Acute histoplasmosis in immunocompetent travelers: a systematic review of literature

Background Histoplasmosis is a fungal infection highly endemic in the American continent. The disease can be severe in immunocompromised subjects. In immunocompetent subjects the clinical manifestations are variable. Aim of this work was to review the cases of acute histoplasmosis in immunocompetent travelers reported in literature. Methods A systematic review of literature was conducted. Electronic search was performed in Pubmed and LILACS. Two reviewers independently extracted data on demographic, clinical and radiological features, and treatment. Cases were classified according to Wheat’s definitions. Results Seventy-one studies were included in the analysis, comprising a total of 814 patients. Twenty-one patients diagnosed at the Centre of Tropical Diseases, Negrar (VR), Italy were also included. The most common travel destination was Central America (168 people, 29.8%); the most common way of exposure to histoplasma was the exploration of caves and/or contact with bat guano (349 people, 60.9%). The multivariate logistic regression model showed association between the development of disseminated histoplasmosis (DH) and activities that involved the exploration of caves and/or the contact with bats’ guano (adjusted OR: 34.20 95% CI: 5.29 to 220.93) or other outdoor activities (adjusted OR: 4.61 95% CI: 1.09 to 19.56). No significant difference in the attack rate between countries of destination was observed (p-value: 0.8906, Kruskal-Wallis test). Conclusions Histoplasmosis often causes no or mild symptoms in immunocompetent individuals, although a severe syndrome may occur. The infection can mimic other diseases, and the epidemiological risk of exposure is an important clue to raise the index of suspicion. Electronic supplementary material The online version of this article (10.1186/s12879-018-3476-z) contains supplementary material, which is available to authorized users.


Background
Histoplasmosis is a fungal infection acquired by inhalation of Histoplasma capsulatum var. capsulatum microconidia. This fungus is endemic in the American continent (North, Central and South America). Cases have also been described in Africa, Asia, and Europe [1].
Histoplasmosis is both a disease originating from soil (infections often occur after human disruption of the soil that aerosolizes the hyphae and conidia of the organism originating from the earth enriched with birds' excreta) and a sapronosis (originating from bats) [2,3]. With rare exceptions, this fungus infects.
its host via the respiratory route: microconidia are inhalated and deposited in the terminal bronchioles and alveoli of the lung.
Clinical manifestations of histoplasmosis depend on the size of the inoculum and the patient's immunological status/underlying conditions. Low level of exposure in the healthy host typically leads to an asymptomatic infection, while the acute syndrome usually follows a heavy inoculum [4]. In the latter case, chest radiographs usually show diffuse interstitial or reticulonodular infiltrates, sometimes associated with hilar or mediastinal lymphadenopathy [4]. Systemic symptoms include fever with possible respiratory impairment, sweats, weight loss, headache and gastrointestinal complaints. A subacute pulmonary syndrome is also possible in case of low inoculum [4]. The most frequent symptoms are cough, fever, malaise and fatigue, typically occurring over weeks to months before seeking medical advice [4]. Solitary lung nodules (histoplasmomas) or multiple pulmonary nodules may develop, characterized by indefinite persistence of foci of infection.
Calcification is usually found in the center of a histoplasmoma or in concentric rings and is generally diagnostic, although it may require years to develop and is not always present.
Pulmonary nodules may slowly enlarge and even cavitate [5]; moreover, they are usually identified as incidental findings on chest radiographs or CT scans, and differential diagnosis includes sarcoidosis, tuberculosis (TB) and malignancies [6][7][8]. Chronic pulmonary histoplasmosis can be observed in individuals with underlying lung conditions, and it is defined as failure to clear H. capsulatum infection [4].
Symptoms and signs include cough, low-grade fever, night sweats and weight loss. Chest radiographs show interstitial or consolidative infiltrates associated with cavitation or fluid-filled emphysematous bullae [4]. This kind of pulmonary syndrome doesn't resolve spontaneously, leading to gradual progression and destruction of lung parenchyma. Mediastinitis, broncholitiasis, pericarditis and a rheumatologic syndromes have been described, too. The disease is particularly relevant for immunosuppressed subjects, who are at risk of developing the disseminated form, which is often fatal if not promptly recognized [4].
On the other hand, the infection is usually mild and self-limiting in immunocompetent people [4], although, apparently in rare cases, the disease can progress to a severe form with high morbidity even in those individuals [9][10][11][12].

