major vector during the great epidemics of plague.

The genomic sequence of Y pestis was determined in strain CO92, which was obtained from a fatal human case of pneumonic plague that was contracted from a cat.⁷ Other bacteria and viruses have transferred many genes, including the genes for adhesion molecules, secretion systems, and insecticidal toxins, to Y pestis.⁷ In fact, the Pla protease that is essential for virulence when Y pestis enters the host via the subcutaneous route was acquired by means of a plasmid; the Pla protease is 1 of 2 proteases unique to Y pestis.

With the modification and acquisition of new sequences, the bacterium may be evolving into a new pathogenic species.⁷ These facts underscore our ability to genetically reengineer bacteria for use in biologic warfare. Some reports in the lay press state that Russian scientists engineered a multidrug-resistant strain of Y pestis, but no scientific evidence supports this claim.^{6, 8} (Refer to the article by Perry and Fetherston⁴ for an extensive discussion of the history and etiologic and epidemiologic features of Y pestis.)

Most patients with endemic plague present with tender, erythematous lymphadenopathy that most commonly affects the groin, and most cases involve bubonic plague (boubon is Greek for groin). Buboes may become pointed and drain spontaneously, or they may require incision and drainage because of pronounced necrosis.⁹ A lesion occurs at the site of a fleabite in as many as 10% of cases.¹⁰ Bubonic plague that spreads to the bloodstream results in septicemic plague, which can cause petechiae, ecchymosis, and acral necrosis.^{10, 11, 12} From the blood, the meninges or the lungs may become infected. Such secondary pneumonic plague is rapidly fatal and transmissible.

Because bacilli in pneumonic plague lesions possess an antiphagocytic capsule (F1 capsular antigen), its transmission by means of a cough or a sneeze can lead to death in a previously healthy individual within 1-2 days of exposure. The median infective inhaled dose is 100-500 bacilli.¹³ Persons within 2-5 feet of the infected individual can inhale respiratory droplets. The ensuing flulike illness rapidly progresses to overwhelming pneumonia with cough and bloody sputum. If not treated within 24 hours of the appearance of symptoms, almost all patients with pneumonic plague die.¹³ Risk factors for pneumonic spread include indoor contact, cool temperatures, increased humidity, and crowded conditions.^{14, 15}

In a biologic warfare scenario, aerosolized bacteria could lead to pneumonic plague. In patients with terminal pneumonic plague and septicemic plague, acral necrosis and petechiae develop with large ecchymoses on the back and dark cyanosis.^{4, 12, 16} See Media Files 10-11. These lesions, or the acral necrosis, may have been the origin of the medieval term "the Black Death."³

In one report, 25% of patients had papules, pustules, vesicles, or eschars near the bubo or in an area that drained the bubo.¹⁷ Rare cases of ecthyma gangrenosum–like lesions and carbuncles due to Y pestis infection are reported.^{9, 18} These lesions appear as small, centrifugally spreading crusts preceded by a border of vesicles that grow to 3- to 4-cm eschars after 72 hours. The most common cutaneous manifestation of plague, the bubo, would not be present in a biologic warfare scenario unless the Japanese plan of releasing infected fleas is resurrected, which is discussed later.

One must have a high index of suspicion to diagnose plague in the absence of buboes. Stains and cultures of blood, bubo aspirates, sputum, cerebrospinal fluid, or even skin scrapings may be helpful in isolating the organism. The diagnosis should be considered in an otherwise healthy person who presents with bloody sputum, shortness of breath, chest pain, and a fulminant course.

Histopathologic examination of the draining lymph nodes shows edema, hemorrhage, necrosis, and a ground-glass amphophilic material, which represents bacteria and necrotic debris. The bacilli can be visualized in formalin-fixed tissues using hematoxylin and eosin, Gram, silver impregnation, and Giemsa stains, but immunohistochemistry, direct fluorescent antibody (DFA) testing, or polymerase chain reaction (PCR) must be used for specific identification of Y pestis.¹⁹

In 2007, an in-house, real-time PCR assay using minor groove-binding probes was developed that can simultaneously detect and identify B anthracis, Y pestis, and F tularensis. This assay may allow for the rapid detection and identification of 3 category-A bacterial species biological threats.²⁰

Y pestis is susceptible to heat, disinfectants, and sunlight, but it can survive in the environment for varying lengths of time. If needed, soap and water is effective for decontamination.²¹

A New Mexico review¹⁰ of 27 patients from 1965-1989 demonstrated 5 presentations: (1) classic bubonic plague (10 cases, no deaths); (2) upper respiratory tract infection syndrome with fever, sore throat, and headache (5 cases, 3 deaths); (3) a nonspecific febrile syndrome of fever, chills, and anorexia (5 cases, 3 deaths); (4) gastrointestinal or urinary tract infection syndrome (4 cases, no deaths); and (5) meningitis (3 cases, no deaths). The 6 fatal cases all had corresponding upper respiratory tract infection or nonspecific febrile syndromes similar to that of gram-negative sepsis. Of the 27 cases, only 10 involved classic buboes; therefore, one cannot rely on the presence of a bubo in diagnosing endemic disease, much less terrorist-induced disease.

