Central line associated blood stream infection (CLABSI) due to Exophiala dermatitidis in an adult patient: Case report and review

Exophiala dermatitidis is a dematiaceous fungus with yeast-like and hyphal growth states that may cause cutaneous and visceral infections. Recently, E. dermatitidis has been linked to central line associated blood stream infection (CLABSI), probably due to its ability to produce extracellular polysaccharides and grow as biofilm. We describe an E. dermatitidis CLASBI. The strain was identified by morphological and molecular methods. E. dermatitidis CLASBI is highly uncommon, but seems to be increasing.


Introduction
Exophiala dermatitidis, previously known as Wangiella dermatitidis, is a dematiaceous, dimorphic fungus that may cause a large spectrum of human diseases [1,2]. As all phaeoid fungi ("phaeo" from the Greek meaning dark), Exophiala presents melanin on its cell wall [3], which enhances its survival in hostile environments [2]. Other aspects for pathogenicity include thermotolerance, adhesion, production of extracellular polysaccharide capsule, and biofilm formation [2]. Phenotypic identification of Exophiala spp. is hampered by its pleomorphic nature. Sequence data of the rDNA internal transcribed spacer (ITS) regions is reliable for species level diagnosis [4,5]. Additionally, proteomic identification using matrix assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF MS) has been described as a useful method for identification of E. dermatitidis [6].
E. dermatitidis infections have been classified into superficial infections; cutaneous and subcutaneous diseases; and systemic deep-seated infections. The latter is not consistently accompanied by fungemia, is most often reported in Asia, approximately 70% of patients are immunocompromised, and the mortality rate is up to 80% [7]. It has been occasionally reported as outbreaks due to contaminated pharmaceutical products [8,9].
CLABSI due to E. dermatitidis has recently been described, therefore, its clinical manifestations and prognosis are not well known yet [10].
We describe a case of CLABSI due to a melanized yeast ultimately identified as E. dermatitidis by morphotype, ITS sequencing, and MALDI-TOF MS.

Case
A 75-years-old man was admitted due to respiratory failure (day 0). The patient had history of colonic and prostatic cancer 11 and 7 years earlier, respectively, both successfully treated without recurrence. He also had a history of hypertension, alcoholism, smoking and chronic obstructive pulmonary disease. During his transfer to the hospital in ambulance, he was given intravenous infusion of furosemide. Upon arrival his physical exam was significant for fever, cough, tachypnoea and tachycardia He was admitted to the intensive care unit (ICU), requiring mechanical ventilation. Laboratory was remarkable for a white blood cell count of 10.5/mm3, C-reactive protein, lactate dehydrogenase and lactic acid were increased. An initial lung CT angiography showed bilateral ground glass pulmonary infiltrate without evidence of pulmonary embolism. With presumptive diagnosis of community acquired pneumonia, bronchoalveolar lavage (BAL) and BCs were done (Bactec aerobic medium; BD Diagnostic Instrument Systems; Bactec 9240). Fungal and bacterial stains and cultures from BAL were negative, and Galactomannan (PlateliaTM Aspergillus Ag) OD index was 0,45. Patient was started on piperacillin-tazobactam, vancomycin and hydrocortisone, and become afebrile at 24 hs.
On day 4, BCs were negative, thus vancomycin was discontinued. On day 6, after a short period of stabilization, he became hypotensive requiring inotropic assistance. fluid culture was negative. During the procedure a liver biopsy was done, later showing cirrhosis. On day 9 the pair of BCs taken at day 6 showed a positive growth index. Direct microscopy showed hyaline, ovoid to elliptical yeasts ( Fig. 1). A new set of BCs were drawn in order to confirm that finding. On day 13, yeasts were found on all BCs samples from days 6 and 9. The patient denied outdoor activity or recent traveling. No skin lesions were found on physical exam, and the portal of entry remained unclear. After 3 days, subcultures on Sabouraud agar showed slow-growing colonies, initially with smooth glossy mucous appearance, that over time became velvety olivaceous black (Fig. 2 A, B). Microscopy revealed pigmented septate branched hyphae with annelidic conidiogenesis, and ellipsoidal conidia of different sizes with a thin wall, forming aggregates (Fig. 3). The isolate was identified as Exophiala spp. MALDI-TOF (Bruker Daltonics) identified the colonies as E. dermatitidis with a 1.689 score. Patient was diagnosed of CLASBI due to non-Candida fungus according to CDC definition. Anidulafungin was started and CVC was removed. The strain was submitted to the national mycology reference center ("Departamento de Micología, Instituto Nacional de Enfermedades Infecciosas Dr. Carlos G. Malbrán"), for further for molecular identification. Sequence data of the rDNA ITS regions of the D1-D2 of the large (28S) ribosomal subunit of the isolate resulted in 98.6% similarity to E. dermatitidis. Antifungal susceptibility testing was performed according the methodology recommended by the CLSI, document M38-A2 (2008), revealing a MIC (μg/mL) for amphotericin B, anidulafungin, and caspofungin of 0.125, 0,008 and 0,008 respectively.
Neither renal ultrasound nor ophthalmologic examination, reveal evidence of disseminated fungal infection. Standard histological stains for fungi were requested on liver histological sample with negative results. Histological diagnosis of cirrhosis was done.
BCs done on day 16 were negative. On day 18 the patient died due to haemoptysis and supraventricular tachycardia.
Due to described outbreaks of Exophiala spp. caused by medication contamination, audit and surveillance of the practices of preparation and administration of intravenous medication was made. No irregularities were found. ICU drug preparation surfaces were cultured in search for fungi, which were negative for Exophiala spp. Surveillance for secondary cases in ICU patients was conducted by incubated all BCs in a prolonged manner (14 days) during a period of 3 months from the index case. No other patient presented E. dermatitidis fungemia.

