Introduction

Invasive fungal infections (IFIs) have turned out to be a major public health concern along with the expansion of at-risk populations besides the lack of universal implementation of definitive diagnostics [1]. Numerous major risk factors for IFIs have been described including surgery, total parenteral nutrition, fungal colonization, renal replacement therapy, hemodialysis, mechanical ventilation, diabetes mellitus (DM), broad-spectrum antibiotics use, red blood cells transfusion, antifungal medication, central venous catheter, and peripheral catheter use [2]. In fact, IFIs can be divided into two categories: endemic mycoses like histoplasmosis and opportunistic mycoses like invasive candidiasis (IC), invasive aspergillosis (IA), cryptococcal meningitis (CM), Pneumocystis jirovecii pneumonia (PJP) [3].

In liver transplant recipients, Pneumocystis is the underlying pathogen in 7% of all pneumonia cases [4]. The European Organization for Research and Treatment of Cancer (EORTC) in a cohort study has indicated that fungemia ranged from 0.15% in patients with solid tumors to 1.55% in hematopoietic stem cell transplantation recipients. It occurred predominantly due to Candida spp. infections (90%), where C. albicans (46.5%), and non-albicans Candida (NAC) (53.5%) were found in patients [5]. IFIs are an important cause of morbidity and mortality among high-risk groups including solid organ transplantation (SOT) recipients and hematological malignancy patients. For instance, mortality rates were the highest for IA (67–82%) as well as cerebral forms of mucormycosis (73.5%) [6].

Even though there are limited choices of antifungals, treating patients with confirmed fungal disease with effective antifungal agents is crucial to reduce morbidity and mortality. Also, several investigations described a significant link between early reliable diagnosis and treatment of IFIs and improved outcomes of patients at risk [7]. The diagnostic includes traditional methods like culture, histopathology, and imaging expertise and newer antigen- and PCR-based diagnostic assays [8].

In this review, we focus on the epidemiology, burden and incidences of IFIs in the Middle Eastern and North African (MENA) region among high-risk groups, to support infectious disease specialists and healthcare workers in this geographic area and assist the provision of optimal care for patients susceptible to IFIs.

Epidemiology of invasive fungal infections in the MENA region

Since the increase of IFIs is strongly associated with the expanding immunosuppressed population and the increase in invasive diagnostics and treatment, an urgent need for surveillance of the changing trends in incidences is required. The knowledge of the current situation allows the assessment of the burden of such infections in the region. Thus, PubMed, Science Direct, Scopus, and Google Scholar databases search was done for epidemiological studies of IFIs from tertiary care hospitals published in the last decade. We used a combination of the keywords for paper retrieving including the following: “invasive fungal infections,” “invasive fungal disease,” “invasive candidiasis,” “candidaemia,” “invasive aspergillosis,” “pneumocystis pneumonia,” “mucormycosis,” “histoplasmosis,” in addition to a MENA country. Indexed original articles and case reports in English and French of any design and sampling strategy were included. Despite the globally growing importance of invasive infections, especially among the high susceptible risk groups, the epidemiological assessment of the status of IFIs is underestimated in the MENA region. Indeed, only very few reports about the estimation of IFIs were found in this region in the last decade. In the next parts of this review, we will discuss the available data concerning IC, IA, CM, Pneumocystis pneumonia, mucormycosis, and histoplasmosis in the region.

Invasive candidiasis

Candida infections accounts for approximately 70 to 90% of total IFIs [9]. Global estimates indicated that ~ 750,000 cases of IC occur annually [10]. Candidemia (Candida bloodstream infection) is the most common clinical presentation of IC and occurs mainly in hospitalized patients with an ascribable mortality of 15–35% for adults and 10–15% for neonates [11]. Only five species contribute to almost 92% of cases of candidemia: Candida albicans, Candida glabrata, Candida tropicalis, Candida parapsilosis, and Candida krusei. C. albicans is the most common etiological agent worldwide [11]. However, an upward trend in the incidence of NAC in IC cases was witnessed worldwide, which may be correlated with an increasing use of triazoles, mainly fluconazole [12]. Furthermore, a recent emerging multi-drug resistant Candida species, Candida auris, has been reported to cause healthcare-associated fungal infections [13]. In comparison with Candida species, several characteristics make the opportunistic C. auris unique in the field of clinical mycology such as his ability to colonize inert surfaces, capacity to cause nosocomial invasive infections, resistance to some commonly used chlorine-based disinfectants, and non-susceptibility to any or all of the systemic antifungal drugs available at this time [14, 15].