Objectives
Aim of our work was to review the literature on acute histoplasmosis in immunocompetent travelers, describing the main characteristics of the infection in this subgroup of patients. Secondary objectives were to investigate the characteristics associated to the development of disseminated histoplasmosis in travelers, and to estimate the attack rate in different areas of the world.

Methods
We conducted a systematic review of case reports and case series available in literature. We analyzed the epidemiological and clinical features, diagnostic work-up, treatment, and outcome of cases of acute histoplasmosis in immunocompetent patients, with history of recent travel (≤ 3 months) in an endemic area.
Electronic search was conducted in PubMed and LI-LACS on 31st December 2016, and all papers published up to that date were included. We included full texts written in English, Italian, French, German, Spanish, Portuguese.
Additional cases were sought from the reference list of included papers and of non-included reviews. Moreover, grey literature was sought through contact with authors. Two authors, SS and AA, independently screened the list of papers generated from the electronic search, using EndNote program, version 6, 2012 Thomson Reuters.
SS and AA extracted the data on the basis of inclusion/ exclusion criteria and case definitions (reported below), and entered all information in a database created with Microsoft Office Excel 2007. Data collected were: age, sex, country of exposure, activities done in the country, attack rate and development of disseminated disease. The Excel data base is available as supplementary material (Additional file 1).
In case of discrepancies in the process of inclusion of papers/data extraction, a consensus was reached through discussion or involvement of a third reviewer (FG).

Inclusion criteria
-Case reports or case series on acute histoplasmosis.
-Patients with history of travel in an endemic area (including travels from an area of no/low endemicity to an area of high endemicity within a country) ≤ 3 months before the onset of symptoms.
Exclusion criteria -Reviews of literature.
-Cases describing patients presenting with radiologic findings compatible with histoplasmosis (i.e. pulmonary nodules) but no apparent temporal relationship with travel.
-Cases native from an endemic area without history of travel to an area with higher endemicity.
-Non-immunocompetence status at the time of the travel, at the onset of symptoms or at the time of diagnosis of histoplasmosis.
-Occupational exposure in native area.
-Cases of reactivation of histoplasmosis acquired in the past.
Case definition (adapted from wheat [13] and EORTC/MSG criteria [14]) Proven Illness consistent with acute histoplasmosis in a patient who had recently (≤ 3 months) travelled to an endemic area and one of the following: -Positive culture from a specimen obtained from the affected site or tissue.
-Histopathologic or direct microscopic demonstration of yeast intracellular forms or yeasts in tissue macrophages.

Probable
-Host factor: recent travel (≤ 3 months) in area endemic for histoplasmosis.
-Clinical picture and/or radiological findings consistent with acute histoplasmosis.
-Mycological evidence (histoplasmin test positivity and/or positivity of serology and/or positivity of urine-serum Histoplasma antigen).

Possible
-Host factor: recent travel (≤ 3 months) in area endemic for histoplasmosis.
-Clinical picture and/or radiological findings consistent with acute histoplasmosis.
-No mycological evidence nor serologic evidence but: Belonging to a histoplasmosis cluster and/or efficacy of oral ex juvantibus itraconazole therapy (resolution of symptoms and normalization of radiological findings).

Activities at risk for developing histoplasmosis
Contact with bats excreta and/or visit to a cave.
Outdoor activities: forest excursions/trekking, camping or other activities not involving visit to caves or direct contact with bats' excreta.
Activities involving contact with chickens or other non-mammal birds.
All activities that put the subject at risk of inhalating Histoplasma microconidia: building construction, renovation, excavation (biologists, building contractors, construction workers, soldiers, volunteers involved in churches or other buildings renovation).

Syndromic classification (according to wheat [13])
Asymptomatic (As): asymptomatic case identified through laboratory screening within a histoplasmosis cluster.
Acute pulmonary histoplasmosis (AP): febrile illness with respiratory symptoms and/or mediastinal involvement and/or accompanying constitutional symptoms and/or gastrointestinal or cutaneous involvement when they are not the main manifestation.
Rheumathologic (R): a case with arthralgia or arthritis as unique symptom.
Disseminated Histoplasmosis (DH): febrile illness with demonstration of bone marrow involvement (leucopenia, thrombocytopenia, anemia) AND/OR disseminated monocytic-macrophage involvement (hepatosplenomegaly AND/OR increased liver enzymes), with or without respiratory symptoms.

