In 2011, we initiated a sentinel surveillance network to assess changes in Clostridioides (formerly Clostridium) difficile antimicrobial susceptibility to fidaxomicin from 6 geographically dispersed medical centers in the United States. This report summarizes data from 2013 to 2016. C. difficile isolates or toxin-positive stools from patients were referred to a central laboratory. Antimicrobial susceptibility was determined by agar dilution. CLSI, EUCAST, or FDA breakpoints were used, where applicable.
KEYWORDS: Clostridioides difficile, epidemiology, fidaxomicin, surveillance
ABSTRACT
In 2011, we initiated a sentinel surveillance network to assess changes in Clostridioides (formerly Clostridium) difficile antimicrobial susceptibility to fidaxomicin from 6 geographically dispersed medical centers in the United States. This report summarizes data from 2013 to 2016. C. difficile isolates or toxin-positive stools from patients were referred to a central laboratory. Antimicrobial susceptibility was determined by agar dilution. CLSI, EUCAST, or FDA breakpoints were used, where applicable. Toxin gene profiles were characterized by multiplex PCR on each isolate. A random sample of approximately 40% of isolates, stratified by institution and year, was typed by restriction endonuclease analysis (REA). Among 1,889 isolates from 2013 to 2016, the fidaxomicin MIC90 was 0.5 μg/ml; all isolates were inhibited at ≤1 μg/ml. There were decreases in metronidazole and vancomycin MICs over time. Clindamycin resistance remained unchanged (27.3%). An increase in imipenem resistance was observed. By 2015 to 2016, moxifloxacin resistance decreased in all centers. The proportion of BI isolates decreased from 25.5% in 2011 to 2012 to 12.8% in 2015 to 2016 (P < 0.001). The BI REA group correlated with moxifloxacin resistance (BI 84% resistant versus non-BI 12.5% resistant). Fidaxomicin MICs have not changed among C. difficile isolates of U.S. origin over 5 years post licensure. There has been an overall decrease in MICs for vancomycin, metronidazole, moxifloxacin, and rifampin and an increase in isolates resistant to imipenem. Moxifloxacin resistance remained high among the BI REA group, but the proportion of BI isolates has decreased. Continued geographic variations in REA groups and antimicrobial resistance persist.
INTRODUCTION
Clostridioides (formerly Clostridium) difficile infection (CDI) continues to be a problem worldwide (1, 2), causing substantial morbidity and mortality. Approximately 450,000 cases of CDI occur annually in the United States, with approximately 29,300 deaths and 83,000 first recurrences (3). The elderly are particularly vulnerable. The U.S. 2012 national estimate for CDI incidence in nursing homes was approximately 100,000 cases, with approximately 20,000 recurrences and 8,700 deaths (4). The appearance of fluoroquinolone-resistant C. difficile strain NAP1/BI/027 isolates, which have been associated with epidemics of complicated CDI cases and increased mortality, has increased the urgency of this problem. The economic burden of CDI is also considerable. Two studies estimate an annual U.S. expenditure in the range of $4.8 billion to $5.4 billion (5, 6). A number of studies performed from Europe confirm a high cost burden as well (7). Thus, development and use of effective treatments for C. difficile infection that provide a durable cure are imperative. In addition, reduced susceptibilities to metronidazole and vancomycin have been noted, including in our previous surveillance efforts (8, 9). Ongoing national surveillance efforts are critical to allow for drug resistance monitoring in C. difficile to both metronidazole and vancomycin, as well as to newly licensed therapies that are entering into clinical practice.
Because diagnosis of CDI is based on detection of the large clostridial toxins ToxA and ToxB by enzyme-linked immunosorbent assay (ELISA) and/or detection of C. difficile-specific DNA sequences, isolation of C. difficile from patient stool is not standard clinical microbiology laboratory practice. Strain isolation is performed in only a few reference laboratories. Lack of available isolates limits epidemiologic surveillance of strain characteristics and drug susceptibility testing. To meet this gap, we previously created a U.S. network to provide the scientific community (researchers, practitioners, clinical laboratories, and regulatory agencies) with accurate information on the changing drug susceptibility and strain epidemiology of C. difficile (9). We reported the drug susceptibility patterns of 925 isolates isolated from patients who developed CDI from 2011 to 2012, located in geographically diverse areas. Geographic variations in susceptibility, restriction endonuclease analysis (REA) group, and binary toxin gene presence were observed. Moxifloxacin and clindamycin resistance was higher among BI and binary toxin-positive isolates. This group also showed reduced susceptibility to metronidazole and vancomycin by EUCAST criteria. From 2011 to 2012, we observed uniform susceptibility to fidaxomicin.
Here, we report on our continued efforts in surveillance and susceptibility testing on 1,889 isolates obtained in a similar fashion over the 2013 to 2016 time period and compare these data in 2-year time intervals to the 2011 to 2012 baseline data.
(This study was presented in part at the 52nd Interscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA, 9 to 12 September 2012, and at the 53rd Interscience Conference on Antimicrobial Agents and Chemotherapy, Denver, CO, 10 to 13 September 2013.)
RESULTS
Susceptibility of the isolates.
The results from the antimicrobial susceptibility testing of all 1,889 isolates are presented in Table 1 and expressed as MIC range, MIC50, MIC90, and percentage resistant based on CLSI and/or EUCAST breakpoints.
TABLE 1.