The formalin-killed plague vaccine does not protect individuals against inhalational plague, and it is no longer in production. More immunogenic live-attenuated vaccine strains have not resulted in increased immunogenicity, and they can sporadically revert to virulent, wild-type bacteria.^{4, 6} New vaccines are being developed for primary pneumonic plague.²² Multiple reports describe different vaccines based on the F1 capsular antigen providing protection in mice, and this will likely be a source of future prevention and treatment.

Attenuated Salmonella-based oral plague vaccines have been demonstrated to protect against the fatal pneumonic form of plague. These vaccines would allow for easier and less costly mass production, distribution, and administration than for purified protein–based vaccines.^{23, 24} Studies from 2007 show that vaccination with live-attenuated Y pestis primes CD4 and CD8 cells that synergistically protect against deadly pulmonary plague infection.²⁵

Most strains of Y pestis are sensitive to streptomycin, gentamicin, tetracycline, and doxycycline.²⁶ Evidence from 2006 suggests that development of antibiotic resistance is less likely to occur with doxycycline. Doxycycline is markedly less expensive than most fluoroquinolones, and, based on clinical case studies, it appears to have similar efficacy in most scenarios. It has been shown to be effective against anthrax, plague, tularemia, Q fever, and brucellosis. Doxycycline should be considered as a first-line antibiotic in the management of bioterrorism agents.²⁷ Although in vitro testing has demonstrated the effectiveness of quinolones, rifampin, third-generation cephalosporins, and amoxicillin, these have not been used to any great degree in human cases.⁶ Isolation should also be considered in persons with septicemic or pneumonic plague.

Plague would most likely be transmitted as an aerosol in the event of a terrorist or biologic warfare attack.¹³ Unlike anthrax, plague may be transmitted from person to person, and it also poses the risk of rapid death. The US government studied Y pestis as a potential offensive weapon in the 1950s. Other countries are suspected of weaponizing plague.¹³

The Japanese used a more creative approach in China during World War II. Human fleas (Pulex irritans) were grown and then infected with Y pestis. These organisms were released into several Chinese cities, where small epidemics of plague ensued. Typically, animal hosts die in epizootic outbreaks before humans are infected, but, in these cases, humans and then the animals began dying of plague. (Inglesby et al⁶ provide an extensive discussion of plague as a biologic weapon.)

A bioterrorism attack with plague might use a naturally occurring or bioengineered drug-resistant strain.²⁸ The first multidrug-resistant isolate of Y pestis was identified in 1995, which contained a self-transmissible plasmid (pIP1202) that conferred resistance to many of the antibiotics recommended for prophylaxis and treatment of plague.²⁹ From 2002-2005 retail meat samples were collected in the United States that contained multidrug-resistant enterobacterial pathogens carrying Y pestis pIP1202-like plasmid backbones. This common plasmid backbone has been demonstrated to be spreading among multidrug-resistant agricultural zoonotic pathogens, representing a significant public health concern.³⁰

Francisella tularensis

F tularensis, a gram-negative pleomorphic coccobacillus, causes tularemia, which is described as a plaguelike illness that affects rodents and, subsequently, humans. Currently, 4 subspecies of F tularensis are recognized (ie, tularensis, holartica, mediasiatica, novicida).³¹ F tularensis exists in nature as 2 serotypes: type A, which has a 5% mortality rate, and type B, which rarely leads to death.^{22, 32}

Deer flies and a variety of ticks are reported to be vectors that can transovarially pass on the bacilli to offspring. Infected rabbits and carcasses are classic sources of infection, but they are not as important as the deer fly and tick vectors.^{33, 34, 35} In Europe, outbreaks of tularemia have occurred in Kosovo, Turkey, Spain, and Scandinavia in the last 10 years.³⁶ In the United States, tularemia is highly endemic in Oklahoma, Missouri, Arkansas, and South Dakota.³² In 1990-2000, 1368 cases were reported to the CDC.³²

Fifty-nine presumed or confirmed tularemia cases were reported from 2000-2006 in Martha's Vineyard, Massachusetts. More than 60% of these cases were presumed due to inhalation of the causative agent. Martha's Vineyard is the site of the only 2 recognized outbreaks of primary pneumonic tularemia in the United States. Important risk factors include landscaping activities such as lawn mowing or brush cutting; however, the fomites that serve as the basis for this increased risk in landscapers remains to be elucidated.³⁷

Tularemia can occur in 6 clinical forms: (1) ulceroglandular, (2) glandular, (3) oropharyngeal or gastrointestinal, (4) typhoidal, (5) septicemic, and (6) pulmonary. Each form starts with the sudden onset of fever, chills, headache, and generalized myalgias and arthralgias after incubation for 2-10 days. An ulcer is generally seen at the bite or inoculation site, and it may persist several months as organisms spread to the local lymph nodes. Untreated, the mortality rate is 8% for all types; treatment reduces the mortality rate to 1-2.5%.^{38, 39} Untreated pneumonic or severe systemic disease has fatality rates of 30-60%.²²

The ulceroglandular form is the most common type of tularemia. The glandular form accounts for 75-85% of naturally occurring cases. The erythematous, indurated, nonhealing, punched-out ulcer lasts for 1-3 weeks. The local lymph nodes may be fluctuant, and they drain spontaneously. Suppuration of the lymph nodes may occur as long as 3 weeks after treatment.⁴⁰ The differential diagnosis of ulceroglandular tularemia includes sporotrichosis, catscratch disease, mononucleosis, lymphangitis, lymphogranuloma venereum, plague, and Pasteurella infections.