Epidemiological studies assessing the status of IC, including candidemia, are underestimated in the developing world, counting MENA countries (Table 1). Several studies have estimated the incidence rates of candidemia in MENA countries. Candidemia incidence rate was estimated to be the highest in Qatar, with an estimated rate of (15.4/100,000) [16] and the lowest in Iran (0.34/100,000) [17] (Fig. 1). IC and other IFIs were compared with reported estimations in different countries at global level (Fig. 2).

Table 1 Epidemiology of invasive candidiasis in the MENA region
Fig. 1
figure 1

Available estimates of the incidence rate (per 100,000 population) of invasive fungal infections in MENA countries [16,17,18,19,20,21,22]

Fig. 2
figure 2

Available estimates of the incidence rate (per 100,000 population) of invasive fungal infections in different countries around the world [23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38]

As for the available epidemiological studies, only a few were found to assess the status of IC and candidemia. In Lebanon, only one study took place in the past 10 years. The study evaluating 2011–2012 data of three hospitals, with a mean of 2.2–2.5 co-morbid conditions per patient, has reported the incidence of IC to be 0.39 cases per 1000 hospital discharges [39]. In Jordan, an overall incidence rate of 0.48 episodes/1000 admissions was recorded in an academic tertiary hospital, with a 30-day mortality rate of 38.7% [40]. In Turkey, several studies have shown that nosocomial IC, including candidemia, ranged between 1.2/1000 and 5.6/1000 admissions [41,42,43]. The incidence of candidemia in ICUs ranged from 1.76 patients/1000 ICU admissions to 11.5 per 1000 neonatal ICU admissions among neonates [44, 45]. However, a prospective study in Iran indicated higher incidence rate of candidemia (15.2/1000 in neonatal ICU admissions), with a 42.5% mortality rate [46]. A systematic review and meta-analysis identified 55 cases of candidemia in Iran where the risk factors were surgery and burns (23.6%), malignancies (20%), use of broad-spectrum antibiotics (18.2%), and diabetes (7.3%) [47].

As for the Arabian Peninsula, it has been recorded in Saudi Arabia that IC rates ranged from 1.55–1.65 cases per 1000 discharges to 26 cases per 1000 ICU admissions [39, 48, 49]. A retrospective study on IC among pediatric patients indicated a group of risk factors: prematurity in 37 (28.7%) of patients, low birth weight (32.6%), central venous catheter (45.7%), malignancy (16.3%), immunotherapy (15.5%), and ventilator support (46.5%) [50]. In 2002–2009, data indicated that malignancy was independently associated with the development of candidemia [51]. In Kuwait, candidemia rate has decreased to a 0.15 cases/1000 in 2016 compared to 0.24 cases/1000 patient-days in 2014 [52]. In Qatar, 201 episodes of candidemia in 187 patients were identified in a single-center study [53].

In North Africa, candidemia in a pediatric ICU reached up to 3 per 1000 inpatient-days in Egypt [54]. A 15-year (1995–2009) retrospective analysis in the Sousse Region, Tunisia, had indicated an increase in the frequency of IC episodes, with an average of 24 episodes per year [55]. A previous study analyzing the data from 1995 to 2010 in the same region has shown an incidence of neonatal IC about 12.2 cases/1000 admissions, with a fourfold decrease of incidence from 2007 to 2010 [56]. A retrospective study of IFI among renal transplant recipients from January 1995 to February 2013 reported only 2 cases of IC [57]. In Morocco, 8-month prospective surveillance from January to August 2011 conducted at the Casablanca University Hospital—pediatric hematology/oncology unit—showed a healthcare-associated infections (HAIs) incidence of 28/1000 patient-days, where Candida accounted for 14% [58].