Antibiotic | MIC range (μg/ml) | MIC50 (μg/ml) | MIC90 (μg/ml) | % Resistant |
|
---|---|---|---|---|---|
CLSI | EUCAST | ||||
Fidaxomicin | 0.004–1 | 0.25 | 0.5 | NA | NA |
Rifaximin | ≤0.004 to >4 | 0.03 | 0.06 | NA | NA |
Rifampin | ≤0.004 to >4 | ≤0.004 | 0.008 | NA | 29.8 |
Tigecycline | ≤0.06–1 | 0.12 | 0.25 | 0.0 | 1.0 |
Vancomycin | ≤0.25–8 | 1 | 2 | NA | 6.5 |
Imipenem | ≤0.12–16 | 4 | 8 | 6.2 | NA |
Moxifloxacin | ≤0.5 to >16 | 2 | 16 | 23.2 | 23.2 |
Metronidazole | ≤0.06–4 | 0.25 | 1 | 0.0 | 1.3 |
Clindamycin | ≤0.5 to >16 | 4 | 16 | 27.6 | NA |
Chloramphenicol | ≤0.5–16 | 4 | 8 | 0.0 | NA |
All isolates were inhibited at a fidaxomicin concentration of ≤1 μg/ml, and the MIC90 was 0.5 μg/ml. There has been no change in fidaxomicin resistance over time from 2011 through 2016.
The number of isolates of C. difficile inhibited at each drug dilution for all of the agents tested is shown in Table 2. Figure 1 shows overall trends in 2-year intervals for antimicrobial resistance according to either EUCAST or CLSI breakpoints, where applicable. Although there is no CLSI resistance breakpoint for vancomycin, the recently published CLSI epidemiological cutoff value (ECV) for non-wild-type C. difficile is ≥4 μg/ml, which is consistent with the EUCAST epidemiological cutoff (ECOFF) value. In this study, the MIC50 and MIC90 were 1 and 2 μg/ml, respectively, in both 2013 to 2014 and 2015 to 2016 (Table 1); whereas, in the 2011 to 2012 baseline period, the MIC50 and MIC90 were 2 and 4 μg/ml. Using these epidemiological cutoff values, 17.9% of isolates in the baseline period had decreased susceptibility to vancomycin, but this proportion diminished over time to a low of 5.5% in 2015 to 2016 (Fig. 1). In 2011 to 2012, 3.6% of isolates demonstrated decreased susceptibility to metronidazole by EUCAST ECOFF values; this proportion also diminished over time to a low of 0.6% in 2015 to 2016. There were no isolates resistant to metronidazole by the CLSI breakpoint. Overall, 33.5% of isolates were moxifloxacin resistant in 2011 to 2012; this decreased over time to 25.7% in 2013 to 2014 and 20.7% in 2015 to 2016 (P < 0.001). Tigecycline resistance remained low by the EUCAST breakpoint; no isolate was considered resistant to tigecycline by the FDA breakpoint. Isolates showed increasing susceptibility to rifampin over time, with 65.3% of isolates resistant by EUCAST breakpoints in 2011 to 2012, decreasing to 39.7% of isolates resistant in 2013 to 2014 and 19.7% of isolates resistant in 2015 to 2016. Clindamycin resistance remained relatively stable. In 2011 to 2012, 23.8% of isolates were resistant to clindamycin; this rose only slightly over time to 27.9% in 2013 to 2014 and 27.3% in 2015 to 2016 (P = 0.0894). Imipenem resistance increased and reached 8.1% in 2015 to 2016 compared to 2.3% in 2011 to 2012 and 4.4% in 2013 to 2014 (P < 0.001).
TABLE 2.
Antimicrobial agent | No. of isolates at each MIC (μg/ml) value: |
||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
0.004 | 0.008 | 0.015 | 0.03 | 0.06 | 0.12 | 0.25 | 0.5 | 1 | 2 | 4 | 8 | 16 | |
Fidaxomicin | 23a | 19 | 17 | 74 | 239 | 687 | 1,026 | 661 | 68 | 0 | |||
Rifaximin | 71 | 165 | 1,167 | 1,073 | 68 | 29 | 15 | 2 | 0 | 3 | 221b | ||
Rifampin | 1,649 | 841 | 76 | 14 | 4 | 4 | 1 | 1 | 3 | 4 | 217c | ||
Tigecycline | 129 | 1,934 | 721 | 26 | 4 | 0 | 0 | 0 | 0 | ||||
Vancomycin | 9 | 131 | 1,461 | 925 | 284 | 4 | 0 | ||||||
Imipenem | 7 | 9 | 13 | 37 | 305 | 1,337 | 967 | 139 | |||||
Moxifloxacin | 47 | 559 | 1,430 | 29 | 95 | 654d | |||||||
Metronidazole | 10 | 57 | 1,052 | 828 | 423 | 385 | 59 | 0 | 0 | ||||
Clindamycin | 124 | 293 | 596 | 1,066 | 285 | 450e | |||||||
Chloramphenicol | 5 | 44 | 199 | 1,396 | 1,139 | 31 |
Some MICs may be lower than lowest the dilution tested. Number shown is equal to or less than that dilution.
Rifaximin: 215/221 MICs were >4.
Rifampin: 215/221 MICs were >4.
Moxifloxacin: 167/654 MICs were >16.
Clindamycin: 186/450 MICs were >16.
Determination of toxin gene profiles.
The majority of isolates from 2013 to 2016 were able to be assayed for a toxin gene profile (Table 3). Compared to the baseline period of 2011 to 2012, a similar proportion of isolates were nontoxigenic (∼3% each year), with the vast majority testing positive by PCR for both tcdA and tcdB (one isolate did not test positive for tcdA with the primers used). The percentage of toxigenic isolates testing positive for binary toxin genes (cdtA positive and cdtB positive) in the baseline 2011 to 2012 period was 33.6% (299/891). This decreased in the ensuing years of surveillance, with only 25.0% in 2013 to 2014 and 23.3% in 2015 to 2016 testing positive by PCR for binary toxin genes. This difference was primarily in isolates with the toxin profile tcdA, tcdB, cdtA and cdtB, tcdC 18-bp deletion. This toxin profile is the most common one associated with the NAP1/BI/027 strain and continued to correlate with BI over time (see Fig. S1A through C in the supplemental material).
TABLE 3.
