0095-1137/09/$08.00⫹0 doi:10.1128/JCM.01614-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.
Bloodstream Infections Due to
Trichosporon
spp.: Species Distribution,
Trichosporon asahii
Genotypes Determined on the Basis of Ribosomal
DNA Intergenic Spacer 1 Sequencing, and Antifungal
Susceptibility Testing
䌤
Thomas C. Chagas-Neto,† Guilherme M. Chaves,† Analy S. A. Melo, and Arnaldo L. Colombo*
Laborato´rio Especial de Micologia, Disciplina de Infectologia, Universidade Federal de Sa˜o Paulo, Sa˜o Paulo, Brazil
Received 19 August 2008/Returned for modification 2 October 2008/Accepted 9 February 2009
The reevaluation of the genusTrichosporonhas led to the replacement of the old taxonTrichosporon beigelii
by six new species. Sequencing of the ribosomal DNA (rDNA) intergenic spacer 1 (IGS1) is currently man-datory for accurateTrichosporonidentification, but it is not usually performed in routine laboratories. Here we describe Trichosporonspecies distribution and prevalence ofTrichosporon asahii genotypes based on rDNA IGS1 sequencing as well as antifungal susceptibility profiles of 22 isolates recovered from blood cultures. The clinical isolates were identified as follows: 15 T. asahii isolates, five Trichosporon asteroides isolates, one
Trichosporon coremiiformeisolate, and oneTrichosporon dermatisisolate. We found a great diversity of different species causing trichosporonemia, including a high frequency of isolation ofT. asteroidesfrom blood cultures that is lower than that ofT. asahiionly. RegardingT. asahii genotyping, we found that the majority of our isolates belonged to genotype 1 (86.7%). We report the firstT. asahiiisolate belonging to genotype 4 in South America. Almost 50% of allT. asahiiisolates exhibited amphotericin B MICs of>2g/ml. Caspofungin MICs obtained for all theTrichosporonsp. isolates tested were consistently high (MICs>2g/ml). Most isolates (87%) had high MICs for 5-flucytosine, but all of them were susceptible to triazoles, markedly to voriconazole (all MICs< 0.06g/ml).
The incidence of invasive mycoses caused by emergent fun-gal pathogens has risen considerably over the last two decades. This fact is related to several factors, including the increased occurrence of degenerative and malignant diseases in different populations, as well as the higher number of patients exposed to organ transplantation, immunosuppressive therapies, che-motherapy, broad-spectrum antibiotics, and invasive medical procedures. It is important to emphasize that emergent fungal infections are usually difficult to diagnose, refractory to con-ventional antifungal drugs, and associated with high mortality rates (8, 10, 26, 32, 34, 47, 51).
Invasive infections caused byTrichosporonspp. are reported mostly for cancer patients that have central venous catheters (22, 50–52). Although trichosporonemia represents a small percentage of all fungal invasive infections,Trichosporonspp. have been reported as the second- or third-most-common agents of yeast fungemia (13, 21, 51).
Phenotypic methods forTrichosporonspecies identification usually generate inconsistent results, and none of the commer-cial tests available include the whole new taxonomic categories in their databases (1, 36, 37). For instance, it has been men-tioned in the literature that an isolate identified by molecular methods asTrichosporon dermatiswas mistakenly identified as
Trichosporon mucoides when Vitek Systems 1 and 2
(Bio-Me´rieux, France) were used (16). Furthermore, Ahmad et al. (1) reported that four isolates previously identified by Vitek 2 asTrichosporon asahii were identified as Trichosporon aster-oidesby molecular techniques. There is a consensus that mo-lecular methods are required for accurate identification of this genus, but they are still costly and not suitable for most routine laboratories (38, 42).
Currently, the genus Trichosporon includes the following species as potential human pathogens:Trichosporon cutaneum,
T. asahii,T. asteroides,T. mucoides,Trichosporon inkin, Tricho-sporon jirovecii,T. dermatis,Trichosporon domesticum, Tricho-sporon montevideense, Trichosporon japonicum, Trichosporon coremiiforme, Trichosporon faecale, and Trichosporon loubieri
(7, 34, 38). Sugita et al. (42) reported that the ribosomal DNA (rDNA) internal transcribed spacer (ITS) region is not suit-able forTrichosporonspecies identification. Therefore, the analysis of the rDNA intergenic spacer (IGS) region is a crucial method which should be used to distinguish phylo-genetically closely related species (42–44).
Considering the limitations of routine laboratories in con-ducting appropriate identification ofTrichosporonspp., there is a lack of data for species distribution and antifungal suscepti-bility of hematogenous infections caused by this particular genus. Girmenia et al. (15) published the largest retrospective clinical study of invasive trichosporonosis documented during a period of 20 years. Despite representing the largest series of invasive trichosporonosis ever published before, only 30 of 287 isolates were identified to the species level.
Several authors have attempted to evaluate the antifungal susceptibilities ofTrichosporonspp. but most of them still used the old nomenclatureT. beigelii (17, 35, 48, 49). Antifungal
* Corresponding author. Mailing address: Laborato´rio Especial de Micologia, Disciplina de Infectologia, Universidade Federal de Sa˜o Paulo, Sa˜o Paulo–SP, Rua Botucatu, 740, CEP 04023-062, Sa˜o Paulo, Brazil. Phone and fax: 55-11-5083-0806. E-mail: arnaldolcolombo @gmail.com.
