Pathophysiology and Clinical Signs of Coccidioidomycosis

Both Coccidioides species cause the disease coccidioidomycosis (CM) also referred to as San Joaquin Valley fever, valley fever, desert rheumatism, or “cocci/coccy.”

Nonfatal disease after exposure opened the possibility of vaccine development (Levine et al.

Prior to the 1950s, there was no effective treatment for coccidioidomycosis (Einstein 1975). The first drug to be found effective was Amphotericin B; however, long-term treatment is complicated by nephrotoxic side effects (Longley and Mendenhall 1960; Fiese 1957; Halde et al. 1957; Lawrence and Hoeprich 1976). The current recommended treatment of CM is fluconazole (Catanzaro et al. 1990; Fierer et al. 1990; Galgiani et al. 1988; Finquelievich et al. 1988; Stevens 1977). Although these drugs are generally well-tolerated, toxicity and drug interactions are still concerns (Stevens and Clemons 2007). Even with treatment, infections may not be cleared for patients who have disseminated disease, although recent work with nikkomycin Z shows promise (Shubitz et al. 2013; Galgiani 2007; Hector et al. 1990). Frequently, lifelong therapeutics and monitoring of disease are required, particularly with coccidioidal meningitis (Antony et al.

Otherwise healthy people and animals living in or visiting endemic areas contract CM via the inhalation of conidia. Rarely, infection has occurred by dermal invasion, in laboratory accidents, and from bandaged subcutaneous lesions (Gaidici and Saubolle 2009; Smith and Harrell 1948; Fischer and Kane 1973). Although most human infections are asymptomatic, symptoms can range from mild to severe (Nguyen et al. 2013). It is argued that the primary reason for this variation is host genotype (Galgiani 2014). This is an inherently unsatisfying argument, given varia­tion in disease presentation as a result of both inoculum levels, as well as isolate/ strain virulence in various laboratory models of infection (Cox and Magee 1998, 2004; Cox and Vivas 1977; Hugenholtz et al. 1958; Friedman and Smith 1957; Berman et al. 1956; Friedman et al. 1953, 1955). Infectious dose, variation among strains, and variation among hosts all play a role in disease outcome. Genome-wide association studies (GWAS) could define genetic-based differences in fungal viru­lence and host response (Muller et al. 2011). Additionally, defining conditions that influence Coccidioides' growth and reproduction will assist with preventing exposure.

4.5.1 Coccidioidomycosis in Primates

Progression and severity of disease in humans are influenced by host response to infection. An increase in disease burden has been detected as an increasing number of naive hosts travel to endemic areas (Fig. 4.3). North American census data shows

Fig. 4.3 Reported cases of coccidioidomycosis in the USA between 1998 and 2015. Data retrieved from https://www.cdc.gov/fungal/diseases/coccidioidomycosis/statistics.html

Arizona population has increased from 3,665,228 in 1990 to 6,731,484 in 2014, with a median age of 36.5 and around one million people aged 65 and older. In California population has increased over the same time period, from 29,760,021 to 38,802,500, with a median age of 35.6 and 4,617,907 over 65.

Among genetic determinants associated with higher possibility of dissemination are mutations in STAT1, STAT3, IL-12Rβ-1, and IFN-γR1. The STAT1 mutation conferred a constitutively active gain of function, which likely results in a dysregulation of IFN-γ mediated inflammation, and was associated with both severe CM and histoplasmosis infections (Sampaio et al. 2013). The STAT3 mutations associated with Job's syndrome result in hyper-IgE levels in serum and dysregulation of IL-17 and Th17 response to infection, and patients are at risk of severe fungal infections (Odio et al. 2015). IL-12Rβ-1 novel missense mutations in two patients with disseminated infection reveals the importance of the IL12:IL23: IFNγ intersection as a risk factor in the human host for severe disease (Vinh et al. 2011). Finally an IFNγ receptor 1 autosomal dominant mutation resulted in severe and long-term disease in a pediatric case of CM, culminating to a coinfection with nontuberculosis Mycobacterium (Vinh et al.