Toxin gene profile | Base pair deletion in tcdC | Percent of isolates with the given profile duringa
: |
||
---|---|---|---|---|
2011 to 2012 | 2013 to 2014 | 2015 to 2016 | ||
tcdA positive, tcdB positive | 0 | 59.0 | 67.4 | 69.6 |
tcdA positive, tcdB positive, cdtA positive, and cdtB positive | 18 | 28.1 | 19.0 | 17.6 |
tcdA positive, tcdB positive | 18 | 5.8 | 5.3 | 4.5 |
tcdA positive, tcdB positive, cdtA positive, and cdtB positive | 39 | 2.3 | 3.4 | 2.8 |
tcdA positive, tcdB positive, cdtA positive, and cdtB positive | 0 | 2.1 | 1.8 | 2.1 |
tcdA positive, tcdB positive, cdtA positive, and cdtB positive | 54 | 0.2 | 0 | 0 |
tcdA negative, tcdB positive, cdtA positive, and cdtB positive | 0 | 0 | 0 | 0.1 |
Nontoxigenic | 2.5 | 3.1 | 3.3 |
In 2011–2012, 914 (98.8%) isolates were profiled; in 2013–2014, 945 (99.7%) isolates were profiled; and in 2015–2016, 941 (100%) isolates were profiled.
REA strain typing.
A random sample subset, stratified by center, was chosen for REA typing in each surveillance period (approximately 40% in each year). The distribution of all isolates by REA group by 2-year intervals is shown in Table 4. There was a significant decrease in BI isolates over time (P < 0.001). BI comprised 25.5% of isolates that were REA typed in 2011 to 2012 but had decreased by one-quarter in 2013 to 2014 (16.9%) and then by one-half (12.8%) in 2015 to 2016. When all centers were analyzed, the proportion of nonspecific REA groups increased over time (P < 0.001), while other REA groups stayed relatively constant. Figure S1 shows relationships between REA group and toxin profile by 2-year intervals, which remained relatively consistent.
TABLE 4.
REA group | Percent of isolates in group during: |
||
---|---|---|---|
2011 to 2012 (n = 322) | 2013 to 2014 (n = 356) | 2015 to 2016 (n = 360) | |
A group | 4.0 | 3.9 | 4.2 |
BI group | 25.5 | 16.9 | 12.8 |
BK group | 2.2 | 2.2 | 2.8 |
CF group | 2.5 | 2.0 | 1.7 |
DH group | 10.2 | 12.6 | 13.3 |
G group | 4.3 | 7.3 | 5.3 |
J group | 3.7 | 2.2 | 1.7 |
K group | 1.9 | 1.7 | 1.1 |
W group | 0.9 | 0.6 | 0.0 |
Y group | 15.8 | 15.7 | 15.6 |
Nonspecific | 28.9 | 34.8 | 41.7 |
We noted a significant difference between the BI REA group and others with respect to antimicrobial resistance (Table 5). Differences in MIC50 and MIC90 were noted between BI and non-BI strains for rifaximin, rifampin, vancomycin, moxifloxacin, metronidazole, and clindamycin. The MIC50 for imipenem for BI isolates was 8 μg/ml versus 4 μg/ml for non-BI isolates, but the MIC90 was the same. No difference was noted for tigecycline. Moreover, for vancomycin, 29.3% of EUCAST ECOFF values were above the cutoff for BI strains and only 7.5% were above the cutoff for non-BI strains. Metronidazole demonstrated a similar difference in EUCAST ECOFF values with 7.4% above the cutoff for BI strains and 0.4% above the cutoff for non-BI strains. Moxifloxacin demonstrated 84% resistance using CLSI breakpoints for the BI group compared to 12.5% for the non-BI group. Finally, the percentage of clindamycin resistance was much higher in BI (48.4%) than in non-BI (23.5%) strains.
TABLE 5.
Antibiotic | REA group | MIC range (μg/ml) | MIC50 (μg/ml) | MIC90 (μg/ml) | % Resistant |
|
---|---|---|---|---|---|---|
CLSI | EUCAST | |||||
Fidaxomicin | BI [188] | ≤0.004–1 | 0.25 | 0.5 | NA | NA |
All non-BI types [850]a | ≤0.004–1 | 0.25 | 0.5 | NA | NA | |
A [42] | ≤0.004–1 | 0.25 | 0.5 | NA | NA | |
DH [126] | 0.03–1 | 0.25 | 0.5 | NA | NA | |
G [59] | 0.008–1 | 0.25 | 0.5 | NA | NA | |
Y [163] | ≤0.004–0.5 | 0.25 | 0.5 | NA | NA | |
Rifaximin | BI | ≤0.004 to >4 | 0.03 | >4 | NA | NA |
All non-BI types | ≤0.004 to >4 | 0.015 | 0.03 | NA | NA | |
A | ≤0.004 to >4 | 0.03 | 0.06 | NA | NA | |
DH | 0.008 to >4 | 0.03 | 0.03 | NA | NA | |
G | ≤0.004 to >4 | 0.03 | 0.03 | NA | NA | |
Y | ≤0.004 to >4 | 0.015 | 0.03 | NA | NA | |
Rifampin | BI | ≤0.004 to >4 | 0.008 | >4 | NA | 74.5 |
All non-BI types | ≤0.004 to >4 | ≤ 0.004 | 0.008 | NA | 35.3 | |
A | ≤0.004 to >4 | ≤ 0.004 | 0.008 | NA | 33.3 | |
DH | ≤0.004 to >4 | ≤ 0.004 | 0.008 | NA | 49.2 | |
G | ≤0.004–4 | ≤ 0.004 | 0.008 | NA | 25.4 | |
Y | ≤0.004 to >4 | ≤ 0.004 | 0.008 | NA | 35.0 | |
Tigecycline | BI | ≤0.06–0.5 | 0.12 | 0.25 | 0.0 | 1.6 |
All non-BI types | ≤0.06–1 | 0.12 | 0.25 | 0.0 | 1.8 | |
A | ≤0.06–0.25 | 0.12 | 0.25 | 0.0 | 0.0 | |
DH | ≤0.06–0.5 | 0.12 | 0.25 | 0.0 | 1.6 | |
G | ≤0.06–0.25 | 0.12 | 0.12 | 0.0 | 0.0 | |
Y | ≤0.06–1 | 0.12 | 0.25 | 0.0 | 1.8 | |
Vancomycin | BI | ≤0.25–8 | 2 | 4 | NA | 29.3 |
All non-BI types | ≤0.25–4 | 1 | 2 | NA | 7.5 | |