† T.C.C. and G.M.C. contributed equally to this work. 䌤Published ahead of print on 18 February 2009.
susceptibilities of bloodstream isolates belonging to this genus have not yet been largely investigated. Rodriguez-Tudela et al. (38) performed antifungal susceptibility testing for Trichos-poron spp. accurately identified by rDNA IGS1 region se-quencing. However, only 9 out of 49 strains were recovered from bloodstream infections (BSI). Thereafter, the same group published a paper investigatingT. asahiisusceptibility testing (39), but once again, only 8 out of 18 isolates were recovered from BSI. To date, only limited data are available for species distribution and antifungal susceptibilities of BSI isolates of
Trichosporonspp. accurately identified by molecular tools. In the present study, we investigated species distribution and antifungal susceptibilities of 22 bloodstreamTrichosporon iso-lates sequentially recovered from different patients hospital-ized in five medical centers between 1995 and 2007.
MATERIALS AND METHODS
Strains.We evaluated a total of 22Trichosporonsp. bloodstream isolates sequentially obtained from different patients admitted to five tertiary care hos-pitals in Sao Paulo, Brazil, between January 1995 and December 2004. The isolates were stored on 1% yeast extract, 2% peptone, 2% dextrose, and 3% glycerol at⫺70°C. Of note, the isolates mentioned above were collected during multicentric surveillance studies of fungemia coordinated by our group (9, 10). The isolates were identified to the species level by sequencing the rDNA IGS1 region and were subsequently submitted to antifungal susceptibility testing by standard methods. Candida parapsilosis (ATCC 22019) and Candida krusei
(ATCC 6258) were used as quality control strains for antifungal susceptibility testing.T. asahii(CBS 2479 and CBS 2530),T. mucoides(CBS 7625),T. inkin
(CBS 5585),T. asteroides(CBS 2481), andT. ovoides(CBS 7556) were used as control strains forTrichosporonmolecular identification. As negative controls,
Candida albicans(ATCC 90028) andC. parapsilosis(ATCC 22019) were used to check for the specificity ofTrichosporon-specific oligonucleotides. The isolates used in the present study are displayed in Table 1.
Screening ofTrichosporonisolates by conventional methods.Yeast isolates were obtained from blood cultures from the referred medical centers by using either Bactec 9000 (BD) or BacT/Alert 3D (BioMe´rieux, France). Yeast isolates from original cultures were plated onto CHROMagarCandida(CHROMagar Microbiology, Paris, France) to check for purity and viability. All strains able to produce arthroconidia, as revealed by their micromorphology on corn meal-Tween 80 agar, and able to hydrolyze urea were considered to belong to the genusTrichosporonand were further identified by molecular techniques (http: //www.cbs.knaw.nl/databases) (3, 11, 18, 23, 38, 40).
DNA extraction.The isolates were grown in Falcon tubes containing 2 ml yeast extract-peptone-dextrose medium for 16 h, 30°C at 200 rpm in a gyrator with shaking for 48 h. DNA was extracted using the fast small-scale isolation protocol described elsewhere (27). RNA was removed by treatment with RNase A (Phar-macia) for 1 h at 37°C. DNA concentration and purity were determined by optical density at 260 nm and ratio optical density at 260/280 nm, respectively.
PCR assay and IGS1 region sequencing.To confirm the correct genus iden-tification, the primer pair TRF (5⬘-AGAGGCCTACCATGGTATCA-3⬘) and TRR (5⬘-TAAGACCCAATAGAGCCCTA-3⬘) was used (44). The samples were amplified in a Thermocycler (model 9600; Applied Biosystems) by using the following cycling parameters: one initial cycle of 94°C for 3 min, followed by 30 cycles of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C and a final cycle of 5 min at 72°C. For the accurate identification ofTrichosporonstrains to the species level, the IGS1 region was amplified by PCR using the following primer pair: 26F (5⬘-ATCCTTTGCAGACGACTTGA-3⬘) and 5SR (5⬘-AGCTTGACTTCGCAG ATCGG-3⬘) and 2⫻PCR Master Mix (Promega). PCR was performed with one initial cycle of 94°C for 5 min, followed by 40 cycles of 1 min at 94°C, 1 min at 57°C, and 2 min at 72°C and a final cycle of 5 min at 72°C. PCR products were sequenced by using the dideoxynucleotide method in an ABI Prism 3100 auto-mated sequencer (Applied Biosystems, CA). The sequencing reaction included the primer pair 26F and 5SR (42) and the BigDye Terminator reagent kit (Applied Biosystems, CA) and was performed according to the manufacturer’s instructions. Nucleotide sequences were submitted for BLAST analysis at the NCBI site (http://www.ncbi.nlm.nih.gov) for species identification. Only quences deposited in GenBank showing high similarities with our query se-quences and an E value of lower than 10⫺5were used in this study.
Phylogenetic analysis.Sequences were aligned using BioEdit software (v7.09). Aligned sequences were used for phylogenetic analysis conducted with Mega 4.1 Beta Software (v4.1 Beta). The method used for tree constructions was the unweighted-pair group method using average linkages (UPGMA). Phylogram stability was accessed by bootstrapping with 1,000 pseudoreplications. To root the phylogenetic tree for all theTrichosporonisolates, the IGS1 sequences of
Cryptococcus neoformans(GenBank accession number EF211300) and Crypto-coccus gatii(GenBank accession number EF211336) were used as outgroups.