These limited but important studies highlight the axis of IL12flL23flFNγ deficiencies and dysregulation as one cause of severe disseminated CM in humans. However, it does not explain a majority of cases or the cause of severe acute disease. These studies remain to be completed, and highlight needed future work. In addition, nonhuman primates are susceptible to infection, and several documented cases of captive macaques, chimpanzees, and baboons have been reported (Johnson et al. 1998; Bellini et al. 1991; Rosenberg et al. 1984; Rapley and Long 1974; Hoffman et al. 2007; Herrin et al. 2005; Ginocchio et al. 2013; Beaman et al. 1980; Breznock et al. 1975; Blundell et al. 1961). These reported infections are frequently severe disseminated valley fever; however it is not known if this is a common manifestation of the disease in nonhuman primates and how this informs us about human disease. Certainly, primates have been used in vaccine research, and immune profiles have been observed when compared to humans.

4.5.2 Coccidioidomycosis in Dogs

Domestic dogs (Canis familiaris) have been proposed as sentinels of disease and a way to map the organism in the environment in endemic regions where the disease is not reported to the CDC or occurrence outside the endemic region (Gautam et al. 2013). The approach has had success in California and Arizona and has been useful in guiding environmental collections of Coccidioides (Barker et al. 2012; Shubitz 2007; Butkiewicz et al. 2005; Shubitz et al. 2005; Grayzel et al. 2016). Canine prevalence data would be a useful tool for monitoring emergence of new regions of endemicity.

Certain breeds may have higher risk of disseminated disease. Data currently are more suggestive than definitive. Early work that studied 100 dogs postmortem found that male dogs, boxers, and Doberman pinschers had higher prevalence of severe disease that resulted in death (Maddy 1958a).

Conversion to a positive serology in dogs occurs at a similar rate compared to humans: around 70% of infections are asymptomatic (Shubitz et al. 2005). Approxi­mately 24% of 124 dogs enrolled in a prospective study that were followed for 2 years converted positive for Coccidioides exposure, with 6% having clinical disease (Butkiewicz et al. 2005). Another benefit to using naturally infected dogs as models to understand disease in humans is that severity and disease manifestations are similar. The disease generally stays localized to the lungs and hilar lymph nodes (Graupmann-Kuzma et al. 2008). Dogs can develop extrapulmonary complications, such as dissemination to bone, meninges, and other internal organs in about 20% of symptomatic cases, which is higher than in humans (Davidson and Pappagianis 1994).

Treatment options are similar to what is available for people. Amphotericin B and azoles are the drugs of choice with similar toxicities and side effects.

Other canids that live in endemic regions that may be naturally infected include Mexican gray wolves (Canis lupus baileyi), coyotes (Canis latrans), kit foxes (Vulpes macrotis), and the endangered San Joaquin kit fox (Vulpes macrotis mutica). A standard method of testing wild animals for any infectious disease is blood collection, which is followed by serological analysis. For CM, an antibody immu­nodiffusion assay or ELISA to detect Coccidioides + IgG or +IgM is standard. However, a negative result may be uninformative, and certainly in wild animals where testing has not been optimized, results should be viewed with caution. No reports of valley fever in wolves were found in the literature; however, wolves were eliminated from most of the endemic region prior to awareness of the disease and were not likely residents of lower deserts (Carroll et al. 2014). One report of an investigation of valley fever among coyotes reveals that three of five coyotes captured in the Tucson AZ area were subclinically infected and two had culturable fungus from tracheobronchial lymph node tissue (Straub et al. 1961). The fungi were

then tested in a mouse model, and only one of the two cultures produced an infection consistent with valley fever. Finally, a survey of San Joaquin kit foxes showed a very low level of positive Coccidioides serology; however, the exact testing method was not reported (McCue and O'Farrell 1988).

A greater understanding of disease prevalence and severity among all canids found in endemic regions could provide valuable information on the genetic basis of host response to disease and distribution in the environment.

4.5.3 Coccidioidomycosis in Cats

The prevalence of valley fever in domestic cats (Felis catus) and other felines is thought to be lower than canines and with a very different presentation (Greene and Troy 1995). An interesting case presentation was reported in a wild mountain lion (Felis concolor) that had been trapped in Texas and transported to Florida for research purposes related to Florida panther reintroduction (Clyde et al. 1990). The animal was otherwise healthy; however, lung function and blood work were identified as abnormal during an examination and minor surgery. A communicable disease needed to be rapidly diagnosed as other mountain lions were to be released into the wild, and this animal had been in proximity to them. The animal was euthanized, and upon postmortem examination, disseminated CM with peritoneal involvement was discovered and likely represented a natural infection in a wild animal. Another peritoneal CM in a mountain lion was reported in a captured animal taking up residence in a tree in a private yard in Kern County, California (Adaska 1999). The animal was extremely lethargic and emaciated. Recent reports of plague in the area were cause for concern, and the animal was shot. Postmortem examina­tion revealed several granulomatous structures throughout the peritoneum consistent with Coccidioides spherules. Infection was confirmed via histopathology and culture.

