Occurrence of coccidioidomycosis in a tertiary hospital in the Brazilian semi-arid region: diagnostic challenges

Abstract

Coccidioidomycosis is a systemic fungal infection that primarily affects the lungs in mammals. It is endemic to certain areas of the Americas. In Brazil, the disease occurs exclusively in the Northeast of the country, and the only etiological agent described is Coccidioides posadasii, typically found in the soil of arid regions with low rainfall. Given the scarcity of cases in Brazil, a descriptive study was conducted based on cases diagnosed with coccidioidomycosis between 2012 and 2022 at a tertiary hospital in the northeastern region of Brazil. The study analyzed medical records to identify cases of coccidioidomycosis diagnosed by clinical and/or laboratory data, discussing their clinical-epidemiological aspects. Thirteen patients were identified from seven municipalities in Ceará, all of whom were male, with a mean age of 19 years. The main risk factor was armadillo hunting, and the clinical manifestations included cough, fever, and chest pain. In this study, two patients were diagnosed by serology, one by microscopy, and ten suspected cases were diagnosed based on the clinical and epidemiological picture. Twelve patients were treated with antifungals. Clinical improvement with hospital discharge occurred in twelve patients, while one patient died. The data corroborate information from the scientific literature that the area covered by the referral hospital is endemic for coccidioidomycosis. However, local resources for diagnosing this disease were found to be deficient, with a high rate of presumptive diagnoses. This study highlights the need for actions by health surveillance systems and greater efficiency in diagnosing coccidioidomycosis in endemic areas of Brazil.

Introduction

The fungal species Coccidioides immitis and Coccidioides posadasii are endemic to the Americas, occurring in the southwestern US, northern Mexico, and focal areas of Central and South America [1]. In the United States, which is the main area of coccidioidomycosis (CM) occurrence in the world, the Centers for Disease Control and Prevention (CDC) reported 20,003 cases of the disease in 2019, with approximately 200 deaths per year between 1999 and 2019 [2].

In Brazil, where only the species Coccidioides posadasii is present, the first cases of CM were described in 1978 and 1979 [3, 4]. From then until 2022, 136 cases have been reported in the literature, all in the Northeast region of Brazil, an area with a semi-arid climate [5]. More recently, a study described 292 Brazilian patients with CM in the states of Piauí and Maranhão [6], suggesting that the number of cases in the country could be even higher than those recorded. Furthermore, there is no reliable official data because, in Brazil, the notification of CM cases is not mandatory [5].

CM is a systemic mycosis that primarily affects the lungs of humans and some mammals. It can affect even immunocompetent individuals, showing a wide range of clinical manifestations, from asymptomatic cases to severe illness, including fatal pulmonary disease or disseminated disease [6]. Most individuals infected by Coccidioides spp. (about 60%) do not show symptoms or develop only mild symptoms (“influenza-like”) and thus do not require medical attention. On the other side of the clinical spectrum, severe or disseminated CM occurs in about 0.5–2% of cases. Therefore, CM can present with different clinical manifestations, and its management requires an individual evaluation of the disease’s extent and the host’s clinical condition [1, 7].

In Brazil, CM occurs in areas of known socioeconomic vulnerability, and patients often cannot find proper diagnosis and treatment in their cities or nearby locations, forcing them to make long journeys to specialized centers in capitals, often hundreds of kilometers away [5]. Almost all cases described in large series in Brazil were diagnosed in referral centers located in Fortaleza and Teresina, the respective capitals of the states of Ceará and Piauí, even though the patients were always from small cities in the countryside [6, 8].

To enhance the understanding of CM epidemiology in Brazil, specifically in the state of Ceará, this study aims to provide valuable insights. It describes a series of novel CM cases that occurred over a decade in a tertiary referral hospital located in the Cariri meso-region of southern Ceará, highlighting the diagnostic methods available in non-referral center services.

Methods

This is a descriptive, retrospective, observational study consisting of a series of cases based on the analysis of hospital data from the medical records of patients admitted to the Regional Cariri Hospital (HRC) from January 1, 2012, to January 1, 2022. The HRC is located in the city of Juazeiro do Norte, approximately 500 km from Fortaleza, the capital of the State of Ceará. It is a tertiary public hospital with 273 beds, comprising 174 infirmary beds, 49 emergency beds, and 50 intensive care unit beds. This hospital serves the populations of 42 municipalities in the health macro-region of Cariri (South and Central-South regions of Ceará), acting as a medical-hospital referral hub for these locations [9].