Discussion
Melanized fungi cause a diverse range of diseases. Although traditionally associated with chronic cutaneous and subcutaneous infections, recently, there seems to be an increasing frequency of systemic forms [3]. The genus Exophiala, order Chaetotheriales, class Ascomycota [3], contains about 40 species [11]. E. dermatitidis is a polyextremotolerant, oligotrophic, metabolically versatile fungus, with noteworthy surviving strategies, and enormous ecological plasticity, capable of survive nutritionally poor or toxic habitats. By modifying its morphology and behavior, E. dermatitidis may move from natural to human habitat and to the human host [11].
Although usually named as a¨yeast¨, E. dermatitidis is polymorphic, thus, able to produce various morphological structures (black yeasts, hyphae, and sclerotic bodies) [3,12]. It has been named as dimorphic due to its capacity of changing between waterborne yeast (submersed growth), and filamentous stage (hydrophobic hyphal growth), this through a yeast-hyphal switch in simultaneously or consecutively occurring phases [1,2,13]. Dried fungal morphotypes (mycelia and conidia) are able of surviving repeated exposures to 60-80°C. When E. dermatitidis gets in an aquatic medium (tap water, pharmaceutical products, bloodstream), far from dying, it develops the submerged morphotype (yeasts), which is able to grow at temperatures of 37°C, produce extracellular polysaccharides around yeast cells (capsule) [11], and grow in a biofilm mode [8]. In the human host, the yeast form is   able to disseminate hematogenously; while the hyphal growth usually causes localized infection [2]. Melanin is a complex polymer deposited in the cell wall, that enhances the survival of fungi in hostile environments, conferring them resistance to multiple ecological and biological threats, including phagocytosis [2]. It also contributes to the organism's ability to elude host immunity through blocking the effects of hydrolytic enzymes and free radicals liberated by phagocytic cells [2,4]. E. dermatitidis has been isolated in various human-made environments, such as dishwashers, saunas and steam baths [13,14]. Also, gastrointestinal tract (GIT) asymptomatic carriage, as well as colonization of the respiratory tract in cystic fibrosis patients have been described [2,15]. Due to its remarkable adaptive plasticity, E. dermatitidis is capable of causing a wide range of infectious diseases, depending on the route and mode of entry, and the host immunological condition.
Cutaneous and subcutaneous infections usually take place throughout direct entry through the skin, but the route of infection of disseminated cases is not always ascertained [2]. When E. dermatitidis gets into the bloodstream, it may cause systemic, disseminated, often neurotropic infections with deep organ involvement (endocarditis, pneumonia and particularly brain abscess). Invasive E. dermatitidis disease is rare and most cases have been reported in Asia [2], but may occur in outbreaks due to direct iatrogenic inoculation. In 2002 an outbreak of E. dermatitidis meningitis or arthritis related to epidural or intra-articular injections of contaminated methylprednisolone acetate was described [9].
Recently described cases of CLASBI due to E. dermatitidis are even more unusual [7]. When bloodstream invasion occurs through a CVC, E. dermatitidis finds the opportunity to adhere, produce extracellular matrix and grow as biofilm, which protects the fungus against host defenses and anti-infective agents [1].
Outbreaks of CLASBI due to intravenous infusion of fungus-contaminated medication through CVC have been described. In an outbreak at an oncology clinic [8], despite the fact that all patients were exposed to a contaminated intravenous solution, only those with CVC developed infection [8].