Other large felids in captivity have been reported infected by Coccidioides including a Indochinese tiger in the El Paso Zoo (Panthera tigris corbetti) and two Bengal tigers (Panthera tigris tigris) (Helmick et al. 2006; Henrickson and Biberstein 1972). In the case of the tiger, other animals tested at the facility did not have positive serology, and few cases of valley fever had been detected at the facility. The tiger had other comorbidities: chronic renal disease and pancreatic adenocarcinoma (Helmick et al. 2006). The other case report detailed that both Bengal tigers were male and contained at the same location in southern California. The tigers also had severe hepatic disease, which may have predisposed them to disseminated CM. In these two cases, infection with Coccidioides may have been asymptomatic for several years and only surfaced when other disease processes occurred.

In domestic cats, valley fever has variable presentation with the first report of disease in two cats appearing in 1963 (Reed et al. 1963). Both animals were euthanized and upon necropsy confirmed to have significant disseminated CM and disease manifestations similar to canine. However, cats are thought to have fewer fungal infections, and few reports of feline CM are found in the literature. Cats tend to be diagnosed with valley fever later in life, with an average age of 6.2 years at diagnosis, and the most common clinical manifestation was skin involvement (Greene and Troy 1995). This study of 48 cats also revealed that respiratory distress was not common. The primary treatment was ketoconazole; however fluconazole or itraconazole was also used. In 44 of the 48 cats treated, 25% failed treatment and were euthanized or died soon after diagnosis. Of the remaining, long-term (10 months) therapy was necessary with relapse after removal from treatment in a few cats.

4.5.4 Coccidioidomycosis in Armadillos

In South America, CM outbreaks have been associated with armadillo hunters (Eulalio et al. 2001; Wanke et al. 1999). Armadillos have lower body temperature than many other mammals and harbor other pathogens, primarily Mycobacterium leprae, the causative agent of leprosy (Duthie et al. 2011; Loughry et al. 2009). An investigation of 26 captured armadillos (Dasypus novemcinctus) revealed that three were subclinically infected. No evidence of gross pathology or histopathology was discovered upon necropsy; however, macerated spleen and lung tissue grew colonies consistent with Coccidioides. These colonies were subsequently used to infect mice, and these mice developed CM.

The role of armadillos in the ecology, distribution, and prevalence of Coccidioides posadasii in the environment is unknown. These animals have been implicated in association with another fungal disease, paracoccidioidomycosis (Bagagli et al. 2006; Arantes et al. 2016). Additionally, whether or not this associa­tion is restricted to South America, or also occurs in Texas, remains to be assessed. An interesting observation however, is the fact that Texas and South America fungal isolates are more genotypically related than Texas isolates are to Arizona isolates, despite being geographically more distant. Armadillos are not found commonly in Arizona and Central California.

4.5.5 Coccidioidomycosis in Rodents

The role of desert rodents, specifically the Heteromyidae, in the life cycle of Coccidioides is an area of research that has provided confusing data. Early researchers noticed a correlation of higher rates of infection with soil disturbance around rodent burrows (Emmons 1942; Stiles and Davis 1942), which suggests a rodent reservoir for Coccidioides (Ashburn and Emmons 1942). Trapping of 1942 animals occurred in Lordsburg, NM, and Wilcox, Tucson, Casa Grande, and Phoe­nix AZ. A range of species were obtained, including pocket mice, kangaroo rats, grasshopper mice, deer mice, pack rats, ground squirrels, rabbits, and harvest mice. No animals from Lordsburg or Wilcox were infected with Coccidioides. Even in Tucson, Phoenix, and Casa Grande, infection rates in trapped wild rodents were low. Seven out of 207 animals (pocket mice and kangaroo rats only) investigated had confirmed Coccidioides infection, with three additional animals having lesions but no fungal growth (Emmons 1943). Wild-caught rodents are susceptible to infection by Coccidioides; however, route and dosage of infection did not mimic a natural infection (Swatek and Plunkett 1957). Onygenales fungi degrade keratin; thus it is possible that Coccidioides is associated with hair and skin in rodent burrows (Untereiner et al. 2004; Sharpton et al. 2009).

