Incidence in Humans

The incidence of human rabies in the United States fell dramatically from 23.5 infections per year in the late 1940s to 1.0 infection per year during the 1980s (Table 60-1). However, 27 infections were reported from 1990 through 1999, and 26 have been reported from 2000 to 2002. From 1980 to 1989, seven infections were acquired outside the United States. From 1990 to 1999, five infections were acquired outside the United States, and from 2000 to 2008, six infections were acquired in other countries.* Of the human rabies infections acquired in the United States, one originated from a skunk (1981) and two originated from dogs (1991 and 1994). The two dog infections were associated with the epizootic in coyotes that developed when rabies spread across the Rio Grande River from unvaccinated dogs in Mexico. The remaining 42 infections originated from bats. Of the 18 human rabies infections acquired outside the United States during these 28 years, 17 originated from dogs. Only an infection in 2008 in California was determined to have originated from a Mexican free-tailed bat³⁴ (Table 60-2).

TABLE 60-1 Human Rabies Infections in the United States, 1946 to 2008

TABLE 60-2 Rabies Deaths in the United States Since 1980 From Infections Contracted Outside the Country

Although reliable rabies vaccines and antisera first became available during this 62-year period, extensive vaccination of domestic animals, particularly household pets, and elimination of unrestrained and stray animals are considered primarily responsible for the decline in the human infection rate.^53,54,151 Such programs reduced the incidence of laboratory-confirmed rabies in dogs from 6949 in 1947 to 75 in 2008.³⁰ The annual cost for these programs is over $300 million, most of which is borne by pet owners.¹⁹²

Rabies in Wild Terrestrial Animals

In the United States and Canada, a vast reservoir of rabies persists in wild animals.¹⁹² During 2008, 49 states and Puerto Rico reported 6841 rabies infections in animals, approximately 93% in wild animals and 7% in domestic animals. However, some states accept only animals responsible for a human or domestic animal incident for rabies testing; others test all submitted specimens. Furthermore, the number of rabid wild animals that die without being detected is estimated to be more than 90% of the total, so the identified infections represent only a small fraction of wild animal rabies.¹²⁵

Several major rabies epizootics are currently recognized. An epizootic of rabies started in Arctic foxes in northern Canada in the late 1940s and early 1950s and swept southward in the middle and late 1950s to involve red foxes in Alberta, Saskatchewan, Manitoba, Ontario, and Quebec. The epizootic crossed the St Lawrence River in 1961, where it involved red foxes in upper New York State, although currently it appears to be limited to the adjacent Canadian provinces, which are experiencing a lower incidence of fox rabies.^11,28,127 The epizootic, which moved westward to involve arctic foxes in Alaska and the Northwest Territories,¹⁹⁰ surrounds the North Pole and may cover the largest land area of any observed outbreak²⁸ (Figure 60-2).

FIGURE 60-2 Rabid foxes reported in the United States in 2008.

(http://www.cdc.gov/rabies/resources/publications/2008-surveillance/foxes.html.)

An outbreak of raccoon rabies started in central Florida in the 1940s and spread at the rate of about 40 km (25 miles) per year, reaching Georgia in the early 1960s and Alabama and South Carolina in the 1970s. In the late 1970s a second outbreak appeared on the Virginia–West Virginia border. That epizootic has now spread north to all of New England; in 1999 it crossed into Canada. It has also spread south to join with the epizootic coming north from Florida in North Carolina.^46,68,69 The second outbreak developed when raccoons were translocated from Florida for restocking for hunters. Although the animal suppliers held legal permits and health certificates, inclusion of some rabid animals among the more than 3500 transported raccoons has been documented.^118,229

As of 2002 more than 50,000 rabies infections in raccoons had been reported in the United States since 1975. The land mass affected by this epidemic is approximately 1 million km² (383,000 miles²) and includes the residences of 35% of the United States human population (Figure 60-3). The raccoon epizootic is considered particularly threatening because many raccoons live in densely populated urban and suburban areas.⁶⁸ However, the only known human rabies infection resulting from this epizootic occurred in 2003.⁴⁷ The spread of rabies from raccoons to humans appears to have been limited in part because raccoons are large animals and their bites are obvious. To some extent, well-vaccinated dogs and other pets form a barrier between wild animals and humans. Perhaps of greatest significance is the nonaggressive behavior of rabid raccoons. In 38 rabid raccoon incidents in Florida over a 5-year period, bites were inflicted only when humans or dogs tried to kill or capture raccoons that seemed tame.²²⁶

FIGURE 60-3 Rabid raccoons reported in the United States in 2008.

(http://www.cdc.gov/rabies/resources/publications/2008-surveillance/raccoons.html.)

Before the raccoon epizootic, most terrestrial rabies in the United States was in skunks. Human rabies resulting from exposure to a spotted skunk in California was reported in 1826.⁶⁷ Four epizootics are recognized, one of which is in the province of Quebec, Canada, and New York state and is associated with the fox epizootic in that area. A larger epizootic started in Iowa in 1955. It has spread east to Ohio, west to Montana, north to the Canadian provinces of Manitoba (1959), Saskatchewan (1963), and Alberta (1971), and south to meet with a third epizootic that originated in Texas and has spread to surrounding South Central states, particularly Oklahoma and Arkansas. The fourth epizootic in skunks is located in northern California⁶⁷ (Figure 60-4).

FIGURE 60-4 Rabid skunks reported in the United States in 2008.

(http://www.cdc.gov/rabies/resources/publications/2008-surveillance/skunks.html.)

