Coccidioidomycosis in Latin America

Abstract

Coccidioidomycosis is a highly prevalent systemic mycosis in Latin America and has been reported (human and zoonotic cases) in México, Guatemala, Honduras, Colombia, Venezuela, Brazil, Paraguay, Bolivia, and Argentina. The incidence of coccidioidomycosis in Latin America is unknown due to lack of clinical awareness and limited access to laboratory diagnosis. Coccidioidomycosis is as prevalent in Mexico as in the endemic regions of the United States. The number of cases reported in Brazil and Argentina has progressively increased during the last decade, including areas that were not considered as endemic. Genetic studies have shown that the prevalent species in Latin America is Coccidioides posadasii. Coccidioides immitis has been reported sporadically in indigenous cases from Mexico and Colombia. Coccidioidomycosis and tuberculosis share some risk factors such as immunosuppression and residing in areas endemic for these conditions, so their coexistence in the same patient is not uncommon in Latin America. In most regions, clinical diagnosis of coccidioidomycosis is based on direct sputum examination and histopathology results from biopsies or autopsies. This would explain why primary coccidioidomycosis is rarely diagnosed, and most cases published are about chronic pulmonary or disseminated disease.

coccidioidomycosis, Latin America, epidemiology, tuberculosis

Introduction

The enormous geographic and climatic diversity of Latin America provides a great diversity of ecosystems that allow the development of different microorganisms, including pathogenic fungi. Coccidioidomycosis is a systemic mycosis that has been reported in many Latin American countries. The region has a high proportion of rural population dedicated to agricultural work, exposed to diverse habitats for fungi, being therefore exposed to inhalation (or transcutaneous inoculation) of Coccidioides arthroconidia. Coccidioidomycosis has an important impact on public health; however, despite the reporting of clinical cases in several countries of the region, clinical and epidemiological information is scarce and fragmentary.

Coccidioidomycosis in Mexico

Coccidioidomycosis (CM) is the most prevalent of the systemic mycosis in Mexico. CM was first diagnosed in Mexico by Dr. Gastón Madrid, who reported a case in 1945.¹ The patient was referred due to the possible diagnosis of tuberculosis (TB) to the 80-bed TB sanatorium in Hermosillo, the capital of Sonora, a state that shares the United States (U.S.)–Mexico border with Arizona. Bacteriology for TB in this patient was negative despite the clinical and radiological features suggestive of TB. A fungal culture was ordered and Coccidioides spp. was isolated from the sputum. Three years later, Madrid reported a series of coccidioidomycosis cases diagnosed in the region.² A few years later^3,Coccidioides spp. was isolated for the first time in Mexico, in the soil from a snake burrow in the outskirts of Hermosillo. In 1974, Madrid publishes the book Coccidioidomycosis, the only monograph on CM published to date in Mexico.⁴

Infection by Coccidioides sp. is as prevalent in Mexico as in the endemic regions of the United States.⁵ The first skin surveys with coccidioidin were carried out in Sonora from the late 1940s to the early 1970s, at different sites in the state. Rates of infection were high all over the state and ranged from 64% to 90% in adults; in children, the rate was reported at 16%.⁵ In 1966 González-Ochoa reported the first national survey on coccidioidomycosis infection; rates were significantly higher in the northwestern states (Baja California, Sonora, Sinaloa, and Chihuahua) of the country with rates that ranged from 30% to more than 50%.⁶ These high rates of infection have been corroborated in later surveys with coccidioidin and spherulin.^7,8

The endemic regions in Mexico are characterized by a dry climate, alkaline soil, summers with very high temperatures (up to 50°C), and annual precipitation indexes as low as 10 cm; this combination of environmental factors facilitates the spread of Coccidioides spores in the air. The high prevalence of infection in the semi-arid northwestern region of Mexico correlates well with the evidence from the United States, where the highest rates of disease are found also in California and Arizona, states with large desert regions that are likely to serve as ecological niches for Coccidioides.⁹

The prevalent species in Mexico is Coccidioides posadasii;¹⁰ the only isolates of Coccidioides immitis have been found in Baja California, a state located in the most northwestern region of the country; no geographical overlap of the species has been reported.¹¹ The clinical burden of disease in Mexico is currently unknown.¹² CM was a reportable disease in Mexico up to 1994, with an average of 1500 cases reported per year; since 1995 there are no data on the clinical burden of CM, when the disease was eliminated from the national epidemiological registry for reportable diseases. From 1988 through 1994 the mean incidence of CM in Mexico was 0.8 per 10.⁵ The states (all of them northern states, located in the U.S.–Mexico border) with the highest rates of clinical cases were Nuevo León, Tamaulipas, Chihuahua, Baja California, and Sonora.⁹

