Donor-Derived Transmission of Candida auris During Lung Transplantation

Abstract

CASE REPORT

A 71-year-old man with end-stage lung disease due to idiopathic pulmonary fibrosis and chronic obstructive lung disease was admitted to the Massachusetts General Hospital in anticipation of double lung transplantation. He had been placed on the United Network for Organ Sharing (UNOS) lung transplant waiting list 3 months earlier due to progressively declining pulmonary function. On admission, he exhibited dyspnea with minimal exertion; auscultation revealed bibasilar crackles. Lungs from a deceased donor with bronchiectasis became available. Before implantation, bronchoalveolar lavage (BAL) of the donor lungs was performed and specimens from right and left bronchi were submitted to the clinical microbiology laboratory for bacterial and fungal cultures. Bilateral sequential lung transplantation (donor cytomegalovirus [CMV] immunoglobulin G [IgG] positive, recipient CMV IgG negative) was performed with a good technical outcome. Intraoperative fiberoptic bronchoscopy showed intact, well-approximated anastomotic sites. Postoperative BAL cultures were obtained. Immunosuppression with methylprednisolone, mycophenolate mofetil, and tacrolimus, and prophylactic cefepime, vancomycin, micafungin, inhaled amphotericin B lipid complex, inhaled tobramycin, and CMV intravenous immunoglobulin were administered per protocol.

Gram staining and Calcofluor white preparations of the BAL specimens from the donor lungs pre- and postimplantation revealed abundant yeast. Two days after transplantation, abundant yeast were isolated in cultures of these specimens. One preimplantation culture also grew rare Pseudomonas aeruginosa; meropenem was substituted for cefepime. To identify the yeast, the isolates were subcultured to differential chromogenic medium (BD BBL CHROMagar Candida; Becton, Dickinson and Company, Sparks, Maryland). Initially white, the colonies became light pink within 48 hours and appeared moist. Because morphological characteristics did not allow identification, analysis with matrix-assisted laser desorption/ionization–time-of-flight mass spectrometry (MALDI-TOF MS) using the VITEK MS system, version 3.0 in vitro diagnostic database (bioMérieux, Durham, North Carolina) was attempted, but the isolates were not identified. Testing with VITEK 2 Yeast Identification Cards version 6.01 (bioMérieux) yielded an identification of Candida haemulonii for each isolate.

Candida auris was suspected based on the inability of many commercial identification systems to differentiate C. auris from other species, including C. haemulonii [1]. Given the clinical and epidemiological implications of C. auris isolation, the laboratory alerted the infectious disease physician, infection control leadership, and a Massachusetts Department of Public Health epidemiologist. Additional testing demonstrated that the yeast grew at 40°C and did not produce pseudohyphae; both features supported the identification of C. auris. Broth microdilution antimicrobial susceptibility testing using a Sensititre YeastOne panel (TREK Diagnostic Systems; Cleveland, Ohio) yielded the following minimum inhibitory concentrations (MICs): fluconazole, 4 µg/mL; voriconazole, 0.03 µg/mL; caspofungin, 0.12 µg/mL; micafungin, 0.12 µg/mL; 5-flucytosine, 0.12 µg/mL; amphotericin B, 2 µg/mL. An Organ Procurement Transplant Network Disease Transmission Advisory Committee donor-derived infection report was filed. The yeast was submitted to the Centers for Disease Control and Prevention (CDC) for definitive identification.

Five days after transplantation, the patient developed a large pneumothorax complicated by ventricular tachycardia, hypotension, and respiratory arrest. Tube thoracotomy was performed, vasopressors were administered, and the patient underwent endotracheal intubation for initiation of mechanical ventilation. The following day, fever (39.2°C) developed and the peripheral white blood cell count was 28 000 cells/μL (normal, 4500–11 000 cells/μL). Blood cultures were sterile, BAL cultures again grew yeast, and chest radiography revealed retrocardiac opacification. Intravenous liposomal amphotericin B and ciprofloxacin were added. Because of worsening hemodynamic instability, venovenous extracorporeal membrane oxygenation (ECMO) was initiated; this was successfully converted to venoarterial ECMO for additional cardiac and respiratory support. Heparin-induced thrombocytopenia was diagnosed on the eighth day after transplantation in the setting of declining platelet counts and detection of heparin PF4 IgG antibodies. During exchange of the heparinized ECMO circuit, circulatory flow was lost and cardiopulmonary resuscitation was unsuccessful. An autopsy was offered but declined by the family.

Three days later, CDC confirmed the identification of C. auris by MALDI-TOF MS using an optimized database (Bruker MALDI Biotyper, Billerica, Massachusetts). Sequencing showed the isolate to be closely related to other C. auris isolates from Illinois (verbal communication, Sharon Tsay, CDC). An inquiry to UNOS revealed that a premortem respiratory culture from the organ donor, who was from Illinois, had grown a yeast reported as C. haemulonii. A CDC investigation of possible epidemiological linkage between the organ donor and patients previously identified to have C. auris infection or colonization is underway.