The glandular form is the second most common form. Glandular tularemia occurs after the skin is inoculated by means of arthropod vectors. This form most commonly affects the inguinal and femoral lymph nodes in adults and the cervical nodes in children.⁴⁰ The oropharyngeal, or gastrointestinal, form occurs after the ingestion of undercooked meat or after direct inoculation from the hands to the mouth.⁴⁰ The oculoglandular form is due to inoculation of the periorbital skin or conjunctiva by arthropod vectors.⁴⁰

Only 10-50 organisms need be inhaled to cause typhoidal tularemia. Although the septicemic form is rare in the United States, it occurs without skin lesions or lymphadenopathy. One must suspect tularemia to diagnose it in time for effective treatment. The disease mortality rate is 30-60%.⁴⁰

Pulmonary tularemia develops in 10-15% of patients with ulceroglandular disease and in about half of those with typhoidal tularemia; patients present with a nonproductive cough with dyspnea or pleuritic chest pain. Chest radiographs reveal a variable parenchymal infiltrate. As many as 30% of patients with pulmonary tularemia succumb to the infection.⁴⁰ The differential diagnosis includes Q fever, mycoplasmal infection, psittacosis, histoplasmosis, and coccidioidomycosis.

Skin or mucous membranes act as the portal of entry for tick bites, other arthropod bites, or abrasions. Rarely, inhaled or ingested organisms cause the disease. At least 108 organisms are necessary to cause gastrointestinal disease; 10-50 organisms can cause cutaneous or pulmonary infection.^{22, 41, 42} The local multiplication of F tularensis causes a tender, red, pruritic papule that rapidly enlarges to form an ulcer with a black base. The organism then spreads to the lymph nodes and causes bacteremia. Pneumonia occurs secondary to bacteremia or primarily by means of aerosolization. Foci of alveolar necrosis and neutrophilic infiltrates develop in the lungs. Chest radiographs depict bilateral patchy infiltrates but not large areas of consolidation.²²

Besides the chancrelike ulcer that forms at the site of bacterial inoculation, some patients develop buboes. Approximately 20% of cases may involve a nonspecific or morbilliform skin eruption.^{17, 39} Erythema nodosum and, rarely, erythema multiforme are reported in some cases.¹⁷ Disseminated intravascular coagulation and bleeding may also manifest in patients who do not receive prompt treatment.²² See Media Files 1-3 for examples of tularemia.

In 1994, tularemia was removed from the list of diseases that require national notification. Because of the increasing concern of F tularensis as a biologic weapon, it was reinstated to this list in 2000. A high index of suspicion and caution must be maintained to diagnose tularemia because specialized tests are typically the only way to confirm a diagnosis and its infectivity. Real-time PCR and a fluorescent-labeled antibody test can be quick and reliable diagnostic tests.⁴³ Culture results are usually negative because the organism does not grow on most media.¹⁷

The histologic features of tularemia ulcers are epidermal and upper dermal necrosis and epidermal acanthosis with spongiosis.⁴⁴ The primary draining lymph nodes of all forms of tularemia show a lymphadenitis characterized by geographic areas of necrosis surrounded by a palisading, neutrophilic, and granulomatous inflammatory infiltrate. Immunohistochemistry and DFA tests have demonstrated the bacteria in formalin-fixed tissue samples.¹⁹ F tularensis is hazardous to laboratory personnel.

Killed vaccines offer little or no protection against virulent bacteria.²² A live-attenuated vaccine is available for individuals at risk (eg, field or laboratory workers), and an intranasal vaccine has been shown to induce gamma-interferon–dependent antibody-mediated protection against tularemia⁴⁵ and has proven protective in mice.⁴⁶ A new killed whole-cell subunit of live-attenuated vaccine has been shown to elicit protective antibody-mediated immunity.^{47, 48} Studies from 2006 have shown that bacteremia does not occur after live-vaccine-strain vaccination in normal, healthy human volunteers.⁴⁹ The therapeutic use of antibodies in Francisella-infected individuals has been demonstrated. Immune serum was passively administered to na�ve mice before intranasal F tularensis live-vaccine-strain infection; this provided 100% protection against lethal challenge, while normal serum or immunoglobulin-depleted immune serum provided no protection.⁵⁰

Streptomycin/gentamicin is the drug of choice in adults.²² A 10-day course is sufficient. Tetracyclines and chloramphenicol are also used,²² but treatment failures are more common with these agents; because of the risk for relapse, a 14-day course should be prescribed. Ciprofloxacin and other fluoroquinolones may be used; however, thus far their use has been limited to treatment of F tularensis type B.^{22, 51} See Tables 1 and 2 for treatment recommendations.

