Insights into the global emergence of antifungal drug resistance

The global prevalence of fungal diseases has escalated in the last several decades. Currently, it is estimated that fungi infect 1.7 billion people annually and result in 1.5 million deaths every year¹. Deaths due to fungal infections are increasing, with mortality often exceeding 50%, further increasing to 100% if treatment is delayed¹. Despite these staggering figures, the contribution of fungal infections to the global burden of disease remains under-recognised. In Australia, over a 5-year period fungal infections cost Australia an estimated $583 million². The median cost for one invasive fungal disease (IFD) is AU$30 957, increasing to AU$80 291 if the patient is admitted to an intensive care unit³. Treatment of fungal infections poses significant challenges due to the small number of safe and effective antifungal drugs available and emerging antifungal drug resistance. Resistance to every class of antifungal drugs has been described and for some drug classes is extremely common^4,5.

More than 90% of all reported fungal deaths result from species that belong to one of three genera: Candida, Aspergillus and Cryptococcus¹. Candida species normally live on the epithelial surfaces of the host and are common in immunocompromised patients who have undergone invasive procedures and can form biofilms on medical devices¹. Candida albicans is the most prevalent species causing disease, followed by C. glabrata, C. tropicalis, C. parapsilosis and C. krusei (recently renamed C. acidothermophilum)⁴. Candida auris is an emerging fungal pathogen that has intrinsic multi-drug resistance. Since first emerging in 2009, C. auris has spread across five continents, including Australia, and resulted in a number of nosocomial outbreaks^6,7. Cryptococcosis, caused by Cryptococcus neoformans and Cryptococcus gattii, is the most common invasive fungal disease globally, at one point with an estimated >1M life-threatening infections per year¹. C. neoformans, isolated from soil or bird excrement, has a worldwide distribution and predominantly causes infections in immunocompromised individuals. C. gattii is isolated from eucalyptus and other trees, so has a more restricted distribution, and causes infections largely in immunocompetent individuals⁸. Aspergillus species commonly cause invasive pulmonary aspergillosis in neutropenic patients, transplant recipients and in patients on immunosuppression (e.g. corticosteroids). Aspergillus species are found worldwide and are common in the environment resulting in continuous lung exposure¹. The most common species are in the A. fumigatus complex, but A. flavus, A. niger, A. terreus and A. nidulans also cause infections in humans.

There are four main classes of antifungal drugs: polyenes, azoles, allylamines and echinocandins (Figure 1, Table 1). Polyenes (e.g. amphotericin B) bind ergosterol in the cell membrane and induce pore formation. Azoles (e.g. fluconazole, voriconazole) and allylamines (e.g. terbinafine) inhibit the enzymes lanosterol 14 α-demethylase (ERG11p/cyp51) and squalene epoxidase, respectively, which are needed for ergosterol biosynthesis in the fungal cell membrane. Echinocandins (e.g. caspofungin) inhibit the synthesis of glucan in the fungal cell wall. Echinocandins have few side-effects but are poorly absorbed and have limited efficacy (e.g. none against C. neoformans)²². Closely related species can exhibit differences in intrinsic susceptibility. For example, 98% of C. albicans isolates are susceptible to fluconazole whereas only 9% of C. krusei are susceptible⁴. Resistance to every family of antifungal drugs has been described and is at least partially responsible for the poor outcomes and high morbidity seen in patients with IFD. Resistance to azoles is common and increasing⁵.

Figure 1.  Mechanisms of action of antifungal drugs. Schematic of the fungal cell wall and plasma membrane. Cell wall components are indicated as follows: blue: chitin, light green: ß-1,3-glucan, dark green: ß1,6- glucan, yellow: proteins, red: α-1,3-glucan and orange: mannans or galactomannans (orange), adapted from Brown et al.¹. Echinocandins inhibit ß1,3-glucan synthase (FKS1/2) to prevent the synthesis of ß1,3-D-glucan in the fungal cell wall. Azoles and allylamines inhibit lanosterol 14α-demethylase and squalene epoxidase, respectively, to inhibit the formation of ergosterol in the cell membrane. Polyenes bind to ergosterol in the cell membrane, causing pores to form and membrane leakage.

Table 1.  Targets of antifungal drugs and resistance mechanisms.

The prevalence of antifungal resistance in Candida is not clear, mainly because of the differences in the levels of intrinsic susceptibilities between Candida species. Primary resistance to fluconazole is rare for C. albicans (1.4%), C. parapsilosis (3.6%) and C. tropicalis (4.1%) but species such as C. krusei are intrinsically resistant to fluconazole (78.3%)¹³. In a recent United States retrospective study acquired fluconazole resistance was reported in 19% of C. albicans infections¹³. At present, resistance to azoles is uncommon in Australia but may be increasing; a recent review of candidemia in Australia showed 16.7% of C. tropicalis isolates are resistant to both fluconazole and voriconazole²³. A major concern is the emergence of infections caused by Candida species other than C. albicans, such as C. auris that has intrinsic multi-drug resistance. Currently C. auris is not established in Australia: the few Australian cases have come about through people being infected overseas (Victoria State Government, August 2018)⁷. Vigilance is warranted towards preventing this new multi-drug resistant pathogen from establishing a foothold in Australia.

A recent systematic review of resistance in Cryptococcus species from 1988–2017 revealed 10.6% of clinical isolates are fluconazole resistant and this rises to 24.1% in patients with relapsed cryptococcal infection²⁴.

The first few instances of azole resistance in A. fumigatus were reported in the USA and Europe in the late 1980s and 90s in association with long-term antifungal treatment (extensively reviewed by Meis et al.²⁵). Since then, the prevalence of resistant isolates has increased and has been found in patients who have not received azole treatment. The acquisition of resistance in patients who have not received azole treatment has led to the hypothesis that resistance can also be gained from the agricultural use of azoles¹⁷. In 1999, the prevalence of azole-resistant A. fumigatus in The Netherlands was 12.8% but current incidence is 20% and resistance has now been reported in 11 different countries^5,26.