A | 0.5–4 | 1 | 2 | NA | 7.1 | |
DH | 0.5–4 | 1 | 4 | NA | 11.9 | |
G | 0.5–4 | 1 | 2 | NA | 6.8 | |
Y | 0.5–4 | 1 | 2 | NA | 4.3 | |
Imipenem | BI | ≤0.12–16 | 8 | 8 | 7.4 | NA |
All non-BI types | ≤0.12–16 | 4 | 8 | 3.9 | NA | |
A | ≤0.12–16 | 4 | 8 | 2.4 | NA | |
DH | 1–16 | 4 | 8 | 5.6 | NA | |
G | 1–8 | 4 | 8 | 0.0 | NA | |
Y | ≤0.12–16 | 4 | 8 | 3.7 | NA | |
Moxifloxacin | BI | ≤0.5 to >16 | 16 | >16 | 84.0 | 84.0 |
All non-BI types | ≤0.5 to >16 | 2 | 8 | 12.5 | 12.5 | |
A | ≤0.5–16 | 2 | 2 | 4.8 | 4.8 | |
DH | ≤0.5–16 | 2 | 2 | 5.6 | 5.6 | |
G | ≤0.5 to >16 | 2 | 16 | 18.6 | 18.6 | |
Y | ≤0.5 to >16 | 2 | 4 | 8.0 | 8.0 | |
Metronidazole | BI | 0.12 to 4 | 2 | 2 | 0.0 | 7.4 |
All non-BI types | ≤0.06–4 | 0.5 | 1 | 0.0 | 0.4 | |
A | ≤0.06–2 | 0.5 | 1 | 0.0 | 0.0 | |
DH | ≤0.06–2 | 0.5 | 1 | 0.0 | 0.0 | |
G | 0.12–4 | 0.25 | 2 | 0.0 | 1.7 | |
Y | ≤0.06–2 | 0.5 | 2 | 0.0 | 0.0 | |
Clindamycin | BI | ≤0.5 to >16 | 4 | >16 | 48.4 | NA |
All non-BI types | ≤0.5 to >16 | 4 | 16 | 23.5 | NA | |
A | ≤0.5 to >16 | 4 | 8 | 16.7 | NA | |
DH | ≤0.5 to >16 | 4 | 16 | 20.6 | NA | |
G | ≤0.5–16 | 4 | 8 | 15.3 | NA | |
Y | ≤0.5 to >16 | 4 | 8 | 17.2 | NA | |
Chloramphenicol | BI | 1–16 | 4 | 8 | 0.0 | NA |
All non-BI types | ≤0.5 to 16 | 4 | 8 | 0.0 | NA | |
A | 1–8 | 4 | 8 | 0.0 | NA | |
DH | 1–8 | 4 | 8 | 0.0 | NA | |
G | 1–8 | 4 | 8 | 0.0 | NA | |
Y | ≤0.5–8 | 4 | 8 | 0.0 | NA |
All non-BI types includes the groups listed separately plus an additional 384 isolates from REA groups with <30 isolates.
Center-specific differences.
Center-specific differences in isolate antimicrobial susceptibility patterns are shown in Fig. S2 in the supplemental material. Increasing vancomycin and metronidazole susceptibility was seen at almost all centers by the 2015 to 2016 interval (Fig. S2A and B). Decreased moxifloxacin resistance was observed at all centers by the 2015 to 2016 interval (Fig. S2C). The proportion of imipenem-resistant isolates was variable among institutions and over time (Fig. S2D). In 3 of the 5 institutions, approximately 11 to 15% of isolates were imipenem resistant by 2015 to 2016, a substantial increase over the time interval (Hines, New York Presbyterian, and Tufts). In contrast, one center (Virginia Commonwealth University Medical Center [VCU]) had a dramatic decrease in the proportion of imipenem-resistant isolates; however, the small number of samples contributed in 2012 from this center compared to the subsequent years may have skewed these data (38 isolates versus 155 and 126 isolates in subsequent intervals, respectively). Two of five centers had relatively minimal changes, maintaining resistance levels of ∼5% or lower (Mayo Clinic and R. M. Alden). Resistance to clindamycin (Fig. S2E) was variable across centers/time. At NYP, there was a spike in the 2013 to 2014 interval that returned to baseline in 2015 to 2016. At Mayo Clinic, there was a substantial increase in clindamycin resistance (7.9% in the baseline interval to 30.1% in 2015 to 2016). A decrease in proportion of rifampin-resistant isolates was observed across all institutions (Fig. S2F).
There were significant differences between centers in proportions of REA types isolated (see Fig. S3 in the supplemental material). For 5/6 of the centers, the proportion of BI isolates decreased in 2015 to 2016 compared to that in 2011 to 2012, but the percentages were variable (Fig. S3B). Different REA types filled the niche evacuated by BI at the different centers. However, there was variability in the proportions of non-BI REA types at each center in a given time interval, and a consistent finding at all centers was an increase in nonspecific REA types in 2015 to 2016 compared to that in 2011 to 2012 (Fig. S3C).
DISCUSSION
This report provides a summary of U.S. national surveillance of fidaxomicin activity against isolates of C. difficile obtained from 6 medical centers from 2013 to 2016 and compares these data to our baseline results from 2011 to 2012. No fidaxomicin resistance was noted among the 1,889 isolates tested from 2013 to 2016 (distributed equally with approximately 450 isolates obtained yearly). There were no differences in fidaxomicin MIC50 and MIC90 values over the three time intervals. Studies from other countries have shown similar results; over 2,500 isolates obtained from patients between 2011 and 2014 across 22 European countries remained highly susceptible to fidaxomicin (10). This was also true in surveillance studies undertaken in Canada from 2013 to 2015 (11) and Australia from 2013 to 2015 (12).
