Abstract
Mycobacterium abscessus is a rapidly growing mycobacterium that causes respiratory tract infections in predisposed patients, such as those with cystic fibrosis and nosocomial skin and soft tissue infections. In order to investigate the clonal relationships between the strains causing epidemic episodes, we evaluated the discriminatory power of the semiautomated DiversiLab (DL) repetitive extragenic palindromic sequence PCR (REP-PCR) test for M. abscessus genotyping. Since M. abscessus was shown to be composed of subspecies (M. abscessus subsp. massiliense, M. abscessus subsp. bolletii, and M. abscessus subsp. abscessus), we also evaluated the ability of this technique to differentiate subspecies. The technique was applied to two collections of clinical isolates, (i) 83 M. abscessus original isolates (43 M. abscessus subsp. abscessus, 12 M. abscessus subsp. bolletii, and 28 M. abscessus subsp. massiliense) from infected patients and (ii) 35 repeated isolates obtained over 1 year from four cystic fibrosis patients. The DL REP-PCR test was standardized for DNA extraction, DNA amplification, and electrophoresis pattern comparisons. Among the isolates from distinct patients, 53/83 (62%) isolates showed a specific pattern, and 30 were distributed in 11 clusters and 6 patterns, with 2 to 4 isolates per pattern. The clusters and patterns did not fully correlate with multilocus sequence typing (MLST) analysis results. This revealed a high genomic diversity between patients, with a discriminatory power of 98% (Simpson's diversity index). However, since some isolates shared identical patterns, this raises the question of whether it is due to transmission between patients or a common reservoir. Multiple isolates from the same patient showed identical patterns, except for one patient infected by two strains. Between the M. abscessus subspecies, the indexes were <70%, indicating that the DL REP-PCR test is not an accurate tool for identifying organisms to the subspecies level. REP-PCR appears to be a rapid genotyping method that is useful for investigating epidemics of M. abscessus infections.
INTRODUCTION
Mycobacterium abscessus is an emerging pathogen responsible for lung disease, most often in predisposed patients, such as those with cystic fibrosis, bronchiectasis, and chronic obstructive pulmonary disease (1, 2), and it has also been described in patients without compromised lung defense (3). M. abscessus is among the top five mycobacteria responsible for respiratory tract infections. It represents half of the infections caused by nontuberculous mycobacteria (NTM) and 95% of infections due to rapidly growing mycobacteria (RGM) in patients with cystic fibrosis in France (4). M. abscessus infections may lead to a rapid decline in respiratory function, and relapses are observed in patients who have undergone organ transplants (5–7). M. abscessus is also responsible for extrapulmonary infections in immunocompetent patients with outbreaks, related to surgical and cosmetic procedures, reported in various countries (8–15). Contamination is assumed to occur from an environmental reservoir (16); however, person-to-person transmission cannot be ruled out. M. abscessus is supposed to be present in water and soils, but it was only recently isolated from potable and municipal water (17–19).
M. abscessus has undergone many changes in its taxonomy over the past 60 years. In 1953, Moore and Frerichs described M. abscessus on the basis of its morphological and biochemical differences from known rapidly growing mycobacteria (20). From 1972 to 1992, the M. abscessus species included Mycobacterium chelonae (21), which was eventually described as a distinct species (22). Molecular approaches, such as sequence analysis of the hsp65, rpoB, sodA, and erm(41) genes, multilocus sequence typing (MLST), and genome sequencing, showed that M. abscessus is a heterogeneous species (23–25). Two or three subgroups or subspecies have been distinguished: M. abscessus subsp. abscessus, M. abscessus subsp. bolletii, and M. abscessus subsp. massiliense, which is close to M. abscessus subsp. bolletii in some studies (26). These subspecies differ also with regard to their antibiotic susceptibilities, especially to macrolides (27–29).
In order to propose adequate prevention measures for M. abscessus infection, it is helpful to know if the strains are transmitted between humans or are only acquired from environmental sources (19). Consequently, the clonal relationships of M. abscessus isolates need to be addressed. Several genotypic methods were described for M. abscessus clonality, with pulsed-field gel electrophoresis (PFGE) being the reference method (15, 30, 31). However, this method is cumbersome and difficult to interpret (32). Multilocus sequence typing (MLST) based on the nucleotide sequence analysis of housekeeping genes provides information on M. abscessus phylogeny and distinguishes subspecies within the M. abscessus group (24). However, it does not assess strain clonality. Random amplification of polymorphic DNA (RAPD) is frequently used in laboratories, but this technique is poorly reproducible. Other methods based on the amplification of repetitive sequences have been described, such as (i) analysis of variable-number tandem repeats (VNTR) (33), which is routinely applied to Mycobacterium tuberculosis with an analysis of mycobacterial interspersed repetitive units (MIRU), (ii) whole-genome sequencing (30), and (iii) the repetitive extragenic palindromic sequence PCR (REP-PCR), which is mainly used to distinguish isolates belonging to Enterobacteriaceae. Few publications have reported the use of REP-PCR for genotyping isolates among mycobacterial species, and rarely for M. abscessus (34). REPs are repetitive extragenic palindromic DNA sequences first described in 1980 in enterobacteria, such as Escherichia coli (35), and are present in many other bacteria (36). They have a role in stabilizing messenger RNAs and are targets for insertion or recombination sites. REP sequences have been described in M. abscessus (http://minisatellites-rec.igmors.u-psud.fr/GPMS/). REP-PCR consists of the amplification of sequences contained between two REP sequences.
DiversiLab (DL) is a commercial REP-PCR-based typing system containing quality-controlled reagents in a kit format, allowing for automated detection, standardization, and analysis using a software package (37). The aim of our study was to evaluate its discriminative power for assessing the clonal relationships of M. abscessus clinical isolates and for differentiating the subspecies within M. abscessus.
MATERIALS AND METHODS
Bacterial isolates and identification.