The emergence of antifungal drug resistance can occur due to changes in the genome^9–11. On exposure to azole antifungal drugs, fungal pathogens can undergo a process termed heteroresistance in which rapid, yet reversible, resistance is conferred by the development of one or more aneuploidies (large scale chromosomal rearrangements) (reviewed by Morrow and Fraser¹²). Cells return to normal ploidy when the azole is removed, as the aneuploidies result in strains with reduced fitness both in culture and in the host¹⁰. In C. albicans, the most commonly occurring aneuploidy is an isochromosome of the left arm of chromosome 5, i(5L). This region contains the ERG11 gene encoding the target of fluconazole, as well as the TAC1 gene encoding the transcriptional activator of drug efflux genes⁹. In C. neoformans, chromosome 1, containing ERG11 and the azole transporter AFR1, is duplicated in response to increasing concentrations of fluconazole, followed by the successive duplication of chromosomes 4, 10 and 14¹⁰.

Resistance to azole drugs in C. albicans, C. parapsilosis, C. krusei, C. tropicalis,C. neoformans and A. fumigatus can also be acquired by single nucleotide mutations in the ERG11 gene (cyp51) encoding the target enzyme 14 α-demethylase^14,15. Mutations that lead to overexpression of the efflux pumps encoded by MDR (major facilitator superfamily pumps) in C.albicans and C. parapsilosis or CDR genes (ATP-binding cassette pumps) also confer resistance in C. albicans, C. glabrata and C. krusei⁴. Mutations in C. glabrataPDR1, a transcription factor regulating the expression of drug efflux pumps, also confers resistance to fluconazole¹⁹. C. auris clinical isolates possess acquired mutations in ERG11 (azole resistance), FKS1 (echinocandin resistance) and FUR1 (5-flucytosine resistance)^6,27. Candida species become resistant to echinocandins due to mutations in the FKS genes encoding the two subunits of the target enzyme 1,3-ß-D-glucan synthase, specifically FKS1 in C. albicans, C. krusei and C. tropicalis and both FKS1 and FKS2 in C. glabrata⁵.

In A. fumigatus, the most common mechanism of azole resistance occurring within a host involves mutations in the cyp51A (ERG11) gene, which prevent the azole from binding to the heme molecule^16,25. In contrast, A. fumigatus resistant clinical strains acquired from the environment most commonly possess a duplication of a 34 bp tandem repeat in the promoter region of cyp51A, combined with a specific substitution that causes overexpression of the gene^5,18. These mutations confer multi-drug resistance and have been found to be correlated with exposure to agricultural azoles in the environment^16,17.

In comparison to resistance to azoles, almost nothing is known about the mechanisms that give rise to resistance to polyenes in fungi. Strains naturally resistant to amphotericin B exist but resistance can also develop on treatment. In Candida species, acquisition of resistance to amphotericin B can occur through mutations of some of the ERG genes of the ergosterol biosynthesis pathway, although most resistant strains have only been characterised by biochemical analysis of membrane sterol composition. The only C. neoformans amphotericin B resistant isolate characterised had a mutation in ERG2, encoding sterol 8-7 isomerase²¹.

Recent studies in C. glabrata and C. neoformans have revealed that a proportion of clinical isolates possess an elevated mutation rate (a ‘mutator’ phenotype), which can contribute to the rapid emergence of spontaneous antifungal drug resistance via increasing the opportunity for selectively advantageous mutations to occur^28,29. Initial studies showed that 55% of C. glabrata clinical isolates contain non-synonymous variation in the MSH2 gene, which encodes a component of mismatch DNA repair²⁸. The presence of non-synonymous variation in MSH2 correlated with multi-drug resistance²⁸. C. glabrata clinical isolates possessing non-synonymous variation in MSH2 have now been detected in clinical populations in many parts of the world with varying prevalence (North America 55%; India 69%; France 44%; Korea 65%)^30–32. However, there is not always an obvious correlation with drug resistance. In addition, an assessment of mutation rate was not performed in these studies, leading to the criticism that the variation in MSH2 may not result in a true mutator phenotype. A recent new green fluorescent protein (GFP) reporter coupled with Fluorescence-Activated Cell Sorting (FACS) technique has been developed to test mutation rates in C. glabrata clinical isolates strains with different MSH2 alleles³³. An elevated rate was not observed for isolates with the MSH2^E231G/L269F allelic variant suggesting that not all non-synonymous variation in MSH2 results in a true mutator phenotype³³. Clinical isolates of C. neoformans with non-synonymous variation in MSH2 have also been identified^29,34. These isolates exhibit a mutator phenotype and an increase in the emergence of spontaneous fluconazole and amphotericin B resistance²⁹. Deletion of MSH2 in both C. glabrata and C. neoformans results in high levels of spontaneous resistance to multiple types of antifungals^28,29.

Resistance to antifungal drugs is clearly becoming an important clinical issue that will escalate in the future unless new classes of antifungals are developed. Early treatment strategies such as prophylaxis or extensive, long-term use of antifungals to avoid relapse frequently selects for drug resistance, as does environmental exposure possibly to agricultural azoles and over the counter use (e.g. fluconazole pessaries for vaginal thrush). An improved understanding of factors influencing the emergence of, and mechanisms of, resistance is required to develop effective future treatment strategies.

Dr Morrissey has been a member of advisory boards for, received investigator-initiated grants from, and given lectures for Gilead Sciences, and Merck, Sharp and Dohme. All funds received are administered by Alfred Health/Monash University.