Statistical analysis
Descriptive statistics were used to analyze the characteristics of the entire cohort. Categorical variables were reported as frequencies and proportions, while quantitative variables were presented as median (interquartile range, IQR). Associations between covariates were explored through univariate logistic models and non-parametric Kruskal-Wallis rank sum test. Lastly, we fitted a multivariate logistic regression model to assess the impact of all covariates on the probability of contracting the disseminated disease. All statistical analyses were conducted using R, version 3.3.3 [15].

Results
The study selection process (PRISMA flow chart) is shown in Fig. 1. The studies included in the analysis through electronic search were 71 [8,, (Table  1) comprising a total of 814 patients (see Fig. 2 for the timeline distribution of studies and cases respectively). In the analysis were also included 21 patients diagnosed at the Centre of Tropical Diseases (CTD) of Negrar, Verona, Italy during the decade 2005-2015 (Staffolani S, Buonfrate D, Farina C, Gulletta M, Gobbi F, Angheben A, manuscript in preparation), for a total of 835 patients. Figure 3 shows the countries of presumable exposure to the infection.
Regarding the cases diagnosed at the Centre of Tropical Diseases, Negrar (VR), briefly, we found 21 cases of acute histoplasmosis in immunocompetent subjects, 17 of them affecting members of a scientific expedition to Ecuador. The other 4 patients did not belong to a cluster and traveled to Panama, Bolivia, Mexico, Cuba respectively. Seventeen out of 21 patients were male (81%). The mean age was 38.5 years for the Ecuador cluster, 46.7 years for the other cases. They were infected through contact with bat excreta, inhalation of contaminated soil, outdoor activities. All cases were classified as possible histoplasmosis, but two probable for positive serology and two proven cases based on histology. The main symptoms were fever (12/21, 57%), respiratory (12/ 21, 57%) and gastrointestinal (9/21, 43%) symptoms. Seven patients (33%) were admitted to the hospital and four of them underwent invasive diagnostic procedures (bronchoscopy, biopsy of other tissues). Eleven out of 17 (65%) patients of the Ecuador cluster and 4/4 (100%) patients not belonging to it had abnormal blood exams. Eleven patients (65%, 7/17 (41%) in the cluster and 4/4 (100%) not belonging to the Ecuador cluster) received oral itraconazole. Nine patients (39.1%) had persistent lung nodules at the 12-month radiological follow up.
Regarding all selected cases of the present review, most patients (488/835, 59%) belonged to a cluster. There were 45 clusters described by 36 authors, plus 2 clusters diagnosed at CTD. The cluster comprising the largest number of individuals (n = 262) was described by Morgan [35].
While we considered all 835 patients for the description of main characteristics of cases, the "Acapulco cluster" [35] was removed from our dataset for univariate and multivariate analyses, as it did not contain information regarding age or the state of disseminated disease. Table 2 displays a summary of the characteristics of the cohort of the 573 immunocompetent travelers considered for univariate and multivariate analyses.

Clinical-syndromic classification
According to the above-mentioned definitions, we found 32 proven, 695 probable and 108 possible histoplasmosis cases.
Information about Histoplasma serology was available for 634 (76%) patients, of whom 413 (65%) were positive. All asymptomatic patients included in the analysis were positive to serology, which was performed on the basis of epidemiological suspicion (travel with patients diagnosed with histoplasmosis). Of 213 (26%) patients screened with histoplasmin skin test, 199 (93%) resulted positive.

Imaging
Imaging included chest X ray and CT scan. This information was available for 376 (45%) patients. Of them, 250 (67%) showed abnormal imaging. In particular: nodular infiltrates were present in 200 patients (80%), patches in 16 patients (6%), miliary pattern in 14 patients (6%), hylar/mediastinal adenopathies in 49 patients (20%), pleural effusion in 4 patients (2%).  2 Timeline distribution of the studies reporting cases of acute histoplasmosis in immunocompetent travelers. Each bubble corresponds to a study, the diameter of the bubble is proportional to the number of cases described in the study, which correspond to the projection of the center of the bubble on the ordinate axis Information about the spleen and liver size was available only for 14 (2%) patients, of whom 5 (36%) had hepatosplenomegaly and one had hepatomegaly.