Many countries have weaponized tularemia for aerosol delivery, including the United States.^{13, 22} Russia has engineered strains that are resistant to antibiotics and vaccines.^{8, 22} As in the plague, biologic manipulation via plasmids is possible, and high morbidity and mortality rates can result. As in plague, the rapid onset of action and nonspecific nature of tularemia make it a formidable biologic threat.

Trichothecene mycotoxins (eg, T-2 mycotoxin) are a particular subset of mycotoxins, and they are the only biologic warfare toxins with cutaneous manifestations. Trichothecenes are produced by 5 genera of fungi.⁵² These agents are of particular interest because of their previous use as biologic weapons (eg, yellow rain).^{53, 54} The trichothecenes include diversely related mycotoxins called sesquiterpenoids that have hemorrhagic and dermotoxic effects.⁵⁵

Naturally occurring cases of human toxicity are rare, but reports from Russia beginning in the 19th century have been found, and confirmed cases from World War II have been documented.⁵⁶ Trichothecene was reportedly responsible for 100,000 deaths due to contaminated grain that occurred in western Russia in 1942-1948; the agent was aerosolized by means of contaminated hay.⁵³ Signs and symptoms included vomiting, diarrhea, skin inflammation, leukopenia, hemorrhage, necrotic angina, and sepsis.^{13, 53} Aerosols caused death within minutes to hours. The toxins affect rapidly proliferating tissues and are cytotoxic to most eukaryotic cells because they inhibit protein and RNA synthesis.

More recently, trichothecene mycotoxins were thought to have caused fatal pulmonary hemorrhage in some infants in the Cleveland, Ohio, area.⁴⁰ In one area of Cleveland, trichothecene mycotoxins may have accounted for 5% of cases of sudden infant death syndrome in 1993-1995. In all cases, the fungus Stachybotrys atra was found growing in water-saturated cellulose in the walls of poorly maintained homes.⁵⁷

At low doses (in nanograms), severe skin irritation with erythema, edema, and necrosis is observed. Airway effects include nasal and pharyngeal pain, nasal discharge, cough, dyspnea, wheezing, chest pain, and hemoptysis.⁵⁸ Other symptoms include burning, itching, tingling, and, at times, pain.⁵⁴ Vesication occurs during some attacks; this finding is not surprising because the T-2 mycotoxin is estimated to be 400 times more potent than mustard in producing skin injury.⁵⁹ T-2 mycotoxins can be absorbed through the skin, and they cause death with a median lethal dose (LD₅₀) of 2-12 mg/kg, compared with an LD₅₀ of 4500 mg/kg for mustards and an LD₅₀ of 37 mg/kg for lewisite.^{59, 60}

A high index of suspicion is required to make the diagnosis of trichothecene exposure. A history of exposure to a yellow mist or smoke (yellow rain) or something similar is vital because of the variability in presentations. Vesication, erythema, edema, and necrosis, along with the features of the history, could assist in the diagnosis. No specific therapy is available, but some authorities believe high doses of systemic steroids may decrease primary and secondary toxin injuries.¹⁷ Microsomal enzyme inducers (eg, phenobarbital) may speed clearance of the toxin after large exposures.⁵⁴ If the toxin is swallowed, superactivated charcoal should be given orally. Ocular exposure should be treated with copious normal saline or water irrigation. While contact with contaminated skin and clothing can produce secondary epidermal exposures, secondary aerosols are not hazardous.

Environmental decontamination should be conducted using a hypochlorite solution under alkaline conditions such as 1% sodium hypochlorite and 0.1M sodium hydroxide with 1 hour of contact time.⁵⁸ Decontamination with soap and water is the most important factor and could diminish morbidity and potential mortality, as was demonstrated in animals in which 80-98% of the toxin was removed.⁶¹

Although still controversial, good epidemiologic, intelligence, and trichothecene assay evidence suggest that trichothecene mycotoxins were used in Southeast Asia in 1974-1981.^{54, 62, 63} A total of 397 attacks delivered by means of aerosol, droplet clouds, aircraft rockets, bombs, canisters, handheld weapons, and booby traps resulted in more than 6300 deaths in Laos, 981 deaths in Kampuchea, and 3042 deaths in Afghanistan. In Laos, the attacks were described as yellow rain; a sticky yellow liquid fell and sounded like rain or looked like a yellow cloud of dust, powder, mist, smoke, or insect spray. The liquid dried rapidly to form a powder. Most attacks involved the use of yellow pigment, but some attacks involved red, green, white, or brown smoke or vapor. More than 80% of attacks were delivered by means of air-to-surface rockets.⁶⁴

Trichothecene mycotoxins possess multiple characteristics that make them serious threats in biologic warfare. These characteristics include its ease of delivery, the difficulty in its decontamination (controversial observation), the lack of treatment or prophylactic measures, and an LD₅₀ that is an order of magnitude below those of other biologic or chemical agents (eg, mustards).

