This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cox, R. A. Articles by Magee, D. M. Search for Related Content PubMed PubMed Citation Articles by Cox, R. A. Articles by Magee, D. M.

Coccidioidomycosis: Host Response and Vaccine Development

Rebecca A. Cox^* and D. Mitchell Magee

Department of Microbiology and Immunology, University of Texas Health Science Center, San Antonio, Texas

SUMMARY

INTRODUCTION

HISTORICAL PERSPECTIVES

BIOLOGY OF COCCIDIOIDES

Ecology

Phylogenetic Species

Classification of Coccidioides as a Select Agent

Critical Comments

CLINICAL MANIFESTATIONS

Epidemiology

Categories of Disease

Primary pulmonary infection.

Cutaneous infection.

Valley fever.

Pulmonary coccidioidomycosis.

Disseminated coccidioidomycosis.

Risk Factors for Severe, Disseminated Coccidioidomycosis

Genetically determined susceptibility.

Gender.

Age.

Pregnancy.

Immunocompromising diseases or conditions.

Critical Comments

HOST DEFENSES IN HUMANS

Innate Immunity

Polymorphonucleur leukocytes.

Monocytes/macrophages.

Natural killer cells.

Dendritic cells.

Adaptive Immunity

Cellular immunity. (i) Cutaneous delayed-type hypersensitivity.

(ii) Cytokine production.

(iii) Cytokine activation of monocytes.

Humoral immunity. (i) Antibodies.

(ii) Immune complexes.

Critical Comments

MURINE MODEL OF COCCIDIOIDOMYCOSIS

Host Defenses in Experimentally Infected Mice

Role of T lymphocytes in protective immunity.

Role of cytokines in host defense.

Role of antibody in host defense.

Critical Comments

VACCINE CANDIDATES

Viable Cells

Nonviable Cells

Cell-Derived Antigens

PBS extract of spherule cell walls.

27K.

Ag2/PRA.

T-cell-reactive protein.

Spherule outer wall.

Urease.

HSP60.

Coccidioides-specific antigen.

GEL-1.

ELI-Ag1.

Critical Comments

SPECIAL CONSIDERATIONS

Adjuvants

MPL adjuvant.

CpG-OGN.

QS-21.

Microparticles.

Route of Immunization

CURRENT AND FUTURE DIRECTIONS

REFERENCES

Top Next References

INTRODUCTION

Top Previous Next References

 
Coccidioidomycosis (San Joaquin Valley fever) is a mycotic disease caused by Coccidioides immitis (68, 98, 125, 214) and the newly proposed phylogenetic species C. posadasii (116). The fungus propagates in soil in the semiarid regions of the southwestern United States, Mexico, and Central and South America, in a region corresponding to the Lower Sonoran Life Zone. The saprobic phase is characterized by mycelia that give rise to infectious arthroconidia, which become aerosolized when the soil is disturbed. Humans acquire the infection by inhalation of the arthroconidia, which differentiate into large, endosporulating spherules once they are in the host.

Coccidioides is a formidable pathogen, capable of causing progressive pulmonary and/or disseminated disease in previously healthy individuals. The disease presents a diverse clinical spectrum that includes inapparent infection, primary respiratory disease (usually with uncomplicated resolution), stabilized or progressive chronic pulmonary disease, and extrapulmonary dissemination which can be acute, chronic, or progressive. The degree of severity varies considerably within each category and depends, in part, on the dose of inhaled arthroconidia, the genetic predisposition of the host, and their immunologic status.

Between 25,000 and 100,000 new cases occur each year in the areas of endemic infection in the United States, but marked increases have occurred during sporadic epidemics (125, 248, 268). Among those who acquire primary infection, persons of African, Asian, and, to a lesser extent, Hispanic descent are more likely to develop disseminated disease than are Caucasians. This genetic predisposition, the geographically localized areas of endemicity, and the resistance of persons who experienced a benign, self-limiting primary infection document the feasibility of developing a vaccine against Coccidioides (16).

Coccidioidomycosis is considered to be a reemerging disease owing to the dramatic increase in the number of cases during the past decade. Major outbreaks occurred in southern California in 1977 and late 1991 through 1994 (105, 221). A new resurgence is indicated by the increase in coccidioidomycosis cases during the past year in Arizona (50). These outbreaks may be linked to climatic conditions, and the 1994 cases occurred after heavy rains, when the fungus propagated in the soil, followed by hot, dry, and windy periods that resulted in the aerosolization of mycelium-derived arthroconidia. The resulting high morbidity and mortality associated with these outbreaks prompted community and health-related organizations to seek funding for intensifying efforts to develop a vaccine for coccidioidomycosis. Financial support from the California HealthCare Foundation, the State of California Department of Health Services, and the Valley Fever Research Foundation led to a coordinated research program involving investigators located in California (Demosthenes Pappagianis and Theo Kirkland), Arizona (John Galgiani), Texas (Rebecca Cox), and Ohio (Garry Cole). New and fundamentally important discoveries have emerged from these research studies, and it is reasonable to predict that a vaccine, composed of multiple immunogens, will be entered into phase I and II clinical trials within the near future.

It is the intent of this review to focus on progress in vaccine development and host responses that are crucial for protective immunity.

HISTORICAL PERSPECTIVES

Top Previous Next References

 
Coccidioidomycosis has been a recognized infection since 1892 (235, 303). The first reported case was in an Argentinian soldier who had exhibited skin lesions for 4 years; the causative organism was thought to be a protozoan. The first two cases in the United States heralded the protean manifestations of this infection, with marked differences in clinical presentation between the two patients (245). The first patient presented with a slowly progressing disease, which lead to his death approximately 9 years after the first appearance of symptoms. In contrast, the second patient presented with a rapidly progressing disease leading to death within approximately 4 months of the onset of symptoms. The species name of Coccidioides was proposed for the infectious agent, which was identified as the same protozoan described by Posadas and Wernicke. Because of the differences in clinical presentation and lesion development, it was proposed that each of these initial U.S. cases was initiated by different species, and it was proposed that the species name of immitis (meaning "not mild") be used for the causative organism of the first case and pyogenes be used for the causative organism for the second case. The fungal nature of the organism was delineated four years later by Ophuls and Moffitt (209), who discovered that cultures of tissues always yielded a mold and that when the mold was injected into the ear vein of a rabbit, the rabbit developed "typical tubercle-like" nodules in the lungs, spleen, and kidneys. Microscopic examination of the nodules revealed protozoan bodies and no mycelium. They further showed the organism was dimorphic, undergoing a change from a mold to the "protozoan" phase in tissue and, conversely, sprouting hyphae when the tissue was examined in a coverslip preparation.