Treatment
Information was available for 744 patients (89%). Of them, 185 (25%) received antifungal therapy. In particular: 105 (57%) patients received oral itraconazole, while 21 (11%) received intravenous amphotericin B. Twelve (57%) of the latter were switched to oral drugs at hospital discharge: one of them was switched initially to oral itraconazole and then to posaconazole, 8 of them were switched to ketoconazole and then to itraconazole, 4 were switched to itraconazole. Twenty-four (13%) patients were

Outcome
A detailed information about the outcome was available for 727 (87%) patients. Most of them (726/727, 99.8%) recovered. In particular: 688 (95%) patients had complete clinical and radiological resolution at the last visit, while 38 (5%) patients had clinical recovery and a partial clearance of the radiological findings. One patient with disseminated histoplasmosis and typhoid fever died because of multi-organ failure and hypercalcemia [85]. Table 3 reports a summary of the univariate and multivariate analyses conducted on all variables of interest. Results from the univariate analysis showed an association between disseminated histoplasmosis and variables such as age (crude OR: 1.03 95% CI: 1.01 to 1.05), Central America (crude OR: 13.01 95% CI: 3.82 to 81.43) or South America (Crude OR: 6.66 95% CI: 1.84 to 42.72) as a destination, and activities that involved inhalation of contaminated soil (crude OR: 3.13 95% CI: 1.72 to 5.62). However, the multivariate logistic regression model did not confirm any association with destination or age, while a strong association was found with activities that involved the exploration of a cave full of bat's guano (adjusted OR: 34.20 95% CI: 5.29 to 220.93) or other outdoor activities (adjusted OR: 4.61 95% CI: 1.09 to 19.56). No significant difference in the attack rate between countries of destination was observed (p-value: 0.8906, Kruskal-Wallis test. Not reported in tables).

Discussion
We were able to identify in the literature 814 cases of histoplasmosis acquired by immunocompetent travelers, for a total of 835 patients including those diagnosed at CTD. Considering that this occurrence is certainly under-diagnosed and under-reported, we assume that the published cases must be the tip of the iceberg. We hope that our systematic review may contribute to raise awareness of the possible occurrence of this condition in travelers visiting endemic countries/areas. It is true that most cases are clinically not severe, nevertheless most of the subjects presented symptoms of various degree, moreover usually long lasting, often implying otherwise unnecessary, invasive diagnostic procedures under the suspicion of more common (and serious) causes such as TB or lung malignancies. Only less than 10% of the subjects were apparently asymptomatic, although it should be recognized the low probability of a pauci-or asymptomatic case to be identified and described in the literature. Of particular note, a relatively large group of patients had Disseminated Histoplasmosis (DH), confirming that this condition, although typical of immunosuppressed individuals, may also occur in immunocompetent subjects. Moreover, given the long course of the infection, an unrecognized/untreated case of histoplasmosis may reactivate later in life, under immune suppressant conditions. Are there any definite risk factors for DH in immunocompetent individuals? We found several associations between disseminated histoplasmosis and other covariates at univariate analysis (Table 2). However, these results appear to be very inconsistent and were not confirmed by multivariate analysis and therefore no conclusion can be drawn from them. At multivariate analysis, the only risk factors resulting significant were exposition to cave-bats and outdoor activities. We found no association between the attack rate and the different endemic areas, hence we cannot suppose that there is a difference in the aggressiveness of the different strains throughout the world. In contrast with this conclusion, some authors [86][87][88][89][90] supposed that different and specific Histoplasma capsulatum genotypes could be responsible of the variability of clinical manifestations, in particular some "aggressive" strains might cause a more severe presentation. It is to be mentioned that our result is based on 47 groups only, and the observation of more groups in endemic areas would possibly provide a more robust conclusion. Seventy-eight per cent of reported cases acquired the infection in the American continent, and 57,2% in Central-South America only; these areas should be therefore considered "highly" endemic, although the disease can be seen globally. From the clinical point of view, when should histoplasmosis be suspected in a returning traveler? Histoplasmosis can be misdiagnosed, not only because of the common low awareness of health-care workers but especially because the clinical picture is mainly characterized by fever (more than 90% of reported cases) associated to respiratory symptoms in around half of cases. More than 80% of patients have a pulmonary form. Therefore, histoplasmosis should be suspected in all travelers presenting with fever/systemic symptoms or community-acquired pneumonia. History should be carefully evaluated, in particular exposure to cave/bats or simply outdoor activities in endemic areas (particularly the American continent) and being part of a cluster of similar illness. Although the most part of cases in our review had also abnormal laboratory exams (two thirds of cases had increased inflammatory markers and half increased transaminases) this is an aspecific finding, especially in pneumonia/febrile systemic illness. However, Histoplasma serology is positive in two third of cases. According to Wheat [4], Histoplasma galactomannan antigen can be positive in 80-95% of cases in the acute pulmonary and disseminated form and could strongly contribute to diagnosis when available.
Finally, radiologic findings can be of help to distinguish histoplasmosis from bacterial/viral pneumonia in immunocompetent travelers; in our review, 80% of retrieved cases with this information showed pulmonary nodules (typically multiple peripheric nodules with halo sign at computed tomography). Our study has some limitations. First, we excluded some papers by language, in particular those written in Chinese and Dutch, and this could have led to the loss of some cases. We did not find a standardized definition of acute histoplasmosis in the immunocompetent subject in the scientific literature, hence we adapted the definition and the inclusion  [13,14]. Most of all, it was difficult to retrieve accurate information about the outcome, that was not always described in details. However, with the limitation of a various and not ever reported follow-up, more than 99% of subjects recovered or had a consistent clinical and radiological improvement.
Inconsistency in the analysis of risk factors for Disseminated Histoplasmosis is most probably due to the low quality of the gathered data, that is probably explained by the vast heterogeneity of the sources across the years of publication and the differences in the statistics reported. We tried to implement different logistic models to verify which one could have a better goodness of fit, but unsuccessfully. We also tried data imputation to improve the completeness of the dataset. Namely, we tried to complete data on age and sex through a Multivariate Imputation by Chained Equations (MICE) [88] with the predictive mean matching method, in order to maintain the original distribution of age and sex of the incomplete data. Unfortunately, this did not have an impact on the quality of results. However, the exposure to cave-bats and outdoor activities resulted significant in the majority of models we fitted, so we assume that there is an association with those variables. We finally decided to report results from the best model in terms of goodness of fit.

