|Home | About | Journals | Submit | Contact Us | Français|
The amplified fragment length polymorphism technique has been applied to identify neurotropic chaetothyrialean black yeasts and relatives from clinical sources. Cladophialophora bantiana, C. emmonsii, C. arxii, C. devriesii, and C. modesta, previously identified on the basis of sequencing and phenotypic and physiological criteria, were confirmed by cluster analysis, demonstrating the clear separation of C. bantiana as a rather homogeneous group from the other species. C. bantiana is a neurotropic fungus causing cerebral abscesses with a mortality of up to 70%. Successful therapy consists of neurosurgical intervention and optimal antifungal therapy. Since the latter is not clearly defined in a large series, we tested the in vitro activities of eight antifungal drugs against clinical isolates of C. bantiana (n = 37), C. modesta (n = 2), C. arxii (n = 1), C. emmonsii (n = 1), and C. devriesii (n = 1), all of which had caused invasive infections. The resulting MIC90s for all neurotropic C. bantiana strains were as follows, in increasing order: posaconazole, 0.125 μg/ml; itraconazole, 0.125 μg/ml; isavuconazole, 0.5 μg/ml; amphotericin B, 1 μg/ml; voriconazole, 2 μg/ml; anidulafungin, 2 μg/ml; caspofungin, 4 μg/ml; and fluconazole, 64 μg/ml. On the basis of these in vitro results and the findings of previous clinical and animal studies, posaconazole seems to be a good alternative to the standard treatment, amphotericin B, for C. bantiana cerebral infections. The new agent isavuconazole, which is also available as an intravenous preparation, has adequate activity against C. bantiana.
The genus Cladophialophora represents anamorph members of the ascomycetes in the order Chaetothyriales in the family Herpotrichiellaceae comprising the black yeasts and relatives (10). These dematiaceous fungi are normally associated with soil or vegetative matter; however, they are increasingly being seen as causative agents of mycoses in humans (27, 37, 48) and domestic (14, 23) and wild (29) animals. Cladophialophora carrionii is the type species and an agent of chromoblastomycosis, a cutaneous and subcutaneous disease. The genus Cladophialophora encompasses several other clinically significant species which are potentially able to cause severe fungal infections in otherwise immunocompetent patients. In human infections, the brain is frequently involved (27, 37, 48). Within the genus, the majority of brain abscesses with fatal outcomes are associated with Cladophialophora bantiana (formerly Cladosporium bantianum, Cladosporium trichoides, Cladosporium trichoides var. chlamydosporum, Torula bantiana, and Xylohypha bantiana), a neurotropic fungus, although severe phaeohyphomycotic infections are also caused by novel Cladophialophora species like C. modesta (40, 44), C. arxii (53, 56), C. emmonsii (Xylohypha emmonsii) (45), C. devriesii (22, 28, 42) C. saturnica (4), and C. boppii (34). Moreover, Exophiala dermatitidis and Rhinocladiella mackenziei, other members of the black yeast group, are also frequently isolated from cerebral infections (8, 27, 37). Central nervous system infection due to C. bantiana is reported worldwide, though a general preference for warmer climates with high humidity is apparent (27). Indeed, many cases are reported from India (19, 30, 33, 55), as opposed to arid climatic zones (8). The first case of C. bantiana (Cladosporium trichoides) infection was reported in 1952, when the fungus was isolated from a human brain abscess and was demonstrated to be neurotropic in laboratory animals (6). A review of 17 cases of brain abscess, published in the English language literature by the mid-1970s, reported that the majority of patients had no underlying disease (41). The most recent series of 48 patients with brain abscess due to C. bantiana showed that 35 patients (72%) had no risk factors and that only 13 patients (28%) survived the infection, despite combined surgical and antifungal treatment (48). The minority of immunocompromised patients are transplant recipients, intravenous drug abusers, or individuals on steroids (2, 13, 15, 24, 31, 35, 36, 51, 54, 57, 59).