Until 1929, disseminated severe coccidioidomycosis, termed coccidioidal granuloma, was the only recognized form of the disease. However, a laboratory accident provided the first insight into milder forms of infection. Harold Chope, a 26-year-old Stanford University medical student who was doing research on Coccidioides in the laboratory of Ernest Dickson, accidentally opened a petri dish containing the mold form of Coccidioides. Chope became ill within 9 days of the incident, with an acute pneumonia, with pleuritic pains, fever, cough, hemoptysis, and a 15-lb weight loss over an 8-day period (114). Four weeks later, erythematous nodules erupted on his shins and endosporulating spherules were observed in the sputum. A diagnosis of coccidioidal granuloma was made, and the prognosis was considered to be grave, but Chope soon recovered completely. In 1937, Dickson presented Chope's case and four additional cases of this newly recognized benign form of this disease to the annual meeting of the California Medical Association. All five cases were characterized by acute pulmonary symptoms and, with one exception, erythema nodosum. Dickson's paper, published in California and Western Medicine (93), was entitled "Valley Fever" of the San Joaquin Valley and fungus Coccidioides, and in it he stated that the cases ‘prove conclusively that fungus Coccidioides is sometimes the cause of a symptom complex of acute illness identical with what has been known locally in the San Joaquin Valley as "Valley Fever." ' Dickson proposed the term "coccidioidomycosis" to include all forms of infection by Coccidioides.

BIOLOGY OF COCCIDIOIDES

Top Previous Next References

 

Early conversion of the arthroconidia into spherule-phase cells begins with isotropic growth characterized by a rounding up and swelling of the cells followed by synchronous nuclear divisions and segmentation, which is initiated by synchronized, centripetal growth of the spherule wall at multiple points (163). The central portion of the young spherule is occupied by a vacuole. Progressive compartmentalization of the cytoplasm that surrounds the vacuole gives rise to uninucleate compartments which round up and differentiate into endospores. The mature spherule measures 30 to 60 µm and can contain 200 to 300 endospores. At maturity, the spherule ruptures, releasing the endospores, which typically measure 2 to 4 µm in diameter. Each of these first-generation spherules is capable of developing into a mature, endosporulating spherule, thereby repeating the parasitic phase cycle. Thus, at any given point in time, the infected host is exposed to immature, mature, and rupturing spherules and newly released endospores, which differ quantitatively, if not qualitatively, in their cell wall and cytoplasmic composition.

An understanding of the regulatory events that underlie the induction of morphogenetic conversion in Coccidioides is rudimentary at best. The development of a defined liquid medium by Converse (59-61) enabled studies to delineate the absolute requirements for spherule induction and maintenance in vitro (29, 59, 60, 63). Increased temperature (between 34 and 41°C) and CO₂ concentration (10 to 20%) induce spherulation. The addition of a surface-active agent such as Tamol N also stimulates spherulation. At 41°C, 100% of arthroconidia convert into spherules, whereas at lower temperatures under the same conditions, the arthroconidia give rise to hyphae. The differentiation of arthroconidia into endosporulating spherules in vitro appears to be identical to that observed in vivo (100) and, in both environments, typically takes between 72 and 96 h.

Molecular-genetic analyses of C. immitis strains began with the use of restriction fragment length polymorphism to analyze and compare the genomes of 14 isolates from California and 1 isolate from Venezuela. In this study, Zimmermann et al. (326) demonstrated that C. immitis contains at least two subgroups. Of the 14 isolates, 2 isolates, including 1 of the standard laboratory strains, Silveira, were placed in group I and the rest were placed in group II. There were no discernible differences in the group I and II isolates in terms of the geographic region where they were isolated or whether they were isolated from a patient with pulmonary or disseminated disease. The investigators noted, however, that group II isolates could be subdivided into additional subgroups. Subsequently, Burt et al. (32-34) identified polymorphic loci and examined 12 of these polymorphic loci in clinical isolates collected from 25 patients in Bakersfield, Calif., 25 patients in Tucson, Ariz., and 20 patients in San Antonio, Tex. Substantial genetic differentiation was observed between the isolates from California and those from Arizona or Texas, with little to no gene flow. There was also a significantly reduced gene flow between isolates from Arizona and Texas, but not as much as was observed between the isolates from these two states and those from California. Koufopanou et al. (176, 177), in a comparison of gene genealogies from 350- to 650-bp fragments of five nuclear genes, separated 17 isolates of Coccidioides that had been collected from California, Texas, Arizona, Mexico, and Argentina into two strains: one from California and the other from all other geographic locations. Strain Silveira was a notable exception in that it did not share polymorphisms with the California species, which corroborated the findings of Zimmermann et al. (326). Exceptions were also observed, since isolates from three patients from California showed the non- California genotype.

Fisher et al. (117, 119) identified seven microsatellite-containing loci for C. immitis, four of which were originally isolated from the California strains and three of which were isolated from the non-California strains. Analyses of 20 clinical isolates from the southwestern United States revealed that six of the seven microsatellites showed nonoverlapping allele distributions between the California and non-California strains. In a subsequent study, these investigators (118) examined 161 isolates from the areas of endemic infection in California, Arizona, Texas, Mexico, and South America. Using nine microsatellite-containing loci, the investigators confirmed that the isolates comprised two primary subgroups, i.e., the California and non-California phylogenetic strains. Subclades were observed with divergence between isolates from Central Valley, Calif., and those from the rest of southern California, which is delineated by the Tehachapi mountain range. Likewise, there was phylogeographic divergence within the non-California clade, with the Texas and South American isolates grouping into a subclade.