Two collections of M. abscessus clinical isolates were studied, (i) 83 isolates previously described (27) (43 M. abscessus subsp. abscessus, 12 M. abscessus subsp. bolletii, and 28 M. abscessus subsp. massiliense) from infected patients and (ii) 35 isolates (32 M. abscessus and 3 Mycobacterium massiliense) obtained over 1 year from four cystic fibrosis (CF) patients at the Necker University Hospital in Paris, France. These isolates were anonymized and used under institutional review board approval for all specimens received at Assistance Publique Hôpitaux de Paris, National Reference Center for Mycobacteria, for diagnostic purposes. The reference strains used were M. abscessus subsp. abscessus strain CIP 104536T (ATCC 19977T), M. abscessus subsp. bolletii strain CIP 108541T, and M. abscessus subsp. massiliense strain CIP 108297T.
The isolates were identified as M. abscessus by the GenoType Mycobacterium CM/AS test (Hain Lifescience, Bandol, France). They were also identified at the subspecies level by analyzing the nucleotide sequences of the hsp65, rpoB, and erm(41) genes, as previously described (25). In addition, the isolates have been submitted to MLST, as previously described (24).
The isolates were subcultured onto blood Trypticase soy agar plates for 5 days in an aerobic atmosphere at 37°C.
REP-PCR analysis.
DNA was extracted from a 10-μl loop of blood agar culture using the UltraClean microbial DNA isolation kit (Mo Bio Laboratories, Solana Beach, CA), which is based on mechanical, chemical, and thermic treatments. Sample DNA was standardized to 25 to 50 ng/μl using the NanoDrop ND1000 spectrophotometer.
The REP-PCR was standardized according to the DiversiLab Mycobacterium kit protocol. Multiple genomic DNA fragments ranging in size from 50 bp to 7,000 bp were amplified simultaneously. Briefly, 2 μl of genomic DNA, 0.5 μl of 10× AmpliTaq DNA polymerase (Applied Biosystems), and 2.5 μl of 10× PCR buffer were added to the REP-PCR master mix to achieve a total volume of 25 μl. Thermal cycling used a calibrated PerkinElmer thermal cycler apparatus with the following parameters: initial denaturation at 95°C for 5 min, 35 cycles of denaturation at 95°C for 30 s, annealing at 66°C for 45 s, extension at 72°C for 60 s, and a final extension at 72°C for 5 min. The PCR products were analyzed on a microfluidic LabChip (DiversiLab; bioMérieux). Electrophoresis was performed automatically on the 2100 Bioanalyzer (Agilent Technologies), and a DNA fingerprint was produced for each isolate. Quality-control assessment was performed by analyzing (i) the quality of migration by including a DNA marker in each well of the LabChip that generates two peaks at 50 bp and 7,000 bp, and (ii) the quality of amplification by measuring the intensity of peaks that should be >50 fluorescence units (FU), with the peak of the DNA marker being >300 FU. Next, fingerprints were compared to each other using the DiversiLab online software (version 3.4). The analysis can be done according to various statistical methods, such as Pearson's method, extended Jaccard, or Kullback-Leibler divergence. We chose to apply Pearson's method because the profiles generated numerous peaks. A clonal relationship was defined according to rules similar to those described for PFGE (38). The interpretative criteria provided by the manufacturer led us to categorize the isolates as indistinguishable, similar, or different (DiversiLab user's guide). Isolates with fingerprints showing more than two band differences were categorized as different. Isolates were categorized as indistinguishable when they shared identical fingerprints, including identical intensities of individual bands. This defines a pattern (P). Isolates with fingerprints differing by one or two bands were categorized as similar and belonging to the same cluster (C).
RESULTS
Diversity of the REP-PCR profiles for isolates obtained from distinct patients.
The REP-PCR results showed a great diversity among the profiles of the 83 isolates obtained from distinct patients (Fig. 1). The global discriminating power (estimated by Simpson's diversity index) (39), which reflects the probability that two unrelated strains appear genetically different, was calculated and measured at 98% (P < 0.001). Fifty-three isolates, i.e., 62% (53/83), were categorized as different according to the interpretation criteria. The results were identical according to the Tenover criteria (a difference of at least three bands) or the criteria recommended by the DL analysis guide (a difference of at least two bands). For the remaining 30 isolates, they grouped in 11 clusters, with 2 to 6 isolates per cluster (Table 1).
FIG 1.
DiversiLab REP-PCR profiles of M. abscessus isolates from distinct patients. (A) Example of profiles and corresponding dendrogram obtained using Pearson's correlation analysis. Each color represents a single pattern, and the red dashed vertical line is the similarity line. Isolates grouped as patterns (P5 and P6) when they had an identical profile or as clusters (C10 and C11) when they showed at most a one- or two-band difference. RPC101 and CR5625 are designated different isolates because their profiles differ by at least three bands. (B) Manual comparison to screen for peak differences. CR5237 and RPC148 show similar profiles (two different peaks), and RPC160 and CR5625 show different profiles (six different peaks).
TABLE 1.
DiversiLab REP-PCR test electrophoretic profiles observed for M. abscessus isolates from distinct patientsa
| Isolate | Organismb | Clusterc | Patternd | ST |
|---|---|---|---|---|
| CR5385 | M. abscessus subsp. bolletii | 1 | None | 90 |
| CR6115 | M. abscessus subsp. bolletii | 1 | None | 43 |
| 21158 | M. abscessus subsp. massiliense | 2 | 1 | 23 |
| CR5729 | M. abscessus subsp. massiliense | 2 | 1 | 23 |
| 01183 | M. abscessus subsp. massiliense | 2 | 1 | 23 |
| 50616 | M. abscessus subsp. massiliense | 2 | 1 | 2 |
| CR5063 | M. abscessus subsp. massiliense | 2 | None | 23 |
| CR7312 | M. abscessus subsp. massiliense | 3 | None | 2 |
| CR5816 | M. abscessus subsp. massiliense | 3 | None | 2 |
| 31177 | M. abscessus subsp. massiliense | 4 | 2 | 2 |
| CR6114 | M. abscessus subsp. massiliense | 4 | 2 | 2 |
| 50075 | M. abscessus subsp. massiliense | 4 | 2 | 2 |
| RPC98 | M. abscessus erm(41) sequevar T28 | 5 | None | 1 |
| RPC102 | M. abscessus erm(41) sequevar T28 | 5 | None | 26 |
| CR7355 | M. abscessus erm(41) sequevar C28 | 6 | None | 91 |
| CR5761 | M. abscessus erm(41) sequevar C28 | 6 | None | 49 |
| RPC110 | M. abscessus erm(41) sequevar T28 | 7 | 3 | 21 |
| RPC118 | M. abscessus erm(41) sequevar T28 | 7 | 3 | 29 |
| CR5701 | M. abscessus erm(41) sequevar C28 | 8 | None | 83 |
| RPC136 | M. abscessus erm(41) sequevar C28 | 8 | None | 61 |
| CR6169 | M. abscessus erm(41) sequevar T 28 | 9 | 4 | 37 |
| RPC142 | M. abscessus erm(41) sequevar T 28 | 9 | 4 | 12 |
| RPC148 | M. abscessus subsp. massiliense | 10 | 5 | 8 |
| 50484 | M. abscessus subsp. massiliense | 10 | 5 | 7 |
| CR5093 | M. abscessus subsp. massiliense | 10 | 5 | 7 |
| RPC11 | M. abscessus subsp. massiliense | 10 | 5 | 7 |
| CR5602 | M. abscessus subsp. massiliense | 10 | None | 7 |
| CR5237 | M. abscessus subsp. massiliense | 10 | None | 7 |
| RPC151 | M. abscessus erm(41) sequevar T 28 | 11 | 6 | 11 |
| RPC160 | M. abscessus erm(41) sequevar C28 | 11 | 6 | 1 |
DL, DiversiLab.
As determined by the erm(41), hsp65, or rpoB nucleotide sequence analysis.
A cluster grouped isolates with profiles with one or two bands of difference or >95% similarity.
A pattern grouped isolates with identical profiles or >97% similarity.
Within each cluster, we manually compared the isolates in pairs by overlapping the profiles (Fig. 1B). This additional step was really useful before the final assessment in order to determine precisely the intensity of the peaks and to evidence smaller peaks that were not taken into account by the statistical general analysis. Consequently, among the 11 clusters, we distinguished 6 patterns (identical REP-PCR profiles), with 2 to 4 isolates per pattern. Patterns 1, 2, and 5 included 4, 3, and 4 isolates of M. abscessus subsp. massiliense, respectively. Patterns 3 and 4 were composed of 2 isolates per pattern identified as M. abscessus subsp. abscessus erm(41) sequevar T28, whereas pattern 6 included 2 isolates of M. abscessus subsp. abscessus, with one with an erm(41) sequevar C28 and the second with an erm(41) sequevar T28. Clinical and epidemiological data were sought for the 17 corresponding patients (Table 2). The first finding was that all but one of the patients with isolates sharing a similar pattern were suffering from cystic fibrosis. There is indeed a significant difference in the prevalences of isolates grouped in patterns with regard to the underlying disease of the patient (cystic fibrosis or other diseases), with 16/54 CF isolates grouped in patterns versus 1/29 from patients without CF, respectively (P = 0.004, Fisher's exact test). For the clusters, prevalences were not significantly different, with 22/54 CF isolates grouped in clusters versus 8/29 from patients without CF (P = 0.34, Fisher's exact test). The second finding was that some patients were followed in the same hospital, suggesting a relationship of some sort, e.g., contaminated water or patient-to-patient transmission. The last finding concerned pattern 6, which included 2 isolates with a different erm(41) sequevars. Since the isolate with the T28 sequevar was obtained from a patient who previously received antibiotics, whereas the isolate with the C28 sequevar was obtained from a patient who did not receive antibiotics, we hypothesized that the resistance conferred by the erm(41) T28 sequevar results from a C-to-T mutation in the erm(41) gene.
TABLE 2.
Clinical and epidemiological data for M. abscessus isolates grouping in patterns
| Pattern | Isolate | Place of isolation in France | Date of isolation (yr) | Clinical findings |
|---|---|---|---|---|
| 1 | 21158 | Lyon | 2002 | CFa |
| CR5729 | Paris | 2007 | CF | |
| 01183 | Paris | 2000 | CF | |
| 50616 | Paris | 2005 | CF | |
| 2 | 31177 | Paris | 2003 | CF |
| CR6114 | Vannes | 2007 | CF | |
| 50075 | Paris | 2005 | CF | |
| 3 | RPC110 | Lille | 2004 | CF |
| RPC118 | Bordeaux | 2004 | CF | |
| 4 | CR6169 | Mayotte | 2007 | Ondine syndrome |
| RPC142 | Paris | 2004 | CF | |
| 5 | RPC148 | Angers | 2004 | CF |
| 50484 | Le Mans | 2005 | CF | |
| CR5093 | Le Mans | 2008 | CF | |
| RPC11 | Garches | 2003 | CF | |
| 6 | RPC151 | Rouen | 2005 | CF |
| RPC160 | Lille | 2004 | CF |
CF, cystic fibrosis.
The clusters and patterns defined by REP-PCR were compared to sequence types obtained by MLST (Table 1). For some clusters (C2, C3, C4, and C10), most of the isolates belonged to the same sequence type (ST). The sequences of the seven genes amplified in the MLST (argH, cya, glpK, gnd, murC, pta, and purH) were compared. We observed 15 to 64 single-nucleotide polymorphisms (SNPs) between strains of the same cluster. It should be pointed out that 2 to 6 of the genes expressed the same allele (data not shown). In some clusters (C1, C5, C6, C7, C8, C9, and C11), the isolates belonged to different STs and the gene sequences differed by 7 to 62 SNPs, with 1 to 4 identical alleles. This difference between the two methods is probably explained by the differences of the sequences targeted by the two methods: MLST targets a limited number of conserved genes, and REP-PCR targets a high number of noncoding genomic fragments.
Similar REP-PCR profiles for repeated isolates from cystic fibrosis patients.
Thirty-five M. abscessus isolates corresponded to 7 to 12 isolates per patient (Table 3). Conversely, identical REP-PCR profiles were observed for isolates from the same patient. The isolates were categorized as one unique pattern per patient, with P7, P8, and P9 described for patients A, T, and E, respectively (Table 3). In contrast, the overall REP-PCR profiles were different for isolates from distinct patients (Fig. 2). This shows that the isolates were not epidemic and that patients did not share a source of infection. However, we observed that the isolates from patients E and V belonged to same cluster (C14), although there was no obvious epidemiological link (different wards, different dates for consultations). One pattern (P9) was observed for patient E, with five isolates identified as M. abscessus subsp. abscessusc erm(41) sequevar T28. Patient V had 3 identical isolates (pattern P10). erm(41) gene sequencing identified 3 isolates as M. abscessus subsp. massiliense (C15), with the isolates of C14 (P10, P11, and P12) belonging to M. abscessus subsp. abscessus erm (41) sequevar T28 (27). Clinical data showed that patient V was treated with cefoxitin, amikacin, and azithromycin for a lung infection from April 2010 to September 2010. This treatment possibly explains the emergence of M. abscessus subsp. abscessus erm(41) sequevar T28, which was indeed mixed with M. abscessus subsp. massiliense in the samples obtained in February and April 2010 before the start of azithromycin treatment. Thereafter, M. abscessus subsp. abscessus erm(41) sequevar T28 was the only subspecies isolated under treatment conditions.
TABLE 3.
Characteristics of M. abscessus isolates from repeated sputum specimens obtained from four cystic fibrosis patients over 1 year
| Isolatea | Date of isolation (day/mo/yr) | Subspecies | DL REP-PCR profile | Cluster | Pattern |
|---|---|---|---|---|---|
| A1 | 05/11/2009 | M. abscessus sequevar erm(41) T28 | 9 | 12 | 7 |
| A2 | 01/12/2009 | M. abscessus sequevar erm(41) T28 | 8 | 12 | 7 |
| A3 | 04/01/2010 | M. abscessus sequevar erm(41) T28 | 7 | 12 | 7 |
| A4 | 23/02/2010 | M. abscessus sequevar erm(41) T28 | 4 | 12 | 7 |
| A5 | 24/02/2010 | M. abscessus sequevar erm(41) T28 | 11 | 12 | 7 |
| A6 | 24/05/2010 | M. abscessus sequevar erm(41) T28 | 12 | 12 | 7 |
| A7 | 24/05/2010 | M. abscessus sequevar erm(41) T28 | 5 | 12 | 7 |
| A8 | 25/05/2010 | M. abscessus sequevar erm(41) T28 | 3 | 12 | 7 |
| A9 | 25/05/2010 | M. abscessus sequevar erm(41) T28 | 2 | 12 | 7 |
| A10 | 26/05/2010 | NDa | 10 | 12 | 7 |
| A11 | 18/06/2010 | ND | 1 | 12 | 7 |
| A12 | 08/11/2010 | M. abscessus sequevar erm(41) T28 | 6 | 12 | 7 |
| T1 | 05/01/2010 | M. abscessus sequevar erm(41) T28 | 20 | 13 | 8 |
| T2 | 08/02/2010 | ND | 18 | 13 | 8 |
| T3 | 19/02/2010 | ND | 19 | 13 | 8 |
| T4 | 08/03/2010 | M. abscessus sequevar erm(41) T28 | 14 | 13 | 8 |
| T5 | 30/03/2010 | M. abscessus sequevar erm(41) T28 | 17 | 13 | 8 |
| T6 | 18/05/2010 | M. abscessus sequevar erm(41) T28 | 15 | 13 | 8 |
| T7 | 30/06/2010 | M. abscessus sequevar erm(41) T28 | 16 | 13 | 8 |
| T8 | 01/10/2010 | M. abscessus sequevar erm(41) T28 | 13 | 13 | 8 |
| E1 | 05/09/2009 | M. abscessus sequevar erm(41) T28 | 28 | 14 | 9 |
| E2 | 18/12/2009 | M. abscessus sequevar erm(41) T28 | 29 | 14 | 9 |
| E3 | 12/02/2010 | M. abscessus sequevar erm(41) T28 | 32 | 14 | 9 |
| E4 | 09/04/2010 | M. abscessus sequevar erm(41) T28 | 26 | 14 | 9 |
| E5 | 04/06/2010 | ND | 30 | 14 | 9 |
| E6 | 27/07/2010 | ND | 31 | 14 | 9 |
| E7 | 24/09/2010 | M. abscessus sequevar erm(41) T28 | 27 | 14 | 9 |
| V1 | 18/02/2010 | M. abscessus subsp. massiliense | 34 | 15 | None |
| V2 | 22/02/2010 | M. abscessus sequevar erm(41) T28 | 23 | 14 | None |
| V3 | 08/04/2010 | M. abscessus subsp. massiliense | 35 | 15 | None |
| V4 | 20/04/2010 | M. abscessus subsp. massiliense | 33 | 15 | None |
| V5 | 22/06/2010 | M. abscessus sequevar erm(41) T28 | 25 | 14 | 10 |
| V6 | 29/06/2010 | M. abscessus sequevar erm(41) T28 | 25 | 14 | 10 |
| V7 | 21/09/2010 | M. abscessus sequevar erm(41) T28 | 22 | 14 | 10 |
| V8 | 26/10/2010 | M. abscessus sequevar erm(41) T28 | 21 | 14 | 10 |
Letters correspond to patients A, T, E, and V.
ND, not determined.
FIG 2.
DiversiLab REP-PCR profiles of M. abscessus isolates from the four patients with cystic fibrosis. (A) Scatter plot of the isolates. Each isolate is represented by its number corresponding to the DL profile, as presented in Table 3. Patient A, DL profiles 1 to 12; patient T, DL profiles 13 to 20; patient V, DL profiles 21 to 25 and 33 to 35; patient E, DL profiles 26 to 32. (B) DL profiles of patients T and V: patient T had DL profiles 13 to 20 belonging to cluster C13; patient V had DL profiles 21 to 25 belonging to cluster C14 and identified as M. abscessus erm(41) sequevar T28, as well as DL profiles 33 to 35 belonging to cluster C15 and identified as M. abscessus subsp. massiliense. Each color represents a single pattern, and the red dashed vertical line is the similarity line.
Based on erm(41) gene sequencing, patients E, A, and T harbored M. abscessus subsp. abscessus sequevar T28. Patients E and T did not receive antibiotics for mycobacterial infection, and M. abscessus subsp. abscessus isolation was considered to represent colonization. Patient A was treated for mycobacterial infection from November 2009 to June 2010. The first month, he received ciprofloxacin, clarithromycin, and amikacin. Antibiotic susceptibility testing performed on the strain isolated in December 2009 showed resistance to ciprofloxacin and inducible resistance to clarithromycin, as expected. The treatment was subsequently modified to amikacin, cefoxitin, and tigecycline for the remaining months without eradication of M. abscessus subsp. abscessus (Table 3).
Differentiation within the M. abscessus subspecies.
Among the two collections of isolates analyzed for their DL REP-PCR profiles, most clusters correlated with subspecies identification done by rpoB or hsp65 and erm(41) gene sequencing, as described above. C1 includes only M. abscessus subsp. bolletii isolates, C2, C3, C4, C10, and C15 include M. abscessus subsp. massiliense isolates, C5, C7, and C9 include isolates of M. abscessus subsp. abscessus erm(41) sequevar T28, C8 and C11 include isolates of M. abscessus subsp. abscessus erm(41) sequevar C28, and C12, C13, and C14 include isolates of M. abscessus subsp. abscessus erm(41) sequevar T28.
The discriminatory power (Simpson's diversity index) of the DL REP-PCR test was calculated within isolates of the same subspecies and was found to have a high score of 93.3% for M. abscessus subsp. massiliense, 97% for M. abscessus subsp. bolletii, and 99.3% for M. abscessus subsp. abscessus (Table 4). However, when taking into account all the subspecies as a whole, the DL REP-PCR test did not group together isolates of the same subspecies. The discriminatory power was estimated to be only 60.5% (Table 4). This means that the REP-PCR test cannot assess subspecies identifications. Moreover, among the M. abscessus subsp. abscessus isolates, the discriminatory power of the DL REP-PCR test was only 35.4% for differentiating M. abscessus subsp. abscessus erm(41) sequevar T28 and M. abscessus erm(41) sequevar C28. This may indicate that although isolates share an erm(41) sequevar, which confers macrolide susceptibility (C28) or resistance (T28), their genomes are not related in the REP-PCR assay.
TABLE 4.
Simpson's diversity indexes calculated by comparing the REP-PCR profiles of M. abscessus isolates from distinct patients, with regard to the various subspecies (M. abscessus, M. bolletii, M. abscessus subsp. abscessus) and the erm(41) sequevars, clusters, and patternsa
| Data group | Simpson's diversity index (%) for (n): |
|||
|---|---|---|---|---|
| M. abscessus (83) | M. abscessus subsp. abscessus (43) | M. abscessus subsp. massiliense (28) | M. abscessus subsp. bolletii (12) | |
| Subspecies | 60.30 | NAb | NA | NA |
| erm(41) sequevar (T28 vs C28) | NA | 35.4 | NA | NA |
| Patterns | 99.70 | 99.90 | 97.80 | 98.50 |
| Clusters | 98.90 | 99.30 | 93.30 | 97 |
Calculated using the online tool (http://darwin.phyloviz.net/ComparingPartitions/index.php?link=Tool).
NA, not applicable.
DISCUSSION
M. abscessus is an emerging pathogen causing respiratory infections and health care-associated infections. It was recently involved in several outbreaks, including >2,000 cases of postsurgical infections (8, 10, 14, 15). Although the reservoir was assumed to be environmental, M. abscessus is not commonly isolated from water or surfaces, and thus the causal link between epidemic cases or a new case and its environment needs to be demonstrated.
The DL REP-PCR assay is a standardized tool for comparing genomes from bacterial isolates, and we evaluated its performance for genotyping M. abscessus clinical isolates. This method has been described as efficient and reproducible for typing many bacterial species (37). It was rarely applied to M. abscessus and so far to only small collections (34). For an evaluation of a relevant genotyping tool, it is recommended to first measure its discriminatory power with a large number of unrelated strains (39). This is the reason why we first evaluated the performance of the DL REP-PCR test in genotyping 83 strains from distinct patients with one isolate per patient. All patients had infections, and the isolates were sent to our reference laboratory. The DL REP-PCR test showed a high clonal diversity among these isolates, with a discriminatory power of >98% (calculated with Simpson's index of diversity). This result is concordant with what was shown previously (40). Although most of the isolates were distinguished by their REP-PCR profiles, one-third of them (30 isolates) grouped into clusters. These clusters were consistent, since they were composed mostly of the same subspecies. Furthermore, they were mainly concordant with the ST obtained by MLST (http://www.pasteur.fr/mlst). The absence of perfect concordance with MLST for few clusters can be explained by the fact that DL genotyping reflects microevolution of isolates, which is useful for outbreak investigations. Indeed, REP sequences are known to be targets of insertion sequences or transposons (41). In contrast, MLST based on housekeeping genes reflects evolution over a longer period of time and is a useful tool for phylogenic studies. Interestingly, we observed that some isolates shared identical REP-PCR profiles, called patterns (no band difference), with each pattern comprising 2 to 4 isolates. The clinical and epidemiological data were reviewed thoroughly, and it appeared that some patterns were observed significantly more frequently in cystic fibrosis (CF) patients than in patients with other underlying diseases. This finding raised the possibility of cross-transmission between CF patients, since they are repeatedly hospitalized in a CF-specialized ward. However, since the sources of infection remain unknown, we cannot exclude that these patients may have been infected from a common source.
After having shown the discriminatory power of the DL REP-PCR test, we then compared repeated isolates from four cystic fibrosis patients. The profiles showed that each patient had his own strain, although two patients shared isolates belonging to the same cluster. Recent publications have highlighted similar results using different genotyping methods, such as VNTR or REP-PCR (40), or even with PFGE (31). These findings demonstrated that individual cystic fibrosis patients are persistently infected with one strain. Some authors have emphasized the possibility of cross-transmission of strains from one patient to another (42). This might have been the case between patients E and V (who were diagnosed and treated at the same hospital). However, epidemiological data did not show where and when this transmission might have occurred.
Although controversial, it has been proposed that M. abscessus contains three subspecies. These subspecies differ by SNPs in the rpoB or hsp65 genes and in their multilocus sequence analysis (MLSA) and MLST profiles (25). However, the DL REP-PCR test did not group together all the isolates belonging to the same subspecies, showing that this genotyping tool cannot identify organisms to the subspecies level within M. abscessus (diversity index of only 60.3%). Additionally, the method did not discriminate between the two erm(41) sequevars (T28 and C28). However, REP-PCR can be used for assessing clonal relationship for isolates belonging to the same subspecies (indexes of >90%).
The good performance we obtained for the DL REP-PCR test for genotyping M. abscessus isolates relies on high reproducibility. This was obtained only because we followed precisely the conditions required by the manufacturer for DNA extraction, the PCR conditions, including accreditation of the apparatus, electrophoresis, and interpretation. The DL REP-PCR test is convenient, fast, easy to use, and useful for disproving clonal relationships between isolates. The method is suitable not only for investigations of epidemics but also for the evaluation of patients in the case of persistent cultures with M. abscessus. The method may also distinguish relapse strains from new infections. The case of patient V showed indeed that a patient can be infected successively by two strains. In this patient, an isolate belonging to the subspecies M. abscessus subsp. massiliense, which is known to be intrinsically macrolide susceptible, was replaced by M. abscessus subsp. abscessus erm(41) sequevar T28, a macrolide-resistant subspecies (27). The replacement was observed during a course of azithromycin, which first eradicated the M. abscessus subsp. massiliense strain. It was recently shown that patients with M. abscessus subsp. massiliense infection are rapidly cured by macrolides, such as clarithromycin or azithromycin (28).
The azithromycin course may also explain the in vivo selection of the M. abscessus subsp. abscessus erm(41) sequevar T28. We cannot be sure that this strain was present before treatment. Since antibiotic susceptibility and identification are usually performed on a sample of bacterial culture and are not performed by taking several single colonies, we suspect that a mix of isolates may have been present. In conclusion, the case of patient V showed that antibiotic pressure may facilitate the colonization or infection of one M. abscessus subspecies over the other because of differences in their intrinsic antibiotic susceptibility patterns.
ACKNOWLEDGMENTS
We thank the microbiologists and technicians from the French reference center laboratories (Pitié-Salpêtrière and Lariboisière Hospitals) and the medical colleagues who sent the M. abscessus isolates.
Our laboratory received a grant from Vaincre la Mucoviscidose (RF20110600574), an annual grant from University Diderot (EA 3964), and a grant from the Institut de Veille Sanitaire (CNR des Mycobactéries et Résistance aux Antituberculeux).
We declare no conflicts of interest.
Footnotes
Published ahead of print 26 March 2014
REFERENCES
- 1.Catherinot E, Roux AL, Macheras E, Hubert D, Matmar M, Dannhoffer L, Chinet T, Morand P, Poyart C, Heym B, Rottman M, Gaillard JL, Herrmann JL. 2009. Acute respiratory failure involving an R variant of Mycobacterium abscessus. J. Clin. Microbiol. 47:271–274. 10.1128/JCM.01478-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Griffith DE, Girard WM, Wallace RJ., Jr 1993. Clinical features of pulmonary disease caused by rapidly growing mycobacteria. An analysis of 154 patients. Am. Rev. Respir. Dis. 147:1271–1278. 10.1164/ajrccm/147.5.1271 [DOI] [PubMed] [Google Scholar]
- 3.Kim RD, Greenberg DE, Ehrmantraut ME, Guide SV, Ding L, Shea Y, Brown MR, Chernick M, Steagall WK, Glasgow CG, Lin J, Jolley C, Sorbara L, Raffeld M, Hill S, Avila N, Sachdev V, Barnhart LA, Anderson VL, Claypool R, Hilligoss DM, Garofalo M, Fitzgerald A, Anaya-O'Brien S, Darnell D, DeCastro R, Menning HM, Ricklefs SM, Porcella SF, Olivier KN, Moss J, Holland SM. 2008. Pulmonary nontuberculous mycobacterial disease: prospective study of a distinct preexisting syndrome. Am. J. Respir. Crit. Care Med. 178:1066–1074. 10.1164/rccm.200805-686OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Roux AL, Catherinot E, Ripoll F, Soismier N, Macheras E, Ravilly S, Bellis G, Vibet MA, Le Roux E, Lemonnier L, Gutierrez C, Vincent V, Fauroux B, Rottman M, Guillemot D, Gaillard JL, Herrmann J-L, OMA Group 2009. Multicenter study of prevalence of nontuberculous mycobacteria in patients with cystic fibrosis in France. J. Clin. Microbiol. 47:4124–4128. 10.1128/JCM.01257-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gilljam M, Scherstén H, Silverborn M, Jönsson B, Ericsson Hollsing A. 2010. Lung transplantation in patients with cystic fibrosis and Mycobacterium abscessus infection. J. Cyst. Fibros. 9:272–276. 10.1016/j.jcf.2010.03.008 [DOI] [PubMed] [Google Scholar]
- 6.Luong ML, Morrissey O, Husain S. 2010. Assessment of infection risks prior to lung transplantation. Curr. Opin. Infect. Dis. 23:578–583. 10.1097/QCO.0b013e32833f9f93 [DOI] [PubMed] [Google Scholar]
- 7.Morales P, Gil A, Santos M. 2010. Mycobacterium abscessus infection in transplant recipients. Transplant Proc. 42:3058–3060. 10.1016/j.transproceed.2010.08.004 [DOI] [PubMed] [Google Scholar]
- 8.Cardoso AM, Martins de Sousa E, Viana-Niero C, Bonfim de Bortoli F, Pereira das Neves ZC, Leão SC, Junqueira-Kipnis AP, Kipnis A. 2008. Emergence of nosocomial Mycobacterium massiliense infection in Goiás, Brazil. Microbes Infect. 10:1552–1557. 10.1016/j.micinf.2008.09.008 [DOI] [PubMed] [Google Scholar]
- 9.Cheng A, Liu YC, Chen ML, Hung CC, Tsai YT, Sheng WH, Liao CH, Hsueh PR, Chen YC, Chang SC. 2013. Extrapulmonary infections caused by a dominant strain of Mycobacterium massiliense (Mycobacterium abscessus subspecies bolletii). Clin. Microbiol. Infect. 19:E473–E482. 10.1111/1469-0691.12261 [DOI] [PubMed] [Google Scholar]
- 10.Duarte RS, Lourenço MC, de Souza Fonseca L, Leão SC, Amorim EdLT, Rocha ILL, Coelho FS, Viana-Niero C, Gomes KM, da Silva MG, de Oliveira Lorena NS, Pitombo MB, Ferreira RMC, de Oliveira Garcia MH, de Oliveira GP, Lupi O, Vilaça BR, Serradas LR, Chebabo A, Marques EA, Teixeira LM, Dalcolmo M, Senna SG, Sampaio JL. 2009. Epidemic of postsurgical infections caused by Mycobacterium massiliense. J. Clin. Microbiol. 47:2149–2155. 10.1128/JCM.00027-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Furuya EY, Paez A, Srinivasan A, Cooksey R, Augenbraun M, Baron M, Brudney K, Della-Latta P, Estivariz C, Fischer S, Flood M, Kellner P, Roman C, Yakrus M, Weiss D, Granowitz EV. 2008. Outbreak of Mycobacterium abscessus wound infections among “lipotourists” from the United States who underwent abdominoplasty in the Dominican Republic. Clin. Infect. Dis. 46:1181–1188. 10.1086/529191 [DOI] [PubMed] [Google Scholar]
- 12.Galmés-Truyols A, Giménez-Duran J, Bosch-Isabel C, Nicolau-Riutort A, Vanrell-Berga J, Portell-Arbona M, Seguí-Prat B, Gumá-Torá M, Martí-Alomar I, Rojo-Arias MÁ, Ruiz-Veramendi M. 2011. An outbreak of cutaneous infection due to Mycobacterium abscessus associated to mesotherapy. Enferm. Infecc. Microbiol. Clin. 29:510–514. 10.1016/j.eimc.2011.03.006 [DOI] [PubMed] [Google Scholar]
- 13.Leão SC, Viana-Niero C, Matsumoto CK, Lima KV, Lopes ML, Palaci M, Hadad DJ, Vinhas S, Duarte RS, Lourenço MC, Kipnis A, das Neves ZC, Gabardo BM, Ribeiro MO, Baethgen L, de Assis DB, Madalosso G, Chimara E, Dalcolmo MP. 2010. Epidemic of surgical-site infections by a single clone of rapidly growing mycobacteria in Brazil. Future Microbiol. 5:971–980. 10.2217/fmb.10.49 [DOI] [PubMed] [Google Scholar]
- 14.Monego F, Duarte RS, Nakatani SM, Araújo WN, Riediger IN, Brockelt S, Souza V, Cataldo JI, Dias RC, Biondo AW. 2011. Molecular identification and typing of Mycobacterium massiliense isolated from postsurgical infections in Brazil. Braz. J. Infect. Dis. 15:436–441. 10.1016/S1413-8670(11)70224-0 [DOI] [PubMed] [Google Scholar]
- 15.Viana-Niero C, Lima KV, Lopes ML, Rabello MC, Marsola LR, Brilhante VC, Durham AM, Leão SC. 2008. Molecular characterization of Mycobacterium massiliense and Mycobacterium bolletii in isolates collected from outbreaks of infections after laparoscopic surgeries and cosmetic procedures. J. Clin. Microbiol. 46:850–855. 10.1128/JCM.02052-07 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Da Mata Jardín O, Hernández-Pérez R, Corrales H, Cardoso-Leao S, de Waard JH. 2010. Follow-up on an outbreak in Venezuela of soft-tissue infection due to Mycobacterium abscessus associated with mesotherapy. Enferm. Infecc. Microbiol. Clin. 28:596–601 (In Spanish.) 10.1016/j.eimc.2009.08.003 [DOI] [PubMed] [Google Scholar]
- 17.Al-Sulami AA, Al-Taee AM, Wida'a QH. 2012. Isolation and identification of Mycobacterium avium complex and other nontuberculosis [sic] mycobacteria from drinking-water in Basra governorate, Iraq. East Mediterr. Health J. 18:274–278 [DOI] [PubMed] [Google Scholar]
- 18.Thomson R, Tolson C, Carter R, Coulter C, Huygens F, Hargreaves M. 2013. Isolation of nontuberculous mycobacteria (NTM) from household water and shower aerosols in patients with pulmonary disease caused by NTM. J. Clin. Microbiol. 51:3006–3011. 10.1128/JCM.00899-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Thomson R, Tolson C, Sidjabat H, Huygens F, Hargreaves M. 2013. Mycobacterium abscessus isolated from municipal water–a potential source of human infection. BMC Infect. Dis. 13:241. 10.1186/1471-2334-13-241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Moore M, Frerichs JB. 1953. An unusual acid-fast infection of the knee with subcutaneous, abscess-like lesions of the gluteal region; report of a case with a study of the organism, Mycobacterium abscessus, n. sp. J. Investig. Dermatol. 20:133–169 [DOI] [PubMed] [Google Scholar]
- 21.Stanford JL, Pattyn SR, Portaels F, Gunthorpe WJ. 1972. Studies on Mycobacterium chelonei. J. Med. Microbiol. 5:177–182. 10.1099/00222615-5-2-177 [DOI] [PubMed] [Google Scholar]
- 22.Kusunoki S, Ezaki T. 1992. Proposal of Mycobacterium peregrinum sp. nov., nom. rev., and elevation of Mycobacterium chelonae subsp. abscessus (Kubica et al.) to species status: Mycobacterium abscessus comb nov. Int. J. Syst. Bacteriol. 42:240–245. 10.1099/00207713-42-2-240 [DOI] [PubMed] [Google Scholar]
- 23.Cho YJ, Yi H, Chun J, Cho SN, Daley CL, Koh WJ, Shin SJ. 2013. The genome sequence of ‘Mycobacterium massiliense’ strain CIP 108297 suggests the independent taxonomic status of the Mycobacterium abscessus complex at the subspecies level. PLoS One 8:e81560. 10.1371/journal.pone.0081560 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Macheras E, Roux AL, Bastian S, Leão SC, Palaci M, Sivadon-Tardy V, Gutierrez C, Richter E, Rüsch-Gerdes S, Pfyffer G, Bodmer T, Cambau E, Gaillard JL, Heym B. 2011. Multilocus sequence analysis and rpoB sequencing of Mycobacterium abscessus (sensu lato) strains. J. Clin. Microbiol. 49:491–499. 10.1128/JCM.01274-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Macheras E, Roux AL, Ripoll F, Sivadon-Tardy V, Gutierrez C, Gaillard JL, Heym B. 2009. Inaccuracy of single-target sequencing for discriminating species of the Mycobacterium abscessus group. J. Clin. Microbiol. 47:2596–2600. 10.1128/JCM.00037-09 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Leao SC, Tortoli E, Euzéby JP, Garcia MJ. 2011. Proposal that Mycobacterium massiliense and Mycobacterium bolletii be united and reclassified as Mycobacterium abscessus subsp. bolletii comb. nov., designation of Mycobacterium abscessus subsp. abscessus subsp. nov. and emended description of Mycobacterium abscessus. Int. J. Syst. Evol. Microbiol. 61:2311–2313. 10.1099/ijs.0.023770-0 [DOI] [PubMed] [Google Scholar]
- 27.Bastian S, Veziris N, Roux AL, Brossier F, Gaillard JL, Jarlier V, Cambau E. 2011. Assessment of clarithromycin susceptibility in strains belonging to the Mycobacterium abscessus group by erm(41) and rrl sequencing. Antimicrob. Agents Chemother. 55:775–781. 10.1128/AAC.00861-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Koh WJ, Jeon K, Lee NY, Kim BJ, Kook YH, Lee SH, Park YK, Kim CK, Shin SJ, Huitt GA, Daley CL, Kwon OJ. 2011. Clinical significance of differentiation of Mycobacterium massiliense from Mycobacterium abscessus. Am. J. Respir. Crit. Care Med. 183:405–410. 10.1164/rccm.201003-0395OC [DOI] [PubMed] [Google Scholar]
- 29.Nessar R, Cambau E, Reyrat JM, Murray A, Gicquel B. 2012. Mycobacterium abscessus: a new antibiotic nightmare. J. Antimicrob. Chemother. 67:810–818. 10.1093/jac/dkr578 [DOI] [PubMed] [Google Scholar]
- 30.Bryant JM, Grogono DM, Greaves D, Foweraker J, Roddick I, Inns T, Reacher M, Haworth CS, Curran MD, Harris SR, Peacock SJ, Parkhill J, Floto RA. 2013. Whole-genome sequencing to identify transmission of Mycobacterium abscessus between patients with cystic fibrosis: a retrospective cohort study. Lancet 381:1551–1560. 10.1016/S0140-6736(13)60632-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Jönsson BE, Gilljam M, Lindblad A, Ridell M, Wold AE, Welinder-Olsson C. 2007. Molecular epidemiology of Mycobacterium abscessus, with focus on cystic fibrosis. J. Clin. Microbiol. 45:1497–1504. 10.1128/JCM.02592-06 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Zhang Y, Yakrus MA, Graviss EA, Williams-Bouyer N, Turenne C, Kabani A, Wallace RJ., Jr 2004. Pulsed-field gel electrophoresis study of Mycobacterium abscessus isolates previously affected by DNA degradation. J. Clin. Microbiol. 42:5582–5587. 10.1128/JCM.42.12.5582-5587.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wong YL, Ong CS, Ngeow YF. 2012. Molecular typing of Mycobacterium abscessus based on tandem-repeat polymorphism. J. Clin. Microbiol. 50:3084–3088. 10.1128/JCM.00753-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Zelazny AM, Root JM, Shea YR, Colombo RE, Shamputa IC, Stock F, Conlan S, McNulty S, Brown-Elliott BA, Wallace RJ, Jr, Olivier KN, Holland SM, Sampaio EP. 2009. Cohort study of molecular identification and typing of Mycobacterium abscessus, Mycobacterium massiliense, and Mycobacterium bolletii. J. Clin. Microbiol. 47:1985–1995. 10.1128/JCM.01688-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Stern MJ, Ames GF, Smith NH, Robinson EC, Higgins CF. 1984. Repetitive extragenic palindromic sequences: a major component of the bacterial genome. Cell 37:1015–1026. 10.1016/0092-8674(84)90436-7 [DOI] [PubMed] [Google Scholar]
- 36.Tobes R, Ramos JL. 2005. REP code: defining bacterial identity in extragenic space. Environ. Microbiol. 7:225–228. 10.1111/j.1462-2920.2004.00704.x [DOI] [PubMed] [Google Scholar]
- 37.Healy M, Huong J, Bittner T, Lising M, Frye S, Raza S, Schrock R, Manry J, Renwick A, Nieto R, Woods C, Versalovic J, Lupski JR. 2005. Microbial DNA typing by automated repetitive-sequence-based PCR. J. Clin. Microbiol. 43:199–207. 10.1128/JCM.43.1.199-207.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Tenover FC, Arbeit RD, Goering RV, Mickelsen PA, Murray BE, Persing DH, Swaminathan B. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233–2239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hunter PR, Gaston MA. 1988. Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J. Clin. Microbiol. 26:2465–2466 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Harris KA, Kenna DT, Blauwendraat C, Hartley JC, Turton JF, Aurora P, Dixon GL. 2012. Molecular fingerprinting of Mycobacterium abscessus strains in a cohort of pediatric cystic fibrosis patients. J. Clin. Microbiol. 50:1758–1761. 10.1128/JCM.00155-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Tobes R, Pareja E. 2006. Bacterial repetitive extragenic palindromic sequences are DNA targets for insertion sequence elements. BMC Genomics 7:62. 10.1186/1471-2164-7-62 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Aitken ML, Limaye A, Pottinger P, Whimbey E, Goss CH, Tonelli MR, Cangelosi GA, Dirac MA, Olivier KN, Brown-Elliott BA, McNulty S, Wallace RJ., Jr 2012. Respiratory outbreak of Mycobacterium abscessus subspecies massiliense in a lung transplant and cystic fibrosis center. Am. J. Respir. Crit. Care Med. 185:231–232. 10.1164/ajrccm.185.2.231 [DOI] [PubMed] [Google Scholar]