Candida antifungal drug resistance in sub-Saharan African populations: A systematic review

Background: Candida infections are responsible for increased morbidity and mortality rates in at-risk patients, especially in developing countries where there is limited access to antifungal drugs and a high burden of HIV co-infection. Objectives: This study aimed to identify antifungal drug resistance patterns within the subcontinent of Africa. Methods: A literature search was conducted on published studies that employed antifungal susceptibility testing on clinical Candida isolates from sub-Saharan African countries using Pubmed and Google Scholar. Results: A total of 21 studies from 8 countries constituted this review. Only studies conducted in sub-Saharan Africa and employing antifungal drug susceptibility testing were included. Regional differences in Candida species prevalence and resistance patterns were identified. Discussion: The outcomes of this review highlight the need for a revision of antifungal therapy guidelines in regions most affected by Candida drug resistance. Better controls in antimicrobial drug distribution and the implementation of regional antimicrobial susceptibility surveillance programmes are required in order to reduce the high Candida drug resistance levels seen to be emerging in sub-Saharan Africa.

Candida species are known to shift from commensal to opportunistic infectious agents when triggered by factors such as immunosuppression, continuous usage of antibiotics and poor nutrition, leading to increased patient morbidity and mortality [1][2][3] . In severely immunocompromised patients, Candida species can spread through the bloodstream and gastrointestinal tract. This can lead to systemic candidiasis, with a reported mortality rate in developed countries of 38% 4 and 44% 5 . Candida is currently the 4 th most commonly isolated microorganism in nosocomial bloodstream infections 6 and has been implicated in >78% of cancerous and precancerous oral lesions 7 .
Various antifungal drugs with different modes of action have been developed over the years. These include polyene antifungals (e.g. nystatin and amphotericin B), which interfere with ergosterol synthesis, thereby causing cell membrane leakage; the imidazoles (e.g. miconazole, clotrimazole, econazole and ketoconazole), which also interfere with ergosterol and other cell membrane sterol synthesis; the echinocandins (e.g. anidulafungin, micafungin and caspofungin), which inhibit β 1-3 glucan synthesis, affecting the fungal cell wall and 5-flucytosine that in turn interferes with fungal RNA and DNA synthesis 8 . The triazoles (including fluconazole, posaconazole, voriconazole and itraconazole) interfere with the synthesis of ergosterol and have been shown to have fewer side effects than some of the other antifungal drug classes 9 .
Resistance to available antifungal therapies is widespread 10,11 , probably due to the widespread and repeated use of these drugs 12 . Different Candida species have varying resistance patterns, which appear to be geographically determined 13,14 . Therefore early recognition of resistance facilitates the selection of an appropriate antifungal drug, with the use of oral antifungals in oropharyngeal candidiasis reserved for cases where there is no response to topical antifungal treatment 15 . Resistance pattern surveillance to avoid an even higher number of improperly treated, and therefore resistant fungal infections, is imperative 16 . This is a cause for concern in the case of immunocompromised patients, who are at a much higher risk of developing opportunistic complications. Importantly, sub-Saharan Africa is the region most affected by HIV, with approximately 25.8 million infected people in 2014 and accounting for almost 70% of the global number of new HIV infections (http://www.who.int/mediacentre/factsheets/fs360/en/).
Programmes on species prevalence and antifungal surveillance have been successfully developed and introduced in Europe, Asia-Pacific, Latin America and North America [17][18][19] . The gap in antifungal drug resistance surveillance in Africa has been documented 20 . Surveillance programmes are crucial tools in the transition from empirical antifungal treatment, which often does not work due to the diverse resistance levels seen in different regions, and the presence of species that are intrinsically resistant to certain antifungal drugs. The non-existence of routine diagnostics laboratories in most African countries has meant that many African patients are treated without the knowledge of which species they harbour and without any updated guideline data that could be used as a reference in prescribing antifungals. Possible causes for the lack of Candida surveillance programmes in Africa include lack of funding, the limited number of research collaborations and the existence of conflict areas within the continent. This prompted the need for a review of the current situation in Africa regarding the drug susceptibility profiles of Candida species available from different regions.

Data collection
A literature search was conducted on 21 published studies that employed antifungal susceptibility testing on clinical Candida isolates from 8 sub-Saharan African countries, with the aim of identifying antifungal drug resistance patterns within different regions of the subcontinent and included resistance data for 14 antifungal drugs. Searches were performed on PubMed and Google Scholar between August and November 2016 using the keywords 'Candida', 'Susceptibility Testing', 'Drug Resistance' and 'Antifungal'.
Data extracted from the individual studies included the regions within the different countries, patient health information, the methods used for antifungal susceptibility testing, the frequency of Candida species and their susceptibilities to antifungal drugs (Dataset 1 21 ).

Inclusion criteria
We attempted to keep the review recent, and with the exception of a few earlier studies which provided an important comparison, only studies conducted in sub-Saharan Africa between the years 1998 and 2016 employing antifungal drug susceptibility testing were included.

Exclusion criteria
Studies conducted in Africa which reported on the prevalence of Candida but did not describe antimicrobial susceptibility were excluded from this review.

Amendments from Version 1
This version addresses the reviewer comments and includes some minor typesetting and referencing corrections. The inclusion criteria was explained in more detail and the Results and Discussion sections were merged into one. A more detailed explanation on how Candida drug resistance in sub-Saharan Africa differs from the rest of the world is also given in the Results/ Discussion section.