In vitro susceptibility testing.Antifungal susceptibility testing was performed using the broth microdilution assay according to an adaptation of the method-ology recommended forCandidaspp. andCryptococcus neoformansand accord-ing to CLSI approved standard M27-A2 (31). The followaccord-ing antifungal drugs, supplied by the manufacturers as pure standard compounds, were tested at the indicated concentration ranges: amphotericin B (AMB), 0.015 to 8g/ml (Sigma Chemical Corporation, St. Louis, MO); 5-flucytosine (5FC), 0.125 to 64g/ml (Sigma Chemical Corporation, St. Louis, MO); itraconazole (ITC), 0.03 to 16
g/ml (Janssen Pharmaceutical, Titusville, NJ); fluconazole (FLC), 0.125 to 64
g/ml; voriconazole (VRC), 0.03 to 16g/ml (Pfizer Incorporated, New York, NY); and caspofungin (CAS), 0.03 to 16g/ml (Merck Research Laboratories, NJ). Briefly, the medium used was RPMI-1640 withL-glutamine, without
bicar-bonate, buffered at pH 7.0 with 0.165 M morpholinepropanesulfonic acid. The yeast inoculum suspension was prepared using a spectrophotometer to obtain a final yeast concentration of 0.5⫻103to 2.5⫻103cells/ml in each well of the
microtiter plate. The assays were incubated at 35°C. Cell cultures were incubated for 48 h as suggested by Brito et al. (6). Quality control strains (C. parapsilosis
ATCC 22019 andC. kruseiATCC 6258) were included on each day of the assay to check the accuracy of the drug dilutions and the reproducibility of the results. The MIC endpoint for AMB was considered the lowest tested drug concentra-tion able to prevent any visible growth. The MIC for triazoles, CAS, and 5FC was considered the lowest tested drug concentration causing a significant reduction (approximately 50%) in growth compared to the growth of the drug-free positive control (31).
Statistical analysis.Differences in mortality rates between adults and infants were analyzed by Fisher’s test.Pvalues of⬍0.05 were considered significant for all the statistics analyzed.
GenBank accession numbers.GenBank accession numbers were as follows: EU934801, EU934808, EU934809, EU934810, EU938047, EU980326, EU938049, EU938050, EU938051, EU938052, EU938054, EU938055, EU938056, EU938058, EU934803, EU934811, EU938046, EU934802, EU934800, EU938059, EU938048, EU938053, EU938057, EU934807, EU934804, EU934805, EU934806, and EU938060.
RESULTS
Demographic and clinical data. We were able to collect clinical information for 18 of 22 patients with Trichosporon
invasive infections. Our series was represented by pediatric (44%) and adult patients exhibiting a large variety of un-derlying conditions, including premature birth, surgery, or-gan failures, inflammatory gastrointestinal disease, and can-cer, who frequently developed systemic bacterial infections (67%) and had a central venous catheter in place at the onset of fungemia. Mortality rates of infants were signifi-cantly lower than those for adults (30% against 87.5%, re-spectively;P⫽0.025) (Table 1).
Molecular identification of Trichosporon isolates. To dou-ble-check for accurate identification of the genus, we per-formed diagnostic PCR using primers TRR and TRF (44). This pair of primers isTrichosporonspecific and amplifies part of the nucleotide sequences of the rDNA small subunit (18S). DNA bands of approximately 170 bp were obtained for all the isolates tested and forTrichosporonreference strains (data not shown). In addition, there was no amplification of DNA iso-lated from Candida spp. (negative control). Therefore, our strains were confirmed to beTrichosporonspp.
TABLE 1. Microbiological and clinical data for 22 episodes ofTrichosporonsp. bloodstream infectionsa
Reference strain and patient
Species accession no.GenBank Age ofpatient Gender ofpatient Associated condition(s) CVC Antifungal therapy Other infection(s) outcomeClinical
CBS 2479 T. asahiigenotype 1 EU934801
1 T. asahiigenotype 1 EU934808 39 yr M Cardiac failure Present AMB Bacteremia Death
2 T. asahiigenotype 1 EU934809 NA NA NA NA NA NA NA
3 T. asahiigenotype 1 EU934810 78 yr M Inflammatory gastrointestinal
disease
Present AMB Staphylococcus
aureusand/or Pseudomonas aeruginosa(BSI)
Death
4 T. asahiigenotype 1 EU938047 3 yr M Burn Present AMB Soft tissue infections Discharged
5 T. asahiigenotype 1 EU980326 16 days F Premature birth⫹low wt Present AMB Enterobactersp.
(BSI)
Death
6 T. asahiigenotype 1 EU938049 55 yr F NA NA NA NA NA
7 T. asahiigenotype 1 EU938050 6 yr M Cancer Present Double-blinded
randomized study
None Discharged
8 T. asahiigenotype 1 EU938051 49 yr M Lung neoplasia Present None None Death
9 T. asahiigenotype 1 EU938052 17 yr M Prune belly syndrome⫹
chronic renal failure
Present FLC⫹AMB Escherichia coli⫹
Enterococcussp. (urine) and/or Staphylococcussp. (BSI)
Death
10 T. asahiigenotype 1 EU938054 NA M NA NA NA NA NA
11 T. asahiigenotype 1 EU938055 NA M NA NA NA NA NA
12 T. asahiigenotype 1 EU938056 33 yr M Kidney transplant Present FLC Death
13 T. asahiigenotype 1 EU938058 77 yr F Inguinal hernia⫹
postoperative enterectomy
Present None Bacterial sepsis Death
CBS 2530 T. asahiigenotype 3 EU934803
14 EU934811 78 yr M Cardiac failure Present None Discharged
15 T. asahiigenotype 4 EU938046 84 days M Premature birth⫹low wt⫹
enterectomy
Present FLC⫹AMB Escherichia coli⫹
Klebsiellasp. (BSI)
Discharged
CBS 2481 T. asteroides EU934802
16 T. asteroides EU934800 2 mo F Galactosemia⫹congenital
cytomegalovirus syndrome
Present AMB Bacteremia Discharged
17 T. asteroides EU938059 3 yr M Testicle carcinoma Present NA NA Discharged
18 T. asteroides EU938048 54 yr F Acute arterial embolism Present FLC Discharged
19 T. asteroides EU938053 15 days F Premature birth⫹low wt Present AMB BCP Discharged
20 T. asteroides EU938057 43 yr M Gunshot wound of abdomen
⫹splenectomy⫹ pancreatectomy
Present FLC⫹AMB Peritonitis⫹sepsis Discharged
21 T. coremiiforme EU934807 10 days F Pulmonary emphysema⫹
pulmonary hypertension
Present FLC Candidasp. (oral)⫹
Stenotrophomonas maltophilia
Discharged
CBS 5585 T. inkin EU934804
CBS 7556 T. ovoides EU934805
CBS 7625 T. mucoides EU934806
22 T. dermatis EU938060 53 yr F Diabetes mellitus⫹systemic
arterial hypertension
Present FLC⫹AMB Candidasp. (BSI) Death
aClinical and epidemiological data for four patients were not available. NA, not available; BCP, bacteremia during pregnancy; CVC, central venous catheter; M, male; F, female.