Protective measures include preventing exposure by wearing a protective mask and clothing or topical skin protectant during an attack.⁵⁸

Variola virus

Variola virus is a serious biologic warfare threat. The Russians had an industrial capacity to produce many tons of smallpox virus annually, and viral stockpiles still remain in Russia and the United States.⁶⁵ The last natural case of smallpox occurred in Somalia in October 1977. The World Health Organization declared the world smallpox-free in May 1980 and stopped smallpox vaccinations worldwide. With today's mobile and susceptible population, one case of smallpox would represent an international public health crisis.

Smallpox typically spreads by means of inhalation and rarely by means of contact with infected clothing or bedding.⁶⁶ It spreads more slowly than chickenpox. In fact, smallpox rarely spreads through schools, although it can infect multiple members of a household. Fortunately, smallpox is transmissible only after the rash appears, and patients are typically moribund at this time and unlikely to have close contact with many other people. The most infectious period is the 7-10 days after the onset of the rash. The transmission of only a few virions is sufficient to cause an infection.⁶⁵ The survival of the virus as an aerosol is inversely proportional to the environmental temperature and humidity.⁶⁵

Two strains of orthopoxvirus infect humans: variola major causes a mortality rate of 20-50%, and variola minor causes death in less than 1% of those who are infected. No animal reservoir exists, although monkeys are susceptible to infection, and this major factor enabled the worldwide eradication of this disease.

Clinically, smallpox is a morphologically monomorphous illness that occurs after an incubation period of 12-14 days. A significant systemic prodrome lasts 2-4 days. The prodrome consists of a high fever, headache, malaise, and backache. After defervescence, an enanthem occurs on the tongue, the mouth, and the oropharynx and precedes the appearance of an exanthem on the face and the forearms by a day. The rash starts as macules that progress to papules, vesicles, deep-seated pustules, and crusts in a characteristic centrifugal pattern over 1-2 weeks. More lesions are present in convex areas than in concave areas, and all lesions on any given area of the body are in the same stage of development. Eventually, the crusts detach in about 3 weeks, leaving depressed, hypopigmented scars.

Scars, particularly those in areas with high sebaceous gland density, remain in 65-80% of survivors.⁶⁵ Death sometimes occurs as a result of toxemia associated with immune complex deposition and hypotension.^{65, 66}

Smallpox has several clinical varieties. The most serious form, hemorrhagic smallpox, had a mortality rate of 96%. Flat smallpox caused the death of 66% of vaccinated individuals and 95% of unvaccinated individuals. Ordinary smallpox, which was present in 80-90% of patients, led to a mortality rate of only 3% among vaccinated individuals but 30% among unvaccinated individuals. Modified smallpox occurred after vaccination and was usually mild. Variola sine eruption occurred in 30-50% of vaccinated persons who had contact with patients with smallpox; serologic studies revealed increased antibody titers after exposure, but no skin lesions appeared, and no transmission ever occurred.^{65, 66}

The clinical differential diagnosis includes chickenpox, erythema multiforme, and, rarely, coxsackie viral infection or molluscum contagiosum, which particularly affected patients who were immunodeficient. However, smallpox lesions develop synchronously and appear monomorphous on a given body area. Although all lesions on the head are the same, they may be in a stage of development different from that of the lesions that appear later on the extremities or the trunk. Chickenpox lesions develop in crops and are more superficial than the deep-seated pustules of smallpox. With chickenpox, the vesicles, pustules, and crusts may simultaneously occur on adjacent areas of skin. Moreover, the rash in chickenpox is denser over the trunk (centripetal), and chickenpox lesions are almost never found on the palms or the soles.

Because no natural case has occurred since 1977, clinical suspicion is paramount in diagnosing smallpox. Because of the implications of a potential worldwide public health emergency, the notification of local health authorities and the CDC is essential. A health care worker who is vaccinated and who adheres to strict universal precautions should obtain the clinical specimens from the patients. Note that urine, saliva, and crusts contain virions and must be appropriately handled. The specimens should be sent to a maximum-containment biocontainment level 4 (BL-4) laboratory, such as the CDC in Atlanta, Ga, or the US Army Medical Research Institute of Infectious Diseases (USAMRIID) in Fort Detrick, Md. In these facilities, the virus can be scanned under the electron microscope for definitive diagnosis. Other definitive diagnostic techniques include immunohistochemical studies, PCR assay to provide genetic material, and newer immunoglobulin M detection assays.⁶⁶

Cutaneous lesion samples for smallpox should include fluid from vesicles for electron microscopy and skin biopsy specimens for histopathology, immunohistochemistry, and DFA testing. Findings from histopathologic examination of vesicular lesions resemble those of herpes simplex, zoster, and varicella. They show multiloculated, intraepidermal vesicles; ballooning degeneration of epithelial cells; intracytoplasmic, paranuclear, eosinophilic viral inclusions; and, occasionally, intranuclear viral changes.⁴⁴ Immunohistochemistry has demonstrated the virus in multiple cell types, including the epidermal keratinocytes and hepatocytes, and in adjacent fibroconnective tissues.¹⁹