An important barrier for the clinical diagnosis in Mexico is the lack of access to laboratory mycological diagnosis. There are just a few centers in the country (some of them dedicated to basic research more than to clinical diagnosis) with the capability and biosafety standards for culture and identification of Coccidioides and/or serological diagnosis. In most regions, clinical diagnosis is based on direct sputum examination and histopathology results from biopsies or autopsies. This would explain why primary coccidioidomycosis is rarely diagnosed, and most cases published are about chronic pulmonary or disseminated CM.

The literature on the interrelationship of diabetes mellitus and coccidioidomycosis suggest that these patients may experience severe, progressive, and complicated coccidioidal infections.¹³ Diabetes is one of the main causes of death in Mexico; the prevalence of diabetes in the country, according to the latest national survey is 9.4%;¹⁴ severe forms of CM have been reported in diabetics in Mexico for more than 30 years.¹⁵ Tuberculosis and CM share the same risks factors and its coexistence is also a common clinical finding in Mexico.¹⁶ The coexistence of CM and human immunodeficiency virus (HIV) has also been reported in Mexico,¹⁷ although not with the frequency that should be expected based on the prevalence of HIV in the country.

Travelers from nonendemic areas should be aware of the potential risk of infection when visiting endemic areas in Mexico. An example of this risk would be an outbreak that included 21 serologically confirmed CM cases (17% attack rate) among a church group from Washington State in the United States. The group had just stayed for 6 days at an orphanage in Tecate, Mexico, a town in the Sonoran Desert adjacent to the U.S.–Mexico border to build a church.¹⁸

A group of clinicians and microbiologist from Baja California, Sonora, and Nuevo León have founded the Mexican Coccidioidomycosis Study Group in 2014. The first objective of the group was to develop the Mexican Guidelines for the Diagnosis and Treatment of Coccidioidomycosis; they were published in the journal of the National Infectious Diseases Society in 2015.¹⁹ Funding has been an issue for the regular meeting of the group.

Coccidioidomycosis Research in México

Coccidioidomycosis in humans (Table 1)

The first molecular study on CM in Mexico was reported in 2004; it dealt with the development of a conventional nested and a real-time polymerase chain reaction (PCR) assay specific for the identification of C. posadasii DNA by targeting a gene coding for the antigen 2/proline-rich antigen (Ag2/PRA). The study included 120 clinical strains isolated from 114 patients and three formalin-fixed paraffin embedded specimens during a 10-year period (1991–2002) in Monterrey, Mexico. The samples were amplified using two microsatellite loci (GAC and 621);²⁰ all strains were correctly identified as C. posadasii, whereas DNA from related members of the family Onygenaceae tested negative. Also, specific DNA was amplified by conventional nested PCR from three microscopically spherule-positive formalin-fixed tissue samples and identified as C. posadasii, whereas 20 human tissue samples tested positive for other dimorphic fungi tested negative.²¹ This assay is used directly in clinical samples,^22,23 as well in clinical isolates and formalin-fixed tissues.^24–28

Table 1.

Molecular studies of Coccidioides spp. performed in clinical, environmental and animal samples from México.

