Histoplasmosis in the Republic of Congo dominated by African histoplasmosis, Histoplasma capsulatum var. duboisii

The Republic of Congo (RoC) is one of the African countries with the most histoplasmosis cases reported. This review summarizes the current status regarding epidemiology, diagnostic tools, and treatment of histoplasmosis in the RoC. A computerized search was performed from online databases Medline, PubMed, HINARI, and Google Scholar to collect literature on histoplasmosis in the RoC. We found 57 cases of histoplasmosis diagnosed between 1954 and 2019, corresponding to an incidence rate of 1–3 cases each year without significant impact of the AIDS epidemic in the country. Of the 57 cases, 54 (94.7%) were cases of Histoplasma capsulatum var. duboisii (Hcd) infection, African histoplasmosis. Three cases (5.3%) of Histoplasma capsulatum var. capsulatum infection were recorded, but all were acquired outside in the RoC. The patients’ ages ranged between 13 months to 60 years. An equal number of cases were observed in adults in the third or fourth decades (n = 14; 24.6%) and in children aged ≤15 years. Skin lesions (46.3%), lymph nodes (37%), and bone lesions (26%) were the most frequent clinical presentations. Most diagnoses were based on histopathology and distinctive large yeast forms seen in tissue. Amphotericin B (AmB) was first line therapy in 65% of the cases and itraconazole (25%) for maintenance therapy. The occurrence of African histoplasmosis in apparently normal children raises the possibility that African histoplasmosis is linked to environmental fungal exposure.


Introduction
Histoplasmosis is an endemic mycosis due to a dimorphic fungus named Histoplasma capsulatum [1]. Its distribution is worldwide [2,3], but Africa is unique in regard to this infection [4] as 2 clinical entities coexist due to Histoplasma capsulatum var. capsulatum (Hcc) and Histoplasma capsulatum var. duboisii (Hcd), the cause of "African Histoplasmosis." Indeed, while the geographic distribution of the former encompasses South and North America, Asia, and Africa, the latter has only been reported in patients living or having lived in Africa [4,5].
While the HIV/AIDS pandemic and the increased use of immunosuppressive agents clearly demonstrated the opportunistic behavior of Hcc. Indeed, there were reports of cases in previously ''non-endemic areas" revealing the global distribution of histoplasmosis in Africa [5] and throughout the world [2,3,6,7]. In countries where patients have limited access to diagnostic testing and antiretroviral therapies (ARTs), histoplasmosis is probably an important cause of mortality in persons living with HIV/AIDS. Histoplasmosis, particularly Hcd, has not been adopted by the World Health Organisation as a neglected tropical disease (NTD) (https:// www.who.int/neglected_diseases/diseases/en/) despite early consideration [8] and was also rejected by G-Finder, despite a formal application (Global action fund for fungal infections, 2016).
In the Republic of Congo (RoC), the health epidemiological profile is characterized by the predominance of infectious diseases, mainly malaria, TB, and HIV/AIDS infections [9]. Many studies have been published since the description of the first case of Hcd infection reported in 1954 in the RoC [10]. However, the real problem of histoplasmosis in the RoC is underestimated, as the available information on this disease comes from case reports, case series, or old reviews published in the literature.
The knowledge gap of the local epidemiology of this disease is a serious limitation for effective infection control and treatment approaches. Therefore, we performed in this study an exhaustive review of the published cases from 1950 to 2019, including some cases published in non-English language, to provide the current status of histoplasmosis in the RoC regarding the epidemiology, diagnostic tools, and treatment.

Ethics statement
This analysis did not deal with individual patient data but with published data, which does not require regulatory approval.

Search strategy and eligibility criteria
Computerized literature searches for publications on histoplasmosis cases or series in the RoC were performed using online databases Medline, HINARI, and PubMed. The search engine used the key words and the detailed medical subject heading (MeSH) terms to identify all published papers: "Histoplasmosis," "Congo," "Brazzaville," "HIV/AIDS," "diagnosis," "epidemiology," "Africa," or "Sub-Saharan Africa." The Boolean operators "AND" and "OR" were used to combine 2 or 3 terms. All included studies were cases and series reports (n = 57) originating from the RoC. Although peer review articles published in the English language were included, most were in French language. No date limitation or any other search criteria were applied to avoid missing papers published in the RoC. Systematically, we searched Google Scholar and gray literature papers regarding the subject, to supplement Medline and PubMed searches. Some cases reports were excluded from some parts of our analysis due to the lack of information related to sex, age, clinical presentation, and treatment.
Further review of relevant individual cited references identified additional cases published in French-language journals such as the Journal of Medical Mycology (formerly Journal de Mycologie Médicale) et Armée, Médecine d'Afrique Noire, Médecine Tropicale, or Bulletin de la Société de Pathologie Exotique.