The treatment options are limited. Cidofovir has been shown to prevent cowpox, vaccinia, and monkeypox in animals.⁶⁷ It was more effective (ie, decreased mortality and decreased cutaneous lesions) than vaccination against a challenge of lethal doses of monkeypox in Macaca monkeys.⁶⁸ Topical cidofovir was effective in mice over the parenteral counterpart and could be considered.⁶⁹

On the basis of the historical data, vaccination within 2-3 days after exposure provides significant protection. For prevention, the CDC has temporarily accepted a ring vaccination containment strategy, which means that in case of an exposure, the following people would be vaccinated: persons directly exposed, persons with face-to-face or household contact with a patient who is infected, personnel directly involved in the care of the patient who is infected, laboratory personnel, and any other people who are likely to be exposed to the infectious materials.⁶⁶ Contacts of close contacts might also be vaccinated. In March 2007, a case of eczema vaccinatum in a child from contact with a military smallpox vaccine recipient was reported, the first since 1988.⁷⁰

Waibel et al⁷¹ suggested in 2006 that the inner arm may be a suitable alternative to the conventional upper outer arm vaccination site, owing to decreased vaccine-site erythema on the inner arm.

Vaccination could present problems today. Immunodeficient states (eg, HIV disease, transplantation) were much less common in the 1960s and the 1970s. New immunomodulating medications could modify the cytokine milieu that is necessary for the local proliferation of vaccinia, and they could also cause eczema vaccinatum or progressive vaccinia. A number of factors, including toll receptors or cathelicidins, are being evaluated and investigated as to why a particular person may be predisposed to these reactions. This is important, especially if a volunteer vaccination program exposes patients unnecessarily. However, a vaccination program could reduce the logistic difficulties of mass vaccination if a biologic attack did occur.

Vaccinia immune globulin could also be considered and may prevent superinfection, especially in patients with inflammatory skin diseases.⁷² Studies from 2002 demonstrated that a 1:10 dilution of old, stored vaccine produces a 97% response rate to a vesicle at the inoculation site.⁷³

Another issue that remains to be addressed is the ability of previously vaccinated individuals to mount an immune response to a new exposure. One study showed that vaccinations given at ages 1, 8, and 18 years conferred a 30-year persistence of neutralizing antibody, but few patients ever received boosters, and, in the United States, widespread vaccination ended 30 years ago.⁶⁵ However, some believe that immunity is long lasting. Referring to an article published in 1913, the study showed that the survival rate in people older than 50 years who were vaccinated as infants was 93% compared with 50% in those who were not vaccinated.

Many reports have been written regarding the experience of the military and volunteer vaccination programs. Unexpected cases of encephalitis, myocardial infarction, and myopericarditis have been described. Changes in vaccine programs have taken these factors into account, and vaccination in patients with cardiac conditions is now contraindicated.⁷⁴ A 2007 study indicates that implementation of prevaccination cardiac risk factor screening was not associated with a reduction in cardiac events.⁷⁵ A self-resolving folliculitis, however, is not an uncommon development.

Three types of hypersensitivity reactions to the vaccine have been reported: urticarial, erythema multiforme–like, and viral exanthem–like. None of the individuals with hypersensitivity reactions developed vacciniform pustules, nor did microscopic examination of their lesional skin show viral inclusion bodies. These 3 hypersensitivity reactions are much more common than life-threatening or serious reactions, and all are characterized by mild or absent systemic symptoms and a benign outcome.⁷⁶ Initially in some centers, semipermeable membrane dressings were used, but this has been shown to prolong the time until scab separation and possible duration of infectivity.⁷⁷ New vaccination programs should take these considerations into account.

Variola would likely be disseminated by means of an aerosol. Although smallpox has been eradicated, at least 2 sites maintain stores of viable variola. All viral stocks were supposedly transferred to the CDC in Atlanta, Ga, or the Russian State Research Center of Virology and Biotechnology in Koltsovo, Russia.⁶⁵ If enemy soldiers or terrorists ever released this stock, the ensuing morbidity and mortality could be considerable. Currently, vaccination is recommended only for scientists, technicians, and health care workers involved in basic science and clinical research trials of vaccinia-vectored recombinant vaccines and for those who work with variola virus in laboratories.⁶⁶

The infectivity of the aerosol, the high mortality rate, and the stability of smallpox virus make it a potential biologic warfare and terrorist threat. The discontinuation of routine vaccination continues to increase the pool of people who are at risk for the disease. Other animal poxviruses could be genetically engineered to be virulent in humans.^{65, 66}

Hemorrhagic fever viruses

Viral hemorrhagic fevers are caused by viruses from 4 families: Arenaviridae, Bunyaviridae, Flaviviridae, and Filoviridae. The Arenaviridae include the etiologic agents of Argentine, Bolivian, and Venezuelan hemorrhagic fevers and Lassa fever. The Bunyaviridae include the members of the Hantavirus genus, the CCHF virus from the Nairovirus genus, and the Rift Valley fever virus from the Phlebovirus genus; the Filoviridae include Ebola and Marburg viruses; and the Flaviviridae include dengue and yellow fever viruses.⁷⁸

Viruses in these families typically cause nonspecific febrile illnesses characterized by rash, hemorrhagic diathesis, and shock. Because of the size, the dose, the infectivity, the transmissibility, and the ease of mass production, the filoviruses and the arenaviruses are the most likely hemorrhagic fever viruses to be used in biologic weapons.⁷⁹ Russia researched the filoviruses (Ebola and Marburg viruses) and the arenaviruses (Lassa virus and New World arenaviruses) until the mid 1990s when their research funding was discontinued. (Refer to the article by Borio et al⁷⁹ and the article CBRNE - Viral Hemorrhagic Fevers for in-depth discussions of other viral hemorrhagic fevers.)