Perhaps the strongest argument for the two species is the lack of evidence of genetic exchange between C. immitis and C. posadasii. It appears, however, that the vast majority of studies performed to date have been done with the same isolates of C. immitis, and at the time of this writing, only 167 strains have been examined. There appears to be considerable overlap in the number of non-California stains (C. posadasii) in the California group (C. immitis) and, conversely, the number of California strains (C. immitis) in the non-California group (C. posadasii) (116). These collective results prompt the question whether posadasii should be used to designate a variety of C. immitis as opposed to a new species of Coccidioides.

As an argument against separation of the species, Pappagianis (213, 220) has emphasized that immunization with one C. immitis strain provides protection against respiratory challenge with a phenotypically or genotypically different C. immitis strain. Although differences in the virulence of C. immitis strains have been documented, those differences do not correlate with differences in their immunogenicity. For example, immunization of mice with viable arthroconidia of strain 46 protected them from intranasal (i.n.) challenge with strain Silveira. Strains 46 and Silveira differ when analyzed by restriction fragment length polymorphism (326) and by single-strain conformational polymorphism analysis. Strain 46 is classified as a California strain and strain Silveira is classified as a non-California strain by genotypic markers (116). Huppert et al. (146) showed that vaccination of mice with killed spherules of strain Silveira protected against i.n. challenge with strain 46, strain Woodville, and five other strains of C. immitis that were considered to be phenotypically atypical. Further support that immunizing strains of C. immitis protect against challenge with other strains is evinced by the solid immunity of persons who recovered from a primary benign coccidioidal pneumonia to exogenous reinfection. That is, such persons are likely to have been exposed to other strains of C. immitis as residents in regions of endemic infection.

While the differences in genotype between C. immitis and C. posadasii are strong, differences in phenotype are not remarkable and also argue against separation of the species. Although C. posadasii was reported to grow more slowly on yeast extract-glucose agar medium with high salt concentration, there was considerable overlap between the C. posadasii and C. immitis isolates, and hence the phenotype could not be used to distinguish the two (116). Minor differences in amino acid sequence of proteins of C. posadasii and C. immitis were reported by Koufopanou et al. (177) and Peng et al. (229), but as yet no differences have been detected in antigenicity, virulence, or morphology of the two groups. To begin to answer this latter question, we collected 12 isolates of Coccidioides (7 C. posadasii strains and 5 C. immitis strains) to compare them in an animal model of virulence (D. M. Magee and R. A. Cox, unpublished data). Since these isolates had been collected from various investigators over time, they were passed through mice before use in the present experiment. Groups of 10 BALB/c mice were infected intranasally with each strain and then monitored for survival over 45 days (Table 1). The analysis of this initial experiment shows that there are three patterns of survival after challenge, showing high, intermediate, and low virulence. Strain Silveira, our standard laboratory isolate, which is included with the proposed C. posadasii designation, exhibited intermediate virulence, and mice challenged with this isolate began to succumb on day 13, with a median 50% survival on day 14. In contrast, strain RMSCC 1040 (from the Roche Molecular Services Culture Collection [RMSCC]), also in the C. posadasii designation, exhibited increased virulence, and mice challenged with this strain began to succumb on day 9, with a median 50% survival on day 10.5. Comparison of the survival curves revealed that survival was significantly decreased compared to that of mice infected with strain Silveira (P < 0.002, Mantel-Haenszel log-rank survival analysis). On the other extreme, strain RMSCC 1038, also in the C. posadasii designation, exhibited low virulence, with 100% survival for over 45 days after challenge. The animals were sacrificed at that time, and all animals contained C. immitis in their lungs; therefore, the increased survival is not due to failure of infection. Thus, while we have Coccidioides strains with different levels of virulence in mice, variation in virulence between the proposed species designations was not demonstrable.

Fierer and Kirkland have questioned the rationale and justification for classifying the fungus as a select agent (110). These investigators argue that (i) it would be relatively easy to isolate Coccidioides from the soil (as would also be the case for Bacillus anthracis); (ii) most primary infections are benign; (iii) the incubation period (usually 10 to 16 days) is too long to disable a military unit; (iv) no vaccine is available to protect those who are using Coccidioides as a biothreat agent; and (v) Coccidioides is not contagious. Although it is difficult to perceive Coccidioides itself as a biothreat, unless a very high inoculum of the fungus could be delivered, one could visualize transfected Coccidioides arthroconidia as vectors for delivering biothreat agent toxins, in which case the availability of a vaccine becomes essential.

The inclusion of Coccidioides as a potential bioweapon, combined with the failure of the National Institutes of Health to include it on the Category A/B/C list for prioritization, may impede the research on this organism. The result is that this organism falls under the strict federal regulations but does not warrant targeted research funding. The cumulative effect could be an overall reduction in the effort to understand the biology and pathogenesis of Coccidioides.

CLINICAL MANIFESTATIONS

Top Previous Next References

 

Coccidioidomycosis is considered to be a reemerging disease because of the dramatic increase in the number of cases during the early part of the past decade (47, 52, 105, 157, 207, 219, 260, 282). Between 1991 and 1994, there was a notable increase of new cases in California, in particular in Kern and Tulare counties in the southern part of the San Joaquin Valley (219). During that time, there were 8,435 reported cases in Kern County, representing a sevenfold increase. This epidemic was attributable to an unusually rainy spring in March 1991 and February-March 1992 following a 5-year drought, the migration of persons previously unexposed to Coccidioides into areas of endemicity in southern California, and new construction projects. In a review of medical records in Kern County by the Centers for Disease Control as Prevention, medical bills totaled $45 million for hospitalization and outpatient care (283). A similar increase in coccidioidomycosis incidence occurred in Arizona between 1990 and 1995, from 7.0 per 100,000 population in 1990 to 14.9 per 100,000 in 1995 (10). The increase was thought to be attributable to the influx of older persons (65 years or older), who were nonimmune upon arrival. There was no apparent correlation between the increased number of cases and climatic conditions. After a period of quiescence, the number of cases is increasing in Arizona, which may herald the onset of a new epidemic (50).