It is possible that a rodent is highly susceptible to CM. Because a sick rodent could be susceptible to predation, a severely ill animal may die in the burrow. Two lines of evidence support this hypothesis. Tissue from infected mice fed to predators did not produce fungal colonies from fecal or pellet material (Swatek et al. 1967), and when infected animals were sacrificed and buried in soil negative for Coccidioides, the soil was subsequently positive for growth (Maddy and Crecelius 1965). Coccidioides does not survive gut passage in foxes, coyotes, or owls (Swatek et al. 1967). Fecal samples from various predators and owl pellets were collected for 5 years near known positive soil sites, but no Coccidioides was recovered (Swatek and Omieczynski 1970; Swatek et al. 1967). Non-predators such as lizards, skinks, a burro, deer, goat, and sheep were also analyzed. Interestingly, Coccidioides does survive in the gut of laboratory and wild mice (Lubarsky and Plunkett 1954). This supports the possibility of transmission of Coccidioides through the gut of rodents, which frequently cannibalize.

Laboratory mice have also shown variable resistance to infection (Kirkland and Fierer 1983). Inbred mice in particular are highly susceptible to severe disease at low dosage of conidia in an intraperitoneal infection model. Comparing DBA/2N (outbred) mice to Balb/c or C57b6 (inbred) mice, the mean lethal dose for outbred was greater than 10⁵ vs. less than or equal to 10³ per mouse. IL-10 deficiencies have been implicated as the source of this susceptibility (Fierer 2007; Fierer et al. 1998).

4.5.6 Coccidioidomycosis in Captive Animals and Other Wildlife

Many wild and domestic mammals are susceptible to CM. The first report of a California sea lion (Zalophus californianus) infected with Coccidioides was a captive animal that was housed in a zoo in Tucson, AZ (Reed et al. 1976). Several years later, a naturally infected dolphin (Tursiops truncatus) was found upon necropsy to be infected by C. immitis (Reidarson et al. 1998). In a recent retrospec­tive report, 36 wild marine mammals that had beached along the Central California coast between 1998 and 2012, including sea lions, sea otters, and harbor seals, were found to be infected with Coccidioides (Huckabone et al. 2015). How marine mammals are exposed remains an open question; however, it is most likely from dust and airborne particulate moving from endemic areas in California to the ocean.

Captive animals in endemic regions are a source of unusual infections but provide information about the range of susceptible hosts. A black rhinoceros (Diceros bicornis) that was moved from Texas to the Milwaukee County Zoo developed severe progressive lameness that was determined upon necropsy to have been caused by disseminated CM (Wallace et al. 2009). Interestingly, between 1984 and 1994, the Phoenix Zoo had several wallabies and kangaroos infected and perish from valley fever (Reed et al. 1996). In another endemic region, a case of a giant red kangaroo being infected, as well as positive soil in its enclosure, was also reported (Hutchinson et al. 1973). One recent case of a koala at the San Diego Zoo succumbing to infection was also reported (Burgdorf-Moisuk et al. 2012). Whether marsupials are particularly susceptible to severe disease remains unknown.

Other reported wildlife native to the Sonoran Desert that has been found to be naturally infected and suffer disease includes bats, desert bighorn sheep, and javelina (peccary). Bats appear to be incidentally infected by Coccidioides (Cordeiro Rde et al. 2012; Krutzsch and Watson 1978). In a recent survey looking for another fungus Histoplasma capsulatum, Coccidioides posadasii was discovered. An exper­imental infection of Macrotus californicus (California leaf-nosed bat) showed this animal to be susceptible to a range of dosages: 50, 100, 200, and 400 conidia (Krutzsch and Watson 1978). However, the pallid bat (Antrozous pallidus) was infected only at the highest dosage of 400 conidia.

In the one reported case of desert bighorn sheep (Ovis canadensis nelsoni), the ram, along with 12 other sheep, was captured in November 1984 from the Marble Mountains in San Bernardino County as part of a relocation effort (Jessup et al. 1989). Animals were penned over the winter in a 17-acre enclosure in the Whipple Mountains near the Colorado River. The herd was released in the February 1985. In mid-September 1985, the ram was observed to have a respiratory infection and died shortly after capture. Postmortem revealed severe disseminated CM, with lungs being heavily infected and damaged. Continued follow-up with the rest of the herd revealed no additional evidence of CM, and no other reports have appeared. Thus it is unknown how frequently this animal is infected and what the burden of disease is for this native desert dweller.

