Candida auris is a globally emerging yeast that causes outbreaks in health care settings
and is often resistant to one or more classes of antifungal medications (
1
). Cases of C. auris with resistance to all three classes of commonly prescribed antifungal
drugs (pan-resistance) have been reported in multiple countries (
1
). C. auris has been identified in the United States since 2016; the largest number
(427 of 911 [47%]) of confirmed clinical cases reported as of October 31, 2019, have
been reported in New York, where C. auris was first detected in July 2016 (
1
,
2
). As of June 28, 2019, a total of 801 patients with C. auris were identified in New
York, based on clinical cultures or swabs of skin or nares obtained to detect asymptomatic
colonization (
3
). Among these patients, three were found to have pan-resistant C. auris that developed
after receipt of antifungal medications, including echinocandins, a class of drugs
that targets the fungal cell wall. All three patients had multiple comorbidities and
no known recent domestic or foreign travel. Although extensive investigations failed
to document transmission of pan-resistant isolates from the three patients to other
patients or the environment, the emergence of pan-resistance is concerning. The occurrence
of these cases underscores the public health importance of surveillance for C. auris,
the need for prudent antifungal prescribing, and the importance of conducting susceptibility
testing on all clinical isolates, including serial isolates from individual patients,
especially those treated with echinocandin medications. This report summarizes investigations
related to the three New York patients with pan-resistant infections and the subsequent
actions conducted by the New York State Department of Health and hospital and long-term
care facility partners.
Clinical C. auris cases were defined as those in which C. auris was identified in
a clinical culture obtained to diagnose or treat disease. Screening cases were defined
as those in which C. auris was identified by polymerase chain reaction testing and
culture, or by culture only, of a sample from an axilla, groin, or nares swab obtained
for the purpose of state public health surveillance (
2
). To assess ongoing colonization with C. auris, additional swabs were collected over
time from patients colonized with C. auris.
Wadsworth Center, the New York State public health laboratory, conducted testing to
confirm presumptive C. auris isolates from various health care facilities in New York
during August 2016–June 2019 by matrix-assisted laser desorption/ionization time-of-flight
mass spectrometry, using both the manufacturer’s and in-house validated library databases.
The laboratory also performed antifungal susceptibility testing for azoles and echinocandins
by broth microdilution and for amphotericin B, by E-test methods* as described previously,
and categorized isolates as resistant based on CDC’s tentative breakpoints (
1
,
2
). A pan-resistant isolate was defined as one with resistance to the triazole class
(fluconazole minimum inhibitory concentration [MIC] ≥32 μg/mL), polyene class (amphotericin
B MIC ≥2 μg/mL [E-test values of 1.5 rounded up to 2]), and echinocandins (anidulafungin
MIC ≥4 μg/mL, caspofungin MIC ≥2 μg/mL, micafungin MIC ≥4 μg/mL), tested at Wadsworth
Center with confirmation by the laboratory at CDC’s Mycotic Diseases Branch (
1
,
2
).
Epidemiologic investigation of patients with pan-resistant cases included collecting
clinical and exposure data, screening close contacts (persons who had an epidemiologic
link to a patient in place or time), and assessing infection control practices in
health care facilities that cared for the patients (
2
,
4
,
5
). When close contacts could be located, the New York State Department of Health attempted
to obtain swabs for culture.
Site visits involved observations of infection control practices, on-site education,
and point prevalence studies. During point prevalence surveys, samples were collected
from the nares, axilla, and groin of consenting patients. When possible, samples from
the environments of facilities where patients with pan-resistant infections were admitted
or resided were collected, with priority given to frequently touched surfaces and
objects in patients’ rooms.