More frequent travel, both domestic and international, to areas of endemicity has contributed significantly to this increase (37, 46, 48, 49, 53, 92, 212, 300). In 2000, nearly 28 million travelers visited Arizona, and of these, nearly 1 million were from abroad. Even greater numbers of travelers visit California and other areas of endemicity. In 2001, more than 300 persons from 30 countries participated in the World Championship of Model Airplane Flying in Lost Hills, Calif., an area where the fungus is highly endemic (46). Cases of coccidioidomycosis in participants who returned to Australia, Finland, New Zealand, and the United Kingdom were reported. Between 1992 and 1997, New York State had a total of 161 cases of coccidioidomycosis, all of which occurred in patients who had traveled to a region of endemicity, mostly in the southwestern United States (53). The Cleveland Clinic had 23 cases between 1980 and 1998 among patients who had traveled to an area where C. immitis was endemic (92). Travelers from the United States to areas of endemicity in Central and South America have also acquired coccidioidal infection. In 1996, members of a church congregation from Washington State traveled to Tecate, Mexico, to assist with construction projects (37); 21 (17%) of the members were diagnosed with coccidioidomycosis following their return. A similar incident occurred with church members from Pennsylvania, who acquired primary infection after traveling to Hermosillo, Mexico, to assist in a church construction project (48).

Cutaneous infection. Coccidioidomycosis can be acquired via a percutaneous route. Most of these occur in laboratory workers as a result of a hypodermic injection of Coccidioides (99, 310, 312). Primary cutaneous coccidioidomycosis is characterized by a painful suppurative lesion at the site of inoculation, often with regional lymphadenopathy. Of the 18 cases reported as of 1977, all but 2 have remained localized (41).

Valley fever. Approximately 5% of all primary infections develop what is termed as the Valley fever complex, usually coincident with the development of delayed-type hypersensitivity reactions (99,114, 275). Erythema nodosum and erythema multiforme comprise the principal manifestations of the valley fever complex and may be accompanied by arthralgias (desert rheumatism) and a mild conjunctivitis. These are considered to be specific cutaneous lesions of acute coccidioidomycosis, in contrast to the nonspecific toxic erythema noted above, and are temporally associated with the acquisition of delayed-type hypersensitivity to Coccioides. Of the two syndromes, erythema nodosum is the more extensively studied. Initially, the lesions appear as bright reddish nodules, but within days they become livid red or purplish, and on healing they appear as bruises. They are commonly limited to the lower extremities. Erythema nodosum occurs most often in Caucasian females and has long been regarded to denote a good prognosis, although exceptions have been documented (99, 122, 157). Curiously, the predisposition of females to developing erythema nodosum is not manifest before puberty. Arthritis involving the joints, most commonly the ankle and knee, develops in approximately one-third of persons with erythema nodosum and/or erythema multiforme. These clinical manifestations of primary coccidioidomycosis are thought to be attributable to a hypersensitivity to the fungus, a concept that is supported by the hyperreactivity of the patient to skin testing with coccidioidin. Fungal cells are not present in the lesions of erythema nodosum or multiforme, nor have they been demonstrated in coccidioidal arthritis or conjunctivitis; as yet, the underlying basis of the valley fever complex has not been determined. Erythema nodosum in other diseases, notably tuberculosis, leprosy, sarcoidosis, and autoimmune disorders, is considered to be a hypersensitivity response, and in many cases, circulating immune complexes, C3, and immunoglobulin G (IgG) are demonstrable in the walls of venules (242). Although circulating immune complexes, composed of anti-Coccidioides IgG and coccidioidal antigen(s), have been demonstrable in the sera of patients with coccidioidomycosis (81, 317), studies have not been done to assess immune complexes in persons with valley fever complex.

Pulmonary coccidioidomycosis. Approximately 5% of persons with primary coccidioidomycosis develop persistent pulmonary coccidioidomycosis, manifested by chronic progressive pneumonia, miliary disease, pulmonary nodules, or pulmonary cavitation (99, 311, 313, 314). Pulmonary nodules are usually benign but can become cavitary. A classic radiologic finding is the presence of a "thin-walled cavity," which typically fails to show a surrounding tissue reaction. Although the latter is not pathognomonic, it is strongly suggestive of coccidioidomycosis (114). Most patients have only a single cavity, whereas in others the cavities are multiple or multilocular (276). In a study of 211 cases, Hyde (149) reported that half of cavities eventually close spontaneously, requiring neither surgery nor chemotherapy. Possible complications of cavitation include hemorrhage, secondary infection, progressive increase in size, and, if located peripherally, bronchopleural fistulae. A few patients develop chronic progressive pulmonary involvement, with symptoms of cough, weight loss, fever, hemoptysis, dyspnea, and chest pain that may persist for years. Radiographic results include inflammatory infiltrates, biapical fibronodular lesions, and multiple cavities.

Peripheral blood eosinophilia has been found in many patients with primary coccidioidomycosis and patients with disseminated disease (106, 114, 135, 259, 306). In one study, 66 of 75 patients with acute symptomatic pulmonary coccidioidomycosis showed eosinophil counts ranging from 3 to 26% of the total peripheral blood count (135). However, one patient was observed to have a peripheral blood eosinophilia as high as 89% (306). Peak eosinophilia generally occurs between the second and third weeks of clinical illness. Schermoly and Hinthorn (259) reported a patient with both 48% eosinophilia in his peripheral blood and 91% eosinophilia in his cerebrospinal fluid. The investigators further noted that a bone marrow biopsy specimen showed a marked proliferation of eosinophils. The patient was treated with amphotericin B, and after 2 weeks of therapy her eosinophil counts were normal. The basis for the increased eosinophilia is not known, but it does not appear to be associated with erythema multiforme or erythema nodosum.