Year of publication	Molecular marker	Samples Number/Type	Year of samples collection	Geographic origin samples	Specie identification	Citation
Human
2004	Ag2/PRA	120 clinical strains	1991–2002	Nuevo León	C. posadasii (100%)	[21]
	2 Microsatellites	3 FFPE
2007	SNPs (proline 157, proline 174, hexokinase 149 and glucose-synthase 192	46 clinical isolates	No data	Baja California, CDMX, Coahuila, Nuevo León	Coccidioides spp.	[11]
2012	Ag2/PRA U region	54 FFPE tissue	1982–2010	Baja California	C. posadasii (83%) C. immitis (17%)	[26]
2013	Ag2/PRA AFLP	32 clinical isolates	No data	Baja California, Campeche, CDMX, Durango, Nuevo León, Torreón.	C. posadasii (100%)	[25]
2014	Ag2/PRA 8 Microsatellites U region	154 clinical isolates	1957–2010	Baja California, Chihuahua, Coahuila, CDMX, Durango, Edo. De México, Guerrero, Michoacán, Nuevo León, San Luis Potosí, Sinaloa, Sonora, Tamaulipas, Zacatecas.	C. posadasii (82%) C. immitis (18%)	[24]
2015	U region	3 clinical isolates	2015–2010	Baja California	C. posadasii (100%)	[31]
2016	U region	1 clinical isolate	2016	Baja California	C. posadasii	[33]
2017	Ag2/PRA	1 clinical isolate	2017	Mexico City	C. posadasii	[22]
	RAPD
Environmental isolates
2012	ITS2 (nested PCR)	32 soil samples	2006–2010	Baja California Valle de las Palmas Zorra (SJZ)	Coccidioides spp.	[37]
2013	ITS2 (nested PCR)	5 soil samples	2009	Baja California (Valle de las Palmas)	Coccidioides spp.	[36]
2015	IT2 (nested PCR)	11 soil samples	2012–2013	Baja California (Valle de las Palmas)	C. posadasii (82%)	[38]
		9 Top soil			C. immits (18%)
		2 Burrows
2014	Ag2/PRA 8 Microsatellites U región	4 isolated culture from soil samples	No data	Coahuila	C. posadasii (100%)	[24]
2014	ITS2 (nested PCR)	10 soil	2011	Baja California (Valle de las Palmas)	Coccidioides spp.	[39]
Animal samples
2014	Ag2/PRA 8 Microsatellites U región	1 isolated culture from dog	1957	Sinaloa	C. posadasii	[24]
2016	U region	3 FFPE tissue from dogs	2010–2015	Monterrey	C. immitis	[40]
2017	Ag2/PRA	1 FFPE tissue from cattle	2014	Baja California	C. immitis	Muñiz-Salazar (2017) unpublished

FFPE, formalin-fixed paraffin-embedded; PCR, polymerase chain reaction; SNP, single-nucleotide polymorphism.

In 2007, Castañon-Olivares analyzed 56 clinical isolates of Coccidioides spp. from Mexican patients by real-time PCR using TaqMan probes to amplify single nucleotide polymorphisms (SNPs) in four target sequences loci (proline 157, proline 174, hexokinase 149, and glucose-synthase 192). In total, 54 isolates identified as C. posadasii and two as C. immitis. Only proline 157, proline 174, and glucose-synthase 192 gave consistent results on SNP differentiation between the two species, while hexokinase 149 gave negative results in 34 samples. Clinical isolates were from sputum, cerebrospinal fluid, bronchoalveolar lavage, pus or tissues, and the tip of a ventriculoperitoneal shunt catheter and patients clinically diagnosed with coccidioidomycosis in Baja California, Coahuila, Nuevo León, and Mexico City.

González-Becuar in 2012,²⁶ analyzed a total of 129 formalin-fixed tissue samples with histopathology diagnosis of CM. The samples were obtained from the collection of the pathology laboratory from the Tijuana General Hospital from 1982 through 2010. In total, 54 (83%) samples identified as C. posadasii, and 11 (17%) as C. immitis, which distributed in five haplotypes for C. posadasii and three for C. immitis.

In 2013, Duarte-Escalante²⁵ analyzed 32 clinical isolates from Mexico (Mexico City, Durango, Torreon, Baja California, Nuevo Leon, and Campeche) and Argentina. All isolates were identified as C. posadasii. Also, the study reported a high genetic variability in the isolates from México (0.5011 ± 0.0382) and Argentina (0.3951 ± 0.0503). A small but significant genetic differentiation between countries (P = .0001) was detected, suggesting that there is a single genetic population in Latin America. This is the first study in Mexico, which explores the genetic structure of Coccidioides using the molecular marker amplified fragment length polymorphism (AFLP).

In 2014, Luna-Isaac et al.²⁴ used nine-microsatellites, Ag2/PRA and Umeyama region to establish the predominant Coccidioides species in Mexico, to delineate the current geographical locations of both species, and to identify a possible correlation between clinical symptoms and a specific genotype. One hundred sixty isolates (155 clinical, four environmental, and one animal) recovered from culture collections of different institutions in Mexico stored from 1957 and 2010. The predominant species was C. posadasii (82%). Among 28 C. immitis strains, 14 unique multilocus genotypes and six clones were observed. While 126 C. posadasii strains showed 29 unique multilocus genotypes and 16 clones. Finally, based solely on the size scores of microsatellite locus 621, which was the most commonly used to discriminate between species, 28 samples identified as C. immitis (416–426 bp) and 126 as C. posadasii (399 bp). Interestingly, samples 5233, 5234, M60_09, M61_09, and TJ3835 showed sequences and fragment sizes that corresponded to C. posadasii using U region that identified as C. immitis based on the Ag2/PRA and 621 loci. This situation has been observed by Tintelnot et al.²⁹ who identified one strain of Coccidioides as one species using restriction fragment length polymorphism (RFLP) and as another using U region. That strain was previously identified as C. immitis and postulated to be a hybrid genotype of both species. Neafsey et al.³⁰ suggest that sexual reproduction between the two species could occur, causing the hybridization and introgression of genomic fragments from one species into the genome of another. On the other hand, Luna-Isaac et al.²⁴ reported the correlation between genotype and clinical presentation was not significant (P > .05). However, 83% of C. immitis isolates were associated with cutaneous dissemination, suggesting this species exhibits a higher tendency to disseminate to the skin, while C. posadasii was related to all forms of clinical presentation. It is important to highlight that this is the only study on the genetic structure of Coccidioides in Mexico.