The mode of infection is either by hematogenous spread from an unrecognized pulmonary focus, through direct extension from adjacent paranasal sinuses, or by penetrating trauma to the head. However, the majority of patients had had no recent evidence of pulmonary or sinus infections (27). Cladophialophora species are prone to identification problems (3, 20). Due to the high degree of phenotypic similarity between recently described new Cladophialophora species and C. bantiana, identification problems are imminent. For most cases published in the older literature, identification down to the species level cannot be repeated or confirmed by molecular methods due to the absence of the original isolates; hence, the etiological agent described in older publications may often have been misidentified. It is now well established that molecular identification methods, which have driven new developments in fungal taxonomy, are more reliable than classical morphological methods (3, 20). Amplified fragment length polymorphism (AFLP) is a technique based on the detection of genomic restriction fragments by PCR amplification, which can be used with the DNA of any organism (58). The purpose of this study was to study the inter- and intraspecific genomic variations of 42 Cladophialophora isolates stored in the CBS collection and recovered from cases with cerebral phaeohyphomycosis and other infections. Antifungal therapy is mainly based on the experience gathered from and published in isolated case reports and mostly involved amphotericin B, itraconazole, and flucytosine singly or in different combinations (48). Animal studies (1) and human experience suggested that amphotericin B has no value in treating cerebral infections, probably due to poor penetration of the central nervous system (CNS), but that triazoles might be of value (26, 38). A recent study of antifungal therapy in a murine model of disseminated infection by C. bantiana confirmed the poor activity of amphotericin B but found that posaconazole and flucytosine extended survival (39). This suggests that new antifungal drugs with broad-spectrum activity and suitable pharmacokinetic profiles compared with those of conventional antifungal agents might be more effective against C. bantiana (1, 39). Only limited data on in vitro antifungal activities against the neurotropic fungus C. bantiana are available. Therefore, with no standard therapy available, unfavorable results in animal experiments, and only a small published series of susceptibility testing with itraconazole and voriconazole (47), the second objective of this study was in vitro testing of this large collection of Cladophialophora strains for their susceptibilities to eight antifungal drugs, including the new triazole isavuconazole.
(Part of this work was presented as a poster at Trends in Medical Mycology, Athens, Greece, October 2009 [4a].)
Table Table11 summarizes the data for and characterizes a total of 42 isolates of Cladophialophora spp. that originated from different human and veterinary clinical sources with cerebral phaeohyphomycosis or other infections. Strains were obtained from the reference collection of the CBS-KNAW Fungal Biodiversity Centre, Utrecht, Netherlands, and were handled under biosafety level 3 conditions. Stock cultures for transient working collections were initially grown on oatmeal agar (OA; Difco Oatmeal; Brunschwig Chemie, Amsterdam, Netherlands) at 24°C for 1 week, and the organisms were identified to the species level by sequencing of the internal transcribed spacer regions of the rDNA region and partial translation of the elongation factor 1-alpha and beta-tubulin genes (3).
The fungal mycelia were grown on 2% malt extract agar plates for 2 weeks at 24°C. A sterile blade was used to scrape the mycelium from the surface of the plate. DNA was extracted using an Ultra Clean microbial DNA isolation kit (Mobio, Carlsbad, CA), according to the manufacturer's instructions. DNA extracts were stored at −20°C prior to use.
Approximately 50 ng of genomic DNA was subjected to a combined restriction ligation procedure containing 50 pmol of HpyCH4 IV adapter, 50 pmol MseI adapter, 2 U of HpyCH4 IV (New England Biolabs, Beverly, MA), 2 U of MseI (New England Biolabs), and 1 U of T4 DNA ligase (Promega, Leiden, Netherlands) in a total volume of 20 μl of 1× reaction buffer for 1 h at 20°C. Next, the mixture was diluted five times with 10 mM Tris-HCl (pH 8.3) buffer. Adapters were made by mixing equimolar amounts of complementary oligonucleotides (5′-CTCGTAGACTGCGTACC-3′ and 5′-CGGGTACGCAGTC-3′ for HpyCH4 IV; 5′-GACGATGAGTCCTGAC-3′ and 5′-TAGTCAGGACTCAT-3′ for MseI), heating to 95°C, and subsequently cooling slowly to ambient temperature. One microliter of the diluted restriction-ligation mixture was used for amplification in a volume of 25 μl under the following conditions: 1 μM HpyCH4 IV primer with one selective residue (underlined) (5′-fluophore-GTAGACTGCGTACCCGTC-3′), 1 μM MseI primer with four selective residues (underlined) (5′-GATGAGTCCTGACTAATGAG-3′), 0.2 mM each deoxynucleoside triphosphate, and 1 U of Taq DNA polymerase (Roche Diagnostics, Almere, Netherlands) in 1× reaction buffer containing 1.5 mM MgCl2. Amplification was done as follows. After an initial denaturation step for 4 min at 94°C in the first 20 cycles, a touchdown procedure was applied: 15 s of denaturation at 94°C, 15 s of annealing at 66°C with the temperature for each successive cycle lowered by 0.5°C, and 1 min of extension at 72°C. Cycling was then continued for a further 30 cycles at an annealing temperature of 56°C. After completion of the cycles, incubation at 72°C for 10 min was performed before the reaction mixtures were cooled to room temperature. The amplicons were then combined with an ET400-R size standard (GE Healthcare, Diegem, Belgium) and analyzed on a MegaBACE 500 automated DNA platform (GE Healthcare), according to the manufacturer's instructions.