Disseminated coccidioidomycosis. The early epidemiological studies by Smith et al. (277) established that approximately 1% of patients with primary coccidioidomycosis developed disseminated disease. This incidence has been higher in more recent studies. In an epidemiologic investigation of the outbreak that occurred in Ventura County following the earthquake in Northridge, Calif., in 1994, Schneider et al. (260) reported that 3.7% of patients developed disseminated disease. Pappagianis and Einstein (221) reported a 4.2% incidence of dissemination as a result of the 1977 dust storm in California, and Pappagianis reported a 5.7% rate in military personnel and their families at Lemore U.S. Naval Air Station between 1961 and 1977 (216). Similar findings were observed in the epidemic of coccidioidomycosis in the San Joaquin Valley between September 1991 and January 1994 (157).

Dissemination, when it occurs, is usually an early event and may occur in the absence of any clinical or radiographic evidence of previous pulmonary infection (99, 114, 122), although exceptions have been noted (252). The process may be acute, subacute, or chronic. Extrapulmonary spread may consist of a single lesion in the skin and subcutaneous tissues, bone, meninges, lymph nodes, spleen, liver, kidneys, pleura, or virtually any part of the body, with the general exception of the gastrointestinal tract (122). If only a single lesion develops, unless it is in the meninges, prognosis is generally favorable. If dissemination is multifocal, the overall mortality rate is greater than 50%.

Lesions in the skin and subcutaneous tissues occur in more than 65% of cases of disseminated disease and may present as small papular nodules, ulcerated nodules, or verrucous granuloma. When bone is involved, complicated sinuses may form communications between the bone and the adjacent soft tissue, leading to a draining sinus through the skin. Sinus tracts also originate in subcutaneous tissues and the viscera. Meningitis occurs in 30 to 50% of cases of disseminated disease, and in some patients this is the only site of extrapulmonary disease. In the absence of treatment, the disease is invariably fatal, with death usually occurring within 2 years of primary infection. Acute miliary dissemination, in which seeding of the fungus is thought to occur early after primary infection, is also almost invariably fatal, with death occurring within 3 to 4 months.

Although the genetic basis for this increased susceptibility remains elusive, investigators have conducted studies to assess genetic polymorphisms that may control or be associated with the predisposition of certain ethnic populations to Coccidoides. Human leukocyte antigen (HLA) molecules are highly polymorphic and present antigenic peptides to /ß T lymphocytes. Scheer et al. reported an increased phenotype frequency of the HLA-A9 and -B9 antigens in patients with disseminated coccidioidomycosis (M. Scheer, G. Opelz, P. Tarasaki, and W. Hewitt; Program Abstr. 13th Intersci Conf. Antimicrob. Agents Chemother., 1973). Persons with the ABO blood group B phenotype have also been reported to be at risk for developing disseminated disease (91). These findings are consistent with, and may merely reflect, the increased phenotype frequencies of the HLA-A9 and blood group B in persons of Filipino and Black descent (2, 204, 228).

In a case-control study of persons from Kern County, Louie et al. (191) compared HLA class II alleles and haplotypes and ABO phenotypes in mild versus severe (disseminated) coccidioidomycosis. No differences were observed in the ABO blood types for the Caucasians or African Americans with regard to whether their disease was mild. Among Hispanics, A and B phenotypes were significantly more frequent in patients with disseminated disease than those with mild, uncomplicated pulmonary disease. The investigators reported that the HLA class II DRB1*1301 allele marks a predisposition to severe disseminated disease in all patients, regardless of their ethnic or racial background. It bears comment that the incidence this allele was increased in Caucasian, Hispanic, and African American patients compared with the ethnicity-matched controls. There was not, however, an increase in the incidence of this allele in patients with disseminated disease versus those with mild pulmonary disease within any of the ethnic groups, as would be expected if the DRB1*1301 allele were truly associated with an increased risk for dissemination. Identification of host genetic variants that predispose persons to severe disseminated coccidioidomycosis would be of great value in vaccine development and should be pursued, particularly in light of the growing availability of intragenic single-nucleotide polymorphisms for mapping human genome sequence variations (250).

While it is assumed that the increased susceptibility of African Americans, Asians, and Hispanics has an immunologenetic basis, studies have not yet documented that supposition. If an immunologic basis exists, it does not appear to reside in an inherent inability of these persons to mount a cellular immune response to Coccidioides. Gifford and Catanzaro (133), in a comparison of Coccidioides antigen skin testing in patients with active coccidioidomycosis, noted that although African Americans were more likely to have disseminated disease than were Caucasians, their skin test reactivity was comparable to those among other ethnic groups. Along this same line, Williams et al. (309) reported that African Americans and Filipinos acquired T-cell reactivity in response to vaccination with the formalin-killed spherule (FKS) vaccine at levels comparable to that observed in Caucasians.

Gender. Males are reported to have a 3.5- to 5-fold-increased occurrence of disseminated coccidioidomycosis compared to females (99, 217). Following the major 4-year epidemic in San Joaquin Valley, Johnson et al. (157) studied 535 patients, 25 of whom had disseminated disease, and found that male gender was a risk factor for dissemination when the data were analyzed by univariate (P < 0.05) but not multivariate analysis. However, Arsura et al. (13), in analyses of 536 cases in Kern County between September and December 1991, found that 76% of the 25 patients who had disseminated disease were male, compared to 52% of males in the total population; this was significant by chi-square analysis (P < 0.02; odds ratio, 2.9). It appears that males are more susceptible than females, but the differences may not be as high as noted in earlier studies.