Some factors may have a probable role in the spread of IFIs among MENA countries. For instance, country socioeconomic status affects device-associated infection rates in developing countries and should be taken into account when comparing device-associated infections from one country to another [59]. In addition, the global share of the health research in the Eastern Mediterranean region is lower than the average percentage worldwide. The rise in academic health publications has been more common in only a few countries [60]. Furthermore, socio-political instabilities influence research output as well, such as in Egypt and Tunisia in the last few years. The concentration of biomedical and health research in the region’s academic institutions is expected to help turn information into public health results, if more suitable conditions are given [60].

Based on data published in the last decade regarding IC causative agents, C. albicans was found to be the most prevalent in Iran (80%) in a local study conducted in Tehran, followed by Turkey (48.3%), Kuwait (37.22–47.2%), and Qatar (30.2%) [61, 62]. According to Ghazi et al. [62], an increase in NAC is being observed in the MENA region, especially in Saudi Arabia (48.1%), Kuwait (52.8%), Egypt (60%), Qatar (69.3%), and Tunisia (~ 76.1%). A decrease in the NAC incidence was observed in Turkey (from 52.43 to 44%). With regard to the NAC distribution, C. tropicalis was prevalent in Saudi Arabia (15.5%), and Tunisia (37.7%), while C. glabrata predominated in Qatar (25.5%), Turkey (13.3%), and Iran (20%) and C. parapsilosis in Kuwait (38.2%) and Egypt (25%) [62].

Concerning C. auris, it was associated with IC, mostly candidemia, in the United Arab Emirates, Kuwait, and Saudi Arabia [52, 63,64,65,66,67,68]. Surprisingly, while C. auris isolation in Kuwait has steadily increased over a study period and almost doubled in 2018 compared to 2017, a noteworthy finding of nearly sixfold increase was in bloodstream C. auris isolates in the same period [68]. Unfortunately, no reports of C. auris were found form other MENA countries. In fact, C. auris is closely related to the species of the C. haemulonii complex, leading to a misidentification with other Candida species such as C. haemulonii or C. famata [69, 70]. Even though most of the region belong to the middle-income countries, the total mass of the publication of research is lower than its global proportion of population or income [60]. Most of the clinical microbiology laboratories in this geographic area are not yet supported with more advanced diagnostic tools that allow the extensive search for this agent, in contrary to the high-income countries, which reported and investigated to a certain extent the C. auris distribution. C. auris infection may be present in other countries as well, but its occurrence has not yet been investigated. Unfortunately, the insufficiency of conventional commercial systems to identify C. auris also leads to delayed intervention and treatment [70, 71]. This will definitely cause a threat in the healthcare settings, since this fungal agent can contaminate and prevail in hospital environments, in addition to transfer between patients, and from health workers and abiotic surfaces of medical equipment to patients [72].

Invasive aspergillosis

IA is an opportunistic fungal infection affecting primarily the immunocompromised population: 70% of the total putative or proven IA happens in immunocompromised patients, in patients with chronic obstructive pulmonary disease (COPD), or patients with diabetes admitted to the ICUs [73]. It is estimated that ~ 250,000 cases occur annually [10]. IA is associated with elevated hospital mortality, extended duration of hospitalization, and high costs [74]. The causative agent Aspergillus spp. is a ubiquitous environmental mold that is found on organic matter, soil, and in the air as conidia [75]. Aspergillus fumigatus, Aspergillus flavus, Aspergillus niger, Aspergillus terreus, and Aspergillus versicolor are the most common causative species, where A. fumigatus is the commonest agent worldwide [76]. Despite the growing concern of aspergillosis, it is not thoroughly reported in the majority of MENA countries (Table 2).