An increase in the number of rabid skunks in the East Coast states has recently occurred, but analysis of these infections indicates they result from raccoon rabies spilling over into skunks and are not indicative of a separate skunk epizootic.¹⁰³

Screening of rabies virus isolates from the epizootics has disclosed five distinctive patterns. Red foxes and skunks in New York and adjacent Canada present one pattern; raccoons from the Atlantic states present a second. The skunks in the south-central states present a third, and a fourth is represented only by a small outbreak in gray foxes in Arizona. However, the fifth pattern is found in skunks from the north-central states and from California, in dogs from the Mexico border states, and in a small rabies outbreak in gray foxes in Central Texas¹⁸⁹ (Figure 60-5).

FIGURE 60-5 Terrestrial rabies reservoirs in the United States in 2008.

(http://www.cdc.gov/rabies/location/usa/surveillance/wild_animals.html.)

Rabies in rodents is an intriguing problem. Rodents are the animals of choice for rabies virus isolation in the laboratory; yet rabies in small free-living rodents is rare. One explanation is that rodents may usually be killed rather than simply infected by the bites of rabid animals, but rodents are carrion eaters and can be infected by that source as well. In recent years the largest number of rodent rabies infections has been in large rodents such as woodchucks that have been infected by rabid raccoons. Rabid beavers have attacked and bitten humans in North Carolina. However, no transmission of rabies to humans by rodents has been documented.²⁰⁰

Rabies in Bats

With the exception of Antarctica, rabies in bats is global. In Canada, the United States, parts of South America, Western Europe, and Australia (where rabies in carnivores, particularly dogs and foxes, has been controlled), bats are the most prominent source of human rabies.¹⁴⁸

Rabies was diagnosed in insectivorous bats in Brazil in the 1920s and in frugivorous bats in Trinidad during the 1930s, although the principal subject of these studies was rabies in hematophagous (“vampire”) bats that was being transmitted to humans. The first definitive diagnosis of rabies in nonhematophagous bats was made in a frugivorous bat that flew into a “chemist’s” shop in Port of Spain, Trinidad, on September 10, 1931. However, the incident that drew widespread attention to bat rabies occurred in Tampa, Florida, on June 23, 1953. The 7-year-old son of a ranch hand was looking for a baseball near some shrubbery when a lactating female yellow bat suddenly flew out of the bushes and bit the boy on the chest, remaining firmly attached until knocked off by the boy’s mother. The ranch owner had heard of rabies in vampire bats in Mexico and insisted the bat be examined for infection. Negri bodies were found in smears of the brain, and the diagnosis was confirmed by mouse inoculation of brain tissue. The boy was given postexposure treatment and did not develop an infection.^17,69,219

The publicity given this event led to many more bats being submitted for rabies examination. Subsequently, rabies has been found in bats in every state except Hawaii, as well as in eight Canadian provinces.^17,165 In 2008, bat rabies was reported from all 48 of the continental states and the District of Columbia⁵⁴ (Figure 60-6). The estimated incidence of rabies in bats in the United States is 0.5% to 1.0%; the incidence in bats that appear ill or injured is much higher, 7% to 50%.^17,201

FIGURE 60-6 Rabid bats reported in the United States in 2008.

(http://www.cdc.gov/rabies/resources/publications/2008-surveillance/bats.html.)

Rabies virus variants from bats are species specific rather than geographically specific¹⁸⁹ and are distinctly different from those of terrestrial animals in the same locations, including the major terrestrial epizootics. Clearly, little exchange of infection between bats and terrestrial animals takes place, although occasional animals infected with rabies virus strains typical of bats are found. Many large areas of the United States, particularly the Pacific Northwest and New England (prior to the raccoon epizootic), report rabies in bats but in no other species. Even though cats and foxes catch and eat bats, only 3 of 136 cat and fox rabies isolates over a 2-year period were antigenically similar to bat rabies strains.^17,79,189

Approximately 70% of human rabies infections and 75% of cryptic rabies deaths in the United States have been caused by the variant associated with silver-haired and the eastern pipistrelle bats, which are reclusive animals rarely found around human habitation. Infections by variants associated with bats that frequent human dwellings are much less common. Infections in other animals by this variant are also disproportionately very high. These bats are small and their bites are difficult to detect. However, in comparison with other rabies virus variants, the variant associated with these two bats replicates better in fibroblasts and epithelial cells, and replicates better at the low temperature of 34° C (93.2° F). These features indicate this variant is better able to replicate in the peripheral tissues involved by most bites.¹⁴⁰

Rabies in Domestic Animals

Since rabies in dogs has been controlled, rabies infections in cats have outnumbered infections in dogs (295 to 75 in 2008).³⁰ A major problem in vaccinating cats is establishing ownership. Farmers value cats for rodent control but do not recognize them as property. Cats wander from farm to farm and contact wild animals with rabies. Capturing feral cats so they can be vaccinated is difficult.⁴² Rabies is not rare in other domestic animals, including cattle, horses, mules, sheep, and goats (Table 60-4).

TABLE 60-4 Human Rabies Infections Identified in the United States and Puerto Rico, 2008

Sources of Human Infection

In the late 1940s and 1950s most human rabies in the United States resulted from bites by dogs or cats. Of 146 infections in the years 1946 to 1961 for which a source of exposure could be identified, dogs were responsible for 120 and cats for 9 (88.4%). Foxes (7), skunks (5), and bats (5) were responsible for the rest.⁵³ However, after rabies in domestic animals was controlled, human rabies resulting from bites by pets disappeared. Since 1966, all but 2 of the 19 human rabies infections resulting from exposures to rabid dogs were acquired outside of the United States.^53,62,151

Before 1965, the CDC had recorded no human rabies occurring within the United States that had been acquired outside the country.¹⁵¹ However, since then the number of infections acquired outside the United States has been significant: 3 of 15 (20%) between 1965 and 1970, 6 of 23 (26%) in the 1970s, 7 of 10 (70%) in the 1980s, and 18 of the 63 cases (29%) since 1980. Lack of knowledge about the risk for rabies in developing countries has led some travelers to disregard animal encounters and not obtain rabies immunoprophylaxis, but some of these infections have been in children who did not inform their parents of the animal contact.

