A Rare Fungal Species, Quambalaria cyanescens, Isolated from a Patient after Augmentation Mammoplasty – Environmental Contaminant or Pathogen?

Some emerging but less common human fungal pathogens are known environmental species and could be of low virulence. Meanwhile, some species have natural antifungal drug resistance, which may pose significant clinical diagnosis and treatment challenges. Implant breast augmentation is one of the most frequently performed surgical procedures in China, and fungal infection of breast implants is considered rare. Here we report the isolation of a rare human fungal species, Quambalaria cyanescens, from a female patient in China. The patient had undergone bilateral augmentation mammoplasty 11 years ago and was admitted to Peking Union Medical College Hospital on 15 September 2011 with primary diagnosis of breast infection. She underwent surgery to remove the implant and fully recovered thereafter. During surgery, implants and surrounding tissues were removed and sent for histopathology and microbiology examination. Our careful review showed that there was no solid histopathologic evidence of infection apart from inflammation. However, a fungal strain, which was initially misidentified as “Candida tropicalis” because of the similar appearance on CHROMagar Candida, was recovered. The organism was later on re-identified as Q. cyanescens, based on sequencing of the rDNA internal transcribed spacer region rather than the D1/D2 domain of 26S rDNA. It exhibited high MICs to 5-flucytosine and all echinocandins, but appeared more susceptible to amphotericin B and azoles tested. The possible pathogenic role of Q. cyanescens in breast implants is discussed in this case, and the increased potential for misidentification of the isolate is a cause for concern as it may lead to inappropriate antifungal treatment.


Introduction
Human beings live in a fungal-rich and fungal-diverse environment. Some emerging less common fungal pathogens are known environmental species (from soil, plants, insects, medical facilities, wastes or other outdoor or indoor environments). These fungal organisms are generally of low virulence and some may exhibit natural antifungal drug resistance [1][2][3], which presents clinical diagnosis and treatment challenges [4][5][6]. While some of the fungal infections can be diagnosed easily, in particular if isolated from blood, cerebrospinal fluid, etc., and with clear infection clinical signs, others present challenges in understanding their role in certain infections. Here we report an interesting case to highlight the challenges clinical pathologists and medical doctors faced in the decision making process of a case involving a rare fungus.
The Quambalariaceae is a family of fungi in the class Exobasidiomycetes. The family contains the single genus Quambalaria, which contains five species, including Quambalaria cyanescens, Q. coyrecup, Q. eucalypti, Q. pitereka and Q. simpsonii [7][8][9]. The first Q. cyanescens strain was isolated from human skin (strain no. CBS 357.73, type strain of the species) and reported as Sporothix cyanescens by de Hoog et al. in 1973 (Table 1) [10]. In 1987, Moore et al. erected the genus Cerinosterus, reset the previous S. cyanescens into this new genus and renamed it as Cerinosterus cyanescens [8]. However, a later study by analysis of partial large subunit (LSU)-rDNA sequences and the nutritional profile revealed that C. cyanescens was a close relative of Microstroma juglandis, but differed from other species within  [8].
Q. cyanescens is one of the rare clinical basidiomycetous pathogens. Most of Q. cyanescens isolated from the humans were reported in the 1990s, including pseudoepidemic nosocomial pneumonia cases reported in a US hospital [13], a possible pulmonary case in a heart transplant patient [14] and potential fungemia in lymphoma patients [11]. However, none of these published human-related cases deposited convincing molecular data.
In this case study, we report the mycology and molecular characteristics of a Q. cyanescens isolate from a 43 year-old female who previously received injected augmentation mammoplasty, and discuss the possible pathogenic role of the organism.

Ethics statement
The present case was from China Hospital Invasive Fungal Surveillance Net (CHIF-NET) study. Study protocol was approved by the Human Research Ethics Committee of Peking Union Medical College Hospital (No. S-263), and written consent was obtained from the patient. The woman was in good health status except for the inflammation of the breast and did not report any other major disease in her clinical history. The blood test results were all within normal values. Clinical examination showed that she was afebrile and no ulceration was present in her left breast. Primary diagnosis was made as left breast infection. Surgical operation was performed to take out the bilateral implants as per patient's request. However, no microbiological examination was done before surgery.

Clinical case
During surgery to remove the implants, it was noted that the yellow-brown semisolid implant had spilled and was mixed with unknown granule, and also there was damage in the mammary tissues. Partial implants and surrounding tissue were sent for histopathologic and microbiological laboratory examination. After surgery, cefmetazole (IV, 1 g bid) was given, combined with metronidazole (IV, 0.915 g, q12h) for 7 days. The patient fullyrecovered and was subsequently discharged on 24 September 2011 before the microbiology laboratory results were finalized. She didn't receive any antimicrobial or antifungal treatment since then, nor were any relapses reported at the 12-and 24-month follow-up visits.

