Clinical Outcomes of Patients Treated for Candida auris Infections in a Multisite Health System, Illinois, USA

Candida auris is an emerging fungal pathogen that is typically resistant to fluconazole and is known to cause healthcare-associated outbreaks. We retrospectively reviewed 28 patients who had >1 positive culture for C. auris within a multisite health system in Illinois, USA, during May 2018–April 2019. Twelve of these patients were treated as inpatients for C. auris infections; 10 (83%) met criteria for clinical success, defined as absence of all-cause mortality, C. auris recurrence, and infection-related readmission at 30 days from the first positive culture. The other 2 patients (17%) died within 30 days. Most patients (92%) were empirically treated with micafungin. Four (14%) of 28 total isolates were resistant to fluconazole, 1 (3.6%) was resistant to amphotericin B, and 1 (3.6%) was resistant to echinocandins. Our findings describe low rates of antifungal resistance and favorable clinical outcomes for most C. auris patients.

2008-April 30, 2019; we excluded pregnant patients, prisoners, and patients <18 years of age. If a patient had multiple positive cultures for C. auris, we included only the first positive culture per hospital encounter. Patients who died before culture result were not included in clinical success evaluation. The study received a non-human subjects research determination from the Advocate Aurora Health Institutional Review Board.
The microbiology laboratory for this system primarily uses matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (Vitek MS, bioMérieux, https://www.biomerieux. com) for organism identification. Because our database does not include C. auris, for isolates not identified by our MALDI-TOF mass spectrometry system, we used Vitek 2 version 8.01 (bioMérieux), which we have been using since December 2017. We sent isolates identified as C. haemulonii, C. duobushaemulonii, and Candida spp. not identified by Vitek 2 to the Illinois Department of Public Health for additional testing using MALDI-TOF mass spectrometry or genomic sequencing to rule out misidentification of C. auris. During June 2018-April 2019, we sent all isolates identified as C. auris to the Illinois Department of Public Health, which forwarded them to CDC for whole-genome sequencing. We performed antifungal susceptibility testing with colorimetric microdilution by using Sensititer YeastOne YO9 (TREK Diagnostic Systems, https://www.trekds.com). Because no C. auris susceptibility breakpoints have been established, we used tentative breakpoints published by CDC for interpretation in this study (22).
We performed manual chart review for all patients. We evaluated patient charts for demographic information, infection source, culture source and susceptibilities, empiric and definitive therapy, length of hospital and ICU stay, clinical success, and reports of adverse events associated with treatment for C. auris infection. We defined clinical success as the absence of 30-day all-cause mortality, 30-day recurrence of the same organism, and 30-day infection-related readmission. We identified adverse drug events associated with antifungal therapy by reviewing patient laboratory results and progress notes from healthcare providers.

Results
We evaluated records from 8 hospitals from the period of January 1, 2008, through April 30, 2019, for C. auris isolates. Cultures were obtained as part of routine clinical practice. A total of 28 patients from 5 hospitals had >1 positive culture for C. auris during the study period. We included 28 C. auris isolates in this study (the first isolate from our health system was collected in May 2018). Members of the cohort had a median age of 70 years (interquartile range 62-78 years), and most (20 [71%]) patients were men. Most (26 [93%]) patients were admitted from a skilled nursing facility; 1 patient was transferred from another hospital, and 1 was admitted from the community. Nine (75%) patients required chronic mechanical ventilation, and 6 (21%) were receiving hemodialysis through a central line. Most isolates were cultured from blood (12 [43%]) or urine (10 [36%]). The median time from admission to collection of the first culture positive for C. auris was 0.14 days (interquartile range 0-0.88 days). The average hospital stay for inpatients was 12 days. Thirteen patients (46%) were admitted to an ICU; the average ICU stay was 3 days.
MICs for the 28 C. auris isolates (Table 1) showed that 4 (14%) were resistant to fluconazole, 1 (3.6%) was resistant to amphotericin B, and 1 (3.6%) was resistant to echinocandins, according to tentative C. auris breakpoints published by CDC (22). One isolate was resistant to fluconazole, amphotericin B, and echinocandins. This isolate was from a patient who was considered to be colonized with C. auris in the urine and did not receive antifungal therapy.
Twelve patients (43%) were treated as inpatients for C. auris infections ( Table 2). Of those patients who were not treated with antifungal therapy, 2 were evaluated in the emergency department and were discharged back to their skilled nursing facility in stable condition before their blood cultures results were available, 3 patients died before culture results were available, and 11 were considered to be colonized with C. auris. Of those patients who were treated for C. auris infections, most were found to have a central  8%]). Of those patients who received an azole as definitive treatment, all were being treated for BSIs. Treatment duration ranged from 5 to 26 days (mean 14 days). The only adverse event noted was an increase in aspartate aminotransferase from 25 to 91 U/L in 1 patient being treated with fluconazole. Fluconazole was continued, and the patient was discharged on fluconazole to complete their treatment course.
Ten (83%) patients met criteria for clinical success. No patients were found to have C. auris recurrence or infection-related readmission within 30 days of first positive culture. Two (17%) patients died within 30 days of first positive culture; both were being treated for CLABSIs.