The municipality of Juazeiro do Norte (7° 12′ 47″ S, 39° 18′ 55″ W) has a tropical, warm, semi-arid climate, with annual rainfall up to 925 mm, distributed unevenly from January to May. The annual mean temperature ranges from 24 °C to 26 °C [10].

All patients admitted to the HRC during the specified period with a CM diagnosis in their medical records were included in the study. A search of electronic medical records identified patients with a CM diagnosis (ICD-10 codes: B38 - Coccidioidomycosis; B38.0– Acute pulmonary coccidioidomycosis; B38.1– Chronic pulmonary coccidioidomycosis; B38.2– Pulmonary coccidioidomycosis, unspecified; B38.3 - Cutaneous coccidioidomycosis; B38.4 - Coccidioidomycosis meningitis; B38.7 - Disseminated coccidioidomycosis; B38.8 - Other forms of coccidioidomycosis; B38.9 - Coccidioidomycosis, unspecified) recorded as the primary or secondary cause of hospital admission and/or discharge.

Medical files containing pulmonology specialist reports were also accessed, which included assessments by specialist doctors required to assist in the diagnostic elucidation of admitted patient cases. Cases where CM was included as a diagnostic hypothesis in the specialist’s assessment were also included in this case series.

All patients with a CM diagnosis were included based on the following criteria of consensus on invasive fungal diseases from the European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium [11]:

• Confirmed diagnosis– Direct visualization of Coccidioides spp. in biological specimens or fungal growth in cultures of biological samples.

• Probable diagnosis– Environmental exposure risk, a compatible clinical picture, and the presence of serum antibodies to Coccidioides spp. or in cerebrospinal fluid.

• Possible diagnosis– Environmental exposure risk and a compatible clinical picture (pneumonia syndrome, including cough, fever, and dyspnea, without clinical improvement with antibiotics used for community-acquired pneumonia (CAP) and with a radiological pattern of pulmonary nodules and/or diffuse cavitated pulmonary nodules).

Cases initially suspected of CM but subsequently confirmed as other definitive diagnoses of different diseases by specific complementary examinations were excluded.

Patient data were obtained directly from medical records and compiled into a standardized form designed by the researchers to collect clinical and epidemiological data. Descriptive statistical analysis was performed using Microsoft Excel and R Studio, version 4.3.0 (2023-04-21).

As the study involved human subjects, it fully complied with Resolution No. 466/2012 of the National Health Council. It was conducted only after receiving approval from the Research Ethics Committee, as documented on the Brazil Platform under numbers 6.061.779 and 5.949.460.

Results

After screening for the ICD-10 code B38 (Coccidioidomycosis), 7 patients were identified. Upon reviewing the pneumology consultation records, 8 additional patients were found where CM was considered the primary diagnostic hypothesis. After a thorough review of these patients’ medical records, one patient was diagnosed with visceral leishmaniasis, and a second patient received empirical treatment (without diagnostic confirmation) for both tuberculosis and CM; both patients were excluded from the study.

Thus, the present study identified 13 patients with a clinical and epidemiological suspicion of CM. Ten of these patients met the criteria for a suspected (possible) diagnosis of CM, two had a probable diagnosis, and one had a compatible microbiological finding [11].

The highest number of admissions due to CM occurred in 2016 and 2018 (Fig. 1a), particularly from August to November (10 out of 13), which is a dry period (Fig. 1b). However, there were also cases of CM during the rainy season (3 out of 13), which occurs from January to April.

Fig. 1.

a. Distribution of CM cases in Cariri (Ceará, Brazil) from 2012 to 2022. b. Distribution of CM cases per month. c. Map of the State of Ceará. The dark yellow areas indicate regions where CM cases have been previously described; the pink-colored areas indicate the municipalities with cases described in the present study

The cases identified were from seven municipalities in the Central-South and South regions of the State of Ceará, as shown in Fig. 1c. In total, eight patients (61.5%) lived in rural areas, while five (38.5%) resided in urban areas. None of the patients reported traveling outside the State of Ceará or having previously lived in another region of Brazil. The likely occurrence of two micro-outbreaks was observed: the first in the municipality of Quixelô (6° 15’ 19’’ S, 39° 11’ 50’’ W), with five patients admitted in September 2016, and the second in Saboeiro (6° 32’ 29’’ S, 39° 54’ 28’’ W), with three patients admitted in August 2018.

All 13 patients in the study were males who had been exposed to soil while hunting armadillos. The mean and median age were both 19 years, ranging from 15 to 23 years. In this series, the mean time from exposure during armadillo hunting to the onset of symptoms was nine days, and the mean time from symptom onset to hospital admission was 11 days. The mean length of stay for these patients was 14.75 days.