Initial laboratory examinations
Microbiology and histopathology examinations were immediately performed on the partial implants and surrounding tissue from the left breast (16 September 2011). No other specimens were sent for microbiological testing. On histopathology examination, breast implants were found to be surrounded by fibrous capsules Table 1. Cont. and infiltrated with inflammation cells and phagocytosis by giant cells and capillary hypertrophy was also observed, which indicated foreign-body reaction. However, no solid evidence of bacterial or fungal infection was found.
In the meantime, bacterial culture was performed on the partial implants and tissue by inoculating them on Columbia agar supplemented with 5% sheep blood, China-blue lactose agar and chocolate agar. However, no fungal culture was performed initially as per surgeon's instructions. No bacteria were recovered. However, a notable amount (from the first to the second sector of the streaked plate) of yeast-like colonies were observed on Columbia blood agar on day 4 of incubation. Preliminary microscopic examination of the colonies showed yeast-like cells with a sympodial conidiogenesis. One pure colony of the isolate was then inoculated onto a chromogenic medium (CHROMagar Candida, CHROMagar Company, Paris, France) for identification, and was assigned as ''Candida tropicalis'' on day 8 based on the production of dark blue pigments. However, the patient had been discharged before the microbiology results were finalized.

Sequence-based identification
The above ''C. tropicalis'' strain was included in the CHIF-NET surveillance study (strain ID no. 11PU348). Genomic DNA was extracted by beating a fungal suspension with glass beads as described before [15]. Amplification of the fungal internal transcribed spacer (ITS) region and the D1/D2 domain of the 26S rRNA gene was performed as previously described with primer pairs ITS1/ITS4 and F63/R635, respectively [15][16][17]. The PCR products were sequenced in both directions using corresponding PCR amplification primer pairs at Ruibiotech Co. Ltd. (Beijing, China) using the DNA analyzer ABI 3730XL system (Applied Biosystems, Foster City, CA). Species identification was performed by comparing the obtained ITS and D1/D2 sequences against those in the Centraalbureau voor Schimmelcultures (CBS) Fungal Biodiversity Center database and GenBank using the BioloMICSNet and BLASTn software, respectively. A sequence similarity of 97% and 99% was applied as species identification 'cut-off' value for the ITS region and D1/D2 domain, respectively [18].

Nucleotide sequence accession numbers
The ITS region and D1/D2 domain sequences of strain 11PU348 were deposited in GenBank with accession numbers KF953496 and KF953497, respectively.

Phylogenetic analysis
The nucleotide sequence alignments within Q. cyanescens, using sequences of Q. cyanescens type strain CBS 357.73 as references, showed this species with little inter-species variation within both the ITS region (99.0% to 100%) and D1/D2 domain (99.6% to 100%) ( Table 1). Of note, the ITS region can clearly discriminate Q. cyanescens and other four Quambalaria species, with highest sequence similarity of less than 97.0%. However, the D1/D2 domain was not able to identify the five species within Quambalaria genus (sequence similarity .99.0%). The NJ analysis of the ITS region and D1/D2 domain yielded similar results (Figure 1).

Phenotypic characteristics on agar
Q. cyanescens isolate 11PU348 grew well at 28uC and 37uC, but failed to grow at 42uC on Sabouraud dextrose agar. By threesector streaking on Sabouraud dextrose agar, the strain had yeastlike colonies which were initially moist, smooth, of various sizes and white colored within 48 h at 28uC (Figure 2a), and turned to be creamy, butyrous and exuding dark-orange pigment after 72 h incubation (Figure 2b). However, the strain grew slower when incubated at 37uC, and tended to be mold-like, especially in the first sector of the streaked plates (Figure 2d and 2e).
After more than 2 weeks' incubation at either 28uC or 37uC, a pure culture of the organism yielded a typical filamentous fungi phenotype that appeared to be restricted, velvety, furrowed, compact and cerebriform, accompanied by a red pigment and a burgundy reverse color. The production of pigments was more obvious at 28uC than at 37uC (Figure 2c and 2f).
On CHROMagar Candida, the colonies of Q. cyanescens 11PU348 were dark blue hybridizing with white, which was very similar to the phenotype of C. tropicalis when incubated at 37uC for 48 h (Figure 2h), but generating dark-orange pigment when incubated at 28uC (Figure 2g).