Discussion
We report patient characteristics and microbiologic data for 28 patients with >1 positive culture for C. auris. We also describe the clinical outcomes for 12 patients treated for C. auris infections. Our observed mortality rate of 17% is lower than previously reported worldwide. A meta-analysis of 742 patients from 16 countries found an all-cause mortality rate of 30% (13). Our lower mortality rate might be a result of empirically selecting echinocandin therapy, to which >95% of our isolates were susceptible. Empiric selection of fluconazole for treatment of C. auris BSI was recently described in a case series in India (19). C. auris is often resistant to fluconazole, and lack of effective empiric therapy can result in poor outcomes. Because echinocandins are commonly empirically selected for treatment of candidemia in the United States, our lower mortality rates might be attributable to lower MICs in our geographic region. For instance, investigators of an outbreak in New York, New York, found a 45% mortality rate among 51 patients with BSIs at 90 days (10). The mortality rate in our study was 25% among patients with BSI. A separate case series from Brooklyn, New York, reported a 22% in-hospital mortality rate among a cohort of 9 patients with BSIs, which is similar to our findings (17). Small sample size and variations in underlying conditions might confound mortality rate comparisons, although another possible explanation for the lower mortality rate in our study is differences in antifungal resistance.
Previous reports of C. auris collections have noted substantially higher rates of antifungal resistance than what was observed in our cohort. In a grouping of 54 isolates from 5 countries, 93% were resistant to fluconazole, 35% to amphotericin B, and 7% to echinocandins (16). The first isolates from the United States demonstrated a similar pattern; 86% were resistant to fluconazole, 43% to amphotericin B, and 3% to echinocandins (18). An even higher rate of fluconazole resistance of 98% was noted in New York, New York (10). Drug-resistance mechanisms are genetically encoded, and some resistance mutations are linked to specific geographic clades (16). An epidemiologic investigation of 133 C. auris isolates from the United States showed that these isolates were genetically related to 1 of 4 major clades (from South America, Africa, East Asia, and South Asia) (20). Previous isolates from Illinois were found to be from the South America clade (15,20). The molecular epidemiology of the isolates we identified was beyond the scope of this investigation because the results of genomic sequencing were not shared with clinicians or laboratory personnel within our health system. Further research would be needed to determine whether these isolates are genetically distinct from clades previously noted to be present in the United States.
Most patients in this study were successfully treated with an echinocandin, which is consistent with treatment recommendations from the CDC (22). The 2 members of our cohort who died within 30 days of first positive culture attained microbiologic eradication with echinocandin therapy before they died. Three patients received azole antifungals as empiric or definitive treatment, and all 3 met our criteria for clinical success. However, treatment failures with fluconazole have been reported despite in vitro susceptibility (8,21). Echinocandin-resistant C. auris, possibly induced by antifungal pressure, has also been reported in the United States and in other countries (16,21). Patients with persistent or recurrent C. auris infections might require repeat susceptibility testing.
Several antifungal agents currently in development have activity against C. auris, including SCY-078, the first orally bioavailable 1,3-β-D-glucan synthesis inhibitor; VT-1598, a tetrazole-based lanosterol 14α-demethylase inhibitor; APX001, which interrupts glycosylphosphatidylinositol biosynthesis by inhibiting the fungal enzyme Gwt1; and CD101, an echinocandin that can be administered once weekly (12). In the future, these agents might become treatment options for C. auris infections, including those caused by isolates resistant to conventional therapies.
Most patients in our study who required treatment for C. auris were previously exposed to skilled nursing facilities and had multiple risk factors for invasive Candida infections, including central venous and urinary catheters (12). No patients were thought to have acquired C. auris infection during their hospital admission. The possibility of transmission within hospitals is a concern, given that C. auris has been shown to colonize the skin, persist on surfaces in the healthcare environment, and cause healthcare-associated outbreaks (10,21). Contact precautions, hand hygiene, and environmental cleaning and disinfection are essential to preventing the spread of C. auris, and these infection control practices are used within our health system (22).
Our study is limited by a small sample size and the inherent limitations of a retrospective, observational case series. In addition, evaluation of clinical outcomes at 30 days from the first positive culture prevented us from capturing any adverse outcomes that might have occurred after 30 days. However, our data provides insight on patient exposures. Most of these patients came from skilled nursing facilities. Targeted infection prevention and antimicrobial stewardship measures within these facilities might help to reduce the emergence and progression of resistance of this organism. Our experiences with C. auris are notable for favorable clinical outcomes and low rates of antifungal resistance. However, the development of resistance remains a concern, and patient response to treatment should be monitored closely.