Data were inspected visually and were also imported into BioNumerics (version 5.1) software (Applied Maths, Sint-Martens-Latem, Belgium) and analyzed by the unweighted-pair group method using average linkages (UPGMA) clustering using the Pearson correlation coefficient. The analysis was restricted to DNA fragments in the range from 80 to 250 bp.
Amphotericin B (AMB; Bristol-Myers-Squib, Woerden, Netherlands), fluconazole (FLU; Pfizer Central Research Sandwich, United Kingdom), itraconazole (ITR; Janssen Research Foundation, Beerse, Belgium), voriconazole (VOR; Pfizer), posaconazole (POS; Schering-Plough, Kenilworth, NJ), isavuconazole (ISA; Basilea Pharmaceuticals, Switzerland), caspofungin (CAS; Merck Sharp & Dohme BV, Haarlem, Netherlands), and anidulafungin (ANI; Pfizer) were obtained from the manufacturers as pure powders. As described in the Clinical and Laboratory Standards Institute (CLSI) M38-A2 guidelines for in vitro susceptibility studies, additive drug dilutions were prepared at 100 times the final concentrations in different solutions (9). The drugs were diluted in standard RPMI 1640 medium (Sigma Chemical) buffered to pH 7.0 with 0.165 M morpholinepropanesulfonic acid (MOPS) buffer (Sigma) with l-glutamine without bicarbonate to yield the following concentrations: amphotericin B, itraconazole, voriconazole, posaconazole, and caspofungin, 0.016 to 16 μg/ml; fluconazole, 0.063 to 64 μg/ml; and isavuconazole and anidulafungin, 0.008 to 8 μg/ml. The plates were stored at −70°C until they were used. Broth microdilution was performed as described by the CLSI, in accordance with the guidelines in document M38-A2 (9). Briefly, all clinical isolates were grown on potato dextrose agar plates (PDA; Difco) at 35°C for up to 7 days for sporulation. Inoculum suspensions were prepared under biosafety laboratory level 3 regulations by slightly scraping the surface of mature colonies with a loop in sterile saline solution with Tween 40 (0.05%). After the heavy particles were allowed to settle, the homogeneous conidial suspensions were transferred to sterile tubes and adjusted spectrophotometrically at a 530-nm wavelength to optical densities (ODs) that ranged from 0.17 to 0.15 (68 to 71% transmission). The inoculum suspensions, including mostly nongerminated conidia, were diluted 1:50 in RPMI 1640 medium. The final concentration of the stock inoculum suspensions of the tested isolates ranged from 0.5 × 104 to 3.1 × 104 CFU/ml, as determined by the use of quantitative colony counts to determine the viable numbers of CFU per milliliter. After inoculation, the microdilution plates were incubated at 35°C and examined visually and spectrophotometrically at 420 nm after 72 h of incubation. The MIC endpoints were defined with the aid of a reading mirror as the lowest concentration of drug that prevents any recognizable growth (100% inhibition) for amphotericin B, itraconazole, voriconazole, posaconazole, and isavuconazole. For fluconazole, a prominent reduction of growth (≥50%) compared to the growth of the drug-free control was used. The minimum effective concentration (MEC) was defined microscopically as the lowest concentration of drug that leads to the growth of small, rounded, compact hyphal forms rather than the long, unbranched hyphal clusters that were seen in the growth control (9). Paecilomyces variotii (ATCC 22319), Candida parapsilosis (ATCC 22019), and Candida krusei (ATCC 6258) were used as quality control organisms. Values for MIC50 and MIC90 were obtained by ordering the MIC data for each antifungal in ascending arrays and selecting the median and 90th quartile of the MIC distribution, respectively. Geometric mean MICs were computed using the Microsoft Office Excel 2003 SP3 program, for which purpose values less than x were set equal to 0.5x.
