Aspergillus fumigatus carrying TR34/L98H resistance allele causing complicated suppurative otitis media in Tanzania: Call for improved diagnosis of fungi in sub-Saharan Africa

Suppurative otitis media (SOM) is characterized by the inflammation of the middle ear and mastoid, tympanic membrane perforation as well as discharge [1]. The tympanic membrane perforation may result in increased exposure of the middle ear to pathogens [2, 3]. In developing countries, SOM is a major cause of preventable hearing loss [4‐6], its incidence ranges from 7 to 46 % and is common amongst children of lower socio-economic status [7, 8]. In Tanzania, SOM constitutes a major cause of otorhinolaryngology clinic visits and contributes significantly to high morbidity and long term hearing loss [9‐11]. While bacterial pathogens (most prominently Pseudomonas aeruginosa and Staphylococcus aureus [12‐15]) have commonly been documented as the cause of SOM in developing countries, the role of fungal infections among patients with SOM is still underestimated [3, 16, 17]. Fungal infections among patients with SOM account for 2.1–25 % of cases [12‐14] and Aspergillus spp. are the commonest cause accounting for 92.1 % of cases [12, 18].

Fungal infections are mainly attributed by compromised immune status, prolonged antibiotic use and immune suppressive therapy [19, 20]. This aspect of mycosis has gained importance over the recent years because of the excessive use of broad-spectrum antibiotics, and an increase in the number of immunodeficiency conditions such as immunoglobulin deficiency, malignant neoplasms, immunosuppressive therapy (corticosteroids and cytotoxic chemotherapy), diabetic mellitus as well as AIDS [3, 8, 21].

Fungi have been least documented as the cause of SOM in Africa probably due to lack of laboratory diagnose for fungal infections in these settings. Here, we determined the prevalence of fungal occurring in the otolaryngology clinic and surgical wards of Bugando Medical Centre (BMC), Mwanza Tanzania complicating cases of SOM.

Out of 410 patients with SOM, 44 (10.7 %, 95 % CI 7.7–13.7) had significant fungal growth. The studied cases are summarized in Table 1. The median age (inter quartile range (IQR) of patients with fungal infections was 29.5 (IQR 16–43) years. Male formed majority of patients with fungal infections (n = 24, 54.6 %). The median duration of illness was 18 (IQR 5–24) weeks. Of the 44 patients with fungal infections, 6 (13.6 %) were HIV positive. On Mann Whitney ranksum test, there were non-significant difference in the median duration of illness between patients with fungal infections and those without fungal infections (18, IQR: 5–24 vs. 12 IQR: 5–24, p = 0.5749).

Out of the 44 positive fungal cultures, 35 (79.6 %) produced pure growth of only one fungal species, while nine (20.4 %) had mixed fungal growth. A total of 23 (52.3 %) patients were positive for moulds and 21 (47.7 %) positive for yeast. Candida albicans was the most commonly isolated fungal (n = 13, 29.6 %) followed by Aspergillus versicolor (n = 8, 18.2 %) (Table1). A total of seven (15.9 %) patients had disease complication at time of enrollment; of them six (13.6 %) had hearing loss. On follow up, seven (15.9 %) had poor treatment outcome. Out of 366 patients with no fungal growth; 30 (8.2 %) had hearing loss as compared to 7/44 (15.9 %) of those with fungal infections (p = 0.0339).

All five strains of Aspergillus fumigatus isolates were resistant to itraconazole and showed reduced in vitro susceptibility to both voriconazole and posaconazole. Sequencing of the cyp51A locus revealed these isolates to carry the TR₃₄/L98H resistance allele. Penicillium citrinum isolates were resistant to voriconazole and had reduced susceptibility to itraconazole and posaconazole. All Penicillium sumatrense were resistant to both itraconazole and voriconazole, and showed reduced susceptibility to posaconazole (Table 2).

There were non- significant association of the poor treatment outcome with increase in age, female gender, being HIV positive and being infected with moulds Table 3.

SOM is highly associated with lower social economic status, potentially due to poor hygiene, low access to medical care, or lack of knowledge [28]. In the current study, the majority of patients with positive fungal growth were from rural area where there are no hospitals with otolaryngology services. Most of these patients presented in BMC after six months of illness.

As seen in other studies [10], C. albicans and Aspergillus spp. were the most commonly isolated fungal from SOM specimens. The ability of C. albicans to maintain the synergistic relationship with bacterial pathogenic flora of skin like S. aureus and ecological niche of these isolates may explain the findings [29, 30].

Fungal have mainly been documented as opportunistic pathogens causing infections in immunocompromised patients. Among the 44 fungal-infected cases of SOM studied here, only six (15.9 %) were HIV positive, which increased the risk of getting poor treatment outcome of SOM by 1.2 fold. The immunocompromised state of these patients could explain this observation. However, this observation was not statistically significant. This necessitates the need to identify other risk factors associated with fungal infections among HIV-negative individuals in future studies.

In the current study, with eight isolates, A. versicolor was the predominant Aspergillus species. A. versicolor is a highly resilient/resistant fungus found in damp indoor environment, able to produce hepatotoxic and carcinogenic mycotoxin sterigmatocystin [31, 32].

Patients infected with moulds had 5.5 times higher risk of getting poor treatment outcome than patients with Candida infections. This could be explained by the fact that the majority (in fact all of the A. fumigatus, P. sumatrense, and P. citrinum isolates) were resistant to at least one commonly used antifungal agent. For the past two decades the increasing occurrence of A. fumigatus isolates resistant to common azoles antifungal agents has been described, including in Tanzania [33]. In our study, all A. fumigatus isolates indeed carried the TR₃₄/L98H allele, which shows that the environmental occurrence of these isolates is also clinically relevant in Tanzania. Additionally, all isolates of P. citrinum and P. sumatrense were resistant to the lead antifungal agent itraconazole or showed reduced susceptibility to voriconazole. However in these species the underlying resistance mechanism is unknown.

In most clinical settings in resource-constrained countries like Tanzania where fungal diagnostics is underdeveloped; treatment of fungal infections relies solely on the empirical use of topical azole agents, which has previously been reported to be in effective for A. versicolor [34], the most frequent moulds isolated here and is probably similarly ineffective for azole-resistant A. fumigatus, a significant emerging problem.

In the current study 15.9 % of patients infected with fungi developed complications, mostly irrevocable hearing loss, proving that hearing loss is a major complication developed by patients with suppurative otitis media [35]. The clinical impact of fungal infections in patients with SOM should drive the effort to improve fungal diagnostics in developing countries.

One of the limitations of this study is the possibility of skin flora contamination during sample collections. However; the quantification of fungal growth significantly minimized the chances suggesting that majority of patients had real fungal infections.

Fungal infections played a significant role in SOM pathology under low resource settings. Diagnosis of fungal infections in developing countries should be improved so that appropriate management can be initiated on time to prevent associated complications.