To our knowledge there have been 6 cases reported of sporadic (not outbreak-related) CLASBI due to E. dermatitidis in adult patients [7,10,[16][17][18][19] (Table 1). Five of these patients had neoplasia, and all the patients had CVC for chemotherapy or nutritional support [7]. The presence of a CVC may be necessary for E. dermatitidis to establish bloodstream infection. Our patient did not have neoplastic active disease. He was a critical care patient with CVC.
E. dermatitidis might be underestimated as etiologic agent of CLASBI owing to it slow growth [15]. Therefore, in ICU patients with suspected CLASBI without microbiological diagnosis, prolonged incubation of BCs may be useful. Rare invasive fungal infections should lead to consider unusual sources of exposition and windows of opportunity for microbial growth [11]. The source and route of infection in our case remains elusive. There were no other cases that could suggest the existence of contaminated infusion, and E. dermatitidis was not documented in the ICU surfaces that were sampled. Although prolonged incubation of BC from all ICU patients was done during a 3 months period, in order to detect a possible hospital outbreak due to E. dermatitidis, there were no additional further cases. The patient did not present cutaneous lesions; but he had the received intravenous infusion in an ambulance, which  could not be tracked. E. dermatitidis has been found in GIT and fungemia in seriously ill patients may occur due to translocation of fungi from edematous or ischemic segments of GIT [13,19]. We didn't perform culture for fungi in feces. We consider that the fungus could have entered bloodstream during the intravenous infusion received in the ambulance, or by translocation from GIT. In any case, it has been shown that 92% of clinical isolates of E. dermatitidis exhibit biofilm formation [1]. The CVC may have favored adherence, biofilm formation and persistence of fungemia. This has been the first case of bloodstream isolation of Exophiala spp. in our Hospital.
Although Exophiala can be identified to the genus level by morphological characteristics, species identification must be complemented by molecular tests [19]. Sequencing of the ITS regions of rDNA is recommended [2,5]. Although MALDI-TOF MS has been described as an accurate method for identification of E. dermatitidis [14], in our case the identification score was found within the "not reliable" species match range (bellow 1.7).
We suggest susceptibility testing to guide treatment in order to choose the drug with the lowest MIC and good penetration at the infection site. Although monotherapy with an echinocandin does not seems to be optimal, the MIC to anidulafungin in our strain was notably low (0,008 mg/L), the patient didn't have evidence of deep organ seated infection, and echinocandins show good activity against biofilms, thus we didn't change therapy. Colistin has shown in vitro activity against Exophiala spp [19] and may be useful combined with antifungals for treating fungal biofilm-related infections, such as CLASBI [1]. Contrasting previous reports, in our case echinocandins MIC was low, and seems to be the lowest MIC described in clinical isolates.
To our knowledge, this is the first reported case of E. dermatitidis CLASBI in Argentina. E. dermatitidis seems to be expanding its borders beyond Asia. Although no definite treatment guidelines have been proposed for the management of E. dermatitidis infections, in case of CLASBI, immediate removal of CVC and targeted antifungal therapy guided by MIC, are key for improving outcome [7].

Conflict of interest
Authors have no conflict of interest. The authors have obtained written and signed consent to publish the case report from the patient's legal guardian.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.