Recently, clinical cases have used molecular markers to identify species of Coccidioides. In 2015, Moreno-Coutiño³¹ reported six Mexican male cases, residents of Tijuana Baja California—three of them with primary pulmonary infection and further cutaneous dissemination, and three cases of primary cutaneous coccidioidomycosis. Half of the cases were identified as C. posadasii using U region.³² In 2016, Arce et al.³³ reported a primary cutaneous coccidioidomycosis caused by C. posadasii in a male patient 41 years of age, a native and resident of Tijuana, Baja California, who for 27 years worked as a surveyor in this region. Recently, Fernández²² reported the case of a Mexican HIV+ patient with a probable endogenous reactivation of coccidioidomycosis. A direct microscopic examination using potassium hydroxide (KOH) on a sample taken from the cheek nodule revealed the presence of spherules. The fungus was isolated, and its identity was confirmed by phenotypic and molecular methods (Ag2/PRA).

Coccidioidomycosis in environmental samples (Table 1)

The screening of environmental samples has had low effectiveness, mainly because of the weak characterization of Coccidioides ecological niche. Coccidioides spp. has been identified from airborne and soil samples by culture and molecular assays. The first environmental report of Coccidioides was by Laniado-Laborin et al.³⁴ who isolated Coccidioides from one air sample in Tijuana. The most studied site in Mexico is Valle de Las Palmas in Baja California. The interest to study this site is because a CM outbreak occurred in 1996 when a Washington State church group traveled to VDP to build an orphanage.³⁵ As mentioned, after they returned to the United States, 17% (21/126) serological cases of coccidioidomycosis were confirmed.

A total of 108 soil samples from Baja California has been analyzed, mainly from Valle de las Palmas (101 samples); 54% of the isolated strains are identified as Coccidioides spp. amplifying the internal transcribed spacer (ITS) region. Most of the soil samples were from heteromyid latrines, and in a lower frequency from topsoil, heteromyids’ active burrows and large mammal dens. No Coccidioides spp. has been found in soils collected in Ensenada, Baja California. Luna-Isaac²⁴ reported the molecular identification of four soil isolates as C. posadasii (M47–06, M49–06, M50–06, and M51–06), sheltered in the culture collection of the Universidad Nacional Autónoma de México (UNAM), but the collection date is unknown.

Romero-Olivares³⁶ constructed a Dikarya-specific gene library for the ITS region of nuclear ribosomal DNA in order to determine the diversity of the mycobiota in Baja California. Although this assay was able to identify species like Penicillium dipodomyicola, Coprinellus radians, Alternaria spp., among others, it was not able to identify Coccidiodes. Thus, a nested PCR was used to determine this species.³⁷ The most recent work analyzes the ITS region by a metagenomic approach in a 454 Roche pyrosequencer.³⁸ The objective was to identify environmental factors determining fungal community structure in two different microhabitats, burrows (influenced by rodent activity), and topsoil and were compared in winter and summer. Differences in composition between microhabitats were mainly correlated to significant differences in environmental factors, such as moisture and clay content in topsoil samples, and temperature and electrical conductivity in burrow samples. Overall, the fungal community structure (dominated by Ascomycota and Basidiomycota) was less variable between seasons in burrow than in topsoil samples. Similarly, 13 Coccidioides spp. went undetected by pyrosequencing. However, a nested PCR approach revealed its higher prevalence in burrows.

The last study reported until now in Valle de las Palmas was performed during 2010–2011by Catalan-Dibene.³⁹ DNA was extracted directly from 24 soil samples to amplify the ITS2 region of rDNA.³⁷ A 170 bp amplicon was recovered from 15 out of 24 samples; however, just 10 samples were confirmed as Coccidioides spp. The rest of the sequences corresponded to members of the genus Aphanoascus (A. Canadensis and A. pinarensis) and Penicillium.