CHAGAS-NETO
ET
AL.
J.
C
LIN
.M
ICROBIOL
the GenBank genome database at the NCBI website (http: //ncbi.nlm.nih.gov) using BLAST to compare gene sequences. All theTrichosporoncontrol strains were unambiguously iden-tified by sequencing. This fact, together with an E value of 0 for all the BLAST matches, supported our accurate final identifi-cation. The 22 clinical isolates were identified as follows: 15T. asahiiisolates, fiveT. asteroidesisolates, oneT. coremiiforme
isolate, and oneT. dermatisisolate (Table 1).
Phylogenetic analysis.UPGMA analysis using 1,000 pseu-doreplications of the IGS1 region sequences of the isolates, as well as reference strains, adequately resolved allTrichosporon
species (Fig. 1). The results obtained by our phylogenetic
anal-ysis reinforce the accurate molecular identification in the present study.
All T. asahii isolates clustered together, although minor differences were observed between isolates related to their genetic relatedness (Fig. 1). We decided to further investi-gate this genetic variability within the species. We con-structed a phylogenetic tree using the UPGMA method including rDNA IGS1 sequences of our T. asahii isolates and a representative sequence of each T. asahii genotype described by Sugita et al. (42) and Rodriguez-Tudela et al. (39) deposited in GenBank. We found that the majority of our isolates belonged to genotype 1 (86.7%), while one
isolate belonged to genotype 3 and another isolate belonged to genotype 4 (Fig. 2). Here we describe the firstT. asahii
isolate belonging to genotype 4 from South America. Un-fortunately, we could not include genotype 6 in our analysis because IGS1 sequences for this genotype were not avail-able from the GenBank database.
Interestingly, the onlyT. coremiiformeisolate found in our study shares a common branch withT. asahii, with a bootstrap support higher than 90, indicating high phylogenetic relation (Fig. 1). All T. asteroides isolates clustered together and showed high genetic relatedness within the group (bootstrapⱖ
99) and share a common branch withT. asahiiandT.
coremi-iforme(bootstrapⱖ99). We have included in our sequencing
analysis one isolate each of reference strainsT. inkinand T. ovoides(CBS 5585 and 7556, respectively). The two species clustered together within the same clade (bootstrap ⱖ 99), demonstrating high phylogenetic relation. Another highly re-lated cluster was composed ofT. dermatis and T. mucoides, showing a bootstrap support ofⱖ99 (Fig. 1).
Antifungal susceptibility testing.MICs of the 22 strains of
Trichosporonspp. tested ranged from 0.25 to 4g/ml for AMB, 0.125 to 8g/ml for FLC, 0.03 to 0.125g/ml for ITC, 0.03 to 0.06g/ml for VRC, 0.5 to⬎64g/ml for 5FC, and 2 to⬎16
g/ml for CAS. Table 2 summarizes ranges and geometric
FIG. 2. Rooted phylogenetic tree ofTrichosporon asahiiIGS1 genotypes based on confidently aligned sequences and obtained by UPGMA analysis and 1,000 bootstrap simulations.
TABLE 2. In vitro activity of six antifungal agents against 22Trichosporonsp. strains tested by CLSI microdilution assay
Trichosporonsp. (no. of isolates)
Cell incubation period (h)
MIC (g/ml) for indicated antifungal agenta
AMB FLC ITC VRC 5FC CAS
Range GM Range GM Range GM Range GM Range GM Range GM
T. asahii(15) 24 0.5–2 0.9 0.25–2 0.8 0.03–0.125 0.07 0.03–0.06 0.03 0.5–32 12.1 2–⬎16 14.6 48 0.5–4 1.4 0.25–8 1.3 0.03–0.125 0.08 0.03–0.06 0.03 4–⬎64 27.9 16–⬎16 16
T. asteroides(5) 24 0.25–1 1 0.125–0.5 0.3 0.03–0.125 0.05 0.03–0.06 0.03 0.5–32 8 2–⬎16 10.6 48 0.5–4 1.5 0.5–1 0.8 0.03–0.125 0.06 0.03–0.06 0.03 4–⬎64 16 16–⬎16 16
T. dermatis(1) 24 0.5 0.5 0.06 0.06 8 16
48 0.5 0.5 0.06 0.06 16 ⬎16
T. coremiiforme(1) 24 0.5 0.125 0.06 0.03 2 8
48 1 1 0.125 0.03 16 ⬎16
means of MICs obtained forTrichosporonsp. isolates against the six antifungal drugs tested.