Table 2 Epidemiology of invasive aspergillosis in the MENA region

IA burden estimates have been done for several MENA countries, and the highest estimated incidence rate in Egypt is 10.7/100,000 [18], followed by Iran with 7.9/100,000 among COPD cases, 0.1/100,000 among lung cancer cases, and 0.03/100,000 among acute myeloid leukemia cases [17]. In Saudi Arabia, incidence rate was 7.6/100,000 [19], followed by Algeria (7.1/100,000) [20], Turkey (4.48/100,000) [21], Iraq (2.62/100,000) [77], Jordan (1.34/100,000) [22], and reached its lowest value (0.6/100,000) in Qatar [16]. Moreover, the incidence of IA in Lebanon was estimated at 1.21 cases/1000 hospital discharges, higher than that found in Saudi Arabia with 0.4 cases/1000 hospital discharges [39]. In Bahrain, medical records during 2009–2013 of a tertiary care hospital of patients with positive Aspergillus cultures revealed 53.3% colonization and 46.7% presumed IA associated with 25% and 32% mortality respectively [78]. In Tunisia, the incidence of IA was 7.5–15% among hematology patients [79,80,81]. A retrospective study of IFIs among renal transplant recipients at Habib Bourguiba Sfax university hospital from January 1995 to February 2013 reported 2 cases of aspergillosis [57]. Between 2002 and 2010, 29 cases of IA were reported in the Sousse Farhat Hached Hospital Hematology Unit, where acute myeloid leukemia (AML) was the most common disease (65.5%) among the severely neutropenic patients [82].

Regarding species distribution, A. fumigatus predominated in IFIs of hospital discharges in both Lebanon and Saudi Arabia (60%) [39] and was responsible for 53% of IA in Bahrain [78]. A. flavus was predominant in 75% of bronchoalveolar lavage (BAL) samples in Iran [83] and in 37.5% of sputum samples and 79.2% among clinical samples of hematology patients in Tunisia [81, 84].

Cryptococcal meningitis

Cryptococcus neoformans is the most common fungal agent that causing meningoencephalitis in the immunocompromised individuals worldwide [85], followed by Cryptococcus gatti [86]. C. neoformans is acquired through inhalation of spores or dried agent in the environment [87]. CM arises in 15% of AIDS-related mortality worldwide [88]. It is estimated that ~ 223,100 incident cases occur globally, with an estimation of 181,100 annual deaths [88].

Considering CM as a rare fungal infection, it could not be estimated in Jordan [22], Iraq [77], and Egypt [18], in addition to being estimated to probably be affecting under 10 patients in Saudi Arabia [19]. In Qatar, CM had a low estimated incidence rate of 0.43 cases per 100,000 considering the low human immunodeficiency virus (HIV) rate in the country [16]. In Turkey, the estimated rate was 0.13 per 100,000 population with the prediction of 106 annual cases occurring among HIV patients [21]. Similarly, the incidence rate in Iran was estimated at 0.14 per 100,000 population [17]. In Algeria, a nearly negligible rate of 0.09 per 100,000 population was estimated [20].

Overall, few cases of CM were reported in the MENA region. A case was observed in a Turkish female who has undergone mastectomy and has had received chemotherapy following surgery [89]. In addition, a rare case of tenosynovitis caused by Cryptococcus luteolus was reported in a 68-year-old type 2 diabetic male [90]. Regarding Iran, a review article reported a case of CM caused by C. neoformans in HIV-positive Iranian female who died 4 days after antifungal therapy due to respiratory failure in Sari. The authors have also included 12 other reported cases of cryptococcosis due to Cryptococcus species between 1969 and 2014 [91]. Moreover, a study showed that the prevalence of cryptococcal infection in HIV-infected patients with a CD4 cell count of < 100 cells/mm3 was very low (< 3%) [92]. In fact, C. neoformans is highly prevalent in pigeon droppings in Iran, allowing its transmission to the environment and consequently to the patients at risk of developing CM [93, 94].