The main respiratory symptoms reported were cough (12/13; 92.31%), chest pain (9/13; 69.23%), and dyspnea (8/13; 61.54%). Expectoration was rare (1/13; 7.69%), and there were no reports of hemoptysis. Fever was reported by 92.31% (12/13) of patients, and other constitutional symptoms included headache (5/13; 38.46%), fatigue (5/13; 38.46%), weight loss (3/13; 23.08%), and anorexia (2/13; 15.38%). Extrapulmonary clinical manifestations included splenomegaly (2/13; 15.38%), skin lesions (1/13; 7.69%), arthralgia (1/13; 7.69%), and hepatomegaly (1/13; 7.69%). None of the patients had comorbidities or immunosuppressive conditions.

The imaging examinations conducted during hospital admission were analyzed for each patient. Only one patient did not have a chest X-ray, and the results of this examination were not found for another patient. However, these two patients underwent chest computed tomography (CT). Only one patient did not undergo a chest CT scan and had only a chest X-ray instead.

Regarding the chest X-ray findings, 8 out of 11 patients (72.73%) showed bilateral involvement. Nodular opacities were the most frequent findings (8/11; 72.73%). Other findings included consolidation (3/11; 27.27%), cavitation (3/11; 27.27%), interstitial opacities (1/11; 9.09%), and costophrenic angle opacity (1/11; 9.09%).

Table 1.

Overall characteristics of the CM cases in the study

Case	Tracking method	Date of admission	LOS (days)	Sex	Age (years)	City of origin	Symptoms	Time between exposure* and symptoms (days)/ time between symptoms and hospitalization (days)	Chest X-ray	Chest tomography	Laboratorial tests (clinical sample)	In-hospital treatment/ Post discharge
1	Pulmonology referral	Nov 7th, 2014	4	M	15	Juazeiro do Norte (7° 14’ 14’’ S, 39° 19’ 20’’ W)	Cough, fever, weight loss, and anorexia	No record / 20	Nodular opacities	Not performed	Not available *** (BAL)	Did not use specific drug therapy
2	B41	Aug 1st, 2016	13	M	21	Várzea Alegre (5° 21’ 0’’ S, 40° 22’ 60’’ W)	Cough, fever, and chest pain	No record / 15	Not available	Consolidation, nodules, nodules with cavitation, centrilobular nodules, lymph node enlargement	Positive KOH smear for CM (sputum). Not available (BAL)	Fluconazole VO, Itraconazole VO/ Itraconazole VO
3	B41	Sep 8th, 2016	29	M	20	Quixelô (6° 15’ 19’’ S, 39° 11’ 50’’ W)	Cough, fever, chest pain, dyspnea, and headache	6 / No record	Consolidation and opacity of the costophrenic angle	Consolidation, nodules, nodules with cavitation	Negative KOH and AFB smears; negative cultures for fungi and mycobacteria (BAL)	Fluconazole VO, Itraconazole VO/ Itraconazole VO
4	B38	Sep 5th, 2016	15	M	20	Quixelô	Cough, chest pain, dyspnea, headache, and fatigue	No record / 8	Nodular opacities	Nodules, centrilobular nodules	Negative KOH and AFB smears; negative cultures for fungi and mycobacteria (BAL)	Amphotericin IV, Itraconazole VO/ Itraconazole VO
5	Pulmonology referral	Sep 5th, 2016	15	M	21	Quixelô	Cough, fever, chest pain, dyspnea, headache, and fatigue	9 / 8	Consolidation	Consolidation, nodules, pleural effusion	Negative KOH and AFB smears; negative cultures for fungi and mycobacteria (BAL)	Amphotericin IV, Itraconazole VO/ Itraconazole VO
6	B38	Sep 10th, 2016	3	M	18	Quixelô	Cough, fever, chest pain, and headache	15 / No record	Not performed	Nodules with cavitation, centrilobular nodules	Not performed	Fluconazole VO, Itraconazole VO/ Itraconazole VO
7	Pulmonology referral	Sep 13th, 2016	7	M	18	Quixelô	Cough, fever, chest pain, dyspnea, headache	21 / 14	Nodular opacities	Nodules, nodules with cavitation	Negative KOH and AFB smears; negative cultures for fungi and mycobacteria (BAL)	Itraconazole VO/ Itraconazole VO
8	Pulmonology referral	Jan 24th, 2017	7	M	23	Farias Brito (6° 55’ 36’’ S, 39° 34’ 25’’ W)	Cough, fever, chest pain, dyspnea, fatigue, and expectoration	No record / 10	Cavitation	Nodules with cavitation	Not performed	Itraconazole VO/ Itraconazole VO
9	B380	Feb 27th, 2017	16	M	17	Assaré (6° 52’ 31’’ S, 39° 52’ 29’’ W)	Cough, fever, chest pain, dyspnea	7 / 13	Consolidation, Nodular opacities, interstitial opacities and cavitation	Consolidation, nodules, nodules with cavitation, pleural effusion, lymph node enlargement	Not performed	Amphotericin IV
10	B38	Aug 2nd, 2018	8	M	17	Saboeiro (6° 32’ 29’’ S, 39° 54’ 28’’ W)	Cough, fever, dyspnea, fatigue, and weight loss	2 / 12	Nodular opacities	Nodules, nodules with cavitation	Positive ID** for reactive CM (serum)	Fluconazole IV/ Fluconazole VO
11	B38	Aug 2nd, 2018	15	M	18	Saboeiro	Cough, fever, dyspnea, and weight loss	15 / 12	Nodular opacities	Nodules, nodules with cavitation	Positive ID** for reactive CM (serum)	Fluconazole IV/ Fluconazole VO
12	B38	Aug 3rd, 2018	18	M	22	Saboeiro	Cough, fever, dyspnea, and headache	2 / 11	Nodular opacities	Nodules, nodules with cavitation, lymph node enlargement	Negative KOH and AFB smears; negative cultures for fungi and mycobacteria (BAL)	Amphotericin IV/ Fluconazole VO
13	B38	Mar 23rd, 2021	31	M	18	Campos Sales (7° 4’ 34’’ S, 40° 22’ 34’’ W)	Cough, fever, chest pain and anorexia	2 / 8	Nodular opacities and cavitation	Nodules, nodules with cavitation, centrilobular nodules, lymph node enlargement	Negative culture for fungi and negative AFB smears (sputum)	Amphotericin IV, Itraconazole VO/ Itraconazole VO