Of note, 47% of allT. asahiiisolates exhibited AMB MICs ofⱖ2g/ml. CAS MICs obtained for all theTrichosporonsp. isolates tested were consistently high (MICⱖ2g/ml). Most isolates (87%) had high MICs for 5FC. With regard to the azoles, the three drugs tested were active against all the spe-cies, while VRC was the most effective antifungal agent. All the isolates had MICs ofⱕ0.06g/ml.
DISCUSSION
This report represents the largest published series address-ing species distribution, prevalence ofT. asahiigenotypes and antifungal susceptibility profiling ofTrichosporonbloodstream isolates accurately identified to the species level by using rDNA IGS1 sequencing.
Surprisingly, in contrast to previously published data, cancer and neutropenia were both present in a limited number of patients (15, 26). Most of our patients were individuals with several underlying associated conditions, such as premature birth, surgery, burns, organ failure, and gastrointestinal inflam-matory disease. Of note, 12 of 18 patients (67%) had an epi-sode of systemic bacterial infection previously or concomi-tantly to the fungemia. As expected, all patients had a central venous catheter in place at the time of the diagnosis. This finding reinforces the concept that invasiveTrichosporon infec-tions may occur in non-cancer patients who present chronic illness and disruption of skin and mucous membranes (12). Crude mortality was significantly higher for adults than for infants.
In our study, we have found that the vast majority of the isolates were identified asT. asahii(approximately 68%) (Ta-ble 1).T. asahiihas extensively been recognized as an emer-gent aemer-gent of fungal invasive infections worldwide (4, 5, 19, 20, 28, 30, 46, 51, 53). The largest multicentric restrospective study addressingTrichosporoninvasive infections found thatT. asahii
accounted for 15 of 30 isolates (50%) identified to the species level (15). In the series of 49Trichosporonisolates published by Rodriguez-Tudela et al. (38), 15 strains were identified asT. asahiiby using IGS1 region sequencing, including eight of nine
Trichosporonisolates recovered from BSI. Considering the dif-ficulties for routine laboratories in correctly identifying Tricho-sporonspp., the true prevalence of fungemia due toT. asahii
may not be accurately estimated. In terms of antifungal sus-ceptibility, in accordance with previously published data, a significant number ofT. asahiistrains exhibited poor suscep-tibility to AMB but good in vitro suscepsuscep-tibility to triazoles (2, 14, 33, 49, 54).
Surprisingly,T. asteroideswas found to be the second-most-common species isolated from BSI due toTrichosporonspp. in our study (approximately 23%) (Table 1), while the isolates did not originate from the same institution (therefore not being characterized as an outbreak).T. asteroidesis a causative agent of superficial infections (42). The first case report of dissemi-nated infection due to this agent was published in 2002, de-scribing an intensive care unit female patient with multiple organ failure andT. asteroides-positive cultures of blood, urine, aspiration fluid, and catheter tip swabs (22). Ahmad et al. (1) reported the isolation of threeT. asteroidesstrains recovered
from blood and one from a catheter tip swab. However, it is important to emphasize that both studies used the rDNA ITS region for Trichosporonsp. identification, a method recently replaced by rDNA IGS1 sequencing (42). In our series, except for one strain of T. asteroides that had an AMB MIC of 4
g/ml, all the other clinical isolates of this species were sus-ceptible to AMB, FLC, ITC, and VRC.
There are very few published instances where T. coremi-iformeis described as the cause of human diseases. Interest-ingly, despite the fact that Sugita et al. (42) refer to this species as nonpathogenic, theT. coremiiformetype strain (CBS 2482) was originally isolated from a patient with a fungal lesion of the head. Recently,T. coremiiformehas been isolated from subcu-taneous abscesses and urine (38). To our knowledge, the present case represents the first report of a BSI caused byT. coremiiforme.
We also recovered in our series an isolate of T. dermatis
from blood (Table 1).T. dermatishas been associated mostly with superficial infections (16). The series of Rodriguez-Tudela et al. (38) documented eight cases ofT. dermatis caus-ing infections. However, only one of these isolates was recov-ered from blood, while all the others were isolated from the skin, nails, and subcutaneous lesions of patients. A recent paper published by Gunn et al. (16) documented a case of trichosporonosis due to T. dermatis in a 13-month-old male with a history of autoimmune enteropathy under immunosup-pressant therapy, maintained with total parenteral nutrition through a central venous catheter. The isolate was initially misidentified asT. mucoidesafter sequencing the ITS region. It is important to emphasize thatT. dermatisis phylogenetically closely related to T. mucoides and cannot be distinguished when ITS sequences are used for comparison. Of note, in contrast toT. dermatis,T. mucoideshas been associated mostly with invasive infection (16). Therefore, previously published data on trichosporonemia must have miscalculated the number of BSI which could have been caused byT. dermatis.This fact emphasizes the necessity of correctly identifyingTrichosporon
spp. by using rDNA IGS1 sequencing in order to avoid under-estimation of the prevalence of a recently considered emergent species.
and invasive cases of trichosporonosis. Further investigations including a larger number of strains are necessary to confirm this finding.