Cases described from the Republic of Congo
Epidemiology. Our exhaustive literature search revealed a total of 57 diagnosed cases of histoplasmosis, including 54 cases (94.7%) of Hcd infection and 3 Hcc, were reported in the RoC between 1954 and 2019 (Tables 1 and 2). All were reported as single case report but 1 series of 11 cases reported by Carme and colleagues in 1990 [11]. The documentation of these cases varies from one study to the other and also from one case to another within the same study. For example, in 13 and 14 cases, sex or age, and clinical presentation (12 cases) were lacking. Also, the treatment of patients was not specified in 17 cases. Thus, those cases were excluded from some parts of our analysis. This total of cases represents a mean incidence of 1 to 3 cases each year, without significant change over the years, notably in regard to the AIDS epidemic (Fig 1).
Of the 57 cases of histoplasmosis, we found males (n = 30; 52.6%) to be as frequently infected as females (n = 14; 24.6%) with the sex ratio (M/F) at 2.1. The mean age and the median age were calculated at 24 ± 17.6 and 22 years, respectively. The patient's age ranged between 13 months and 60 years.
An equal number of cases were observed in adults in the third or fourth decades (n = 14; 24.6%) and in children aged �15 years (Tables 1 and 2). Eight patients were under 10 years. The HIV testing results were mentioned for 7 children and none were HIV-positive. In 10 children, Hcd infection was disseminated.
Of the 54 Hcd cases, 2 cases were diagnosed in Congolese (from the RoC) expatriates living in France, a man arrived 15 years previously [23] and a woman for some months [21]. A single case of Hcd-disseminated infection was diagnosed in the RoC in a foreign national, a Chadian man living in the RoC for an indeterminate time [10]. In contrast, 51 (94.4%) of the 54 cases were definitely acquired in the RoC, as these patients had never traveled outside the country. No documented exposure to caves was reported in those cases and no clustering of cases was evident. Only 5 cases (9.3%) reported being farmers as their occupation in our series. The result for HIV testing in 54 cases of Hcd infection is mentioned for 39 patients of whom 20 were HIV-infected (51.3%). In adults, 9 cases (15.8%) of Hcd infection were associated with  [32]. One of those cases had a disseminated infection in the context of HIV infection [30]. The first case was diagnosed in 1984 in a 22-year-old young Zairean woman (people of the Democratic Republic of Congo, formerly Belgian Congo or Zaïre) living in the RoC for an indeterminate period [31]. The second case was reported in 1991 in a Congolese AIDS patient living in France with an infection suspected to be acquired in Central Africa [32]. In the last case, Lapèze and colleagues reported a case of Hcc-disseminated infection occurring in a French patient living in the RoC for 3 years with a history of traveling and living in Indonesia (6 years), Chad (6 years), and Gabon (2 years).
Although there is no particular regional distribution, the possible origin of infection was notified for 29 patients (50.9%) out of the 57 cases of histoplasmosis reported in the RoC. Patients originated from urban areas (Brazzaville, the capital city) in 26 cases (45.6%) and only 3 cases have mentioned the rural origin of patients' regions (Gamboma, Owando, Kinkala). Further related areas of rural origin were not mentioned in 28 cases (49.1%) and these cases were excluded from the map (Fig 2).
Clinical presentation. The clinical features of the 54 cases of Hcd infection are presented in Table 3. Skin lesions (46.3%) including cutaneous and subcutaneous lesions, lymphadenopathy or lymph nodes (37%), and bone lesions (26%) were the most reported clinical presentations. As the most frequent type of lesion, cutaneous lesions resemble Molluscum contagiosum (15%) and usually take the form of umbilicated papules or nodules on the skin of limbs and face. Liver and spleen involvement was reported in 5 cases (9.3%). Lung and mucosal infection were reported with equal frequency (7.4%). Of the 14 pediatric cases of Hcd infection, the patients presented with bone lesions in 3 cases [14] of which 2 patients had major bone damage [16,27]. On X-ray, the images of bone lysis are observed, often resulting in the form of poorly defined bone cyst.  Treatment. Treatment information was described in 40 (70.2%) of 57 cases of histoplasmosis. Twenty-six patients (65%) were treated with intravenous amphotericin B (AmB) as first-line therapy. The use of itraconazole (ITRA) was reported only in cases of Hcd infection (n = 10) while 5 patients were treated with ketoconazole (KETO). In 6 cases of Hcd infection, the antifungal drug was combined with surgical treatment for example excision of necrosis and paravertebral and epidural abscesses. Terbinafine was used as a maintenance therapy after AmB in 1 case of Hcd infection with bone lesion [22]. All patients presented with Hcc infection (n = 3) were treated with AmB. Two patients did not receive any antifungal treatment, either due to the misdiagnosis of the disease in favor of tuberculosis or delayed diagnosis of a severe case leading to the rapid death of the patient.  Clinical outcome depended on a number of factors, such as the type of disease (Hcd or Hcc infection), early or prompt diagnosis, and accessibility to the effective drugs. Most cases with skin lesions (n = 20) including those with cutaneous and subcutaneous lesions and disseminated diseases (n = 10) resolved with AmB and ITRA, while response to KETO was poor leading to the death of patients (n = 4).