As of June 28, 2019, a total of 801 patients with C. auris were detected in New York,
identified through clinical cultures (349) or skin or nares screening swabs only (452)
(
3
). Testing of the first available clinical isolates with susceptibilities revealed
that 276 of 277 (99.6%) were resistant to fluconazole, 170 of 277 (61.3%) were resistant
to amphotericin B, and none was resistant to echinocandins (
1
,
6
). Testing of subsequent available isolates obtained from infected patients with susceptibilities
revealed 330 of 331 (99.7%) were resistant to fluconazole, 210 of 331 (63.4%) were
resistant to amphotericin B, and 13 of 331 (3.9%) were resistant to echinocandins
(
1
,
6
). Three patients’ subsequent isolates were pan-resistant.
The first two patients with pan-resistant C. auris infections (patient A and patient
B) were aged >50 years and residents of long-term care facilities; each had multiple
underlying medical conditions, including ventilator dependence and colonization with
multidrug-resistant bacteria (Table). The two patients developed C. auris infections
in 2017 (patient A) and 2018 (patient B), and multiple samples obtained from them
had C. auris-positive cultures. Patient A had C. auris isolated from a central venous
catheter tip and later from blood and urine cultures; patient B had C. auris isolated
from a urine sample and a tracheal aspirate. All isolates were resistant to fluconazole;
seven of 13 (54%) isolates from patient A and three of five (60%) isolates from patient
B were resistant to amphotericin B; no isolates were initially resistant to echinocandins.
Neither patient was known to have received antifungal medications before the diagnosis
of C. auris infection, but both patients were treated with prolonged courses of echinocandins
after C. auris was identified. Patient A was also treated with amphotericin B. Cultures
taken after echinocandin therapy from both patients yielded C. auris isolates resistant
to fluconazole, amphotericin B, and echinocandins. Both patients died; the role of
C. auris in their deaths is unclear.
TABLE
Characteristics of three Candida auris cases with emergence of pan-resistance to antifungal
agents — New York, 2019
Characteristic
Patient A
Patient B
Patient C
Underlying condition
Chronic ventilator dependence
Chronic ventilator dependence, alcohol dependence
Acute mechanical ventilation, alcohol dependence, chronic skin condition
Antifungal medication received
Echinocandin, amphotericin B
Echinocandin
Echinocandin
Date pan-resistance confirmed
February 2019
March 2019
June 2019*
Sample type for pan-resistant isolate
Blood
Urine
Rectal swab
Time from first isolation of C. auris to collection of pan-resistant sample
22 mos
13 mos
2 mos
Time from isolation of pan-resistant C. auris to patient’s death
2 wks
3–4 wks
10 mos
MICs for pan-resistant isolates (μg/mL)†
Triazole class
Fluconazole
>256
>256
>256
Voriconazole
2
2
2
Posaconazole
0.25
0.5
0.25
Polyene class
Amphotericin B
2
2
2
Echinocardin class
Caspofungin
16
2
16
Anidulafungin
4
4
4
Micafungin
4
4
4
No. of facilities at which screening was conducted
1
2
1§
No. of contacts with C. auris/No. tested (%)
4/35 (11)
2/50 (4)
0/15§(0)
No. of contacts with pan-resistant C. auris
0
0
0§
No. of environmental surfaces and equipment with C. auris/No. tested (%)
14/36 (39)
3/28 (11)
1/11§ (9)
No. of environmental surfaces with pan-resistant C. auris
0
0
0§
Abbreviation: MIC = minimum inhibitory concentration.
* Isolate was from April 2017.
† Tentative CDC MIC breakpoints (
µg/mL): fluconazole, ≥32; voriconazole: N/A; amphotericin B, ≥2; caspofungin, ≥2;
anidulafungin ≥4; micafungin, ≥4.
https://www.cdc.gov/fungal/candida-auris/health-professionals.html
.
§ Data from an assessment of contacts and environments in March 2017, approximately
1 month before collection of the pan-resistant isolate; laboratory surveillance of
a sampling of Candida isolates from urine was also conducted.