We document the first description ofT. asahiigenotype 4 in South America. Until the present study, strains belonging only to genotypes 1, 3, and 6 have been described for this continent (39). Sugita et al. (42) reported that genotypes 2 and 4 were detected only in Japan. However, Rodriguez-Tudela et al. (39) inferred that all genotypes may be found in all geographic regions. Until the present moment, genotype 2 has not been described in South America. As reported by Rodriguez-Tudela et al. (39), this genotype differs from genotype 1 by only one nucleotide (thymine instead of adenine), suggesting a likely recent evolution of genotype 2 in Japan.
We found high genetic relatedness of our T. coremiiforme
isolate and the T. asahiicluster (Fig. 1). Indeed, T. coremi-iforme was considered for a long time to be a variety of T. asahii(45). Despite the fact that the two species are almost indistinguishable (99.7% similarity), when ITS region se-quences are used for comparisons, the IGS1 regions of T. coremiiformeandT. asahiiare 78.8% similar (42). Therefore, they should be considered different species.
We state here that allT. asteroidesisolates were included in the same cluster, sharing a common branch withT. asahiiand
T. coremiiforme(Fig. 1). Of note, T. asteroides andT. asahii
DNA sequences are similar by 98.7% (for the ITS region) and 75.1% (for the IGS1 region) (42). It is important to emphasize that all three species mentioned above belong to the same clade (ovoides), according to the taxonomic classification of the genusTrichosporonproposed by Middelhoven et al. (29). In addition, the two reference strains,T. inkinandT. ovoides, which clustered together and presented a high bootstrap (ⱖ99) also belong to the clade ovoides (29).
Finally, the clinical isolates T. dermatis and T. mucoides
shared the same cluster with a high bootstrap (ⱖ99). As
men-tioned before, both species have undistinguishable ITS se-quences and biochemical assimilation profiling (16). They be-long to the clade cutaneum, according to the classification of Middelhoven et al. (29).
This paper reinforces the relevance of correctly identifying and testingTrichosporonBSI isolates with different antifungal compounds. We have documented the great diversity of dif-ferent species causing fungemia and a high frequency of T. asteroidesisolation from blood, a frequency lower than that of
T. asahiionly. The differentTrichosporonspecies also exhibited peculiar in vitro antifungal susceptibility profiles, including a significant number ofT. asahiiisolates showing low suscepti-bility to AMB.T. asahiigenotype 1 was confirmed to be the most prevalent agent ofT. asahiiinvasive infections in Brazil, as published elsewhere. However, other genotypes can occur in South America, such as genotype 4, as described in the present study. Further investigations using a larger number of Tricho-sporonisolates identified by molecular methods will contribute to the understanding of the epidemiology and species distri-bution of the genus worldwide.
ACKNOWLEDGMENTS
This study was supported by Fundac¸a˜o de Amparo a Pesquisa do Estado de Sa˜o Paulo (FAPESP), Brazil (grant 2005/02006-0). G.M.C. and A.S.A.M. received postdoctoral fellowships from FAPESP (grant
2005/04442-1) and CAPES (PRODOC 2005), respectively. We are very greatful to T. Guimara˜es, D. Bergamasco, D. Santos, R. Rosas, and C. Pereira for the collection of clinical data.
REFERENCES
1.Ahmad, S., M. Al-Mahmeed, and Z. U. Khan.2005. Characterization of Trichosporon species isolated from clinical specimens in Kuwait. J. Med. Microbiol.54:639–646.
2.Asada, N., H. Uryu, M. Koseki, M. Takeuchi, M. Komatsu, and K. Matsue.
2006. Successful treatment of breakthrough Trichosporon asahii fungemia with voriconazole in a patient with acute myeloid leukemia. Clin. Infect. Dis.
43:e39–e41.
3.Barnett, J. A., R. W. Payne, and D. Yarrow.2000. Yeasts: characteristics and identification, 3rd ed. Cambridge University Press, Cambridge, United Kingdom.
4.Bayramoglu, G., M. Sonmez, I. Tosun, K. Aydin, and F. Aydin.2008. Break-through Trichosporon asahii fungemia in neutropenic patient with acute leukemia while receiving caspofungin. Infection36:68–70.
5.Biasoli, M. S., D. Carlson, G. J. Chiganer, R. Parodi, A. Greca, M. E. Tosello, A. G. Luque, and A. Montero.2008. Systemic infection caused by Trichosporon asahii in a patient with liver transplant. Med. Mycol.46:719– 723.
6.Brito, L. R., T. Guimaraes, M. Nucci, R. C. Rosas, L. Paula Almeida, D. A. Da Matta, and A. L. Colombo.2006. Clinical and microbiological aspects of candidemia due to Candida parapsilosis in Brazilian tertiary care hospitals. Med. Mycol.44:261–266.
7.Chagas-Neto, T. C., G. M. Chaves, and A. L. Colombo.2008. Update on the genusTrichosporon. Mycopathologia166:121–132.
8.Colombo, A. L., A. S. Melo, R. F. Crespo Rosas, R. Salomao, M. Briones, R. J. Hollis, S. A. Messer, and M. A. Pfaller.2003. Outbreak of Candida rugosa candidemia: an emerging pathogen that may be refractory to ampho-tericin B therapy. Diagn. Microbiol. Infect. Dis.46:253–257.
9.Colombo, A. L., M. Nucci, B. J. Park, S. A. Nouer, B. Arthington-Skaggs, D. A. da Matta, D. Warnock, and J. Morgan.2006. Epidemiology of candi-demia in Brazil: a nationwide sentinel surveillance of candicandi-demia in eleven medical centers. J. Clin. Microbiol.44:2816–2823.