Coccidioides in animals (Table 1)

The available published studies on animal CM in Mexico are scarce; heteromyids, prairie dogs, dogs, and cattle are the only species reported. Consequently, the actual prevalence of CM infection in domestic and wild mammals in Mexico is unknown.

In 2014, Catalan-Dibene³⁹ detected antibodies against Coccidioides in two species of heteromyids (Peromyscus maniculatus and Neotoma lepida) sampled in Valle de las Palmas, Baja California.

Ramirez-Romero⁴⁰ analyzed 765 dog biopsies with a presumptive diagnosis of neoplasm between April 1, 2010, and March 31, 2015, from Monterrey, Nuevo Leon. In total, three cases of CM were found (3/765 = 0.39%). The presumptive diagnoses in these cases were osteosarcoma, lymphoma, and neurofibroma, respectively. Molecular analysis of formalin-fixed paraffin-embedded (FFPE) positive samples amplifying the U region³² showed C. posadasii in one of these cases.

Sera from urban dogs and wild rodents (prairie dogs) were tested employing the Ouchterlony double immunodiffusion technique, with coccidioidin produced at UNAM as an antigen. One hundred domestic dog sera were tested, and three (3%) tested positive for anti-Coccidioides antibodies; of 158 sera from prairie dogs, 113 (71.5%) tested positive.

Only a few studies of CM in cattle have been reported in Mexico. Six cases of CM in beef (3) and dairy (3) cattle slaughtered and inspected in 1995 and 1996 were reported in Mexicali, Baja California. The identification was performed just by histopathological analysis. After that, there are no reports in this area until 2017, when a post-mortem examination in 1 out of 319 cattle slaughtered at Mexicali, using histopathological and molecular (Ag2/PRA) testing identified C. immitis (Muñiz-Salazar unpublished data).

Coccidioidomycosis and tuberculosis: Differential diagnosis

CM and tuberculosis (TB) share very similar pathogenic mechanisms: they usually start with a primary lung infection that may lead to a chronic condition in which they may present the same symptoms and radiologic manifestations.^41–45 In endemic regions for both diseases, the clinicians may face quite a challenge when considering these two infections in the differential diagnosis. Support from specialized TB and mycology laboratories is needed but, even when this is available, confirmation of either (or both) infection can be difficult. In the following section, we present a clinical approach for the differential diagnosis of tuberculosis and/or coccidioidomycosis in Monterrey, Mexico.

Diagnostic work-up for TB and CM

Monterrey is located close to the United States–Mexican border, an endemic region for coccidioidomycosis;⁴³ also it is one of the cities with the highest incidence of TB in Mexico.⁴⁶ Clinical guidelines and recommendations to diagnose CM^47,48 or TB^44,49 are readily available; in individual cases, the diagnosis of either disease can be straightforward in some cases but in others, it can be extremely complicated. Due to this difficulty, a number of diagnostic modalities may be needed. Cultures are still considered as the gold standard tests for both infections, but their sensitivity is far from perfect, creating the need for complementary methods to support the diagnosis.^44,47 Diagnosis of TB is based on the initial finding of acid-fast bacilli (AFB) either in sputum or bronchoalveolar lavage (BAL) stains and the posterior isolation of Mycobacterium tuberculosis in culture; in the past decade, molecular methods have improved the rapid diagnosis of TB, and nowadays they are widely used and recommended as the preferred initial tests.⁴⁴ On the other hand, diagnosis of CM based on sputum microscopy and culture is much less sensitive. BAL stains and culture also lack sensitivity but bronchoscopy is usually indicated in severe cases when a specific diagnosis is urgently needed.⁴⁷ Serodiagnosis is not recommended for TB⁴⁴ but can be very useful for the diagnosis of CM.^47,49 Several tests are commercially available for clinical laboratories but others need to be sent to reference laboratories. Serology is probably the most commonly ordered diagnostic test, but their results must be interpreted judiciously since misinterpretation can lead to false positive or negative diagnosis; a positive serology test by itself does not make the CM diagnosis and a negative one does not rule out the disease, mainly in the most seriously ill patients.⁴⁷ Skin testing, when available, can also be used to detect either infection, but they are not useful for the diagnosis of the disease.^44,47 Interferon-gamma release assays (IGRAs) for the detection of TB infection, are basically interpreted in a similar way than the tuberculin test but with a higher sensitivity and specificity.⁴⁴ However, as mentioned, skin tests and IGRA tests do not distinguish active disease from latent infection.