Age. Sievers (266), in his study of disseminated coccidioidomycosis in the American Indian population, found that children younger than 5 years and adults older than 50 years were significantly more susceptible than persons aged 6 through 49 years. An increased susceptibility of older persons was observed in the outbreak in Ventura County following the Northridge earthquake (260), in the 3-year coccidioidomycosis epidemic that started in San Joaquin Valley in 1991 (12, 157) as well as a subsequent follow-up covering the period from January 1995 through December 1996 (248), and in analyses of cases of coccidioidomycosis in Arizona (182). In the last study, elderly persons who had recently relocated to Arizona appeared to be at the highest risk.

Pregnancy. Until recently, the increased susceptibility associated with pregnancy was considered to be unequivocal (99, 136, 137, 238, 274). In areas of high endemicity, approximately 0.1% of pregnancies have been complicated by coccidioidomycosis, with a resulting mortality rate of 88% (291). In a review of 65 women who had coccidioidomycosis during pregnancy, VanBergen et al. (291) reported that 37 (57%) developed disseminated disease and, of these, 29 (78%) died. The later the gestation stage, the more likely it is that dissemination will occur (217). These studies have prompted some to consider abortion or early delivery in gravid females with coccidioidomycosis, in particular those in the third trimester of pregnancy. More recent studies suggest that the risk of dissemination and death in pregnancy have been overstated (15, 38, 293). Wack et al. (293) examined 47,120 pregnancies among three health care centers in Tucson, Ariz., during a 6-year period and reported that only 10 were complicated by coccidioidomycosis. None of these were fatal. The two women who developed fulminant, disseminated coccidioidomycosis were thought to have acquired their coccidioidal infection during the third trimester. In a retrospective analysis of coccidioidomycosis in pregnant women during the California 1991 to 1994 outbreak, Caldwell et al. (38) reported that disseminated disease occurred in 3 (9%) of 32 patients. None of the patients died. The investigators noted that although the 9% incidence of disseminated coccidioidomycosis in pregnant women was lower than that reported in earlier studies, it is higher than that observed in the general population and in females of reproductive age. One consistent finding is that the increased risk of developing disseminated coccidioidomycosis occurs in women who acquired their primary infection during pregnancy and not in those who had a previous coccidioidal infection (99).

It has been theorized that the high risk of severe coccidioidomycosis in pregnant females is attributed to the immunosuppressive state that accompanies gestation. Another mechanism was offered by Drutz et al. (101) and Powell et al. (236, 237), who reported that progesterone and 17 ß-estradiol, at levels comparable to those in the sera of pregnant females, stimulate the growth of the spherule/endospore phase in vitro. While this offers an attractive and plausible explanation to account for the unique susceptibility of gravid females to coccidioidomycosis, there have been no reports of whether these hormones have a stimulatory effect on fungal growth in vivo.

Immunocompromising diseases or conditions. Coccidioidomycosis is a frequent complication for persons who are immunologically compromised by human immunodeficiency virus (HIV) infection (3, 6, 115, 197, 315). A prospective study conducted during a 4-year period in Phoenix and Tucson indicated that the risk of active coccidioidomycosis in HIV-infected persons ranged from 8% to 41% (6). It is not known what percentage of HIV-infected persons with coccidioidomycosis have a primary infection with Coccidioides or reactivation of a previously quiescent infection as a consequence of their pronounced immunodeficiency. In support of the former, Ampel (3) did not observe any association between the length of time the HIV-infected person resided in a coccidioidomycosis-endemic area and a history of a previous infection with the fungus. The course of coccidioidomycosis in HIV-infected persons can vary widely, ranging from a median survival of only 1 month in patients with a diffuse bilateral reticulonodular infiltrates to several months or longer in persons who have unilateral focal pulmonary infiltrates (125).

Persons with other immunosuppressive diseases or conditions, such as Hodgkin's disease, malignant neoplasms, and collagen vascular disease, recipients of immunosuppressive drug therapy, and organ transplant recipients also have a high risk of dissemination of disease and death (30, 95, 292). There have been several reports of reactivation of previous coccidioidal infection in organ transplant recipients (30, 95, 292).

It is also possible that an acute bacterial or viral infection can reactivate a quiescent coccidioidal infection. To cite one example, coccidioidomycosis was diagnosed in a 48-year-old Caucasian man who reactivated a prior benign coccidioidal infection while in England after developing a group A, beta-hemolytic Streptococcus infection (307).

HOST DEFENSES IN HUMANS

Top Previous Next References

 

Transformation of arthroconidia into spherules renders the latter impervious to phagocytosis and killing by PMNL (123, 126), owing in part to the increased size of the spherules (60 to 80 µm) relative to PMNL (12 µm). Although Galgiani (126) reported that PMNL appear to directly adhere to spherules in vitro, Frey and Drutz (123) reported that spherules possess an extracellular fibrillar matrix that impedes their physical contact with PMNL. Rupture of the spherules and release of the endospores triggers an influx of PMNL (122, 123). The newly released endospores are encased in a fibrillar matrix, but over time they become single cells that are readily phagocytized by PMNL. Ingestion of the endospores triggers an oxidative burst, albeit to a lesser extent than that induced by arthroconidia (23, 123), and the level of intracellular killing is no more impressive than that observed in the interaction of PMNL and arthroconidia (23, 31, 123).

Monocytes/macrophages. Kashkin et al. (158) reported that peritoneal macrophages from nonimmune guinea pigs phagocytized but did not kill arthroconidia. The relative inefficacy of nonimmune macrophages in killing Coccidioides was confirmed by Beaman et al. (20-24). These investigators examined the in vitro interaction between arthroconidia and endospores with alveolar macrophages from nonimmune rhesus macaques, resident peritoneal macrophages from DBA/2 mice, and peripheral blood monocytes from healthy, skin test-negative persons. Both arthroconidia and endospores are phagocytized by monocytes/macrophages, but fewer than 1% of the phagocytized cells were killed. In contrast, Ampel and Galgiani (7) reported that peripheral blood monocytes from healthy, skin test-positive or -negative persons inhibited or killed arthroconidia. The differences in the results of these studies could be attributable to differences in the methods used to quantify fungal viability. Beaman et al. (22-24) used conventional assays for determining fungal CFU, whereas Ampel and Galgiani (7) used a newly developed microtiter system for determining fungal CFU (102) and a radiolabeled N-acetylglucosamine (precursor to chitin) uptake experiment to assess the inhibition of fungal viability.