Only a handful of isolates with decreased susceptibility to fidaxomicin have been reported as follows: one isolate in Florida, isolated in 2016 from a patient with CDI with an MIC of 16 μg/ml (13), and 4 isolates in Mexico from patients with recurrent CDI with MICs of 2 μg/ml (14). Interestingly, fidaxomicin is not in clinical use in Mexico. The clinical significance of reduced fidaxomicin susceptibility is not clear. Like vancomycin, fidaxomicin reaches very high levels in stool (15).
The proportion of isolates that were vancomycin resistant by EUCAST criteria decreased significantly (P < 0.001) over the course of our surveillance, in contrast to the studies cited above which demonstrated low rates that varied little over time. In this study, the MIC90 for vancomycin was 2 μg/ml in both 2013 to 2014 and 2015 to 2016. In our 2011 to 2012 study, the MIC90 was 4 μg/ml, which contrasted with historical values of 1 μg/ml (9).
In this study, we also observed a decrease in metronidazole resistance over time (P < 0.001), with both the MIC50 and MIC90 values decreasing from 1 μg/ml and 2 μg/ml in 2011 to 2012 to 0.25 μg/ml and 1 μg/ml in 2015 to 2016. We observed the same decrease in isolates that were resistant by EUCAST (overall 2.1% from 2011 to 2016), with a rate of 0.6% in 2015 to 2016. Conversely, others have noted increases in resistance over time (8). The annual rate of metronidazole-resistant organisms reported in the European ClosER study (2011 to 2014) (10) ranged from 0.1% to 0.5%. In a 2013 to 2014 study from Australia, no metronidazole-resistant organisms were reported. In the CAN-Diff study (2013 to 2015) (11), CLSI breakpoints were reported rather than EUCAST ECOFFs; 58/1,310 isolates were above the EUCAST ECOFF. A relationship between ribotype 027 (REA group BI) and higher metronidazole MICs was observed in both the ClosER and CAN-Diff studies as well as in our study. Unlike fidaxomicin and vancomycin, which reach very high levels in the gastrointestinal tract (8), metronidazole gastrointestinal tract concentrations can be low, so increased metronidazole MICs may be clinically relevant. Furthermore, strain heteroresistance may result in clinical failures. In the recently updated Infectious Diseases Society of America (IDSA)/Society for Healthcare Epidemiology of America (SHEA) guidelines, metronidazole is no longer recommended as a first-line therapy for uncomplicated CDI (16).
We observed an increasing proportion of isolates resistant to imipenem, with 8.1% of isolates found to be resistant in the 2015 to 2016 interval compared to 2.3% at baseline, with differences between centers. This increase in resistance is contrary to European surveillance of isolates obtained from July 2011 to July 2014, where rates decreased from 7.2% to 2.2% over time (10). However, a recent single-center study from Spain reported that 20% of 70 isolates isolated from 2015 to 2016 were imipenem resistant (17). In this study, clindamycin resistance remained high (27.6%), but moxifloxacin resistance decreased over time from 33.5% in 2011 to 2012 to 20.7% in 2015 to 2016.
As has been observed in other studies, the proportion of isolates with BI/NAP1/027 decreased over time. This decrease was seen at five out of six centers contributing isolates. Dingle et al. has noted the profound effect of restricted fluoroquinolone usage on the proportion of BI/027 strains in the UK (18).
There are a number of limitations in this study that must be stated. These include a lack of information on antimicrobial exposure prior to the onset of C. difficile infection, information on the specific treatments used for C. difficile disease, clinical outcome data for treatment of C. difficile, and estimates of fidaxomicin usage for treatment of C. difficile in the medical centers that contributed the isolates to the survey. These important limitations should be explored for future surveillance and analysis of relationships of exposure to C. difficile isolates.
In summary, we report the results of ongoing surveillance efforts for C. difficile epidemiology and antimicrobial susceptibility, showing changing trends in the prevalence of REA groups, including a decrease in the epidemic BI group and changing susceptibility to a number of different agents. Importantly, 5 years after licensure, isolates remain highly susceptible to fidaxomicin.
MATERIALS AND METHODS
Medical centers.
From 2013 to 2016, a total of 1,889 C. difficile isolates were referred by 6 medical centers for processing to the Special Studies Laboratory (SSL) at Tufts Medical Center. The medical centers were Hines VA Hospital, Hines, IL; Mayo Clinic, Rochester, MN; New York Presbyterian Hospital/Weill Cornell Medical Center, New York, NY; Tufts Medical Center, Boston, MA; RM Alden Research Laboratory, Culver City, CA; and Virginia Commonwealth University Medical Center (VCU), Richmond, VA. Duke supplied isolates for 2011 only.
Clinical samples.
Isolates of C. difficile from toxin-positive stool samples were obtained from 6 different locations around the United States, from institutions that had excellent anaerobic bacteriology laboratories and investigators willing to collaborate (Table 6). Each institution that performed isolation of C. difficile was invited to send at least 75 isolates collected throughout the year. These were forwarded in chopped meat broth (Anaerobe Systems, Morgan Hill, CA) to the SSL at Tufts Medical Center for antimicrobial sensitivity testing and toxin gene profiling. Alternatively, some centers sent approximately 100 frozen stool samples from toxin-positive patients; these were processed for isolation of C. difficile at the SSL as outlined below.
TABLE 6.
Medical center | No. of isolates referred |
||
---|---|---|---|
2011 to 2012a | 2013 to 2014 | 2015 to 2016 | |
Hines VA Hospital | 137 | 137 | 125 |
Mayo Clinic | 139 | 151 | 136 |
NY Presbyterian Hospital | 179 | 155 | 192 |
R. M. Alden Research Laboratory | 163 | 194 | 206 |
Tufts Medical Center | 208 | 156 | 156 |
Virginia Medical College | 38b | 155 | 126 |
Total isolates | 925 | 948 | 941 |
Duke University Medical Center provided 61 isolates in 2011.
2012 isolates only.
Processing of samples.