* All patients in the series had a report of soil exposure while hunting armadillos as the main risk exposure factor for CM, LOS: Length of stay, ** ID: radial immunodiffusion, *** not available: with no results located in medical records, B38 ICD − 10: Coccidioidomycosis, B41 ICD − 10: Paracoccidioidomycosis, BAL: Bronchoalveolar Lavage, AFB: Acid-Fast Bacillus smear, VO: Oral route, IV: Intravenous route.

In the evaluation of the chest computed tomography (CT) scans, all examinations showed bilateral involvement. Nodules and cavitated nodules were present in 10 out of 12 scans (83.33%). Other findings included consolidations (4/12; 33.33%), centrilobular nodules (4/12; 33.33%), lymph node enlargement (3/12; 25%), pleural effusion (2/12; 16.67%), and lung mass (1/12; 8.33%). Figure 2 illustrates some of the radiological findings described.

Fig. 2.

Image referring to one of the cases with reagent serology for Coccidioides a - chest Rx. Pulmonary nodules (white arrow). b, c, and d – chest TC. Pulmonary nodules (white arrow) and nodule with cavitation (black arrow)

Regarding microbiological examinations, nine (69.23%) of the 13 patients had sputum or bronchoalveolar lavage samples collected. Seven samples were subjected to fungal and mycobacterial testing (direct examination with KOH and acid-fast bacillus (AFB) smears), eight samples were cultured for fungi, and four samples were cultured for mycobacteria. However, microscopy for fungi in sputum was positive in only one sample, which showed fungal structures compatible with spherules. No samples were positive for mycobacteria. The samples were not analyzed at a mycology reference center in the capital of Ceará, but instead, they were sent for analysis to a local laboratory that does not have biosafety level 3 for pathogenic bioagents nor recognized expertise in fungi such as Coccidioides spp.

The patients in the case series presented with mild leukocytosis (leukocytes– mean of 13,333 cells/µL; neutrophils– mean of 10,888 cells/µL) with neutrophilia, along with an increase in C-reactive protein (mean of 5.84 mg/dL).

There were also few specific complementary tests for the diagnosis of CM. Only three patients underwent CM serology using the Double Radial Immunodiffusion method, after sending blood samples to a reference laboratory at the Oswaldo Cruz Foundation (Fiocruz) in the State of Rio de Janeiro. Of these, two tested positive for antibodies; the patient with a negative initial serological test was not subjected to a second test. No patient underwent PCR testing on biological samples.