Figure Figure11 depicts a dendrogram of the AFLP analysis with the type strains of C. bantiana, C. emmonsii, C. devriesi, C. modesta, and C. arxii demonstrating that C. bantiana is phylogenetically distinct from the other Cladophialophora spp. The AFLP patterns of C. bantiana strains from different geographical regions (Table (Table1),1), such as the United States, Japan, India, Belgium, France, South Africa, Mexico, and Brazil, clustered together. The structuring of this data set suggests that C. bantiana populations dispersed quickly across the world. There was one main cluster of C. bantiana, including the type strain (CBS 173.52), with similarities of more than >75%. The C. emmonsii strains segregate in one cluster, including the reference strain (CBS 979.96) and one clinical taxon originating from a subcutaneous lesion, which had >50% similarity to each other. The remaining taxa from clinical sources were represented by single strains, and the AFLP pattern shows that they are completely distinct from the C. bantiana and C. emmonsii clade (<20% similarity).
Table Table22 summarizes the results of the geometric mean MICs, MIC ranges, and MIC50 and MIC90 distributions of eight antifungal agents for 37 C. bantiana isolates. Five other clinical Cladophialophora species have individual MIC results similar to the C. bantiana MIC90 data. For each antifungal, the MIC50 and geometric mean MIC values differed by <1 log2 dilution step, indicating that in all cases the MIC50s obtained by inspection reasonably reflect the central tendency of the antifungal susceptibility of the population. Overall, all of the isolates showed a uniform pattern of low MICs for itraconazole, posaconazole, and isavuconazole. The widest ranges and the highest MICs were seen for fluconazole (range, 16 to 64 μg/ml). Amphotericin B MICs ranged from 0.125 to 2 μg/ml; and itraconazole, posaconazole, and isavuconazole had MIC ranges of <0.016 to 0.25 μg/ml, <0.016 to 0.25 μg/ml, and 0.008 to 1 μg/ml, respectively. There were no differences in the activities of itraconazole and posaconazole; and they were generally more active than amphotericin B, fluconazole, voriconazole, and isavuconazole. C. modesta, C. arxii, C. emmonsii (Xylohypha emmonsii), and C. devriesii were also inhibited by low concentrations of azoles, similar to C. bantiana. Isavuconazole exhibited potent activity against C. bantiana and the other neurotropic strains with a MIC90 of 0.5 μg/ml. The voriconazole MIC90 (2 μg/ml) was 2-log2-dilution steps less active than isavuconazole (0.5 μg/ml), which in turn was 2-log2-dilution steps less active than itraconazole and posaconazole (MIC90, 0.125 μg/ml). The only available isolate of C. modesta had a high MIC (2 μg/ml) for voriconazole and isavuconazole, similar to the MIC90 of voriconazole for C. bantiana. Of the two echinocandins, anidulafungin had the best activity, with the geometric mean MEC being 5 log2 dilution steps more active than caspofungin, although the MEC90 of 2 μg/ml would not qualify anidulafungin as an agent which can be used as monotherapy.
Primary cerebral phaeohyphomycosis in humans without obvious predisposing factors is rare, but it is increasingly recognized as an infectious disease associated with high mortality and a poor prognosis (13, 26, 37, 48). The most common agent of cerebral phaeohyphomycosis that belongs to the order of Chaetothyriales is C. bantiana, which has been reported in both healthy and immunocompromised hosts. In addition, other species causing similar clinical presentations have been described. Patients with bone marrow and solid organ transplantation and patients who are on steroid therapy are particularly affected (2, 13, 15, 24, 31, 35, 36, 51, 54, 57). Cladophialophora bantiana, a neurotropic fungus, has rarely been isolated from sources other than clinical samples. C. bantiana is distributed worldwide, but infections with this organism are especially encountered in subtropical and humid climate areas (8). Although C. bantiana has been recovered from the environment (12) and clinical infections are linked to traumatic inoculation (46), the environmental niche of C. bantiana is largely unknown; and only one such strain of C. bantiana (CBS 647.96), recovered from sawdust, was available in this study. An unambiguous connection between a clinical and an environmental strain still has to be proven.