Until the 1980s, identifying the source of a number of human rabies infections in the United States was impossible. For many infected persons no animal exposure incident—even an opportunity for animal exposure—could be identified. An infectious source could not be found for 84 of 230 (35%) human rabies infections occurring in the United States between 1946 and 1961,¹⁵¹ or for 6 of 38 (16%) human infections between 1960 and 1970.⁵²

Only since the 1980s has monoclonal antibody typing or RT-PCR nucleotide analysis allowed the source of human rabies infections to be determined when no animal exposure incident could be identified.^17,79,189 However, such studies have made it unmistakably clear that bats are now the major source of human rabies in the United States.

Of the 45 human rabies infections acquired within the United States since 1980, 42 are attributed to bats. A study of human rabies of bat origin in the United States and Canada from 1950 through 2007 identified 56 infections, of which 22 (39%) reported a bite, 9 (16%) had a direct contact but no bite, 6 (11%) found one or more bats in their homes (two in the room in which they slept), and 19 (34%) had no history of a bat contact.⁷⁶

How the infection is transmitted has been uncertain. In 1956 and 1959, two men died of rabies after exploring Frio Cave near Uvalde, Texas. The walls and ceiling of that cave hold astonishing numbers of bats—300 to 400 per square foot. Neither man had been bitten, and the infections were attributed to aerosol transmission of the rabies virus. Subsequently, when experimental animals of various species were placed in the cave in cages that only allowed the virus to be transmitted as an aerosol, a significant percentage developed rabies.^74,75 Additionally, aerosol transmission of rabies to humans has occurred at least twice in laboratories.^24,228 The CDC recommends rabies vaccination for spelunkers.⁵⁶

However, nursery caves such as Frio Cave contain an astounding number of bats. Saliva and urine constantly rain down on anyone entering the cave, and the blanket of guano on the floor ranges from several inches to several feet in thickness. In Frio Cave air circulation is so poor that the bats warm the cave, the air is humidified by their respiration, and the concentration of ammonia from their urine is so high that the cave usually cannot be entered without respirators.⁷⁴ Similar infections in other caves have not been reported.

Unrecognized bites appear to be the source of infection for most individuals who have had no recognized bat encounters. Bat teeth are so small and sharp that a bite may not be felt. Even the recognized bites are not particularly painful, although at least one of the individuals known to have been bitten was intoxicated with ethanol at the time.⁹⁸ For centuries, South American vampire bats have been reported to bite sleeping victims without awakening them.

Limiting human rabies of bat origin is best addressed by informing the public of the risk.¹⁶⁵ Reducing the bat population is not an acceptable approach. Significant population reduction would be difficult and, if achieved, would be an ecologic disaster because bats play such a major role in insect control (Table 60-3).

TABLE 60-3 Human Rabies Infections Contracted Within the United States Since 1980

The CDC and other institutions now advocate the following measures:

Epidemiology

Rabies is found throughout the world, and although more common in tropical or temperate climates, is by no means limited to those areas. Arctic foxes with rabies have been found in Alaska, Northern Canada, Greenland, Norway’s Svalbard Islands, and much of Siberia. An epizootic in the Thule district of Greenland in 1958 and 1959 resulted in death from rabies of 50% of the sled dogs in that area. More than 1000 dogs in the Egedesminde district died in another epizootic in 1959 and 1960. For reasons that are not known, transmission of rabies to humans is rare in these areas, even though exposures are common.⁷⁵ Perhaps the rabies variant is more infective for foxes and dogs.

Rabies is not found in a few areas of the world, all of which are landmasses or peninsulas isolated by water. Rabies does not occur in Hawaii (the only state in the United States in which rabies is not found), some of the Caribbean islands, Pacific Oceania (including Australia [although Australian bat lyssavirus is found there] and New Zealand), or Antartica.^101,147 The two human genotype 1 rabies infections that have occurred in Australia are thought to have been acquired outside that country^24,58,101 (Figure 60-7).

FIGURE 60-7 Areas of the world in which rabies is present (orange), areas in which rabies is absent (green), and areas for which information is not available (yellow).

(Modified from http://www.nathnac.org/pro/factsjeets/rabies.htm, and http://www.ktl.fi/portal/sip,o/julkaisut/oppaat_ja_kirjat/matkailijan_terveysopas_old/4_matkailijoiden_rokotukset/1415.)

Great Britain had been free of rabies since an extensive dog confinement and vaccination program in 1903, although concern about reintroduction of rabies was raised by the Channel Tunnel and the reduction of border controls between members of the European Community.^69,202 A 55-year-old Scotsman with a fatal infection in 2002 was the first locally acquired lyssavirus infection on that island in 100 years, but the virus was not of genotype 1.⁹⁴

Rabies has been endemic in Japan since the 10th century. However, following World War II, members of the U.S. Army Veterinary Corps determined that no reservoir of rabies existed in the wild animal populations of Japan, Taiwan, and the Philippines—perhaps in part because wild animals were hunted for food during the war. Extensive campaigns to eliminate stray dogs (which in some areas of Japan reduced the canine population by 70% to 80%) and to vaccinate those remaining succeeded in eradicating the infection from Japan and Taiwan. Endogenously acquired rabies has not occurred in those islands since the late 1950s.^5,191

The success of canine rabies eradication programs depends on the society in which the programs are initiated. Such programs achieve little success in nations that are predominantly Hindu or Buddhist, because the people do not support elimination of animals that have no apparent owner. They often put out food for stray dogs. In contrast, Malaysia, a peninsula that is predominantly Muslim, has been largely free of rabies since the early 1950s.²⁵

Elimination of stray dogs must be combined with vaccination programs. Dogs that are eliminated because they cannot be associated with human ownership are quickly replaced. The annual turnover of the dog population in developing countries has been found to range between 30% and 40%.³¹

Vaccination of domestic animals for rabies is limited largely to industrialized nations. In many developing countries, vaccination of animals is considered unaffordable and rabies control resources are expended on postexposure immunoprophylaxis of humans. Even though rabies immunoprophylaxis is administered to 800 to 900 persons per million inhabitants annually in such countries, the human death rate from rabies is still high, an average of nearly five deaths for each 1 million population annually.³² In the United States that death rate would result in approximately 1500 rabies deaths a year.