No epidemiologic links were found between the two patients. They resided in and were
patients at different health care facilities in the same borough of New York City,
and neither patient had any known domestic or international travel. Point prevalence
surveys, environmental sampling, and infection control assessments were performed
at facilities where the two patients had resided to determine whether spread of the
resistant isolates occurred (
2
,
4
,
5
). No pan-resistant isolates were identified among contacts or on environmental surfaces
from the index patients’ rooms or common equipment (after discharge and terminal cleaning)
at the three facilities that had cared for these two patients; however, non–pan-resistant
C. auris was isolated from other patients and the environment at two of these facilities
and from the environment at the third facility. Additional infection control and cleaning
interventions were implemented by the facilities based on gaps identified during infection
control assessments.
After identification of patients A and B in 2019, a retrospective review of all New
York C. auris isolates and additional antifungal susceptibility testing at CDC identified
a third patient (patient C), from whom a C. auris isolate from 2017 was found to be
resistant to the three major antifungal classes. Patient C was also aged >50 years
and had multiple comorbidities and a prolonged hospital admission and long-term care
admission at facilities that were different (including in another borough) from those
that cared for patients A and B. The initial isolate of C. auris from patient C was
from a February 2017 blood culture; treatment with an echinocandin for 2 weeks followed.
Serial isolates obtained from February to early April 2017 were resistant to fluconazole,
had varying susceptibility to amphotericin B (11 of 17 [65%] total isolates resistant),
and were initially susceptible to echinocandins; the isolate resistant to all three
classes of antifungals was obtained from a rectal swab collected in late April 2017
to assess ongoing colonization following resolution of active infection. Patient C
was discharged to a long-term care facility (different from the facilities that cared
for patients A and B) on contact precautions. Subsequent serial surveillance cultures
from several body sites were obtained, and all remained negative for >6 months until
the patient died from underlying medical conditions. Patient C was not known to have
had any recent foreign or domestic travel and did not have any known contact with
patient A or patient B.
Isolates from all three patients were initially sensitive to echinocandins; resistance
was detected after treatment, indicating that it emerged during treatment with the
drugs. No evidence of transmission of the resistant isolates following these events
was found.
Discussion
The precise mechanism of resistance in these isolates is unknown, although echinocandin
resistance in other species of Candida is linked to mutations in the drug target protein
Fks1 (
7
). Approximately 3 years into the New York outbreak, these pan-resistant isolates
still appear to be rare, but their emergence is concerning. In other countries with
earlier emergence of C. auris, higher levels of echinocandin resistance and pan-resistance
have been reported (
8
). An isolate from Illinois with development of echinocandin resistance after echinocandin
treatment was recently described, although that isolate was susceptible to azoles
(
9
). The pan-resistant cases reported here were all from New York, where the South Asia
clade (clade 1) predominates (
5
). This clade is known to exhibit increased antifungal resistance compared to other
clades of C. auris (
8
). Surveillance for additional pan-resistant isolates in New York is ongoing.
Echinocandins are the treatment of choice for C. auris infections (
1
). Most New York C. auris strains are fluconazole-resistant, and most strains of C.
auris have been susceptible to echinocandins (
1
). However, because of the potential for development of resistance, patients on antifungal
treatment for C. auris should be monitored closely for clinical improvement, and follow-up
cultures should be obtained. Repeat susceptibility testing should also be conducted,
especially in patients previously treated with echinocandins. Consultation with an
infectious disease specialist is recommended, especially given the possibility of
emergence of pan-resistance.
These findings illustrate the need to continue surveillance for C. auris, encourage
prudence in the use of antifungal medications, and conduct susceptibility testing
on all clinical isolates, including serial isolates from a single patient, especially
those treated with echinocandins.
Summary
What is already known about this topic?
Candida auris is an emerging yeast that is often drug-resistant.
What is added by this report?
Three chronically ill patients in New York were identified as having pan-resistant
C. auris after receipt of antifungal medications. No transmission of the pan-resistant
isolates was found in patient contacts or the facility environments.
What are the implications for public health practice?
Three years after the first identification of C. auris in New York, pan-resistant
isolates remain rare. Continued surveillance for C. auris, prudent antifungal use,
and susceptibility testing for all C. auris clinical isolates (especially after patients
have been treated with antifungal drugs) are needed.