The first report of CM in the Arabian Peninsula was in Saudi Arabia in 1990, showing the occurrence of this disease in a child having systemic lupus erythematosus [95]. The only reported case from Kuwait was in 1995, considered the second in the Arabian Peninsula, in a 22-year-old man presented with confusional psychosis caused by C. neoformans serotype A [96]. However, in Oman, a study conducted between January 1999 and December 2008 showed that CM accounted for 22% of AIDS-defining opportunistic infections [97].

Regarding the North African region, 0.6% of CM cases were estimated among annually HIV-infected patients in Egypt [18]. The first cases of CM were reported in two HIV-positive patients in Libya [98]. In Morocco, a study that was conducted from January 2005 to May 2015 showed that CM accounted for 20% of neurological disorders involved that was suggestive of HIV infection in 68.8% cases [99]. A 2016 cross-sectional study in Tunisia declared that CM was responsible for 11% out of 70.4% deaths associated with HIV [100]. Among renal transplant recipients between January 1995 to February 2013, 2 cases of cryptococcosis were reported [57].

Pneumocystis pneumonia

Pneumocystis spp., a ubiquitous yeast-like fungus that affects a wide variety of mammals, is a host strict agent where P. jirovecii infects primarily humans [101]. The main transmission route is through inhalation; however, some evidence suggests a direct-contact transmission route [102]. PJP is widely known as an HIV-positive-related opportunistic fungal infection. However, because of the extensive use of immunosuppressant drugs, non-HIV risk groups have emerged, particularly among cancer patients [103].

Concerning incidence rate estimates, it could not be estimated in both Iraq and Saudi Arabia due to the lack of data on PJP infections [19, 77]. In Qatar, the estimated rate was 0.8 per 100,000 population [16], followed by Turkey 0.79 per 100,000 [21], Algeria 0.18 per 100,000 [20], Egypt 0.15 per 100,000 [18], and Jordan 0.1 per 100,000 [22] (Table 3).

Table 3 Epidemiology of Pneumocystis jirovecii pneumonia in the MENA region

Case reports and epidemiological data on PJP were reported in few countries. A Lebanese report indicated a prevalence of 10.9% PJP cases among HIV patients [104]. In Turkey, the prevalence of PJP among AIDS/HIV patients ranged from 10 to 46.7% [104,105,106], while in Iran, PJP ranged from 4.5 to 39.3% [107,108,109]. Regarding the Arabian Peninsula region, PJP occurring in 9 episodes among primary immunodeficiency disorder patients in Kuwait [110] was found in 5.1% of total HIV-infected patients in Bahrain [111], while accounted for 25% of opportunistic infections in HIV/AIDS patients in Oman with no mortality reported due to this infection [97]. Concerning North Africa, a retrospective analysis in 2013 conducted on HIV-related hospitalizations showed that PJP was responsible for 8.8% respiratory diseases in Libya [112] and accounted for 11.1% death in HIV-infected patients in Tunisia [100]. Another retrospective study from January 1995 to February 2013 on IFIs reported 4 cases of pneumocystosis among renal transplant recipients [57].

Mucormycosis

Mucormycosis is an IFI ordinarily seen in individuals with underlying predisposing risk factors including DM and hematological malignancies. Disseminated forms are usually seen in individuals with such risk factors, although rhino-sinusoidal and cutaneous forms may occur in all individuals [113]. Causative agents belong to the subphylum Mucoromycotina. Rhizopus arrhizus is the most common agent causing mucormycosis globally. It is acquired through inhalation of sporangiospores [114]. The global estimates of mucormycosis were around > 10,000, with an estimate of disseminated mucormycosis around 100,000 [10].