Correct identification by mycological procedures remains difficult, due to the high degree of phenotypic similarity between these groups of fungi. For most cases published in the older literature, the etiological agent may often have been misidentified; however, reidentification down to the species level cannot be performed by molecular methods due to the absence of the original isolates. Recently, molecular data have shown that Cladophialophora contains a number of hitherto unknown species closely related but significantly different from C. bantiana (3). A study of the variability and molecular determination of the neurotropic species C. bantiana and sequencing data for the internal transcribed sequence indicated a low degree of variability (20). Interestingly, C. bantiana consistently contains an invariable intron of 558 bp at position 1768 in the small-subunit rDNA gene, while it was absent and present in all C. emmonsii isolates and C. psammophila, respectively, which are closely related to C. bantiana (3). Remarkably, this intron might be involved in evolutionary events, because it is not found outside the genetically homogeneous species C. bantiana. Other neurotropic species of Cladophialophora closely related to C. bantiana are involved in brain infections. A fatal cerebral infection with Cladophialophora modesta was reported in a 25-year-old black male after possible traumatic inoculation (40), and Cladophialophora arxii was found as the cause of cerebral phaeohyphomycosis in a 30-year-old African-American female who underwent cardiac transplantation for postpartum cardiomyopathy (44). In contrast, Cladophialophora devriesii was involved in a case of disseminated disease without involvement of the central nervous system (42), which years before presented as subcutaneous mycosis (22). C. (Xylohypha) emmonsii was the cause of a subcutaneous mycosis (45), and C. boppii was responsible for a pulmonary infection in a lung transplant recipient (34).
AFLP is one of a series of techniques for phylogenetic studies, plant and animal genetic mapping, and genotyping and is well suited for distinguishing closely related organisms at the species to strain level (32). The AFLP method relies on selective amplification of restriction fragments from a digest of genomic DNA and has many advantages compared to other marker technologies, including randomly amplified polymorphic DNA, restriction fragment length polymorphism, and microsatellites. AFLP not only has higher reproducibility, resolution, and sensitivity at the whole-genome level than other random amplification techniques but it also has the ability to amplify between 50 and 100 fragments at one time. In addition, no prior sequence information is needed for amplification (58). Bakkeren et al. (5) have shown that the phylogenetic trees obtained from AFLP analysis are quite similar to those obtained by the use of ITS sequences in Ustilago species and, in addition, permit distinction of closely related isolates that cannot be resolved by ITS sequence comparison.
Our AFLP results are in line with previous sequencing data and also show obvious differences among clinically important Cladophialophora species as agents of cerebral infection. On the basis of the AFLP patterns, we did not see any misidentification for those taxa. All strains in this study were originally identified as C. bantiana, C. emmonsii, C. arxii, C. devriesii, and C. modesta on the basis of sequencing and phenotypic and physiological criteria, which is in the line with the results of cluster analysis demonstrating a clear separation of C. bantiana as a rather homogeneous group from C. modesta, C. emmonsii, C. arxii, and C. devriesii. Although dematiaceous fungi as agents of central nervous system infections are generally susceptible to most antifungal agents in vitro, treatment is difficult because frequent relapses and failures are observed (48). Antifungal therapy is based on the experience from single patient case reports or small series which mostly involved amphotericin B, itraconazole, and flucytosine. In animal models, amphotericin B prolonged the survival of mice infected with C. bantiana, but the infection did not completely disappear (1). Although the in vitro activity of amphotericin B against C. bantiana has been demonstrated in this and previous studies, the drug was ineffective in many cases of cerebral phaeohyphomycosis with or without flucytosine (15, 19, 24, 25, 40, 48, 49, 52). In vitro results indicate that the amphotericin B MICs for most nondermatophyte opportunistic filamentous fungal isolates