DISCUSSION

Since initial description in 2009 in Tokyo [2], C. auris has been isolated from patients on 5 continents. Rather than resulting from spread of a single epidemic strain, the global emergence of C. auris has been attributed to the independent appearance of 4 distinct clades particular to South Asia, East Asia, South Africa, and South America [3]. Candida auris was first identified in the United States in 2013 from a fungemic patient in New York [4]. As of February 2017, CDC had confirmed >30 cases of C. auris, most in New York. Three had been reported in Illinois (where the organ donor in this case resided). Our patient represents the first case of C. auris reported in Massachusetts and the first in a solid organ transplant (SOT) recipient.

Several characteristics of C. auris make its worldwide emergence an important public health concern. First, while antifungal MICs observed in our case were relatively low, up to 50% of C. auris isolates have demonstrated elevated antifungal MICs, some to all available classes [3], albeit susceptibility testing with commercial systems and reference methods may produce discrepant results [5]. The Clinical and Laboratory Standards Institute has not yet developed species-specific breakpoints for C. auris; extrapolation from experience with treatment of other candidal infections suggests that such elevated MICs may translate into clinical resistance. Moreover, treatment failure (including persistent and recurrent fungemia) and development of resistance have been reported in the setting of seemingly adequate antimicrobial therapy [4, 6]. In addition, the mortality associated with C. auris infections has been exceedingly high, ranging from 44% to 60% [3, 4], possibly reflecting greater pathogenicity compared to other Candida species [7]. Second, C. auris can colonize body sites for prolonged periods and eradication from environmental surfaces is difficult [4, 8]. Environmental persistence has been linked to instances of nosocomial transmission, including an outbreak of 50 cases of C. auris infection spanning 16 months in a UK cardiothoracic center [8]. CDC recommends contact precautions and a hospital-grade disinfectant with a sporicidal claim for terminal room cleaning [9]. Third, microbial identification is challenging, often requiring molecular characterization. Commercial biochemical identification platforms in widespread use in clinical laboratories may misidentify C. auris [5, 10], as occurred in this case; VITEK 2 testing in our laboratory misidentified the isolate as C. haemulonii, and misidentification (by an unknown method) of the donor’s premortem isolate as C. haemulonii went unrecognized by the testing laboratory.

The majority of reported cases of C. auris infection have involved patients with comorbid medical conditions [3, 4, 8, 11]. However, few data exist about C. auris infections among transplant recipients, with only 1 reported case after bone marrow transplantation [4]. Our report of C. auris, the first in a SOT recipient, also represents the first description of donor-derived transmission. Candida auris was repeatedly isolated from respiratory cultures in the days following transplantation, but was thought to more likely represent respiratory colonization than infection. Moreover, prophylactic micafungin would likely have been effective against this isolate. But although it is unclear whether C. auris contributed to this patient’s demise, his poor outcome, taken together with previous reports of high mortality in patients with C. auris infection despite appropriate antifungal therapy, prompt consideration of whether isolation of this fungus might serve as a marker for severe illness or underlying immune dysfunction.

The possibility of donor-derived transmission of C. auris has several implications for organ transplantation. Yeast isolated from donor lung surveillance cultures are generally thought to be colonizers and are not uniformly identified to the species level, potentially leading to missed cases and onward nosocomial transmission. Likewise, the current inability of many clinical laboratories to identify C. auris may lead to delays and failures to recognize this pathogen. Updates to commercial identification systems are required; in the interim, laboratories must be vigilant about the pitfalls of misidentification. Multidrug-resistant strains may require modification of peritransplantation antifungal prophylaxis regimens. As most C. auris strains recovered from the United States have exhibited lower echinocandin MICs, prophylaxis should include an echinocandin when there is concern for potential donor-derived transmission [4]. As demonstrated by the United Kingdom nosocomial outbreak [8], C. auris poses a significant threat to immunosuppressed transplant recipients; isolation from a donor organ warrants careful consideration before transplantation. At present, there are few data to guide such decisions. While there have been no reports of transplant-related deaths directly attributable to C. auris, failures to correctly identify the organism may have precluded recognition as an important pathogen in other cases. If the prevalence of C. auris increases, current protocols for identification of yeast recovered from peritransplant surveillance cultures may need to be revisited, with consideration given to species-level identification. The benefits of organ transplantation must, on a case-by-case basis, be weighed against existing knowledge about the risks of C. auris infection when deciding whether to proceed with transplantation.

Notes

Acknowledgments. We thank Evan Mojica for technical assistance and Drs Sharon Tsay and Shawn R. Lockhart for their work on the epidemiological investigation and for confirmatory identification of C. auris at Centers for Disease Control and Prevention.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References