Microscopic morphology
Yeast-form of Q. cyanescens 11PU348 showed the typical sympodial conidiogenesis, and had smooth-walled, obovoidal, solitary or bearing secondary conidia. The filamentous form of the strain showed hyphae which were regular, hyaline, smooth-walled, branched and suberect. The conidia formed by sympodial growth of the conidiogenous cells (primary conidia) mostly give rise to several secondary conidia.

Discussion
Q. cyanescens is rarely identified in the clinical microbiology laboratory, and its pathogenic role is still uncertain. A review of literature shows that this fungus was recovered primarily from individuals who were immunocompromised or debilitated [11,14], including a possible pulmonary case in a heart transplant patient [14], fungemia in lymphoma patients [11]. However, none of the above studies provided unequivocal clinical evidence of infection. In addition, Jackson et al. reported a pseudo-epidemic of Q. cyanescens pneumonia in a US hospital introduced by contamination of bronchoscopy suites [13], which suggests that the species may be an environmental contaminant in human patients.
We note that the pathogenic role of Q. cyanescens in this clinical case is questionable. There was no corroborating direct microscopic, histopathologic or serological evidence of fungal infection. Although the isolation was obtained from a specimen which showed histological signs of acute inflammation, this could be due to either real infection or foreign-body reaction. In addition, no samples (except routine bloods) other than the implant and the surrounding tissue removed during surgery were sent for laboratory examination, nor was repeat isolation attempted, as the present study was done retrospectively. The patient fully recovered after removal of implants, without any antifungal therapy administered. Although no other micro-organisms were isolated from this patient, and no fungal organisms were isolated from other patients who underwent plastic surgery during the same time-period, the possibility of environmental contamination cannot be excluded.
If this described case was due to a real infection, the slow progression of the inflammation, and the fact that the patient was both afebrile and asymptomatic with all blood test results within normal values, is consistent with an infection caused by a low virulence micro-organism. A previous experimental study in a murine model demonstrated that Q. cyanescens does have a low virulence potential [11].
Misidentification of Q. cyanescens 11PU348 was noted during confirmative identification process in CHIF-NET study [17]. Initially, two matrix-assisted laser desorption ionization-time of flight mass spectrometry systems (Vitek MS, bioMérieux, Marcy   Figure 2g and 2h: (i) Q. cyanescens 11PU348; (ii) C. glabrata sensu stricto 10H1043; (iii) C. albicans ATCC 90028; (iv) C. parapsilosis sensu stricto ATCC 22019; (v) C. krusei ATCC 6258; (iv) C. tropicalis 10H1048. C. glabrata sensu stricto 10H1043 and C. tropicalis 10H1048 were selected from CHIF-NET study [15]. doi:10.1371/journal.pone.0106949.g002 l'Etoile, France; Bruker Biotyper, Bruker Daltonics, Bremen, Germany) failed to identify strain 11PU348. Subsequent ITS sequencing identified the strain as Q. cyanescens. The main reason for the misidentification in the initial identification (that reported to clinic) was the yeast-like colonies with dark blue appearance at 48 h on CHROMagar Candida at 37uC, which was very similar to the appearance of C. tropicalis (Figure 2h). Although the patient in this case was cured by removal of the breast implant, the high MICs to 5-flucytosine and all echinocandins of Q. cyanescens were notable. Therefore, accurate identification of Q. cyanescens is important to avoid ineffective antifungal treatment. Mass spectra data of Q. cyanescens were neither represented in Vitek MS nor in Bruker Biotyper identification databases. Hence both MALDI-TOF MS systems assigned ''no identification'' to this isolate and importantly, did not misidentify the strain to another species. Although we were not able to identify Q. cyanescens against the current commercially available library of spectra, our result will nevertheless contribute to the existing spectral building.
In the most recent study, the ITS and D1/D2 sequences were used to cluster the Quambalaria genus and replaced Q. cyanescens species from another genus [8]. But in the present study, we found that the D1/D2 domain was not able to distinguish the different species within Quambalaria genus ( Figure 1; Table 1). Compared with the D1/D2 domain, the ITS region was accurate in the identification of Q. cyanescens and other species within this genus ( Figure 1; Table 1).

Conclusions
In conclusion, Q. cyanescens is a rare clinical basidiomycetous pathogen. Here we report a Q. cyanescens strain isolated from a patient after augmentation mammoplasty in China. The possibility of its real pathogenic role was discussed. The high MICs to 5flucytosine and all echinocandins highlight the importance of accurate identification so that appropriate therapy can be prescribed. To date, ITS sequencing remains the only available method to obtain an accurate identification result on this organism, while the pathogen is potentially misidentified as C. tropicalis by CHROMagar Candida.