clustered between 0.5 and 2 μg/ml. Very few data concerning the correlation between the MIC and the outcome of treatment with amphotericin B are available for dematiaceous fungi. Generally, filamentous fungi are not susceptible to fluconazole and most MICs were >16 μg/ml. Two case reports described the improvement of C. bantiana brain abscesses after treatment with fluconazole for up to 6 weeks and surgical excision (11, 55). Surgical intervention was probably the cause of the improvement. Fortunately, the antifungal armamentarium has been extended with new triazoles, potent agents that are active against drug-resistant strains and that have less toxicity. There are only limited data in the literature regarding the susceptibility of C. bantiana isolates to antifungals (47). Less successful outcomes of treatment with itraconazole (44) and voriconazole (17, 24, 49) for cerebral abscesses due to C. bantiana were observed, although itraconazole had low MICs (MIC90, 0.125 μg/ml) in this study. The explanation might be the less optimal penetration into the CNS. Whereas voriconazole has a good penetration into the CNS, the MICs of C. bantiana were in the higher ranges (MIC90, 2 μg/ml). Some authors report successful treatment of C. bantiana brain abscesses with voriconazole (38), while others reported clinical failure (17, 24, 49). Although it is an excellent drug for the treatment of cerebral aspergillosis, voriconazole might not be the first choice for the treatment of C. bantiana infections. In contrast to another study with only seven strains of C. bantiana (47), which gave a MIC range of 0.12 to 1 μg/ml of voriconazole, we found a large range of activity (MICs, 0.125 to 4 μg/ml). Therefore, in vitro susceptibility testing may be warranted before voriconazole is used to treat a CNS infection due to C. bantiana. Posaconazole and itraconazole (MIC90, 0.125 μg/ml) demonstrated the best in vitro activities, followed by isavuconazole (MIC90, 0.5 μg/ml). Some clinical experience suggests that therapy with itraconazole was successfully used in treating a C. bantiana cerebral infection (25, 35), eumycetoma (7), and C. arxii osteomyelitis (53). Treatment with posaconazole is supported by our in vitro results, other in vitro data from small series (n = 5, range 0.06 to 0.5 μg/ml) (16), data for a clinical case (18), and data from murine infection models in which posaconazole prolonged the survival of mice and reduced the level of brain fungal burden compared to that achieved with itraconazole and amphotericin B (1, 39). Moreover, the apparently good penetration into the CNS (50) supports the use of posaconazole for this difficult-to-treat infection. Most melanized fungi appear to be resistant to echinocandins, probably due to the reduced presence of β-glucan in the cell walls (16, 43). We found that caspofungin had poor activity but that anidulafungin had activity against C. bantiana showing a geometric mean MIC of only 0.073 μg/ml, although several isolates had high MECs (MEC90, 2 μg/ml). Another echinocandin, micafungin, was not active in animal studies when it was used as monotherapy but seemed to be promising in combination with posaconazole and flucytosine (39). The investigational agent isavuconazole shows broad-spectrum activity against many opportunistic and true pathogenic fungi (21). Here, we show that isavuconazole is also active against C. bantiana, but its true value needs to be confirmed in animal models.
In conclusion, itraconazole, posaconazole, and isavuconazole demonstrated in vitro activity against neurotropic isolates of C. bantiana. Some positive and negative correlating clinical experiences are available for itraconazole, voriconazole, and posaconazole; but the most important intervention for cerebral phaeohyphomycosis caused by C. bantiana probably remains complete neurosurgical excision of the abscesses.
This study and the work of Hamid Badali were funded by the Ministry of Health and Medical Education of the Islamic Republic of Iran (grant number 13081) and the School of Medicine, Mazandaran University of Medical Sciences, Sari, Iran. This study was partially supported by an unrestricted grant from Basilea Pharmaceuticals, Basel, Switzerland.
J.F.M. received grants from Astellas, Basilea, Merck, and Schering-Plough. He has been a consultant to Basilea, Merck, and Schering-Plough and received speaker's fees from Gilead, Janssen Pharmaceutica, Merck, Pfizer, and Schering-Plough. C.H.W.K. received a grant from Pfizer. None of the other authors has a potential conflict of interest.
Published ahead of print on 26 April 2010.