Frequently, a combination of clinical data, radiologic presentation, and results of diagnostic tests should be used in order to make a specific diagnosis.^44,47 Guideline algorithms for the diagnosis of TB or CM are designed to search for a specific diagnosis, either TB or CM, but they do not consider into their algorithms the possibility of the other infection.^44,47–49 Complicating matters even more, the two infections may and do coexist.⁵⁰ Even though we try to be aligned with international recommendations, we must adapt them to our local needs and resources. A modified algorithm for the differential diagnosis when facing a patient in whom TB or cocci are possible is shown in Figure 1.

Figure 1.

Flow chart for the usual approach used at the University Hospital of Monterrey in patients whose differential diagnosis includes chronic Coccidioidomycosis or TB. (Tb: tuberculosis; CM: coccidioidomycosis; broncho: bronchoscopy)

The approach suggests to first rule out TB since this infection is more common and is contagious. We begin the diagnostic work-up with the sputum smear for AFB. If AFB is positive, we start treatment for TB, while we wait for the diagnosis to be confirmed by the cultures. If AFB is negative, a molecular test is performed: if positive, TB is diagnosed, if negative, a bronchoscopy is performed. TheBAL is simultaneously tested for TB and cocci with stains, cultures, and molecular TB tests. Any positive test is considered a definite diagnosis. Transbronchial biopsies are performed mainly on those with interstitial infiltrates or on those with no cavities. For those with stains and negative molecular tests, we order skin test for both diseases. Negative results may help to rule out any of the two infections. Positive results are carefully considered since they do not confirm active disease. In some cases, Cocci serology is also sent to a reference laboratory. For patients in which results are all negative and cultures are still being processed, we may prescribe empirical therapy if the clinical conditions deserve it. Empirical therapy is chosen analyzing all the available data for each case on an individual basis. In most cases, the empirical treatment prescribed is for presumed TB. According to the clinical response and the cultures results, treatment is further adjusted. For those patients with HIV infection, diabetes or on any kind of immunosuppressive therapy, an earlier more aggressive diagnostic approach is used including invasive procedures to obtain tissues for histopathology study and cultures. Thereare no data to support if our approach is better or worse compared to the one used in other centers with different epidemiology and different resources.

Coccidioidomycosis in Central America

Coccidioidomycosis is an endemic fungal infection found in certain arid and semi-arid regions of Central America.⁵¹ Reports of its presence in the region are scarce, and documented reports of autochthonous human CM are found in only two countries, Guatemala and Honduras.

Two early attempts to investigate the presence of disease in Central America using coccidioidin skin test surveys exist. The first one by Andrade in 1945,⁵² included 1200 persons, children, and adults in Guatemala City. A second study in 1950 included 500 persons in the Canal Zone of Panama.⁵³ The reactivity to the coccidioidin in both studies was less than 1% of the subjects tested.

A few years later in 1951, the first human case of CM in Central America,⁵⁴ was documented, by two pathologists from Costa Rica who received lymph node tissue belonging to a 53-year-old truck driver born in Nacaome, Honduras, and who lived and worked for the last 24 years in the Comayagua Valley. In 1946, he developed a cough, chest pain, hemoptysis, night sweats, and weight loss. In 1950 he traveled to Costa Rica, developing at that time bilateral supraclavicular lymphadenopathy, with a bloody purulent discharge. Spherules were documented in the purulent secretion and in sputum. A chest X-ray suggested mediastinal lymphadenopathy. Tuberculin test and CM skin tests were reactive. He received intramuscular colloidal copper, with little improvement.

It should be mentioned that in 1959, a dog from Nicaragua was diagnosed with coccidioidomycosis in Norway, after its arrival,⁵⁵ raising the possibility of human coccidioidomycosis there.