After arrival of stool samples, C. difficile was isolated by plating on selective C. difficile medium (cycloserine-cefoxitin-fructose agar with taurocholate) and observing for characteristic colonial morphology (19, 20). This was followed by using the rapid identifying method API20A (bioMérieux Inc., Durham, NC) on the isolate. If identification by this methodology was not conclusive, methods outlined in the Wadsworth Anaerobic Bacteriology Laboratory Manual were followed (20). Confirmation of purity and identification of isolates as C. difficile submitted by participating centers was accomplished by following the same procedure. All isolates were maintained in chopped meat broth until testing, after which a cell paste swabbed from fresh plates was suspended directly into skim milk medium and frozen at −80°C for future reference (21).
Antimicrobial susceptibility testing.
MICs of the isolates were determined using the agar dilution method per Clinical and Laboratory Standards Institute (CLSI) recommendations, M11-A8 (21, 22), for the following antimicrobial agents: fidaxomicin, vancomycin, metronidazole, rifaximin, rifampin, tigecycline, imipenem, moxifloxacin, clindamycin, and chloramphenicol. Fidaxomicin (obtained from Optimer, Cubist, and/or Merck depending on time of isolate testing) was solubilized in dimethyl sulfoxide (DMSO) and further diluted with distilled water to a final concentration in the agar of <0.5% DMSO. Rifampin (Sigma-Aldrich, St. Louis, MO) was dissolved in methanol and further diluted with distilled water. Rifaximin (Salix Pharmaceuticals, Inc., Morrisville, NC) was dissolved in methanol and then diluted in 0.1 M phosphate buffer (pH 7.4) plus 0.45% sodium dodecyl sulfate. All other antibiotics, provided by their respective manufacturers, were dissolved in distilled water.
Antibiotic-containing plates were prepared on the day of the test. The medium was brucella agar (Becton, Dickinson, Sparksville, MD), supplemented with 5 μg of hemin/ml, 1 μg vitamin K1/ml, and 5% (vol/vol) laked sheep blood. Serial 2-fold dilutions of the antibiotics were added to the molten agar, poured into 100-mm Petri dishes and allowed to solidify and dry. Non-antibiotic-containing control plates were prepared in the same manner. The isolates were grown to log phase in brain heart infusion supplemented (yeast and hemin) broth and the turbidity adjusted to that of a 0.5 McFarland standard. For C. difficile ATCC 700057, the density was 1 × 107 to 4 × 107 CFU/ml. The inocula were deposited on the agar surface using a Steer replicator, resulting in a final concentration of 104 CFU/spot. The plates were incubated at 35 to 37°C in an anaerobic chamber with an atmosphere of 85% N2, 5% CO2, and 10% H2. MICs were read after 48 h of incubation. The following ATCC reference organisms were used as controls in all tests: C. difficile ATCC 700057, Bacteroides fragilis ATCC 25285, Bacteroides thetaiotaomicron ATCC 29741, and Staphylococcus aureus ATCC 29213. The rates of resistance of the antimicrobial agents were determined using currently accepted breakpoint recommendations by the CLSI and/or by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) (22, 23). Resistance breakpoints for tigecycline were those recommended by the FDA (24). The antibiotics tested, range of concentrations, and breakpoints are listed in Table 7. Optimer, the sponsor at the time, choose the antibiotics to be tested.
TABLE 7.
Antimicrobial agent | Range of test (μg/ml) | Breakpoint for resistance |
|
---|---|---|---|
CLSI | EUCAST | ||
Fidaxomicin | 0.004–4 | NAa | NA |
Rifaximin | 0.004–4 | NA | NA |
Rifampin | 0.004–4 | NA | >0.004 |
Vancomycin | 0.25–64 | ≥4b | >2 |
Metronidazole | 0.06–16 | ≥32 | >2 |
Moxifloxacin | 0.5–16 | ≥8 | >4 |
Clindamycin | 0.5–16 | ≥8 | NA |
Imipenem | 0.12–8 | ≥16 | NA |
Tigecyclinec | 0.06–64 | ≥16 | >0.25 |
Chloramphenicol | 0.5–16 | ≥32 | NA |
NA, resistance breakpoint not established when study was performed.
CLSI epidemiological cutoff value for vancomycin.
FDA resistance breakpoint used for tigecycline.
Toxin determination.
Toxin gene profiles for C. difficile were performed using PCR methodology as described by Persson et al. (25, 26). Isolates were tested in duplicate, with nonconcordant tests repeated. The following controls were included with each test: (i) a nontoxigenic strain of C. difficile VPI 11186 (ATCC 700057) with a PCR profile of tcdA negative, tcdB negative, cdtA negative, and cdtB negative, in which the entire PaLoc is absent (27) and in which no functional cdt locus is present; (ii) VPI 10463 with a PCR profile of tcdA positive, tcdB positive, cdtA negative, and cdtB negative with an intact tcdC region (28, 29); and (iii) R20291 with a PCR profile of tcdA positive, tcdB positive, cdtA positive, and cdtB positive with 18-bp deletion in the tcdC hypervariable region (30). Loss of PaLoc was confirmed in nontoxigenic isolates by performing a PCR using Lok1/Lok3 primers as previously described (31).
Restriction endonuclease analysis typing.
A randomly selected sample of isolates, stratified by center (∼30 isolates from each center per year), were referred to the Microbiology Reference Laboratory at the Hines Veterans Administration Hospital (Hines, IL) for restriction endonuclease analysis (REA) strain typing. Following purification, total cellular DNA was cut with HindIII restriction enzyme, and fragments were separated by electrophoresis on a 0.7% agarose gel as previously described (32, 33). HindIII restriction patterns with a 90% similarity index were included in the same REA group (letter designation).
Data storage and analysis.
Results from the various tests were sent to the SSL for analysis. All data were stored and analyzed using Microsoft Excel spreadsheets. Fisher’s exact test was used to determine equality of proportions using a P value of 0.05 (GraphPad 7.01).