The magnitude of the rabies threat in developing countries is illustrated by the experience in Thailand. Rabies has been of particular interest in that country since Princess Banlusirisarn was bitten by a rabid dog within the palace grounds near Bangkok in 1911 and subsequently died. (No rabies vaccine was available in Thailand, then called Siam, at that time.) The princess’s death was instrumental in establishing the Institute Pasteur of Bangkok in 1913, which was renamed the Queen Saovabha Memorial Institute (QSMI) in 1922. The Institute, a WHO Collaborating Center for Research on Rabies Pathogenesis and Prevention, has been and remains the site of many sophisticated rabies investigations.

In the 1990s the Institute’s postexposure rabies clinic treated about 18,000 patients with new animal bites each year—an average of almost 50 new patients a day! Furthermore, these patients were only 28% of the estimated 64,000 Thais who receive postexposure therapy annually, many of whom were residents of rural or remote portions of the country and were treated by local physicians.²²⁴ However, the number of human deaths from rabies in Thailand declined from about 400 a year in the 1970s to 70 in 1999, even though dog rabies has not been controlled.⁶⁸

Other developing nations have similar rabies problems. WHO agencies have estimated that 87 countries and territories with a total population of about 2.4 billion people are afflicted with endemic canine rabies.³²

Sources of Human Infection

Although domestic animals are rarely the sources of human rabies in the United States and other developed countries, in developing countries the vast majority of human rabies—99% by some estimates—is the result of exposure to rabid dogs.^{33,69,213,219} In Thailand, although rabies has been found in an array of exotic tropical animals, including tigers and leopards, between 1979 and 1985, 90.6% of human infections resulted from dog bites and an additional 6% followed unknown events. The remaining 3.6% followed cat attacks.¹⁸¹

Other animals, particularly bats, transmit rabies. Hematophagous, or “vampire,” bats are a major source of animal and human rabies in South and Central America, the only areas where such bats are found. Their range extends between northern Mexico and northern Argentina—basically between the tropics of Cancer and Capricorn—and fossils indicate vampire bats have inhabited those areas for 2.5 million years.⁹³ These animals consume 20 to 25 mL of blood at a feeding, and although cattle are their preferred food source, a study in Colima, Mexico, found human blood in the stomachs of 15.7% of 70 vampire bats.¹³

Human rabies of vampire bat origin was first recognized in 1929 in Trinidad when Negri bodies were found in the brains of 17 individuals, mostly school-age children, who died with acute ascending paralysis. Subsequently, small epidemics have been recognized almost every year in that country. Interestingly, almost all of the rabies transmitted by vampire bats in Latin America is paralytic in type rather than furious.²¹⁵

Human rabies resulting from vampire bat bites has been reported almost every year from Mexico, but was first reported from South America in 1953 when 9 of 43 diamond miners who slept outdoors died of a mysterious illness. Autopsies of five of the miners disclosed rabies. In an outbreak in two rural communities in the Amazon Jungle of Peru during the first four months of 1990, 29 of 636 residents (4.6%) died after a rapidly progressive illness characterized by hydrophobia, fever, and headache. Rabies virus was isolated from the brain of the only individual on whom necropsy was possible. Of the 29 victims, 96% had a history of bat bites, although bats also had bitten 22% of unaffected community members.⁶

Human infection is not the only major problem resulting from rabies transmitted by vampire bats in Central and South America. Migrating epizootics of vampire-bat–transmitted bovine paralytic rabies kill thousands of animals annually; the estimated cost in 1980 was $500 million.^6,68 Efforts to control these epizootics have included vaccination of cattle and attempts to limit the vampire bat population by administering anticoagulants, usually warfarin.

It is interesting to note that meat is often consumed from cattle slaughtered at the first, virtually pathognomic sign of disease, paralysis of the hindquarters. Even normal appearing animals may have infected brains. Four of 1000 (0.4%) apparently healthy cattle selected at random at the Mexico City slaughterhouse of Ferrería were found to be infected by rabies when investigated with fluorescent antibody staining and animal inoculation of brain tissue. However, no cases of human rabies from this source have been reported.¹³ Human rabies has resulted from consumption of brain tissue from a rabid dog and a rabid cat in Vietnam.²¹⁸

Mongooses are the major source of rabies in South Africa and in some Caribbean Islands such as Puerto Rico.^14,88 The yellow mongoose is the main reservoir of rabies in South Africa.²¹⁹ The small Indian mongoose, imported many years ago in an effort to control rodents,¹⁰⁷ is an important reservoir and vector of rabies in Cuba, the Dominican Republic, Grenada, Haiti, and Puerto Rico.²¹⁹ In Grenada, from 1968 to 1984, mongooses accounted for 787 (73%) of 1078 cases of animal rabies on the island. Of 208 human exposures requiring antirabies therapy, mongooses were responsible for 119 (57%). The possibility of eliminating the animals by hunting or trapping appears remote in view of the island’s topography and the animal’s skill at adapting to its surroundings. However, mongooses take oral baits, so oral vaccination appears feasible if an appropriate vaccine can be identified.¹³³ About 20% to 40% of mongooses have naturally acquired antirabies antibodies, possibly from having survived infection.⁸²