Nine years after the first human case from Honduras was found, two veterinarians reported several infected animals with coccidioidomycosis in Guatemala. Two bulls⁵⁶ and seven dogs,⁵⁷ none of them came from the endemic areas documented later. Also, during the same year (1960), the first human case in Guatemala was documented by Eduardo Pérez-Guisasola,⁵⁸ and the clinical findings were described by García Valdés.⁵⁹

The patient was a 16-year-old black male, who lived in Gualán Zacapa, part of the Motagua Valley. His chest X-ray showed pulmonary infiltrates and skin and bone dissemination. Pérez-Guisasola, during the next 6 years, found and documented six more persons with the disease.⁶⁰ Based on the climatic characteristics of Guatemala, and the region where the patients lived, Pérez-Guisasola proposed the Motagua Valley, an arid region in the western part of the country, as the potential endemic area. He carefully described the clinical, microbiologic and epidemiologic origin of the infected persons. Patient number one, was 34-year-old 9-month pregnant female, from Guatemala City (outside the endemic zone), with disseminated disease. Patient number two was a 64-year-old diabetic female from Salamá (near to an endemic zone) with disseminated disease. Patient number three was a 19-year-old military recruit, stationed in Teculután (Motagua Valley) with primary coccidioidomycosis that progressed to chronic pulmonary disease. Patient number four was a 64-year-old male carpenter from Quetzaltenango, outside the endemic zone, who traveled to the endemic zone, and had disseminated disease. Patient number five was a 20-month old male with disseminated disease, living in Río Hondo in the endemic zone. Finally, patient number six was a 24-year-old Mayan male from Chicojom, Baja Verapaz (near to an endemic area) with disseminated disease.

In 1967, Ruben Mayorga, based on the previous reports, decided to perform a skin test survey of the Motagua Valley^61–63 and other areas in search of reactive responders. He included 7725 school children; 9734 subjects were from Guatemala and 991 from Nicaragua. He classified his results in five study areas, designated from A to E; the first four areas were located in Guatemala and the last area, E, was located in Nicaragua. Area A, with the highest concentration of reactive results (42.4%), corresponded to the endemic area of the Motagua river. Areas B to D were in the surrounding areas, outside the Motagua Valley, with less than 4% reactivity.

Phylogenetic analysis of C. posadasii isolates from Central America

Recently,⁶⁴ five genomes from Guatemala were analyzed together with 81 genomes from other regions (26 C. immitis and 60 C. posadasii). Of the C. posadasii isolates, three major clades were clearly separated and well supported by the bootstrap analysis: one clade included all isolates from Arizona; another included four Guatemala isolates and the third clade, contained nearly all other non-Arizona isolates. There was a fifth Guatemalan isolate included in the non-Arizona clade. This isolate was included in the Texas-Mexico-South America clade and came from an HIV-infected migrant worker, who abandoned his antiretroviral treatment and probably was infected while crossing the border in Texas. He died from disseminated coccidioidomycosis, shortly after he returned to Guatemala. This initial work demonstrates that C. posadasii is found in Central America and is part of a different clade. Phylogenetically it is the youngest clade, probably originated from the Arizona clade.

There are approximately six new cases of coccidioidomycosis documented yearly in Guatemala, mostly in HIV-infected individuals. In Honduras there are only sporadic cases reported. Based in the number of persons reactive to coccidioidin found in the endemic areas, these numbers should be higher, but due to the lack of awareness and mycological skills in the region, a significant number of persons with CM probably are not diagnosed.

Epidemiology of coccidioidomycosis in Argentina, South America

Although this mycosis was described initially in Buenos Aires, Argentina in 1892,^65–67 the number of CM cases reported in South America is much lower than those diagnosed in North America, and the real incidence is unknown perhaps because it is not a reportable disease, because it may be misdiagnosed as tuberculosis, which is very endemic, and also because there is a high percentage of subclinical cases. However, available data suggest that its prevalence is increasing in some areas of South America.^68–70

As in North America, the geographical endemic areas for CM in South America are characterized by arid and semi-arid areas with sandy, alkaline soils and extreme temperatures in Colombia, Venezuela, Brazil, Paraguay, Bolivia, and Argentina.

In Colombia, five cases were reported from 1958 to 2015 and low infection rate was detected using coccidioidin skin tests (3–13%).⁷¹ Venezuela reported 114 cases from 1948 to 2004 and an infection rate of about 50% in major endemic areas of this country.⁷²

The area that has most recently been described as endemic for CM is the northeast of Brazil, where the first autochthonous case was reported in 1978 in a patient from Bahia.⁷³ Epidemiological studies show autochthonous cases in Ceará, Bahia, Piauí, and Maranhão.⁷⁰ In 2016 the Brazilian Ministry of Health reported 829 hospitalizations due to CM (incidence 7.12/1000 hospital admissions).⁶⁹ In the state of Ceara in Brazil, a coccidioidin skin survey showed a prevalence of infection of 26.4%.⁷⁴

In the southernmost areas of South America coccidioidomycosis is endemic in the west of Paraguay, southeast of Bolivia, and in the arid geographic areas from the province of Jujuy to the province of Río Negro in Argentina.^68,75,76

Both Paraguay and Bolivia report only occasional cases; however in the northwest of Paraguay, the infection rate has been reported as high as 44% in aboriginal population and oil company employees.⁷⁵ Table 2 summarizes the epidemiological studies performed in South America using the coccidioidin skin test.