Supplementary Material
ACKNOWLEDGMENTS
We thank Barbara Rapino, Tara Rapino, Anne Kane, and Roselia Martinez for their assistance as well as Adam Cheknis and Susan Sambol for REA typing.
This study was supported by Optimer, Cubist, and Merck sequentially during the study.
Footnotes
Supplemental material for this article may be found at https://doi.org/10.1128/AAC.00391-19.
REFERENCES
- 1.Burke KE, Lamont JT. 2014. Clostridium difficile infection: a worldwide disease. Gut Liver 8:1–6. doi: 10.5009/gnl.2014.8.1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Borren NZ, Ghadermarzi S, Hutfless S, Ananthakrishnan AN. 2017. The emergence of Clostridium difficile infection in Asia: a systematic review and meta-analysis of incidence and impact. PLoS One 12:e0176797. doi: 10.1371/journal.pone.0176797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lessa FC, Mu Y, Bamberg WM, Beldavs ZG, Dumyati GK, Dunn JR, Farley MM, Holzbauer SM, Meek JI, Phipps EC, Wilson LE, Winston LG, Cohen JA, Limbago BM, Fridkin SK, Gerding DN, McDonald LC. 2015. Burden of Clostridium difficile infection in the United States. N Engl J Med 372:825–834. doi: 10.1056/NEJMoa1408913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hunter JC, Yi M, Ghinwa YM, Dumyati K, Farley MM, Winston LG, Johnston HL, Meek JI, Perlmutter R, Holzbauer SM, Beldavs ZG, Phipps EC, Dunn JR, Cohen JA, Avillan J, Nimalie D, Stone ND, Gerding DN, McDonald LC, Lessa FC. 2016. Burden of nursing home-onset Clostridium difficile infection in the United States: estimates of incidence and patient outcomes. Open Forum Infect Dis 3:ofv196. doi: 10.1093/ofid/ofv196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Desai K, Swati G, Dubberke E, Prabhu VS, Browne C, Mast TC. 2016. Epidemiological and economic burden of Clostridium difficile in the United States: estimates from a modeling approach. BMC Infect Dis 16:303. doi: 10.1186/s12879-016-1610-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dubberke ER, Olsen MA. 2012. Burden of Clostridium difficile on the healthcare system. Clin Infect Dis 55(Suppl):S88–S92. doi: 10.1093/cid/cis335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Reigadas Ramírez E, Bouza ES. 2018. Economic burden of Clostridium difficile infection in European countries. Adv Exp Med Biol 1050:1–12. doi: 10.1007/978-3-319-72799-8_1. [DOI] [PubMed] [Google Scholar]
- 8.Spigaglia P. 2016. Recent advances in the understanding of antibiotic resistance in Clostridium difficile infection. Ther Adv Infect Dis 3:23–42. doi: 10.1177/2049936115622891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Snydman DR, McDermott LA, Jacobus NV, Thorpe C, Stone S, Jenkins SG, Goldstein EJ, Patel R, Forbes BA, Mirrett S, Johnson S, Gerding DN. 2015. U.S.-based national sentinel surveillance study for the epidemiology of Clostridium difficile-associated diarrheal isolates and their susceptibility to fidaxomicin. Antimicrob Agents Chemother 59:6437–6443. doi: 10.1128/AAC.00845-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Freeman J, Vernon J, Pilling S, Morris K, Nicholson S, Shearman S, Longshaw C, Wilcox MH, Pan-European Longitudinal Surveillance of Antibiotic Resistance among Prevalent Clostridium difficile Ribotypes Study Group. 2018. The ClosER study: results from a three-year pan-European longitudinal surveillance of antibiotic resistance among prevalent Clostridium difficile ribotypes, 2011–2014. Clin Microbiol Infect 24:724–731. doi: 10.1016/j.cmi.2017.10.008. [DOI] [PubMed] [Google Scholar]
- 11.Karlowsky JA, Adam H, Kosowan T, Baxter MR, Nichol KA, Laing NM, Golding G, Zhanel GG. 2018. PCR ribotyping and antimicrobial susceptibility testing of isolates of Clostridium difficile cultured from toxin-positive diarrheal stools of patients receiving medical care in Canadian hospitals: the Canadian Clostridium difficile Surveillance Study (CAN-DIFF) 2013–2015. Diagn Microbiol Infect Dis 91:105–111. doi: 10.1016/j.diagmicrobio.2018.01.017. [DOI] [PubMed] [Google Scholar]
- 12.Knight DR, Giglio S, Huntington PG, Korman TM, Kotsanas D, Moore CV, Paterson DL, Prendergast L, Huber CA, Robson J, Waring L, Wehrhahn MC, Weldhagen GF, Wilson RM, Riley TV. 2015. Surveillance for antimicrobial resistance in Australian isolates of Clostridium difficile, 2013–14. J Antimicrob Chemother 70:2992–2999. doi: 10.1093/jac/dkv220. [DOI] [PubMed] [Google Scholar]
- 13.Peng Z, Jin D, Kim HB, Stratton CW, Wu B, Tang YW, Sun X. 2017. Update on antimicrobial resistance in Clostridium difficile: resistance mechanisms and antimicrobial susceptibility testing. J Clin Microbiol 55:1998–2008. doi: 10.1128/JCM.02250-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Martínez-Meléndez A, Tijerina-Rodríguez L, Morfin-Otero R, Camacho-Ortíz A, Villarreal-Treviño L, Sánchez-Alanís H, Rodríguez-Noriega E, Baines SD, Flores-Treviño S, Maldonado-Garza HJ, Garza-González E. 2018. Circulation of highly drug-resistant Clostridium difficile ribotypes 027 and 001 in two tertiary-care hospitals in Mexico. Microb Drug Resist 24:386–392. doi: 10.1089/mdr.2017.0323. [DOI] [PubMed] [Google Scholar]