At the Canadian International Water and Energy Consultants (CIWEC) Clinic in Kathmandu, Nepal, 51 travelers required immunoprophylaxis following rabies exposure during a 2-year period. Although 36 of these encounters were with dogs, 10 were with monkeys at Swayambhunath, the “Monkey Temple,” a Buddhist shrine popular with tourists. The bites were inflicted when monkeys leaped for food carried by visitors.¹⁸⁵

Human-to-human transmission of rabies is rare. Eight documented infections were in individuals who received corneal transplants (two from the same person) from individuals whose neuroparalytic disorder was not recognized as rabies.^57,92,111 In 1996, Fekadu and colleagues reported two apparent instances of human-to-human rabies transmission in Ethiopia. A 41-year-old woman, who died of rabies 33 days after her 5-year-old son died of the same infection, had been bitten on a finger by her son. Another 5-year-old boy, who developed rabies 33 days after his mother died of that infection, had been repeatedly kissed on his mouth by his mother, apparently passing infected saliva to him. However, these infections were not confirmed by laboratory studies.⁹⁰

In July 2004, the CDC reported four cases of human rabies transmitted by organ transplants from a single donor. The male donor was hospitalized in Texas with “severe mental status changes,” a low-grade fever, and neurologic imaging findings indicative of a subarachnoid hemorrhage. That lesion expanded rapidly in the 48 hours after admission and lead to cerebral herniation and death. An autopsy was not performed. Only after the organ recipients’ deaths from rabies was the donor’s history of having been bitten by a bat discovered.

The donor’s lungs were transplanted to a male who died of intraoperative complications. The liver and one kidney were transplanted to males, and one kidney was transplanted to a female, all of whom died of rabies 27, 37, and 39 days later. The fourth victim had a liver transplant from another donor, but a segment of iliac artery from the first (rabid) donor was inserted during the procedure. This recipient died of rabies approximately one month after the transplant.^59,64

In the same year, three German individuals who had received a lung, a kidney, and a kidney/pancreas transplant from the same individual died of rabies. When the cause of their deaths was recognized, three additional transplant recipients who had received a liver and two corneas were given postexposure prophylaxis and survived. The corneas were removed. The donor had been scratched by a dog while visiting India and had not received postexposure therapy. The nature of her disease was not recognized until the three transplant recipients died.¹¹⁹

One case of human rabies appears attributable to transplacental infection. However, a number of mothers dying of rabies encephalitis have given birth to healthy babies, presumably because the virus travels through nerves—viremia has never been documented—and cannot reach the placenta or fetus.^91,111

Mortality

Rabies in humans, once it has become clinically apparent, is uniformly fatal. No other infection is so lethal or progresses so rapidly. In the 1970s intensive support allowed three humans with clinical rabies to survive.^61,105,159 Three rabies survivors have been reported subsequently. The first five infections were vaccination failures, and four of the five survivors had severe residual neurologic deficits, severe enough to be fatal 34 months later in one person.^2,136 The rabies virus was not cultured from any of these patients and at least one may have had a reaction to neural-derived vaccine rather than an actual infection.¹¹⁷

In October 2004, a 15-year-old female in Wisconsin became the first human to survive clinical rabies without vaccination. This young lady, who developed symptoms 1 month after she failed to report a recognized bite by a bat, was admitted to the Medical College of Wisconsin, Milwaukee, 5 days later. On the second hospital day, the CDC confirmed the presence of rabies virus–specific antibody in serum and cerebrospinal fluid. Because she had evidence of an adequate immune response, because brain pathology in humans succumbing to rabies largely reflects secondary complications rather than a clear primary process, and because clinical reports have included the hypothesis that death results from “neurotransmitter imbalance” and autonomic failure, her physicians, with her parents’ approval, elected to treat her with antiexcitatory and antiviral drugs. Rabies vaccine and rabies immune globulin were not administered.

Coma was induced with ketamine and midazolam, she was intubated and maintained on a ventilator, and received intravenous ribavirin and amantadine. Ketamine is a dissociative anesthetic, but it is also an N-methyl-D-aspartate (NMDA) antagonist, and the NMDA receptor has been speculated to be one of the rabies virus receptors. Later she also was given benzodiazepines and supplemental barbiturates. She recovered slowly, was removed from isolation on the 31st day, and was discharged from the hospital on the 76th day. Attempts to isolate the rabies virus, detect viral antigens, or identify rabies RNA in two skin biopsies and nine salivary specimens were uniformly unsuccessful. Five months after her initial hospitalization, she was alert and communicative, but had choreoathetosis, dysarthria, and an unsteady gait. Twenty-five months after hospitalization, she continued to have fluctuating dysarthria and gait difficulties, as well as an intermittent sensation of cold feet. She had no difficulties with activities of daily living, including driving. In high school, she took college-level courses in English, physics, and calculus, scored above average on a national college achievement test, graduated in 2007, and planned to attend a local college. She had no problems with peer relations or mood disorders.^62,113,227

At the time, hope was held that this therapy might be a breakthrough in the treatment of clinical rabies. However, a number of individuals have subsequently been treated by the same protocol, and none survived.^29,110,136

In February 2009, a 17-year-old female, who two months previously had come in contact with flying bats in a Texas cave, where several of the bats had hit her body but she had not detected any bites or scratches, presented to a community hospital emergency department with severe frontal headache of 2 weeks’ duration, photophobia, vomiting, neck pain, dizziness, and paresthesia of her face and forearms. She had intermittent disorientation and a Glasgow Coma Scale score of 14, nuchal rigidity, and fever as high a 38.9° C (102 °F). Her cerebrospinal fluid (CSF) had a white blood cell count of 163/mm³ with 97% lymphocytes. Bacterial cultures of CSF grew no organisms. After 3 days, her symptoms resolved and she was discharged.

Six days later she presented to another hospital with photophobia, vomiting, and muscle pain. Her CSF had a protein of 160 mg/dL (normal is 15 to 60 mg/dL) and a WBC count of 185/mm³ (normal is <5/mm³), of which 95% were lymphocytes and the rest macrophages. She was transferred to a tertiary care children’s hospital.