Discussion
Human histoplasmosis is due to 2 different fungal varieties in which geographical distribution differs, and that lead to different clinical presentations. However, molecular analysis reveals that strains of both varieties isolated in Africa belong to the same clade, suggesting that the orientation toward one or the other is driven by environmental conditions or mode of contamination rather than by intrinsic genetic characteristics [33]. To the best of our knowledge, no animal model of Hcd infection has yet been set up, that could support different virulence between the varieties. A major result of our study is the large dominance of Hcd infection in contrast to Hcc infection in the RoC. The first case of Hcd has been described in 1952 by Dubois and colleagues in a European man coming from the Democratic Republic of Congo (formerly, Belgian Congo or Zaïre), a border country of the RoC [10,34].
In regard to the total population of the country (5.2 million inhabitants), the number of Hcd infection (54 cases) gives a higher prevalence of this infection, unlike other countries in central Africa [5]. Considering the difficulties in diagnosis in our country, it is likely that this prevalence is underestimated. In contrast, the number of Hcc was few, as only 3 cases of uncertain geographical origin have been described from the RoC. In addition to a limitation in the availability of the most key diagnostic tools, one can imagine particular geoclimatic local conditions can play a role. Indeed, a remarkable annual rainfall, a high degree of humidity, and a reduced variation in diurnal temperature characterize the RoC climate. We did not observe any particular geographical distribution within the country in the cases reviewed. The possible source of infection was notified for 29 patients (50.9%), and a rural origin of the patients was only notified for 3. Whether or not local environments favor the "expression" of Hcd infection rather than Hcc infection required further studies. Finally, it should also be mentioned that facilities in the diagnosis of histoplasmosis are limited mainly to Brazzaville, the capital and largest city. Indeed, most of infected patients were from urban areas (Brazzaville) but there might be a bias because the population is predominantly urban (62.2%), and 56.5% of the total population lives in Brazzaville [35]. Also, the diagnosis can only be achieved in medical centers from urban areas. This contrasts with previous reports suggesting that Hcd infection is mainly a rural mycosis in other countries [26,36]. However, the increase in HIV-positive patients in the RoC probably promotes cases of histoplasmosis being seen in Brazzaville. Nevertheless, it should be noted that in central Africa, Zaïre was the main focus of Hcd infection with 26 cases. In contrast, all of the cases (n = 7) described before 1984 originated from a rural area (the plateaux and Pool region) of the RoC [15]. Although 5 cases were reported in farmers, the overall risk of Hcd infection is not well understood. Moreover, Carme and colleagues [14] reported that most patients with Hcd infection are farmers, most consistent with a soil-related acquisition. In the study from Democratic Republic of the Congo, which shares a border with the RoC, Pakasa and colleagues reported that 4 patients infected with Hcd had collected guano from bat roosts to fertilize gardens [37]. In the environment, Histoplasma may be found in socalled microfoci in endemic areas for histoplasmosis. The main characteristic of these microfoci is contamination with bird/or bat guano, notably caves, but the extent of its natural occurrence remains largely unexplored. Of note, in a recent review, no difference was found between the 2 varieties in terms of environmental and wilderness-related risk factors [38].
Interestingly, only 3 cases of Hcc infection were recorded and none of these cases seemed to be acquired in the RoC. Additional studies are warranted to explore more precisely the conditions leading to infection with one variety rather than the other.
Moreover, 14 cases of Hcd infection were recorded in children and none of them was HIVpositive. This finding agrees with that of previous reports [18,[24][25][26]. A review of the risk factors by Lopez and colleagues revealed that environmental fungal exposure was the most important contributing risk factors to the acquisition of Hcd infection in children [39].
In Hcd infection cases from the RoC, skin lesion including cutaneous or subcutaneous and mucosa involvement and lymph nodes were the most common as noted by Carme and colleagues [14] and Develoux and colleagues [5]. The clinical manifestations of Hcd infection are more commonly the skin, subcutaneous tissues, lymph nodes and bones unlike Hcc infection which generally involves lungs and mucosa and rarely shows cutaneous or bone lesions. The cutaneous manifestations in Hcd infection are papular, nodular, ulcerative, eczematoid lesions [40]. However, the severity of the disease is related to the immunity status of patients. Nine Hcd infection cases (51.3%) in adults were associated with HIV/AIDS in the RoC. Similarly, Oladele and colleagues reported that the disease manifestations vary depending on immune status of patients, the number of fungal particles inhaled (particularly for Hcc), and the virulence of the infective strain [4]. A single RoC case of Hcc-disseminated infection was reported with a painful oral lesion on a background of leukoplakic mucosa [30].