One mechanism that Coccidioides might use to survive intracellularly is the inhibition of phagosome-lysosome fusion (21, 24), a strategy used by many intracellular pathogens to evade the antimicrobial effects of phagocytes (11, 206, 232, 249, 284, 305). Coincubation of monocytes/macrophages with immune T lymphocytes or gamma interferon (IFN-) significantly enhanced their anticoccidioidal activity (18, 19) (see below).

Natural killer cells. Natural killer (NK) cells are a major component of innate immunity. Under normal conditions, NK cells are confined primarily to the peripheral blood, spleen, and bone marrow, but they can migrate to sites of inflammation in response to chemokines. On activation, NK cells secrete cytokines, notably IFN-, and chemokines that induce inflammatory responses and control the growth of monocytes and granulocytes (203). Before adaptive immunity has fully developed, NK cells are thought to the main source of IFN-, in response to macrophage-derived interleukin-12 (IL-12) and IL-18.

Petkus and Baum (231) reported that incubation of spherule/endospore-phase cells for 4 h with human peripheral blood lymphocytes, depleted of monocytes/macrophages, significantly reduced the viability of the fungal cells. Incubation of the lymphocytes with Leu-11 (CD56) antibody, which binds the Fc receptor of NK cells, and complement reduced the anticoccidioidal effect of the lymphocytes by 80%. Supernatants from peripheral blood lymphocytes coincubated with K562 cells or Coccidioides were cytotoxic for Coccidioides, as judged by an inhibition of fungal CFU. These results suggest a direct cytotoxicity of NK cells and NK-cell factor. However, because Leu-11 is not specific to NK cells, additional studies are needed to confirm that NK cells are directly toxic to Coccidioides. One approach might be to utilize the NK-92 cell line (CDRL-2407; American Type Culture Collection), which is a human NK-cell line that is highly toxic to a broad range of NK-cell targets (193, 289).

Dendritic cells. Dendritic cells (DCs) are potent antigen-presenting cells (APCs) and play a pivotal role in innate immunity and adaptive immunity (190). On initial infection, precursor DCs are recruited from the blood to inflammatory sites, where they transform to immature DCs. In the initial interaction, the pathogen binds to pattern recognition receptors, notably Toll-like receptors (TLR), which recognize structurally conserved pathogen-associated microbial products. This initial recognition and binding leads to the induction of proinflammatory cytokines, which include tumor necrosis factor alpha (TNF-), IL-1, IL-6, and IL-8.

Recent studies by Richards et al. (243) showed that a spherule-phase antigen, toluene spherule lysate (TSL), induced maturation of peripheral blood-derived DCs from healthy, nonimmune subjects. DC maturation was demonstrated by increased cell surface expression of HLA-DR, CD40, CD54, CD80, CD83, and CD86. The TSL-pulsed DCs also stimulated proliferation in allogeneic lymphocytes, to a greater level than did nonpulsed (immature) DCs, and they stimulated autologous nonadherent peripheral blood mononuclear cells to produce IFN-. The potential immunotherapeutic use of DCs was established by Richards et al. (244), who showed that the anergy demonstrated by peripheral blood lymphocytes from patients with disseminated coccidioidomycosis could be reversed by the addition of DCs pulsed with coccidioidal antigen. Although the latter studies were conducted in vitro, additional studies of the restoration of immunity by DC immunotherapy in animal models could reveal a new avenue for adjunctive therapy in severe coccidioidomycosis.

Beginning with the early studies by Smith et al. (276, 279-281), a profile of adaptive immune responses in persons with different entities of coccidioidomycosis emerged. Persons with primary, asymptomatic, or benign disease characteristically have strong skin test reactivity to coccidioidin and low or nondemonstrable levels of anti-Coccidioides complement fixation (CF) antibody. The converse pattern develops in patients who develop severe, chronic, or progressive pulmonary or disseminated disease. Typically, these persons, in particular those with disease involving two or more organ systems (lungs, central nervous system, bones and/or joints), are hyporesponsive or anergic to coccidioidal skin testing but have high levels of anti-Coccidioides IgG antibody to the CF antigen. Recovery from active disease, either spontaneous or in response to antifungal therapy, is in many patients associated with a reacquisition of T-cell reactivity to Coccidioides antigens and decreased CF antibody titers. However, the responses of patients with inactive disease do not coincide with those of persons who were able to overcome their primary infection without consequence; instead, they tend to be intermediate between those of the latter patients and patients with active disseminated disease.

Cellular immunity. (i) Cutaneous delayed-type hypersensitivity. The classical antigen preparation that was used in the early skin test and serologic studies was coccidioidin. This antigen was prepared by Smith (276) as a soluble broth culture filtrate of mycelial cells grown for 2 months in a synthetic asparagine-glycerol-salts medium. With the development of a medium for the cultivation of the spherule-endospore phase in vitro (61), Levine et al. (184) produced an aqueous lysate of spherules of strain Silveira that had been grown in Converse medium and then incubated in distilled water for up to 40 days at 34°C. The soluble aqueous lysate was designated spherulin (SPH) and used as a skin test antigen at a dose of 2.8 µg (Usual Test Strength), which corresponded to coccidioidin 1:100. Skin test reactivity persists in most persons who recover from primary infection, and such persons are endowed with immunity to exogenous reinfection (274, 276, 279). The persistence of coccidioidin reactivity in persons who have recovered from their primary infection may be attributable to reexposure to the fungus, as would probably occur in persons who reside in areas of endemicity, or to the presence of viable Coccidioides cells in calcified lesions (36). Although skin test reactivity persists in most persons, some gradually lose their coccidioidin sensitivity (215, 266). Whether this is accompanied by a loss of resistance to Coccidioides is not known but is a question of immense importance since it may relate to long-term protection in response to vaccination.