A total of ten (76.92%) patients were treated with antibiotics indicated for Community-Acquired Pneumonia either at the beginning of or prior to hospitalization, before starting antifungal therapy. None of the patients received drug therapy for tuberculosis. Antifungal therapy was initiated in 12 (92.31%) patients: 10 due to lack of clinical improvement with antibiotics and 2 as the initial therapy of choice. One patient experienced spontaneous improvement even in the emergency department without the use of antimicrobials or antifungals.

During the hospital stay, the antifungal treatments used included Amphotericin B (27.78% of therapies), intravenous Fluconazole (11.11%), oral Fluconazole (16.67%), and oral Itraconazole (44.44%), all administered in standard doses. At hospital discharge, 11 patients were prescribed oral azoles for prolonged use, with oral Fluconazole used in 27.27% of cases and oral Itraconazole in 72.73%, also in standard doses. These patients were referred for follow-up at their local primary health care units.

Of the patients who received antifungal treatment, 11 out of 12 (91.67%) were discharged due to clinical improvement, with instructions for continued outpatient antifungal therapy. One patient died during hospitalization. This patient’s death was recorded as being due to coccidioidomycosis on the death certificate but was not subjected to a postmortem examination.

Discussion

In this study, all patients were young, male, and had recent exposure to armadillo hunting. This patient profile is consistent with the predominant characteristics of coccidioidomycosis (CM) cases described in Brazil. In a case series of 30 patients in Ceará, the average age was 30 years, all were male, and 76.67% were armadillo hunters [12]. Similarly, a recent study of 292 CM patients from Maranhão and Piauí found a male-to-female ratio of 57.4:1, with 91.1% being armadillo hunters [6]. Therefore, the epidemiological data from this study align with existing literature, which identifies young men who hunt armadillos as a high-risk group for CM infection in northeastern Brazil.

Prior to this study, only 10 cases of CM had been described in the Central-South and South regions of Ceará between 2007 and 2017. These patients were diagnosed in a referral hospital in Fortaleza, located approximately 390 km from their places of origin [12]. This highlights the difficulty of diagnosing CM in smaller health units in northeastern Brazil, forcing patients to travel long distances to referral centers for diagnosis.

Regarding the municipalities of origin, it is noteworthy that only Saboeiro had been previously mentioned as a location of CM occurrence. This study reports CM cases in previously unrecognized areas in Ceará, increasing the number of municipalities affected by CM in the state from 20 to 26 [8, 12].

Unlike most previously reported cases in Brazil, which were diagnosed at referral centers [6, 8, 12], the patients in this study were referred to a regional referral hospital. This underscores the limitations of managing CM in health services closest to patients’ residences.

As anticipated, more CM cases were observed from August to November, a period of high temperatures and low rainfall in the region. A negative correlation between increased rainfall and CM cases has been described [6], with a predominance of cases from September to January [5].

The symptom profile was similar to other CM series in Brazil, predominantly presenting with cough, fever, chest pain, and dyspnea [6, 12]. Clinicians in endemic areas should consider CM in differential diagnoses when these symptoms are present.

The most common radiological findings on chest tomography were pulmonary nodules and nodules with cavitation, with bilateral involvement observed in all patients. This presentation is consistent with that described by Capone et al. [13] in an analysis of chest CT scans in CM patients in northeastern Brazil.

Despite analyzing nine samples (sputum or bronchoalveolar lavage) using microscopy and culture for fungi, only one sample showed spherical fungal structures suggestive of Coccidioides. The lack of microbiological confirmation in most cases may be due to errors in sample collection, transport, or storage. It is also noteworthy that the samples were processed by a local laboratory without expertise in mycology, potentially affecting the accuracy of fungal identification. C. posadasii requires careful handling as it is classified as a Biosafety Level 3 pathogen, and processing in facilities lacking appropriate safety measures poses risks to laboratory staff [14, 15].

Although point-of-care serological tests for CM are now available, only three patients were evaluated using immunodiffusion (ID). Given Brazil’s vast territorial distances, implementing local testing schemes is crucial to aid diagnosis and ensure timely initiation of therapy.

The main limitation of this study is the low rate of microbiological or serological confirmation. Unfortunately, hospitals in endemic areas in northeastern Brazil lack the necessary infrastructure and support for these tests. This scenario hampers our understanding of CM epidemiology in Brazil and calls for urgent improvements in the management of endemic mycoses in the country.

Funding

No funding

Data availability

The datasets generated during and/or analysed during the current study are not publicly available due to ethical reasons, but are available from the corresponding author on reasonable request.

Declarations

Competing interests

On behalf of all authors, the corresponding author states that there are no conflicts of interest.