Initially she had decreased strength of the left lower and upper extremities, but that resolved. Four days later she reported loss of sensation and strength of the right extremities and her vomiting increased. A lumbar puncture demonstrated increased intracranial pressure. The history of bat exposure was elicited that day. The following day, serum and CSF, saliva, and nuchal skin samples were submitted to the CDC for rabies testing. No rabies virus antigens or RNA could be identified in the skin or saliva, but four serum and CSF specimens contained rabies virus antibodies. Serum immunoglobulin G (IgG) reactivity increased to a peak dilution of 1 : 8192 and immunoglobulin M (IgM) to 1 : 32. The CSF IgG was positive to a dilution of 1 : 32.

After the initial results of testing were confirmed, the young woman received one dose of rabies vaccine and 1500 units of human rabies immune globulin. To avoid potentiating the immune response, more vaccine was not given. She was managed supportively, never required intensive care, and was discharged 16 days later after her symptoms had resolved. Twelve days later she returned to the emergency department with headache and vomiting, which resolved after a lumbar puncture. Her CSF pressure was still elevated. Subsequently, she has not returned for follow-up.

Testing for evidence of 30 other causative agents of encephalitis or aseptic meningitis was uniformly negative. Her illness has been classified as presumptive abortive human rabies.⁶⁰ Other than the 15-year-old girl from Wisconsin and the 17-year-old from Texas who received only a single dose of vaccine, no person who has not been vaccinated has survived clinically evident rabies. (Subclinical human infections probably occur, as discussed later.) The clinical phase of rabies encephalitis rarely lasts more than a few days to a few weeks, and infected persons are severely incapacitated.^3,92 The catastrophe of rabies is compounded by the young ages of many of its victims; 40% to 50% are 15 years old or younger.⁹²

Incubation Period

For many years, some human rabies infections have been thought to follow prolonged incubation periods. In 1987, a 13-year-old boy who had immigrated from the Philippine Islands 6 years earlier died with rabies determined by nucleotide analysis to be of Philippine dog origin. He had not been out of the United States since he arrived.⁴⁹ The second documented Australian rabies patient, a 10-year-old girl of Vietnamese origin, had experienced no identifiable animal contact since she had left North Vietnam 6 years and 4 months earlier. She had spent some time in Hong Kong, and the virus responsible for her death was of an immunotype found in China, although the brain tissue was partially decomposed and nucleotide sequencing was limited.^24,101,120 Joshi and Regmi have reported an individual who had an apparent incubation period of 1100 days (3 years).¹²³ The first documented patient with rabies reported in Australia, a 10-year-old boy who died in 1987, probably resulted from a monkey bite inflicted in northern India 16 months earlier.⁵⁸ An 18-year-old Mexican man who died in Oregon in 1989 was infected with rabies virus of a strain to which he could not have been exposed for at least 10 months, although no history of any type of exposure could be obtained.⁵² Even longer incubation periods of 10 years and years have been reported, but these occurred in areas where rabies is endemic and a second exposure in the intervening period could not be ruled out.¹⁸⁹

Confirmation of such prolonged incubation periods was achieved in three immigrants into the United States from the Philippines, Laos, and Mexico. Nucleotide analysis disclosed rabies viral amino acid compositions essentially identical to the patterns from rabies viruses isolated from dogs in their native countries and unlike rabies viruses found in the United States. These individuals had been in the United States for 6 years, 4 years, and 11 months before the onset of clinical disease.^189,192

In 2004, a 22-year-old man from El Salvador who had been in Los Angeles for 15 months died of rabies that, by nucleotide analysis, was typical of rabies viruses found in dogs in El Salvador and unlike viruses found in the United States.¹²⁶

Such prolonged incubation periods may explain the inability to recall any animal exposure by some patients. However, the possibility that in the past rabies infections resulted from an unrecognized source, such as an undetected bite by a bat, cannot be dismissed.

Almost 99% of human rabies infections clinically manifest in less than 1 year, typically 2 to 12 weeks.^15,24,92 The median incubation period in the United States for persons diagnosed between 1960 and 1990 was 24 days for those 15 years old and younger, but 43.5 days in people older than 15. Fixed (laboratory) strains of virus tend to produce shorter incubation periods than do wild or “street” strains. In 1960 in Brazil, 60 people were injected with vaccine that had been inadequately inactivated. Sixteen developed rabies, and the incubation periods ranged from 4 to 16 days.⁹²

The size of the viral inoculum clearly influences the incubation period. Experimental animals injected with large numbers of viruses develop clinical infections significantly faster than do those receiving small inocula. Small inocula resulted in greater central nervous system histologic damage and more widespread infection outside the central nervous system, particularly in salivary glands.⁸⁷

Pathogenesis of Central Nervous System Infection

Immediately after a bite (or investigational injection), rabies virus can be identified at the site with fluorescent antibodies and remains near the wound or injection site for hours to weeks, depending on the animal species. Viral antigen can be demonstrated in muscle, and viral particles budding into the sarcoplasmic reticulum and from the sarcolemma have been demonstrated by electron microscopy.⁶⁷ The virus appears to enter both motor and sensory nerves, probably through motor endplates and neuromuscular spindles.^12,177

Passage of the virus through peripheral nerves was demonstrated in 1887 when rats⁸⁰ and rabbits⁸¹ were protected from rabies following injections of the virus in their hind legs by sectioning the sciatic nerve. After entry into peripheral nerves, the virus travels at a rate of about 5 to 10 mm per hour to neuronal cell bodies such as dorsal root ganglia.^15,67 Replication can begin at this site, and prolonged ensconcement at this site has been suggested as one explanation for prolonged incubation periods.¹⁷⁷