Conclusions
Presumably, cases of histoplasmosis in immunocompetent travelers are largely misdiagnosed and under-reported, because these subjects are usually asymptomatic or present mild symptoms. However, it is important to consider this infection in the differential diagnosis in case of epidemiological risk of exposure, also in light of possible (though rare) progression of the disease also in immunocompetent individuals. In addition, the risk of reactivation of the infection in case of subsequent acquired immunosuppression is present, even after a long time [91]. Studies are needed to understand whether antifungal therapy prescribed for the acute, mild form of the disease could prevent further reactivations. Finally, a high index of suspicion might avoid unnecessary hospitalization and invasive procedures.

Funding
This work did not receive any specific funding. Abbreviations, OR odds ratio, CI confidence interval a Relative frequencies are here presented, and the median (interquartile range) is reported for age b Adjusted for every other covariate through a multivariate logistic regression model c Includes: Asia, Oceania d Each activity was entered in the model as a single dummy variable, as most people did more than one activity during the trip. Caves/Bat's guano: contact with bats excreta and/or visit to a cave. Outdoor activities: forest excursions/trekking, camping or other activities not involving visit to caves or direct contact with bats' excreta. Bird_chickens: activities involving contact with chickens or other birds. Inhalation_soil: all activities involving building construction, renovation, excavation (biologists, building contractors, construction workers, soldiers, volunteers involved in churches or other buildings renovation)

Availability of data and materials
The data base created for the collection of cases and for the statistical analysis is available as supplementary material.
Authors' contributions SS, AA independently screened the papers. When the opinion regarding the inclusion of a paper was discordant, a consensus was reached through discussion with a third reviewer, FG. MG and GG conducted the statistical analysis. DB, SS and AA wrote the manuscript. FB and ZB made the final revision of the manuscript. All Authors read and approved the final version of the manuscript.
Ethics approval and consent to participate All the 21 patients seen at CTD gave their informed (verbal) consent for the publication of anonymous data. We submitted a study protocol for the retrospective review of CTD cases to the Ethics Committee (Comitato Etico per la sperimentazione clinica delle Province di Verona e Rovigo), and received a waiver of consent (protocol 20,901 of 26 April 2017), since patients' data were de-identified.
Consent for publication N.A.

Competing interests
The authors declare that they have no competing interests (Additional file 2).