Estimated incidence rates were scarce. The highest estimated rate was in Iran at 9.2 per 100,000 population [17], then in Iraq and Algeria at 0.2 per 100,000 [20, 77]. In Lebanon, an average incidence of 0.83 cases/10,000 admissions (range 0–2.22) was seen over a 10-year period in hospitalized patients, in addition to 20% mucormycosis-related deaths [115]. As for Turkey, 223 cases occurred in hematological malignancies patients and 214 cases in patients with diabetic ketoacidosis [21]. A 17-year assessment indicated 60% cases of rhino-cerebral and 1.9% disseminated forms of mucromycosis diagnosed in 151 patients, with 49% diabetes 39.7% hematological malignancies co-morbidities among them [116]. In Iran, a systematic review indicated that the commonest form was rhino-cerebral (48.9%) in addition to eight disseminated forms were identified which accounted for 75% of total mortality [117]. Another 10-year assessment indicated the prevalence of the rhino-cerebral form (95%) among the total diagnosed mucormycoses [118]. In addition, a case of rhino-orbital mucormycosis caused by Rhizopus oryzae was reported in an acute lymphoblastic leukemia patient [119]. Among children with hematological malignancies, mucoromycetes were responsible for 11.5% of IFIs and accounted for 53.3% mortality [120]. An overall incidence rate was estimated at 4.27 per 100 leukemia patients, which decreased from 2001 to 2011 [121].

Concerning the Arabian Peninsula, mucormycosis was estimated to probably being affecting less than 10 patients in Saudi Arabia [19]. A retrospective analysis has reported rhino-cerebral mucormycosis as a cerebral venous thrombosis (CVT) [122]. Gangrenous necrotizing cutaneous mucormycosis in a female neonate who has had injuries on the scalp, abdomen, and back was reported in an ICU in Oman [123].

In North Africa, proven mucormycosis developed in 45 patients in which 90% of the cases had hematological malignancies in Egypt; in addition to mucormycosis-related complications in 5 cases with disfigurement and perforated hard palate. Mucormycosis-related mortality was 33% [124]. In December 2010, 3 out of 5 acute leukemia patients due to Rhizomucor outbreak [125]. Another study during 2010 reported 10 cases of mucormycosis, in which 80% had pulmonary mucormycosis and only 20% sinus involvement [126]. As for Tunisia, two patients had died after reporting rhino-cerebral, rhino-orbital, auricular, pulmonary, and cutaneous mucormycosis in one acute leukemia and 3 diabetic patients. The responsible Mucorales were R. arrhizus in 3 cases and Lichtheimia corymbifera in 2 cases. [127]. One mucormycosis was reported among 321 renal transplant recipients [57]. Surprisingly, considering the rarity of hemophagocytic lymphohistiocytosis, it was reported to be complicated by sinonasal mucormycosis in a diabetic child [128]. Retrospective analysis of data between 1992 and 2007 included 17 diabetic patients with rhino-orbito-cerebral mucormycosis which resulted in 65% mortality primarily due to delay in diagnosis and the lack of surgical treatment [129] (Table 4).

Table 4 Epidemiology of Mucormycosis in the MENA region

Histoplasmosis

Histoplasma capsulatum, a dimorphic fungus and causative agent of human histoplasmosis, causes respiratory and systemic mycoses in immunocompromised individuals [130]. The global burden of disseminated histoplasmosis is estimated to be ~ 100,000 annual cases [10]. The co-occurrence of histoplasmosis and tuberculosis in advanced HIV poses a diagnostic problem especially in histoplasmosis endemic regions [131]. Histoplasmosis is a known endemic mycoses of North America [132]; however, some reports were made from North Africa. In Morocco, a 39-year-old HIV-infected patient presented sinusitis and cutaneous histoplasmosis [133]. As well, a case of disseminated histoplasmosis caused by H. capsulatum was diagnosed in the bone marrow of a 34-year-old HIV-infected woman at Rabta Hospital of Tunis [134].

Concluding remarks

Despite the limited number of investigations dealing with the epidemiology of IFIs in this geographic area, the currently available data underlines that IFIs are not negligible. There is an urgent need for surveillance and implementation of recommendations and procedures in tertiary care centers in order to stem the morbidity and mortality related to IFIs, particularly among high-risk populations. Furthermore, a regional reference center for invasive mycoses and antifungals at the MENA region, with broad missions, is needed to promote the implementation of antimicrobial stewardship programs in this geographical area.