Hemorrhagic fever viruses are generally transmitted by means of contact with infected animal reservoirs or vectors. Person-to-person spread may occur after direct contact with patients who are infected or with their blood and body fluids. Patients with end-stage disease often have vomiting, diarrhea, shock, and hemorrhage, which increases the risk of transmission.⁷⁹ Person-to-person aerosol spread rarely, if ever, occurs. Infections can be prevented by reducing contact with vectors; by practicing standard hospital infectious control, patient isolation, and disinfection procedures; and by reporting cases to public health officials. Outbreaks of viral hemorrhagic fever occurred in Cuba in 2007 and provided evidence of a decreased risk of hemorrhagic fever in people of Negroid race compared with those of Caucasoid race.⁸⁰

In 2006, the first documented fatal case of hemorrhagic fever caused by West Nile virus in the United States was reported in Florida.⁸¹

Clinical findings

Cutaneous manifestations of viral hemorrhagic fevers are mainly due to vascular instability and bleeding, which manifest as cutaneous flushing, petechiae, purpura, ecchymoses, and edema.⁵⁵ The filovirus diseases frequently result in an exanthem. Patients with Lassa fever have a high incidence of facial edema, which is probably due to the extensive capillary leak syndrome. Lassa fever causes an extensive capillary leak syndrome that leads to edema without petechiae or hemorrhage. The South American arenaviruses cause petechiae, purpura, ecchymoses, palatal hyperemia, and mucosal hemorrhage.

Viral hemorrhagic fevers present as acute febrile illnesses characterized by malaise, myalgia, and prostration dominated by generalized abnormalities of vascular permeability. The initial signs can include flushing, conjunctival injection, periorbital edema, petechiae, and hypotension. Early in the course of the disease, the signs and symptoms are nonspecific and make the differentiation of this disease from endemic febrile illnesses difficult. These other illnesses include malaria; enteric fever; typhoid fever; meningococcemia; rickettsioses; other viral infections; and noninfectious conditions, such as vasculitis, thrombotic thrombocytopenic purpura, hemolytic uremic syndrome, and heat stroke.

Full-blown disease progresses to shock and generalized mucous membrane hemorrhage. Neurologic, hematopoietic, or pulmonary involvement is often present. Diffuse bleeding often occurs as a result of widespread vascular damage, hepatic dysfunction, and/or disseminated intravascular coagulation. Life-threatening blood loss rarely occurs.

A case of severe, persistent dengue myositis affecting the shoulder and hip girdle muscles was reported in 2006. Muscle enzyme levels, including creatine-phosphokinase, were elevated, and the myositis remained after symptomatic treatment of the dengue fever. The myalgias were amenable to opiate therapy and resolved completely only after corticosteroid treatment.⁸²

All of the viruses are small, lipid-enveloped RNA viruses that spread by means of an aerosol, and most have natural nonhuman animal reservoirs, although the vector of Ebola virus remains unknown.⁵⁵

The most reliable diagnostic sign of a Marburg infection is a nonpruritic, centripetal, pinhead-sized erythematous eruption on the trunk and the extremities that appears at days 5-7 of the illness. This eruption develops after the onset of a flulike syndrome accompanied by a soft palate and hard palate enanthem. Scrotal or labial erythema and dermatitis may accompany the diagnostic cutaneous eruption. After 24 hours, the cutaneous eruption develops into large, well-demarcated, coalescent macules and papules that sometimes hemorrhage. In severe cases, a dark, livid erythema with or without cyanosis develops on the face, the trunk, and the extremities. Desquamation of the palms, the soles, the dorsal aspect of the feet, and the extremities occurs 1-2 weeks later. Death is due to a combination of hemorrhage, capillary leakage, shock, and end-organ failure. The mortality rates for Marburg fever and Ebola fever are 23% and 70%, respectively.^{55, 79}

In the 6 outbreaks before the October 2000 outbreaks, Ebola viruses accounted for 851 cases in humans, with 595 deaths (70% mortality rate). The Uganda 2000 outbreak involved another 425 suspected cases with 224 fatalities (53% mortality rate).⁸³ The early findings in Ebola fever include conjunctival injection and adenopathy. Around day 5, a morbilliform eruption is followed by petechiae, ecchymoses, and even hemorrhage. Most commonly, the erythematous, nonpruritic, morbilliform, and centripetal rash desquamates by day 7. Psychosis, delirium, seizures, and coma are often noted. With progressive disease, hemorrhage exudes from mucous membranes, venipuncture sites, and body orifices. Death occurs 6-16 days after the hemorrhage begins. The mortality rate in pregnancy is 100%. Death is due to a combination of hemorrhage, capillary leakage, shock due to vasodilatation and disrupted hemodynamic regulation, and end-organ failure.