Table 2.

Epidemiological studies in Coccidioidomycosis performed in South America.

Country	Period	Cases reported (n)	Infection rate* (%)
Colombia	1958–2015	5	3–13
Venezuela	1948–2014	114	11–46
Brazil	1978–2015	829	26
Paraguay	1950–1967	3	15–44
Bolivia	1948	1	…
Argentina	1892–2015		2–40

*

Infection rate in each country varies according to the geographical area of the country.

In Argentina, epidemiological studies revealed highest infection rates in Catamarca province, with values about 40% of reactors. Other areas with high infection rates were described in the northwest of Córdoba (34%); the west of Santiago del Estero (19.8%); and La Rioja (19.13%); all of these provinces are adjacent to Catamarca.

Only a small number of CM cases had been reported in Argentina before the year 2000, despite the large extent of the endemic area. In 2009 the first retrospective review of all CM cases reported in the country documented 128 cases from the original case reported by Posadas in 1892 to the year 2009. From 1892 through 1939 only six cases were registered; between 1940 and 1999, 59 more cases were reported, and finally, the remaining 63 cases (49% of total cases) were reported from 2000 to 2009, with the Catamarca province having the highest number of cases.⁶⁸ The National Mycology Network reported 48 new CM human cases from 2010 through 2015. Almost 50% of these cases occurred in patients who lived or visited a geographic region in the Central Valley of Catamarca. Annual disease incidence rates in this province increased from historical values below 0.5/100 000 to 2/100 000 inhabitants from 2000 to 2009;⁶⁸ incidence from 2010 to 2015 is reported between 1 and 2 cases/100 000 inhabitants.

Probably the dramatic changes in the ecosystem that occurred between the years of 1973 and 2007 in the central valley of Catamarca (an increase of 4 times the urbanized areas and 38 times the cultivated areas) was a factor in the increase of CM in the region.⁷⁷ Coccidioides has been isolated from the soil in Argentina (Province of Cordoba).⁷⁸

In addition, increased diagnosis capability could be a factor in this incidence increase because the Reference National Mycology Laboratory of Argentina has provided reagents for CM immunodiffusion tests to the laboratories of the laboratory network since the year 2000.⁷⁹

With respect to zoonotic CM, data collected by the Reference National Mycology laboratory registered only three cases before to 2005 in dogs from Catamarca. In 2005, an increase in the number of CM cases in dogs was reported in this province. Animals showed fever, cough, and lameness and osteomyelitis lesions. Up to 2014, the total number of canine CM cases in Catamarca Province was 86. All cases were confirmed by detecting anti-Coccidioides antibodies by immunodiffusion test with titer >1:4 plus clinical and radiological findings. Figure 2 shows the evolution of reported dog and human cases by year from 1997 to 2014 in Catamarca Province.

Figure 2.

Evolution of  reported dog and human cases from 1997 to 2014 in Catamarca Province, Argentina*. (*Data obtained from the Reference National Laboratory for Mycology of Argentina.)

Coccidioides posadasii is the most prevalent etiological agent of CM in South America^68,70; there have been two reports of C. immitis cases in South American patients, one from Venezuela⁸⁰ and the other by Argentina,⁸¹ but in these cases, it was impossible to exclude travel by the patients to regions where C. immitis is prevalent. In 2015, Coccidioides immitis DNA has been detected in a specimen from a patient from Pinto, Colombia (the patient never traveled abroad). Although this species is primarily found in California, it has been reported in other geographical areas as shown by this case from Colombia; there is the possibility that C. immitis endemic areas may be extending from the formerly proposed geographical borders.⁷¹

Coccidioidomycosis is a highly prevalent systemic mycosis in Latin America and has been reported in México, Guatemala, Honduras, Colombia, Venezuela, Brazil, Paraguay, Bolivia, and Argentina. However, despite the reporting of clinical cases in several countries of the region, clinical and epidemiological information is scarce and fragmentary. The incidence of CM in the region is unknown, mainly due to lack of clinical awareness and limited access to a laboratory. Clinical training in this condition and the strengthening of the network of mycology laboratories in the region is necessary to adequately characterize this mycosis in the region.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.