Upon reaching the CNS, the virus is widely disseminated with extreme rapidity, almost simultaneously with entry, but the manner in which the virus disseminates throughout the CNS is not known. Viremia has not been documented. Plasma membrane budding from infected to uninfected neurons and dissemination through intercellular spaces or cerebrospinal fluid have been suggested. Clusters of viral particles at neuromuscular junctions, reduction of viral infectivity by nicotinic acetylcholine receptor competitors, and other data suggest that the virus recognizes cholinergic binding sites and perhaps enters peripheral and central nerve fibers through those sites. The large numbers of muscle cholinergic binding sites in foxes, which are exquisitely sensitive to rabies, and the small number of such sites in opossums, which are highly resistant to rabies, support this hypothesis and possibly explain the mechanism of sensitivity or resistance.¹⁵ The glycoprotein that coats the viral particle is a major determinant of neuroinvasiveness, and alteration of this protein can markedly alter the kinetics of CNS viral spread.¹²

Viruses can be isolated from cerebrospinal fluid—a significant antibody concentration in this fluid is considered diagnostic of CNS rabies infection—and spread by this route could be quite fast.¹⁸³ Additionally, rabies virions have been identified in intercellular spaces in the CNS by electron microscopy. Rabies antigen can be found in essentially all parts of the CNS and, although limited mostly to neurons, can also be found in oligodendrocytes.¹¹⁵

The rabies virus can infect a wide variety of cells in culture; no explanation for its localization to neurons in vivo has been found.¹²⁴

After wide CNS involvement, the virus passes centrifugally through neural axoplasm to a wide variety of tissues, including salivary glands, corneas, and skin of the head and neck, sites at which identification of the virus may aid in the diagnosis of clinical illnesses. The route of spread to the periphery was demonstrated over 90 years ago, when Bartarelli sectioned nerves to salivary glands and found that the glands subsequently did not contain rabies virus.²² Even within the salivary gland, the virus appears to spread by neural networks and not between adjacent epithelial cells.⁶⁷

An element of immunopathology is produced by disseminated rabies infection. Among persons exposed to rabies, those immunized with early vaccines who subsequently developed infections did so more rapidly than did unvaccinated individuals, a phenomenon termed “early death.” Experimental confirmation of this feature has been achieved by injecting mice with a lethal quantity of rabies virus and immunosuppressing a portion of them. The immunosuppressed animals survived 20% to 25% longer than unsuppressed animals, but their survivals were shortened to those of the control animals when they were injected with antirabies antibody. Additionally, cytolytic T cells appear to be a significant component of the protective response to rabies virus. Avirulent strains of rabies virus induce rabies specific cytolytic T cells, but virulent strains do not.¹⁴⁵

Pathologic alterations in the CNS infected by rabies are surprisingly mild, unless supportive care has kept the patient alive for several weeks, which allows much more extensive, necrotic lesions to develop.^165,177 Leptomeningeal congestion is the only grossly visible change typically found. Mild edema may be present if the patient has been hypoxic. Pressure grooves are rare. The meninges may be cloudy if severely inflamed.¹⁵⁵

Typical histologic features are perivascular cuffing by mononuclear inflammatory cells, microscopic collections of reactive glial cells known as Babes nodules (named for the man who first described them in 1892⁷), and areas of neuronal degeneration and neuronophagia. Some leptomeningeal inflammation is usually present. Spongiform degeneration similar to that found in prion diseases has been described in animals, particularly in skunks.⁶⁷

Van Gehuchten and Nelis described a striking proliferation of the capsular cells surrounding ganglionic neurons that pushed these cells apart and separated them by a dense cellular layer. The neurons also contained degenerative changes. Subsequent studies have found the Van Gehuchten–Nelis changes to be present in almost everyone dying with rabies—and to be a much more consistent and reliable diagnostic feature than are viral inclusions.^204,205

Negri bodies are the best-known histologic feature of rabies and were the first viral inclusions to be found (Figure 60-8). Negri described these cytoplasmic inclusions in 1903 and considered them parasites that caused the disease.¹⁴⁶ Entirely independently, Bosc described identical inclusions in two separate papers published the same year, but he is rarely recognized.^34,35 Negri bodies are eosinophilic cytoplasmic inclusions. Some contain a small, basophilic inner body, or innerkörperchen, and are known as lyssa bodies. They are found almost entirely within neurons; although most common in Ammon’s horn and Purkinje cells of the cerebellum, they may be seen in any part of the central nervous system, particularly in humans. Inclusions may be seen in other tissues, such as the salivary gland, skin, cornea, and pancreas, as well, but are not seen in the ependyma and choroid plexus.

FIGURE 60-8 Negri bodies (arrow) in the cytoplasm of neurons in the hippocampus of a dog. (Hematoxylin and eosin, ×1000.)

(Courtesy Mérieux Institute, Inc.)

The appearance of the inclusion bodies varies in different animal species, and uninfected animals such as cats commonly contain cytoplasmic inclusions that easily could be confused with rabies bodies.¹⁹⁵

By electron microscopy, three types of bodies have been identified: one composed of a granular matrix of viral protein and typical virus particles, a second composed of matrix and tubular structures, and a third composed of matrix alone. Invaginations of cytoplasm into the inclusion give rise to the innerkörperchen, indicating that Negri bodies and lyssa bodies are both diagnostic of rabies.¹⁵⁵

Clinical Features

Pain, paresthesias, or symptoms such as burning, itching, or numbness at the site of the bite or in the bitten limb are the most common early symptoms. Paresthesias may result from proliferation of rabies virus in the spinal cord at the level at which the nerves enter from the bite site.⁹² In Thailand, this initial symptom often takes the form of severe itching that can lead to frenzied scratching and extensive excoriations.²¹⁵ This symptom is so well known among the Thai people, and animal bites are so common in that country, that any cause of itching or dysesthesia, even contact dermatitis, can lead to anxiety months to years after an animal exposure.²²⁴