- 15.Sears P, Crook DW, Louie T, Miller M, Weiss K. 2012. Fidaxomicin attains high fecal concentrations with minimal plasma concentrations following oral administration in patients with Clostridium difficile infection. Clin Infect Dis 55(Suppl):S116–S120. doi: 10.1093/cid/cis337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.McDonald LC, Gerding DN, Johnson S, Bakken JS, Carroll KC, Coffin SE, Dubberke ER, Garey KW, Gould CV, Kelly C, Loo V, Shaklee Sammons J, Sandora TJ, Wilcox MH. 2018. Clinical practice guidelines for Clostridium difficile infection in adults and children: 2017 update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis 66:987–994. doi: 10.1093/cid/ciy149. [DOI] [PubMed] [Google Scholar]
- 17.Arca-Suárez J, Galán-Sánchez F, Cano-Cano F, García-Santos G, Rodríguez-Iglesias MA. 2018. Antimicrobial susceptibility and molecular typing of toxigenic clinical isolates of Clostridium difficile causing infections in the south of Spain. Anaerobe 54:146–150. doi: 10.1016/j.anaerobe.2018.09.006. [DOI] [PubMed] [Google Scholar]
- 18.Dingle KE, Didelot X, Quan TP, Eyre DW, Stoesser N, Golubchik T, Harding RM, Wilson DJ, Griffiths D, Vaughan A, Finney JM, Wyllie DH, Oakley SJ, Fawley WN, Freeman J, Morris K, Martin J, Howard P, Gorbach S, Goldstein EJC, Citron DM, Hopkins S, Hope R, Johnson AP, Wilcox MH, Peto TEA, Walker AS, Crook DW, Modernising Medical Microbiology Informatics Group. 2017. Effects of control interventions on Clostridium difficile infection in England: an observational study. Lancet Infect Dis 17:411–421. doi: 10.1016/S1473-3099(16)30514-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.George WL, Sutter VL, Citron D, Finegold SM. 1979. Selective and differential medium for isolation of Clostridium difficile. J Clin Microbiol 9:214–219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Jousimies-Somer HR, Summanen P, Citron DM, Baron EJ, Wexler HA, Finegold SM. 2002. Wadsworth-KTL anaerobic bacteriology manual, 6th ed Star Publishing Company; Belmont, CA. [Google Scholar]
- 21.Clinical and Laboratory Standards Institute. 2012. Methods for antimicrobial susceptibility testing of anaerobic bacteria. Approved standard. CLSI publication number M11-A8. Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 22.Clinical and Laboratory Standards Institute. 2019. Performance standards for antimicrobial susceptibility testing; 29th informational supplement. CLSI document M100-S29 Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 23.The European Committee on Antimicrobial Susceptibility Testing. 2019. Breakpoint tables for interpretation of MICs and zone diameters. Version 9.0. [Google Scholar]
- 24.Tygacill. 2017. Pfizer product monograph. Pfizer, Kirkland, Canada: https://www.pfizer.ca/sites/g/files/g10050796/f/201710/TYGACIL_PM.pdf. [Google Scholar]
- 25.Persson S, Jensen JN, Olsen KE. 2011. Multiplex PCR method for detection of Clostridium difficile tcdA, tcdB, cdtA, and cdtB and internal in-frame deletion of tcdC. J Clin Microbiol 49:4299–4300. doi: 10.1128/JCM.05161-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Persson S, Torpdahl M, Olsen K. 2008. New multiplex PCR method for the detection of Clostridium difficile toxin A (tcdA) and toxin B (tcdB) and the binary toxin (cdtA/cdtB) genes applied to a Danish strain collection. Clin Microbiol Infect 14:1057–1064. doi: 10.1111/j.1469-0691.2008.02092.x. [DOI] [PubMed] [Google Scholar]
- 27.Moncrief JS, Zheng L, Neville LM, Lyerly DM. 2000. Genetic characterization of toxin A-negative, toxin B-positive C. diff isolates by PCR. J Clin Microbiol 38:3072–3075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sullivan NM, Pellett S, Wilkins TD. 1982. Purification and characterization of toxins A and B of Clostridium difficile. Infect Immun 35:1032–1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Carter GP, Lyras D, Allen DL, Mackin KE, Howarth PM, O'Connor JR, Rood JI. 2007. Binary toxin production in Clostridium difficile is regulated by CdtR, a LytTR family response regulator. J Bacteriol 189:7290–7301. doi: 10.1128/JB.00731-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Stabler RA, He M, Dawson L, Martin M, Valiente E, Corton C, Lawley TD, Sebaihia M, Quail MA, Rose G, Gerding DN, Gibert M, Popoff MR, Parkhill J, Dougan G, Wren BW. 2009. Comparative genome and phenotypic analysis of Clostridium difficile 027 strains provides insight into the evolution of a hypervirulent bacterium. Genome Biol 10:R102. doi: 10.1186/gb-2009-10-9-r102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Braun V, Hundsberger T, Leukel P, Sauerborn M, von Eichel-Streiber C. 1996. Definition of the single integration site of the pathogenicity locus in Clostridium difficile. Gene 181:29–38. doi: 10.1016/S0378-1119(96)00398-8. [DOI] [PubMed] [Google Scholar]
- 32.Clabots CR, Johnson S, Bettin KM, Mathie PA, Mulligan ME, Schaberg DR, Peterson LR, Gerding DN. 1993. Development of a rapid and efficient endonuclease analysis typing system for Clostridium difficile and correlation with other correlation with other typing systems. J Clin Microbiol 31:1870–1875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Killgore GE, Thompson A, Johnson S, Brazier J, Kuijper E, Pepin J, Frost EH, Savelkoul P, Nicholson B, van den Berg RJ, Kato H, Sambol SP, Zukowski W, Woods C, Limbago B, Gerding DN, McDonald LC. 2007. Comparison of seven techniques for typing international epidemic strains of C. difficile: restriction endonuclease analysis, pulsed-field gel electrophoresis, PCR-ribotyping, multilocus sequence typing, multilocus variable number tamdem-repeat analysis, amplified fragment length polymorphism, and surface layer protein A gene sequence typing. J Clin Microbiol 46:431–437. doi: 10.1128/JCM.01484-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.