10.da Matta, D. A., L. P. de Almeida, A. M. Machado, A. C. Azevedo, E. J. Kusano, N. F. Travassos, R. Salomao, and A. L. Colombo.2007. Antifungal susceptibility of 1000 Candida bloodstream isolates to 5 antifungal drugs: results of a multicenter study conducted in Sao Paulo, Brazil, 1995–2003. Diagn. Microbiol. Infect. Dis.57:399–404.
11.De Hoogs, G. S., J. Guarro, J. Gene, and M. J. Figueras.2000. Atlas of clinical fungi, 2nd ed. Guanabara, Rio de Janeiro, Brazil.
12.Ebright, J. R., M. R. Fairfax, and J. A. Vazquez.2001. Trichosporon asahii, a non-Candida yeast that caused fatal septic shock in a patient without cancer or neutropenia. Clin. Infect. Dis.33:E28–E30.
13.Fleming, R. V., T. J. Walsh, and E. J. Anaissie.2002. Emerging and less common fungal pathogens. Infect. Dis. Clin. N. Am.16:915–933. 14.Fournier, S., W. Pavageau, M. Feuillhade, S. Deplus, A. M. Zagdanski, O.
Verola, H. Dombret, and J. M. Molina.2002. Use of voriconazole to suc-cessfully treat disseminated Trichosporon asahii infection in a patient with acute myeloid leukaemia. Eur. J. Clin. Microbiol. Infect. Dis.21:892–896. 15.Girmenia, C., L. Pagano, B. Martino, D. D’Antonio, R. Fanci, G. Specchia,
L. Melillo, M. Buelli, G. Pizzarelli, M. Venditti, and P. Martino.2005. Invasive infections caused byTrichosporonspecies andGeotrichum capitatum
in patients with hematological malignancies: a retrospective multicenter study from Italy and review of the literature. J. Clin. Microbiol.43:1818– 1828.
16.Gunn, S. R., X. T. Reveles, J. D. Hamlington, L. C. Sadkowski, T. L. Johnson-Pais, and J. H. Jorgensen.2006. Use of DNA sequencing analysis to confirm fungemia due toTrichosporon dermatisin a pediatric patient. J. Clin. Microbiol.44:1175–1177.
17.Hata, K., J. Kimura, H. Miki, T. Toyosawa, T. Nakamura, and K. Katsu.
1996. In vitro and in vivo antifungal activities of ER-30346, a novel oral triazole with a broad antifungal spectrum. Antimicrob. Agents Chemother.
40:2237–2242.
18.Hazen, K. C.1995. New and emerging yeast pathogens. Clin. Microbiol. Rev.
8:462–478.
19.Herbrecht, R., J. Waller, P. Dufour, H. Koenig, B. Lioure, L. Marcellin, and F. Oberling.1992. Rare opportunistic fungal diseases in patients with organ or bone marrow transplantation. Agressologie33:77–80. (In French.) 20.Hoy, J., K. C. Hsu, K. Rolston, R. L. Hopfer, M. Luna, and G. P. Bodey.1986.
Trichosporon beigeliiinfection: a review. Rev. Infect. Dis.8:959–967. 21.Krcmery, V., Jr., F. Mateicka, A. Kunova, S. Spanik, J. Gyarfas, Z. Sycova,
and J. Trupl.1999. Hematogenous trichosporonosis in cancer patients: re-port of 12 cases including 5 during prophylaxis with itraconazol. Supre-port Care Cancer7:39–43.
22.Kustimur, S., A. Kalkanci, K. Caglar, M. Dizbay, F. Aktas, and T. Sugita.
(ed.).2002. Tratado de micologia medica, 9th ed. Sarvier Publishers, Sa˜o Paulo, Brazil.
24. Reference deleted.
25.Marty, F. M., D. H. Barouch, E. P. Coakley, and L. R. Baden.2003. Disseminated trichosporonosis caused byTrichosporon loubieri. J. Clin. Microbiol.41:5317–5320.
26.Matsue, K., H. Uryu, M. Koseki, N. Asada, and M. Takeuchi.2006. Break-through trichosporonosis in patients with hematologic malignancies receiv-ing micafungin. Clin. Infect. Dis.42:753–757.
27.Melo, A. S., L. P. de Almeida, A. L. Colombo, and M. R. Briones.1998. Evolutionary distances and identification of Candida species in clinical iso-lates by randomly amplified polymorphic DNA (RAPD). Mycopathologia
142:57–66.
28.Meyer, M. H., V. Letscher-Bru, J. Waller, P. Lutz, L. Marcellin, and R. Herbrecht.2002. Chronic disseminated Trichosporon asahii infection in a leukemic child. Clin. Infect. Dis.35:e22–e25.
29.Middelhoven, W. J., G. Scorzetti, and J. W. Fell.2004. Systematics of the anamorphic basidiomycetous yeast genusTrichosporonBehrend with the description of five novel species:Trichosporon vadense,T. smithiae,T. de-hoogii,T. scarabaeorumandT. gamsii. Int. J. Syst. Evol. Microbiol.54:975– 986.
30.Nakajima, M., T. Sugita, and Y. Mikami.2007. Granuloma associated with
Trichosporon asahiiinfection in the lung: unusual pathological findings and PCR detection ofTrichosporonDNA. Med. Mycol.45:641–644.
31.NCCLS/CLSI.2002. Reference method for broth dilution antifungal suscep-tibility testing of yeasts, 2nd ed. Approved standard M27-A2. National Com-mittee for Clinical Laboratory Standards, Wayne, PA.