Systemic symptoms usually develop later and are largely nonspecific. Local symptoms may not appear at all. One-third or less of all patients initially have symptoms that suggest the etiology of their infection to physicians who do not commonly encounter the disease. Complaints include malaise, chills, fever, or fatigue. Symptoms that suggest an upper respiratory infection are common and include sore throat, cough, and dyspnea. Gastrointestinal symptoms include anorexia, dysphagia, nausea and vomiting, abdominal pain, and diarrhea. Headache, vertigo, irritability, or anxiety and apprehension suggest CNS involvement. However, even advanced rabies often has nonspecific features.^53,151 Hypoventilation and hypoxia are common during the prodrome and early acute neurologic phase, and the cause is not understood. Cardiac involvement is common and is manifested by tachycardia out of proportion to the fever; hypotension, congestive failure, or even cardiac arrest may ensue.⁹²

Two clinical forms of human rabies are recognized. The furious, encephalitic, or agitated form that is associated with periodic episodes of hyperactivity, restlessness, or agitation is considered most typical. This form of rabies is characterized by periodic opisthotonic spasms or convulsions, particularly in response to tactile, auditory, visual, or olfactory stimuli (aerophobia and hydrophobia). Episodes of disorientation, sometimes with hallucinations or violent behavior, often alternate with periods of lucidity, which can be particularly horrifying because the patient recognizes the nature of his or her illness. The terror associated with hydrophobia has been labeled “powerful but indescribable.” Episodes of priapism, increased libido, insomnia, nightmares, and depression may suggest a psychiatric disorder. Patients maintained with supportive therapy progressively deteriorate, become comatose, lose peripheral nerve function, lose brainstem function, and die.^92,215

Hydrophobia has been described in only 32% of recent U.S. patients,⁹² although one recent U.S. rabies victim, an 11-year-old boy, was so afraid of water he would not even take a bath.⁵⁰ Experienced observers in Thailand have described hydrophobia as a violent, jerky contraction of the diaphragm and accessory muscles of inspiration that is triggered by attempts to swallow liquids and by other stimuli. Usually it is not associated with neck or throat pain or with laryngopharyngeal spasms. It has been likened to respiratory myoclonus (Leeuwenhoek’s disease). When patients lapse into coma, hydrophobia is typically converted into cluster breathing with long apneic periods.²¹⁵

The variability of the clinical manifestations of rabies may result from heterogeneity in wild or “street” rabies viral populations (as contrasted with “fixed” viral strains maintained in laboratories), even in the viruses infecting a single animal. Rabies viruses isolated from a boy who died after being bitten by a fox and propagated by intracerebral inoculation of white mice produced three distinctly different forms of disease in the mice.¹⁰²

The second clinical form of rabies, the paralytic, dumb, or Guillain-Barré–like form, is characterized by progressive paralysis without an initial furious phase. Even though the paralytic form of the disease does not appear to be as familiar to some health care providers, 20% to 30% of human rabies victims present in this manner.^3,224 Other animals also have furious or dumb rabies presentations.¹⁴⁸ Paralytic rabies is more common after rabid vampire bat bites, in persons injected with fixed virus strains, and in persons who have received postexposure vaccination.⁹² Distinction from Guillain-Barré syndrome may be difficult, although individuals with that disorder usually do not have urinary incontinence, which is common in rabies-infected persons.¹¹⁶

Individual case reports make clear that every patient is different. The most common misdiagnoses are psychiatric or laryngopharyngeal disorders.¹¹⁶ Physicians at QSMI, who have vast clinical experience, consider inspiratory spasms to be the most reliable clinical sign of rabies, particularly in comatose patients, regardless of whether the disease was initially furious or paralytic. Such respirations also have been described as rapid, irregular, or jerky, termed apneustic.

In addition, most of the QSMI patients have myoedema, particularly in the region of the chest, deltoid muscles, and thighs.^219,224 However, this phenomenon, a brief, unpropagated, localized muscle contraction that appears in response to percussion with a tendon hammer, has not been confirmed as an important sign of paralytic rabies.¹¹⁶

Because the signs and symptoms are so nonspecific and often are rather mild at onset, many patients with rabies are not hospitalized the first time they are seen by a physician. (Two patients who died in the United States, one between 1960 and 1980 [the year of death was not stated] and the second in 2008, were never hospitalized.³²) When admitted, most patients have a fever, which may be mild, but commonly is above 39.4° C (103° F). Of the 38 individuals who died of rabies in the United States between 1960 and 1980, 24 (63%) had difficulty swallowing, but only one-half of those had definite hydrophobia or aerophobia; 27 (72%) manifested excitement or agitation, 24 (63%) had paralysis or weakness, and 12 (32%) had hypersalivation. Dysesthesias at the exposure site were described by 19 (79%) of the 24 individuals who had an identifiable bite or similar exposure.³

Of the 28 patients diagnosed before death, 26 had a history of an animal exposure. Only four of eight patients diagnosed after death had this history. All 12 patients with hydrophobia were diagnosed before death.³

The duration of illness for patients not given supportive care averages 7.3 days and ranges from 2 to 23 days. For patients who are given supportive care, the average duration of illness is 25.3 days, with a range of 7 to 133 days.^3,195

“The following [therapeutic] measures have … been tried in clinical rabies, but without any evidence of effectiveness: administration of vidarabine; multisite intradermal vaccination with cell-culture vaccine; administration of α-interferon and rabies immunoglobulin by intravenous as well as intrathecal routes; and administration of anti-thymocyte globulin, high doses of steroids, inosine pranobex, ribavirin and the antibody-binding fragment of immunoglobulin G.”²¹⁹ In a recent review of the management of rabies in humans, a combination of some specific therapies was suggested. The authors point out that essentially no individuals with clinical rabies survive, and the best therapy often is palliative.¹¹⁷