As many as 80% of persons who develop solitary pulmonary lesions manifest skin test reactivity to coccidioidin, whereas less than one-third of those who develop progressive or chronic pulmonary disease manifest reactivity (44, 82, 254, 274, 276, 304). Skin test reactivity in persons with extrapulmonary disease varies, depending on the extent of disease involvement. Approximately 70% of those who have a single extrapulmonary site of involvement manifest reactivity to coccidioidin 1:100, whereas fewer than 30% of patients with multifocal disease are reactive to coccidioidin or SPH. Low or nondemonstrable skin test reactivity denotes a poor prognosis for recovery. Smith (274, 276) reported that 75% of patients who were skin test reactive to coccidioidin recovered from their disease whereas only 17% of patients who were skin test negative recovered.

The specificity of the cutaneous anergy has been examined in a number of studies (44, 82, 133, 274). In most patients, the anergy is specific to Coccidioides, as evidenced by skin test reactivity to a panel of recall antigens such as Candidin, mumps antigen, trichophyton, and streptodornase-streptokinase. The exceptions occur in patients who have severe disseminated disease involving multiple foci of infection. Approximately half of these patients fail to respond to recall antigens. In addition, Rea et al. (239, 240) reported that some patients fail to respond to contact sensitization with dinitrochlorobenzene, a result that documents a pronounced suppression of cell-mediated immune responsiveness.

(ii) Cytokine production. TNF- is a cytokine produced by a large variety of cells, including macrophages, DCs, CD4⁺ and CD8⁺ T cells, and B cells (27, 109, 285, 299, 320). Cumulative evidence has established that TNF- is responsible for many of the biological and physiological consequences of acute infection, immunological reactions, and tissue injury (181). In addition to its oncolytic activity and ability to induce cachexia, TNF- can activate neutrophils, enhance the cytolytic activity of macrophages, augment NK-cell activity, promote T- and B-cell proliferation, and modulate endothelial cell surface antigens. TNF- plays multiple roles in immune and pathologic responses in tuberculosis (120, 132, 141, 162, 202, 246, 295). On one hand, TNF- is required for the control of acute infection and the formation and maintenance of granulomas, but on the other hand, it has been implicated as a major component in host-mediated destruction of lung tissue.

Ampel (4) reported that autoclaved spherules and arthroconidia of Coccidioides induced the production of TNF- by adherent mononuclear cells from healthy human donors. TNF- production was increased in cells from healthy, skin test-positive persons when the supernatants were assayed for cytotoxicity against the TNF--susceptible L929 cell line. No differences were evident, however, when the supernatants were assayed by an enzyme-linked immunosorbent assay (ELISA), which, unlike the L929 assay, is specific for TNF- (198). Dooley et al. (97) reported that both FKS and live spherules induced TNF-, IL-1ß, and IL-6 production by peripheral blood mononuclear cells and plastic-adherent monocytes/macrophages of healthy persons and coccidioidomycosis patients. The production of the proinflammatory cytokines was comparable in 15 healthy SPH skin test-positive subjects, 13 healthy skin-test negative persons, and 16 patients with active coccidioidomycosis.

Studies have also been done to assess the production of the Th1-asociated cytokines IL-2 and IFN-. In a study of 20 healthy subjects who were skin test positive to SPH and 15 healthy, skin-test negative persons, Ampel et al. (5) showed that peripheral blood mononuclear cells from skin test-positive but not skin test-negative donors secreted both IL-2 and IFN- in response to in vitro stimulation with a toluene-induced lysate of spherules, designated TSL (Table 2). Comparisons of cytokine production, lymphocyte proliferation, and skin test reactivity revealed that there was a tendency toward a direct correlation, but the heterogeneity of responses precluded a significant correlation. Of note, lymphocytes from five subjects who were skin test positive to SPH were essentially nonresponsive to TSL in vitro. The basis for this unexpected deviation between in vivo and in vitro T-cell responses is not known and, as noted by the investigators, is unlikely to be attributable to differences in the antigenic composition of SPH and TSL, given that the other 15 skin test-positive subjects were reactive to both preparations. An ensuing study by Corry et al. (66) compared the production of the Th1 cytokines IFN- and IL-12 and the Th2 cytokines IL-4 and IL-10 by lymphocytes from healthy, SPH skin test-positive and -negative subjects and patients with active pulmonary or disseminated coccidioidomycosis. The results established that IFN- production was significantly lower in cells from the patients with disseminated disease than in those from healthy, skin test-positive persons. By contrast, lymphocytes from patients with pulmonary disease secreted levels that were comparable to those of healthy, SPH-reactive donors. The study groups did not differ in their production of the Th1 cytokine IL-12 or the Th2 cytokines IL-4 or IL-10. Additional studies are clearly needed to further examine the production of Th1 and Th2 cytokines in coccidioidomycosis, with emphasis on relating cytokine responses to clinical progression or regression. It would be at least as important to address the question whether genetically determined susceptibility is associated with, and perhaps attributable to, an inability to maintain Th1 responses, as has been shown in the murine model (discussed below).

View this table: [in this window] [in a new window]   TABLE 2. Percentage of CD3+ lymphocytes producing intracellular IFN- after incubation with tissue culture medium alone, IL-12, the Coccidioides antigen TSL, or TSL plus IL-12a

(iii) Cytokine activation of monocytes. Beaman and Pappagianis (24) reported that human peripheral blood monocytes phagocytized but did not kill Coccidioides arthroconidia or endospores and that the lack of killing was associated with an inhibition of phagosome-lysosome fusion by the fungal cells. When monocytes were incubated in the presence of lymphocytes from immune persons, there was a significant increase in phagosome-lysosome fusion and killing of the fungal cells. Incubation of the monocytes with recombinant human IFN- or recombinant TNF- augmented the fungicidal capabilities of the monocytes (18). The mechanism by which IFN- or TNF- activate human monocytes to an anti-Coccidioides level is not known, but in studies of human alveolar macrophages from tuberculosis patients, IFN-