32.Padhye, A. A., S. Verghese, P. Ravichandran, G. Balamurugan, L. Hall, P. Padmaja, and M. C. Fernandez.2003.Trichosporon loubieriinfection in a patient with adult polycystic kidney disease. J. Clin. Microbiol.41:479–482. 33.Paphitou, N. I., L. Ostrosky-Zeichner, V. L. Paetznick, J. R. Rodriguez, E. Chen, and J. H. Rex.2002. In vitro antifungal susceptibilities ofTrichosporon
species. Antimicrob. Agents Chemother.46:1144–1146.
34.Pasqualotto, A. C., T. C. Sukiennik, L. C. Severo, C. S. de Amorim, and A. L. Colombo.2005. An outbreak ofPichia anomalafungemia in a Brazilian pediatric intensive care unit. Infect. Control Hosp. Epidemiol.26:553–558. 35.Perparim, K., H. Nagai, A. Hashimoto, Y. Goto, T. Tashiro, and M. Nasu.
1996. In vitro susceptibility of Trichosporon beigelii to antifungal agents. J. Chemother.8:445–448.
36.Pincus, D. H., S. Orenga, and S. Chatellier.2007. Yeast identification—past, present, and future methods. Med. Mycol.45:97–121.
37.Ramani, R., S. Gromadzki, D. H. Pincus, I. F. Salkin, and V. Chaturvedi.
1998. Efficacy of API 20C and ID 32C systems for identification of common and rare clinical yeast isolates. J. Clin. Microbiol.36:3396–3398. 38.Rodriguez-Tudela, J. L., T. M. Diaz-Guerra, E. Mellado, V. Cano, C. Tapia,
A. Perkins, A. Gomez-Lopez, L. Rodero, and M. Cuenca-Estrella.2005. Susceptibility patterns and molecular identification ofTrichosporonspecies. Antimicrob. Agents Chemother.49:4026–4034.
39.Rodriguez-Tudela, J. L., A. Gomez-Lopez, A. Alastruey-Izquierdo, E.
Mel-lado, L. Bernal-Martinez, and M. Cuenca-Estrella.2007. Genotype distri-bution of clinical isolates of Trichosporon asahii based on sequencing of intergenic spacer 1. Diagn. Microbiol. Infect. Dis.58:435–440.
40.Sidrim, J. J. C., and M. F. G. Rocha.2004. Micologia me´dica a` luz de autores comtemporaˆneos. Guanabara, Rio de Janeiro, Brazil.
41.Soll, D. R.2000. The ins and outs of DNA fingerprinting the infectious fungi. Clin. Microbiol. Rev.13:332–370.
42.Sugita, T., M. Nakajima, R. Ikeda, T. Matsushima, and T. Shinoda.2002. Sequence analysis of the ribosomal DNA intergenic spacer 1 regions of
Trichosporonspecies. J. Clin. Microbiol.40:1826–1830. 43. Reference deleted.
44.Sugita, T., A. Nishikawa, and T. Shinoda.1998. Rapid detection of species of the opportunistic yeastTrichosporonby PCR. J. Clin. Microbiol.36:1458– 1460.
45.Sugita, T., A. Nishikawa, T. Shinoda, and H. Kume.1995. Taxonomic position of deep-seated, mucosa-associated, and superficial isolates of
Trichosporon cutaneumfrom trichosporonosis patients. J. Clin. Microbiol.
33:1368–1370.
46.Te´llez-Castillo, C. J., M. Gil-Fortuno, I. Centelles-Sales, S. Sabater-Vidal, and F. Pardo Serrano.2008.Trichosporon asahiifatal infection in a preterm newborn. Rev. Chilena Infectol.25:213–215. (In Spanish.)
47.Tuon, F. F., G. M. de Almeida, and S. F. Costa. 2007. Central venous catheter-associated fungemia due to Rhodotorula spp.—a systematic review. Med. Mycol.45:441–447.
48.Uchida, K., Y. Nishiyama, N. Yokota, and H. Yamaguchi.2000. In vitro antifungal activity of a novel lipopeptide antifungal agent, FK463, against various fungal pathogens. J. Antibiot. (Tokyo)53:1175–1181.
49.Uzun, O., S. Arikan, S. Kocagoz, B. Sancak, and S. Unal.2000. Susceptibility testing of voriconazole, fluconazole, itraconazole and amphotericin B against yeast isolates in a Turkish University Hospital and effect of time of reading. Diagn. Microbiol. Infect. Dis.38:101–107.
50.Walsh, T. J.1990. Role of surveillance cultures in prevention and treatment of fungal infections. NCI Monogr.9:43–45.
51.Walsh, T. J., A. Groll, J. Hiemenz, R. Fleming, E. Roilides, and E. Anaissie.
2004. Infections due to emerging and uncommon medically important fungal pathogens. Clin. Microbiol. Infect.10(Suppl. 1):48–66.
52.Walsh, T. J., J. W. Lee, G. P. Melcher, E. Navarro, J. Bacher, D. Callender, K. D. Reed, T. Wu, G. Lopez-Berestein, and P. A. Pizzo.1992. Experimental
Trichosporoninfection in persistently granulocytopenic rabbits: implications for pathogenesis, diagnosis, and treatment of an emerging opportunistic mycosis. J. Infect. Dis.166:121–133.
53.Walsh, T. J., K. R. Newman, M. Moody, R. C. Wharton, and J. C. Wade.
1986. Trichosporonosis in patients with neoplastic disease. Medicine (Balti-more)65:268–279.
54.Wolf, D. G., R. Falk, M. Hacham, B. Theelen, T. Boekhout, G. Scorzetti, M. Shapiro, C. Block, I. F. Salkin, and I. Polacheck.2001. Multidrug-resistant