In a 2007 study of 124 patients with dengue infection in India, cutaneous involvement was seen in 46.8% of patients (maculopapular/morbilliform, 48.3%; ecchymotic, 27.6%; petechial, 13.8%; macular/scarlatiniform, 10.3%). Maculopapular eruptions were seen more in dengue fever, while petechiae, ecchymosis, and mucosal lesions were seen more in dengue hemorrhagic fever. Involvement was generalized in 48.3% of patients.⁸⁴ Clinical manifestations of dengue virus infection differ between adults and children, with epistaxis, liver enlargement, and oliguria occurring more often in children. Signs and symptoms present in both adults and children with dengue virus infection include gum bleeding, hematemesis, uterine bleeding, diarrhea, coryza, flush face, and shock. Anorexia and abdominal pain are the predominant symptoms in both children and adults.⁸⁵ In a report of 232 cases of dengue hemorrhagic fever in Taiwan in 2002, rash was present in 53.7% of patients. Significant risk factors were age older than 65 years, diabetes mellitus, hypertension, and uremia.⁸⁶

Diagnosis and management

Clinical suspicion must be maintained to diagnose hemorrhagic fever, especially in the United States, where the diseases are not endemic. A history of exposure to a suspicious aerosol (eg, on an airplane, in a subway) is helpful. A CBC count may be helpful in checking for leukopenia and thrombocytopenia, while high liver enzyme levels and thrombocytopenia can serve as indicators of disease severity in dengue fever.⁸⁷ Proteinuria and hematuria could also be present. Immunofluorescent antibody testing, enzyme-linked immunosorbent assay, or PCR assay of the serum may help in the diagnosis. Public health authorities, with the CDC, assist in processing the specimens and in transporting patients suspected of having hemorrhagic fever.⁷⁹

Histopathologic features include a light lymphocytic perivascular infiltrate with mild endothelial swelling. Purpuric lesions demonstrate extravasated red blood cells.⁴⁴ Pathologic features of most viral hemorrhagic fevers include necrosis of liver and lymphoid tissues and diffuse alveolar damage. Hepatocytes from patients with Ebola or Marburg infections have viral inclusions visible under light and electron microscopy. During outbreaks of Ebola hemorrhagic fever in Africa, large numbers of fatal cases were diagnosed by immunohistochemistry on skin punch biopsy samples.¹⁹

The mainstays of treatment are supportive care and maintenance of the patient's blood pressure, oxygenation, and fluid and electrolyte balance. Arenaviruses, but not filoviruses, are sensitive to ribavirin. Future treatment options may lie in prevention with vaccination. A live-attenuated vaccine against Argentine hemorrhagic fever has been proven safe and effective in placebo-controlled trials, but it is not licensed for use, even in endemic regions.⁸⁸ The vaccine also protects monkeys against aerosol exposure to Bolivian hemorrhagic fever.

Harmless Ebola-like particles identical in shape to the Ebola virus have been discovered, yet no genetic material was found in the particle. This particle could provide the ideal vaccine material. The potential vaccine would have an advantage compared with DNA vaccines because the antigen is readily accessible and would not have potential for reversion, unlike some live-attenuated vaccines (eg, Y pestis vaccines). Studies have already been conducted on monkeys exposed to the Ebola virus, with and without the vaccine protection. The vaccinated monkeys lived for 6 months without any evidence of disease, whereas the unvaccinated monkeys all died within 1 week. These findings have dramatic ramifications on the prevention of both natural and terrorist-caused cases of Ebola.⁸⁹

Tetravalent dengue vaccines have entered successful phase 1 and 2 clinical trials in the United States and Southeast Asia. Vaccine-induced, antibody-dependent enhancement of wild-type dengue virus infections is currently under investigation.⁹⁰

Biologic warfare considerations

Little is known about the potential uses of hemorrhagic fever viruses in a biologic warfare or terrorist scenario. However, hemorrhagic fever viruses cause high morbidity rates and, in some cases, high mortality rates. The viruses replicate well enough in cell cultures to permit their weaponization. Results of multiple studies of nonhuman primates have shown that infection can occur by means of aerosolization. In monkeys, only a few virions of the Marburg virus are necessary to cause infection.⁷⁹

Among the arenaviruses, the cause of Lassa fever could elicit widespread fear. In one Nigerian outbreak, the involved hospital became deserted, taxi drivers would not stop in the village, and rumors abounded in the city. Such fear could do as much harm as the actual disease.⁷⁹

Filoviruses could make excellent offensive biologic warfare agents because they are highly infectious and deadly, and they can be stabilized for aerosol dissemination. The Russians were experimenting with filoviruses and arenaviruses as biologic warfare agents.⁵⁵

The authors and editors of eMedicine gratefully acknowledge the contributions of previous author, Dr. Thomas McGovern, to the development and writing of this article.