The javelina, or the collared peccary (Pecari (Tayassu) tajacu), is a common resident of the desert southwest and behaviorally likely to be exposed to large inocula. However, only a single report was found in this animal (Lochmiller et al. 1985). The 25 animals in the report were captured throughout Texas and housed in an outdoor enclosure at Texas A&M University. In January 1984, a female exhibited neurological symptoms that resulted in euthanasia, and a blood analysis indicated an infection. The two other animals in the same pen were also euthanized as prevention. Necropsy revealed several granulomas in the lungs, kidney, and spleen verified as Coccidioides via histopathology. It is unclear how the infection occurred, as the animals that were ill were harvested outside the endemic range, and Brazos County (where animals were held captive) is considered to be outside the endemic range. Of interest is that both reports occurred in 1984-1985. It is unknown if this was concurrent with particularly high level of disease in humans as well, as cases were not nationally reported before 1994.

4.5.7 Coccidioidomycosis in Livestock

Greater assessment of CM in livestock has been reported, likely because of potential economic impact. Reports in pigs, cattle, sheep, horses, and llamas have revealed high rates of infection with rare cases of disseminated disease, although llamas and horses appear to have more frequent severe disease than other livestock.

A first report of CM in cattle (Bos taurus) was in 1918, after a slaughterhouse observed infected thoracic lymph tissue and submitted the tissue to a USDA pathologist (Giltner 1918). The pathologist grew Coccidioides from the tissue and tested the organism in several animals: guinea pigs, rabbits, dogs, cattle, sheep, and swine. In calves, subcutaneous infection resulted in lesions at the site of infection but no apparent disease. However, intravenous inoculation resulted in rapid death and involvement of lung tissue.

Later observations by a veterinary meat inspector in Los Angeles slaughterhouses of 3173 cattle from the southwestern USA revealed a similar trend of natural infections resulting in thoracic lymph node involvement (Maddy 1954b). Between 1947 and 1951, Coccidioides infected 1.8% of all cattle (calves excluded), 2.9% of steers and heifers, and 7.3% of steers and heifers from San Joaquin Valley feedlots. Of interest was an assessment of animals shipped direct for slaughter from other areas. Upon arrival, 23 animals from west Texas (Amarillo, Lubbock), 36 animals from eastern New Mexico (Clovis, Tatum), 17 from southeastern Colorado (Springfield), and 4 from southern Oregon (Medford) were found to be infected. No animal had evidence of acute CM.

Skin test surveys of 11,643 range cattle in Arizona revealed a high rate of infection that overlapped with the disease prevalence and distribution seen in humans (Maddy et al. 1960a). Between 1954 and 1959, cattle in Arizona between

I and 6 years of age were skin tested using the coccidioidin skin test developed for humans (Palmer et al. 1957; Edwards and Palmer 1957). Pinal County had the highest rate of infection with 42% of cattle being skin test positive. However, assessments made during slaughter at a southwestern feedlot reveal that although coccidioidal granulomas are visible in thoracic viscera, no other signs of disease were noted (Reed 1960). In addition, laboratory infection experiments with dosages ranging from 5 x 10⁵ to 1.5 x 10⁶ arthroconidia and mycelia intratracheally showed that cattle did not develop disease, with few granulomas in lungs or thoracic lymph tissue (Maddy et al. 1960b). Serological testing was not confirmatory; however infected animals did eventually convert to skin test positive using coccidioidin. Thus cattle are susceptible to infection but do not develop severe disease.

Reports of new world camelids infected and suffering disseminated valley fever are few; however, it appears that this animal does develop severe disease (Fowler et al. 1992). The first case report was a single 8-year-old female llama (Lama glama) with severe disease which was euthanized and upon necropsy was discovered to have disseminated valley fever (Muir and Pappagianis 1982). The rest of the herd was tested for infection using a complement fixation antibody test, and 3 of the

II other llamas showed evidence of infection but no disease. This initial disease report was expanded in a report on 19 retrospective cases between 1981 and 1989 from California and Arizona (Fowler et al. 1992). All animals but one had disseminated valley fever with multiple affected organs. Clinical signs varied widely, with and without cough or dyspnea, and dissemination to kidney, liver, intestine, adrenal gland, and meninges. Dermal infection was more common in llamas from California, but no gender or age correlations were observed. Diagnosis in llamas is confounded by highly variable clinical presentation and lack of reliable serological testing. Only one additional recent report was found, which described ocular disease in a 7-year-old male llama, which disseminated and resulted in euthanasia (Coster et al. 2010). Of particular interest in this case is a lack of travel or residence in the endemic region. Llamas appear to have higher rate of complicated disease than other agricultural animals, but overall disease burden is not high.