REVISED
This review included seven studies from two regions in South Africa [22][23][24][25]36,40,41  A new multi-drug resistant species, C. auris is taking the world by storm. First discovered in Japan 46 , this species has been found in nine other countries on four continents. The Centre for Disease Control and Prevention (CDC) has issued a warning for increased awareness of C. auris in healthcare settings. This nosocomial pathogen is frequently misdiagnosed, shows resistance to different classes of antifungals routinely administered, and is associated with high mortality rates 47 . The isolation of this species in South Africa 40 appears to be the only report in Africa at the time of writing this paper.
Regional differences in Candida susceptibility profiles have been observed. In South Africa, earlier reports of baseline data demonstrated a high susceptibility (100%) of C. albicans to fluconazole, along with intrinsically resistant non-albicans species 22,24 , with more recent studies in South Africa showing an emerging resistance to azoles 23,25,48 . The reasons for this change in susceptibility patterns is not clear, but it is worth noting that the earlier studies were done before the 2002 introduction of fluconazole as prophylaxis to patients attending HIV-AIDS clinics in South Africa 49 .
Studies from abroad have reported cross-resistance to fluconazole in patients receiving itraconazole prophylaxis 50 and other previously administered azole therapies, such as ketoconazole and miconazole 51,52 . Similar cross-resistance was recently reported in South Africa where 37% of C. parapsilosis isolates were susceptible to fluconazole and voriconazole, and 44% of fluconazole-resistant isolates were voriconazole cross-resistant 41 .
Studies from Bamenda 25 and Douala 26 in Cameroon showed high resistance of C. albicans isolates to azoles (>50% and 70% respectively), with low resistance reported from Mutegene 27 and Bamenda 25 for amphotericin B (4.9 and 4.3%) and 5-flucytosine (10.7 and 6.5%), respectively. The Douala study, on the other hand, reported increased C. albicans resistance to amphotericin B (52.6%) and 5-flucytosine (70% The prevalence of Candida species isolated from sub-Saharan Africa appeared to follow the same trend as in other regions, with C. albicans being the predominant species, followed by C. glabrata. However, in a study from Cameroon 27 and two studies from Nigeria 28,29 , C. tropicalis was the second most prevalent species. When comparing the drug resistance patterns with those in the rest of the world, C. albicans resistance to azoles was seen to be noticeably higher in Southwest South Africa 25 and two distinct regions in Cameroon 25,26 , while C. glabrata resistance to azoles was found to be generally low. However, in a study from South Africa 41 , C. glabrata resistance to fluconazole was found to be similar to the high resistance levels seen in North America 19 . C. tropicalis continent, notably in at-risk population groups, such as HIV patients. With the emergence of inherently drug resistant nonalbicans species, more studies on Candida prevalence and drug susceptibility are needed throughout sub-Saharan Africa. This is most critical in resource-poor areas where there is little or no information available, such as southern (with the exception of South Africa) and central African countries and countries bordering the Sahara.
We would like to conclude by adding that Candida identification to species level is rarely made in clinical settings in Africa, and patients are treated empirically based on their clinical symptoms. The introduction of routine antimicrobial susceptibility testing before initiation of therapy can be relatively expensive, but is certainly a long-term cost effective solution in preventing the progression of drug resistance. Changes in drug susceptibility over time serve as a reminder for the need to test clinical Candida isolates for sensitivity to antifungal drugs in the effort to improve patient care and reduce patient morbidity and mortality. Author contributions CA conceived the study and contributed to the writing of the manuscript. PA wrote the first draft of the manuscript. Both authors were involved in the revision of the draft manuscript and have agreed to the final content.

Competing interests
No competing interests were disclosed.

Grant information
This study was supported by the Research Office of the University of the Western Cape (project registration no. ScRIRC2012/10/72).
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
resistance to fluconazole was seen to be noticeably high in East African countries, with up to 50% resistance seen in Tanzania 31 . Such high levels of resistance have not been documented in other regions.
Fluconazole is widely used in public health settings in the African continent and is used empirically in the treatment of systemic or localized Candida infections 54 , as it is less toxic and regarded as more effective than imidazole antifungals, such as ketoconazole or amphotericin B, even though it is a teratogenic drug 55,56 . Although still somewhat effective in other regions, the use of azoles as first line drugs for systemic infection should be revisited in certain areas of South and West Africa, due to their increasing inefficacy. Regular monitoring of Candida at a regional level could therefore be an important tool to aid in the prescription of antifungals based on the prevalent species and their susceptibilities to antifungal drugs in areas where routine microbiological laboratory testing is not available.
The sale of antimicrobial medications is largely unregulated in Africa and is exacerbated by the influx of fake and adulterated drugs with little or no active ingredients, often available both in pharmacies and on the streets. This problem is aggravated by practitioners who prescribe antimicrobial medications empirically based on clinical presentation, without prior knowledge of which microbial agent(s) are causing infections in their patients. These issues pose a serious public health threat, as they are largely responsible for more and more antimicrobial drugs being rendered ineffective in treating life-threatening infections. This is especially true in the case of the antifungal armamentarium, which is already very limited 57 , especially in resource-poor settings. Limitations of this study include the paucity of available data from African regions, differences in sample sourcing, as well as techniques for isolation and susceptibility profiles from different regions, all of which complicate a comparison of outcomes of the cited studies.
The regional differences in antifungal drug susceptibility of Candida species, often seen within the same country, are an important finding that justifies the implementation of Candida species prevalence and susceptibility testing programmes in the African association between candidal infection and carcinoma, particularly oral I would, however, suggest that, similarly to what authors do in the Introduction, they would mention in Discussion, how is the situation of antifungal drug resistance in sub-Saharan Africa different from Candida that in the rest of the world, in a general outlook and as a result from this literature review.
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. Competing Interests: