Previous Article | Next Article ![]()
Journal of Clinical Microbiology, February 2004, p. 693-699, Vol. 42, No. 2
0095-1137/04/$08.00+0 DOI: 10.1128/JCM.42.2.693-699.2003
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Servicio de Microbiología,1 Servicio de Enfermedades Infecciosas, Hospital Universitario Ramón y Cajal, Madrid 28034,3 Laboratorio de Infecciones Intrahospitalarias, Centro Nacional de Microbiología, Instituto Carlos III, Majadahonda, Madrid 28220, Spain2
Received 14 April 2003/ Returned for modification 12 June 2003/ Accepted 27 October 2003
|
|
|---|
|
|
|---|
On the other hand, information about the genomic relationship within this species and the degree of relationship among nosocomial strains remains scarce (3, 15). Nowadays, pulsed-field gel electrophoresis (PFGE) is considered to be the reference genotyping method for S. maltophilia and has been used not only to identify outbreaks and possible reservoirs and modes of transmission involving this species but also to understand population structure from isolates from cystic fibrosis patients (26). In addition, this technique provides a broad look at the whole chromosome of the microorganisms and is useful for determining the genetic relatedness among isolates of a given species by comparison of their macrorestriction profiles (2). In a previous work, Berg et al. (3) determined the diversity of 40 clinical and environmental S. maltophilia isolates and analyzed phenotypic profiles and molecular types by several molecular methods. In their study, the most discriminatory method was PFGE under DraI digestion. In our study, we have determined the genetic relatedness and epidemiological links among 139 isolates recovered in the same hospital over a long period by using profiles generated by PFGE under XbaI restriction. Moreover, detection of cross-transmissions among different patients is also presented.
|
|
|---|
PFGE. Chromosomal DNA was prepared in agarose plugs as described previously (31). After digestion with endonuclease XbaI (30 U; Roche Diagnostics, Barcelona, Spain), restriction fragments were resolved by PFGE with the CHEF-DRII system (Bio-Rad, Hemel Hempstead, United Kingdom). A second enzyme, SpeI (20 U; Roche), was used to confirm identical restriction XbaI profiles. The electrophoresis conditions were as follows: (i) for XbaI, pulse times were ramped from 10 to 60 s over 24 h at 5.4 V/cm with a second ramp from 5 to 20 s over 5 h at 5.4 V/cm at 12°C (30); and (ii) for SpeI, pulse times ranged from 25 to 45 s over 20 h at 6.0 V/cm at 12°C (22). Standard lambda ladders of 48.5-kbp concatemers (Roche) were run as molecular weight markers. Restriction fragments were visually compared and interpreted according to previous established criteria (27). Isolates with identical profiles were considered to represent a single strain.
Computer-monitored fingerprinting and discriminatory power analysis. Computer analysis of the PFGE banding patterns was performed with Bio-Rad Molecular Analyst software. The images analyzed included two reference lanes representing concatemer phage lambda ladders. Bands were automatically assigned by the computer and were manually corrected after observation in the computer screen. Only fragments exceeding 97.0 kbp were included in the analysis. Fingerprinting profiles were scored for the presence and absence of bands at given molecular weights, and strains that differed by one band were assigned different PFGE profiles, codified by a number. Only one profile was represented in the dendrogram for the isolates with identical XbaI profiles. The Dice correlation coefficient was used to analyze the similarities of the banding patterns (21). Strains with an identical PFGE banding pattern (100% similarity coefficient) were considered isogenic strains. Moreover, as previously stated (5, 31), closely related strains within the same clone were those with a similarity coefficient ranging from 80 to 95%. Clustering was based on the unweighted pair group method with arithmetic averages (UPGMA). The tolerance position was 1%.
The discriminatory power of PFGE between isolates from Ramón y Cajal hospital and isolates from the control group was evaluated by using Simpson's index of diversity (16), which expresses the probability that two unrelated strains will be placed in two different typing groups.
Patient clinical data. For all patients, demographic information and the presence of repetitive S. maltophilia-positive cultures and any other organism in the same positive culture were recorded. Upon identification of an outbreak situation, defined by indistinguishable PFGE profiles of isolates, the medical charts of patients colonized or infected with the outbreak strains were retrospectively reviewed. Hospital-acquired infections were classified according to the Centers for Disease Control and Prevention definitions (12). Clinical infection or colonization status by S. maltophilia was considered according to clinical judgment, high or moderate bacterial counts, tissue invasion, repetitive S. maltophilia isolation, monomicrobial culture, and antibiotic treatment response. Respiratory infection was considered when evidence of pulmonary infiltrates with X rays, fever (>38°C), cough, and respiratory function deterioration was observed (18, 20, 29). Nosocomial acquisition was considered as being 72 h after hospital admissions or within 30 days of a surgical patient's discharge (12).
|
|
|---|
The dendrogram generated by computer-aided genotype analysis based on the unique PFGE patterns of all 139 S. maltophilia strains ranged from 28 to 80% similarity (Fig. 1). Two major clusters (designated clusters I and II, three minor clusters (III, IV, and V), and two independent branches (represented by profiles 6 and 62) were observed with a 36% similarity coefficient. The major clusters, I and II, included 37.7 and 38.5%, respectively, of the strains studied, whereas the minor clusters III, IV, and V only contained 3.6, 2.2, and 5.8%, respectively, of the strains studied. Profiles 6 and 62 did not fit into any of the five clusters even at similarity values of 28 and 26%, respectively. The origin of S. maltophilia clinical isolates for each defined XbaI profile is included in the legend to Fig. 1.
![]() View larger version (42K): [in a new window] |
FIG. 1. Phylogenetic analysis of digitized 99 PFGE XbaI profiles of 132 clinical and 7 environmental S. maltophilia isolates recovered in our hospital during a 4-year period. The dendrogram was constructed with PFGE data by similarity and clustering analysis by using the Dice coefficient and UPGMA with the Molecular Analysts software. A percent genetic similarity scale is shown above the dendrogram. Profile types are marked on the left in arabic numerals, and the clusters (cutoff value of 36% similarity) are marked on the right in roman numerals. The clinical origins of the isolates are as follows: respiratory specimens, profiles 1 to 51; blood, profiles 52 to 67; wounds, profiles 70 to 77; organic fluid, profiles 69 and 79; ocular specimens, profiles 68 and 78; urine, profiles 80 to 83; stool, profiles 84 and 85; environmental isolates, profiles 1, 86 to 91; and others, profiles 92 to 99.
|
It is of note that five isolates with different profiles (profiles 1, 4, 38, 52, and 61) were responsible for cross-transmission in 15 patients (see below). Profiles 1 (clone A), 4 (clone B), and 38 (clone C) were recovered from respiratory sources and were detected in six, three, and two patients, respectively. Moreover, profile 52 (clone D) was recovered from blood cultures in two patients, and profile 62 (clone E) was recovered from blood cultures in two patients. Strains belonging to these profiles represented different cross-infection situations among different patients and were associated with clusters II, II, V, and II and a nonclustered profile, respectively. These results were confirmed by SpeI restriction.
The degree of heterogeneity among control stains was also high. No new cluster or independent branches were observed when control strains were integrated in the dendrogram (data not shown).
Demographic data and patient characteristics. One hundred thirty-two isolates were recovered from 105 patients. We had access to demographic data for 97 patients (61 male and 36 female). Thirty-five percent of patients (34 of 97) were hospitalized in different ICUs: Medical ICU (n = 15), Neurosurgical ICU (n = 6), Digestive Tract Surgical ICU (n = 7), Cardiovascular Surgery ICU (n = 3), and Pediatric ICU (n = 3). The other patients were from the Pulmonary Diseases (n = 12), Digestive Tract Surgery (n = 13), Infectious Diseases (n = 5), and Hematology (n = 4) Wards, while the rest of the patients were located in different units. S. maltophilia appeared as a monomicrobial culture in 13 of 19 (68.4%) blood cultures, 21 of 79 (26.6%) respiratory samples, 1 of 14 (7.1%) samples from wounds and cutaneous tissues, and 6 of 12 (50.0%) samples from other locations. In combination with other organisms, S. maltophilia appeared mainly associated with normal flora in respiratory samples (22%) and with gram-positive organisms (mainly coagulase-negative staphylococci) in blood (26%) and wound (61%) cultures.
Detection of cross-transmission. A total of five episodes of cross-transmission were detected during the period studied: (i) three episodes involved three different S. maltophilia strains (clones A, B, and C) recovered from respiratory samples from 11 patients grouped as 6, 3, and 2 patients, and (ii) two episodes involved two different S. maltophilia strains (clones D and E) implicated in episodes of bacteremia in 4 patients. Risk factors, previous defined by others (7, 9, 11, 19) for S. maltophilia acquisition in all of these patients are shown in Table 1. Figure 2 shows progression of the respiratory outbreaks, the number of patients involved, the patients who underwent a bronchoscope procedure, and the patients' origin.
|
View this table: [in a new window] |
TABLE 1. Comorbidities and potential risk factors for S. maltophilia infection and colonization in patients sharing similar clone typesa
|
![]() View larger version (24K): [in a new window] |
FIG. 2. Progression of clones A, B, and C of S. maltophilia. Boxes represent one isolate each; the number below corresponds to the number of the patient. The time period, ward of precedence, bronchoscopy procedure, infection, and patient deaths are also indicated.
|
![]() View larger version (113K): [in a new window] |
FIG. 3. XbaI profiles obtained in S. maltophilia isolates from blood from 16 patients. Two episodes of cross-transmission were suspected (clone E, lanes 8 and 9, and clone D, lanes 14 and 15). Lane M, molecular size marker.
|
|
|
|---|
The low interstrain variability of phenotypic typing methods in S. maltophilia (23, 34) makes molecular techniques, particularly PFGE, widely accepted methods for epidemiological typing of this species (3, 20, 26). As previously demonstrated in S. maltophilia, PFGE under XbaI restriction is appropriate for tracing isolates and generating stable profiles in long-term-colonized patients. In the present study, the discriminatory ability of the XbaI PFGE technique reached a Simpson coefficient value of 1.0, allowing us to establish the molecular relationships among clinical S. maltophilia isolates recovered in a single hospital. Despite being from the same institution, the majority of strains tested (71.7%) displayed different PFGE profiles, and only identical genotypic patterns were observed in isolates recovered from 15 patients (14.3%), suggesting a common source of these strains. Moreover, the PFGE fingerprinting analysis for the strains tested revealed a great discriminatory ability, resulting in a Simpson coefficient value of 0.996. Berg et al. (3), using PFGE under DraI restriction, also demonstrated a high intraspecific diversity in S. maltophilia clinical and environmental isolates from different locations and niches. A similar conclusion was obtained with arbitrarily primed PCR (32), ERIC-PCR (6), randomly amplified polymorphic DNA (35), and amplified-fragment length polymorphism (15) techniques in S. maltophilia isolates from different environments. However, the high intraspecific diversity decreased, because closely related strains could be detected in chronically colonized cystic fibrosis patients with this pathogen (5).
It is of note that unlike isolates from cystic fibrosis patients (5, 31), isolates serially recovered from the same patient showed identical PFGE profiles (100% similarity). In a previous work from our group (31), which included only isolates from cystic fibrosis patients, the presence of strains with a similarity coefficient ranging from 80 to 95% was observed within the same patient. This fact represents the dynamic situation of S. maltophilia isolates in the pulmonary environment in cystic fibrosis patients, which may persist for more than 6 years (31). On the contrary, nosocomial patients present a lower length of persistence of S. maltophilia isolates (8.5 ± 6.5 days in the present study).
On the basis of XbaI PFGE profiles, 99 distinct profiles were identified among 139 isolates studied. Isolates were grouped into five phylogenetic clusters when a cutoff of 36.0% in the genetic similarity scale was considered. The resolved profiles showed a great genetic distance, and the genetic diversity extended from low (28.0%) to high (80.0%) similarity. These results clearly demonstrated the low homogeneity level of S. maltophilia strains, irrespective of the time frame of collection in the same clinical setting. Interestingly, the major clonal lineages (I and II) were nearly grouped 84.0%% of the strains analyzed.
As previously noted, prolonged hospitalization and broad-spectrum antimicrobial therapy may facilitate the selection of S. maltophilia from respiratory or gastrointestinal locations (32, 33). In our study, the great number of distinct S. maltophilia profiles may reflect a wide environmental distribution of this species, allowing acquisition from different environmental sources. Moreover, S. maltophilia strains from different patients may have been acquired independently, discarding the presence of specific nosocomial clones. In S. maltophilia, the epidemiological relationship among different isolates needs to be analyzed, because unexpected results can be obtained. This was the case in a Croatian hospital in which, over a 4-month period (28), nine different profiles were observed in S. maltophilia isolates recovered from 20 patients. Six of these profiles were observed in different groups, including up to four hospitalized patients in the same or different units. Moreover, in an Italian university hospital, an epidemiological investigation of ICU patients revealed that although most patients were infected or colonized by different S. maltophilia clones, strains with identical genotypes were isolated, and two separate outbreaks were identified (7). Similarly, in our investigation, three consecutive respiratory outbreaks were detected in an approximately 1-year period. As previously noted, the acquisition of S. maltophilia isolates could be due to defective sterilization of the bronchoscope rather than dissemination of the organisms from other environmental sources or between patients. This hypothesis was reinforced when an S. maltophilia isolate with an indistinguishable profile from clone A was isolated in a fiberbronchoscope. However, in other clones, the sources and vehicles of infection could not be detected, as has occurred in most S. maltophilia outbreaks (8). Recently, an outbreak of Pseudomonas aeruginosa infections after bronchoscopic procedures was demonstrated (25), but to our knowledge, the involvement of these procedures as a potential source for S. maltophilia transmission has not been published. In contrast, faucet aerators and contaminated faucet water, electronic ventilator temperature sensors, and disinfectants have been recognized as contributing to disseminate epidemic S. maltophilia isolates (8).
In general, S. maltophilia is recovered from mixed cultures, particularly from respiratory tract and skin and soft tissue infections, which makes it difficult to establish an unequivocal role of this organism as a pathogen (8). In our study, S. maltophilia appeared as a single etiological agent in 26.6% of respiratory samples and 7.1% of wounds. This value reached 68.4% in blood cultures, similar to that found by Jang et al. (17). The pathogenicity of S. maltophilia is still controversial; however, in a systematic retrospective case control study excluding polymicrobial bacteremia, the mortality rate for S. maltophilia (26.7%) was similar to that observed for other nosocomial pathogens causing bloodstream infections (24). As in other clinical studies (14), most of our S. maltophilia isolates (79 of 139) were recovered from respiratory specimens. However, clinical evidence of true infection was difficult to determine. Previous risk factors for S. maltophilia pneumonia have been reported (29, 32, 33), including neutropenia, immunosuppression, use of H2 antagonists, previous antibiotic exposure, hospitalization and surgery, catheterization, mechanical ventilation support, prolonged hospitalization, and ICU stay (7-9, 11, 19). Most of these risk factors were observed in our 11 patients involved in respiratory cross-transmission (Table 1).
In conclusion, our study revealed a high genetic diversity among S. maltophilia isolates despite their origin in a single hospital. This result may be related to the high potential environmental distribution of this pathogen. However, a few clones could have been transmitted among different patients, producing outbreaks and epidemic situations. Transmission of S. maltophilia isolates was in fact demonstrated in six patients after the use of a bronchoscope device. Molecular typing investigations of S. maltophilia isolates are useful for control strategies to decrease infections due to this emerging pathogen.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»