CM in naturally infected sheep (Ovis aries) has been reported (Maddy 1954a; Beck 1929). Upon slaughter, in both cases, the animals had lesions in the mediastinal and bronchiolar lymph nodes. In the earlier case, Coccidioides was grown from the tissue and verified in a guinea pig model of infection (Beck 1929). In the second case, caseous lymphadenitis was suspected, which could be caused by a contagious bacterial infection common in sheep and goats, so tissue was sent for pathology (Maddy 1954a). Coccidioides was determined to be the causative agent, although no evidence of illness was present, and no validation was performed in a rodent model of infection. One experimental infection of sheep has been conducted (Giltner 1918). Two animals were infected, one intravenously and the other subcutaneously. Both animals appeared well nourished, and no evidence of outward disease was reported. However, the intravenous infection resulted in multiple organ involvement, and fungus grew from the liver, lymph, and lung tissue. The animal infected subcutane­ously had no fungal structures in any tissue.

One experimental case of infection in a swine (Sus scrofa) is reported (Giltner 1918). Two animals were infected via the right marginal ear vein. Upon necropsy, miliary lung nodules and lesions in spleen and liver were observed. A single report of a young pig succumbing to infection was found (Prchal and Crecelius 1966). A survey of both young (6-month-old butcher hogs) and older (3-year-old breeding sows) animals in Tucson, Arizona, revealed no disease but many granulomatous lesions in the bronchiolar lymph nodes (Prchal and Crecelius 1966). Coccidioides was confirmed first by microscopy, followed by infection in mice. It appears that although animals are susceptible to infection, complicated disease rarely develops.

CM in horses (Equus caballus) is reported with higher frequency that other agricultural animals, possibly because the animals are often considered pets, rather than livestock. One review states that pulmonary involvement and weight loss is a common manifestation of disease, along with osteomyelitis and lesions in thoracic lymph nodes and liver (Ziemer et al. 1992). All 15 animals died or were euthanized, and treatment was not effective in the 4 animals where it was attempted. These cases were also interesting in that several of the animals resided in areas not considered to be highly endemic. Subsequent treatment of animals was more successful due to earlier diagnosis and treatment (Higgins et al. 2006). A case of a 14-day-old foal with acute CM that was euthanized due to severe disease suggests that some horses are more susceptible to infection and suffer disease as a result (Maleski et al. 2002).

Interestingly, the wild Przewalski's horse (Equus przewalskii) appears to be even more susceptible to severe disease than domesticated horses (E. caballus). These animals appear to be resistant to most common infectious disease of horses; how­ever, they are susceptible to disseminated CM (Terio et al. 2003). Thirty Przewalski horses housed in the San Diego Wild Animal Park as part of a breeding and reintroduction program died over a 16-year period, with 10 deaths attributed to CM. No other exotic equids at the same facility had reported valley fever deaths. The cause of susceptibility to valley fever could be from a defect in Coccidioides-specific immune response, but results were inconclusive.

Specific non-mammalian hosts infected intraperitoneally, including crayfish, goldfish, and amphibians, developed mycelia in various tissues, and lizards can develop spherules in the lung (Swatek and Plunkett 1957). A naturally infected Sonoran gopher snake with pulmonary lesions and histological and microscopic evidence of Coccidioides infection was reported (Timm et al. 1988). Although morphology consistent with Coccidioides was observed, the fungus was not validated by mouse infection, and genotyping was not yet available. A second case of an infected Sonoran lyre snake at the Phoenix Zoo suggests that natural infection of reptiles deserves more attention (Reed et al. 1996). Again microscopy consistent with Coccidioides was observed, but in this case the fungal agent was confirmed by infection in a mouse. An incidental finding of a lung granuloma in a Gila monster was reported by the Arizona Veterinary Diagnostic Laboratory, but the animal perished from another cause (Reed et al. 1996).

No avian species have been reported to have infection. No detailed investigation of invertebrate associations, such as nematodes in soil or soil-burrowing insects, has been conducted. An interesting account of mice injected with chromium-51 and buried in soil was relayed at an international CM meeting (Egeberg 1985). Follow­ing the movement of the radiation, insects were implicated in the digestion of mouse tissue in both cases.

4.6