Oral and laryngeal lesions are possible with Hcc infection. Most oral lesions occur in the Hcc-disseminated form of the disease and may affect any area of the oral cavity but are most commonly found on the tongue, palate, and buccal mucosa [41,42]. Patients may present with painful ulcers that persist for several weeks [30,41]. Hcd-disseminated forms (51.3%) were not frequently reported in HIV-infected patients. In the current decades, the rate of disseminated forms (60.6%) has significantly increased without correlation with HIV infection [5]. The bone lesions in Hcd infection are of a lytic type found at cranial, maxillary, femoral, tibial, and spinal bones where they simulated Pott disease (tuberculous osteomyelitis) [28]. All organs and tissues can be involved [43]; indeed, we found that skin, mucosa, bone, and lung were mostly involved in the Hcd infection.
The WHO-endorsed essential diagnostic tests (2019) for fungal diseases are direct microscopy and histopathology, fungal culture, blood culture, and Histoplasma antigen [44]. In the RoC, most of the reported cases were diagnosed by histopathology [20,21,45] and mycological examination (direct microscopy and culture) [14,21]. It is important to recall that due to the coexistence of the two H. capsulatum varieties in Africa, differential diagnosis cannot be made only based on culture, as it is similar for both. Only the demonstration of the ovoid budding yeast (Fig 4) with thick cell walls and much larger (6 to 12 μm in diameter) and intracellular fat droplets supports the identification of Hcd [46]. In addition, Hcc appears as a 2 to 4-μm narrow-based budding yeast on histopathology of a tissue specimen [46]. Thus, the histological appearances of Hcc must be differentiated from other microorganisms, principally Candida glabrata, Cryptococcus neoformans, Leishmania spp., Blastomyces dermatitis, Pneumocystis jirovecii, Talaromyces (formerly Penicillium) marneffei, Toxoplasma gondii, and Trypanosoma cruz [46]. Although, many of these organisms are not endemic and would only be seen in travelers abroad. Mostly, the characteristics that help distinguish these yeasts include predominant cellular location (intracellular for H. capsulatum, predominantly extracellular for C. glabrata), shape and size variation (uniform versus heterogenous), histopathologic response (granulomatous versus suppurative), and culture. Also, the use of specific histochemical stains such as Giemsa, hematoxylin and eosin, Gomori methenamine silver, and periodic acid-Schiff facilitates the differentiation of these pathogens [46]. Detection of Hcc is also possible in approximately 40% of blood smears of patients with disseminated histoplasmosis [43][44][45], though this was not yet reported in the RoC. While antigen detection and PCR emerge as valuable adjunctive diagnostic tools, those are not yet available in the RoC.
The treatment of histoplamosis is a challenge for physicians in most low-and middleincome countries including the RoC, where WHO-listed essential systemic antifungal drugs namely itraconazole, voriconazole, flucytosine, and AmB are not always available. Currently, only fluconazole, ketoconazole (withdrawn across most of the world because of toxicity), topical miconazole, griseofulvin, and nystatin are available in the RoC.
In our review, most patients were treated with intravenous AmB. The lack of itraconazole availability is a serious limitation as only fluconazole can be given for maintenance therapy in the RoC. Fluconazole has a lower efficacy compared to itraconazole in the treatment of histoplasmosis due to its lower bone penetration and lower intrinsic activity [47]. Symptoms of Hcd infection in children are similar to those that occur in adults, with some exceptions, thus treatment recommended for adults is considered for children.

Conclusions
Our study shows that histoplasmosis appears to have a higher rate of prevalence in the RoC compared with other countries in central Africa. Surprisingly, 94.7% of the cases were due to Hcd, and Hcc cases may have been infected out of the RoC. While both varieties cannot be distinguished based on molecular analysis, there is no clear explanation for the reasons leading infection with one or other. Currently the differentiation between the two varieties only relies on direct examination of biopsy specimens or pathologic fluids, and set-up of more standardized techniques allow this differentiation is desirable. This review also highlighted the importance of pediatric cases of histoplasmosis due to Hcd. This observation should also drive future studies to better understand the mode of contamination, which remains uncertain.
While most of the patients were treated with intravenous AmB, itraconazole should be more widely available in Africa as it represents a valuable alternative for ambulatory treatment.