Introduction
In orotracheally intubated critically ill patients, bacteria from the oropharyngeal or gastric microbiota can rapidly colonize the lower respiratory airways passing over the endotracheal tube cuff (ETT cuff) and colonizing the inner ETT surface by forming biofilms [
1,
2]. Thus, these patients are especially vulnerable to developing a respiratory infection caused by a nosocomial pathogen such as
Staphylococcus aureus, in either its methicillin-sensitive (MSSA) or methicillin-resistant (MRSA) form.
S. aureus has recently been identified as the second most frequently isolated microorganism responsible for intensive care unit (ICU)-acquired pneumonia, of which 29% of cases are MRSA [
3]. The current clinical guidelines for hospital-acquired pneumonia (HAP) and ventilator-associated pneumonia (VAP) published by the Infectious Diseases Society of America and the American Thoracic Society (IDSA/ATS) recommend either intravenous vancomycin (VAN) or linezolid (LNZ) as first-line treatment for MRSA ICU-acquired respiratory infection [
4], while the International ERS/ESICM/ESCMID/ALAT guidelines prefer LNZ than VAN [
5]. However, the microbiological confirmation of MRSA cultures takes at least 48–72 h, a sufficient time-lapse for ETT biofilm formation.
ETT biofilm formation is currently considered one of the multiple factors that can lead to VAP or its relapse [
2]. While several preventive strategies have targeted ETT biofilm eradication [
6‐
9], none of them has achieved 100% success. The silver-coated ETT and mucus shaver have demonstrated improvements in reducing ETT biofilm in randomized clinical trials but still present certain limitations which may delay their implementation in clinical practice [
7,
8]. Although biofilms exhibit antimicrobial tolerance [
10], little is known about how antimicrobials affect biofilm formation during endotracheal intubation. In a previous study in pigs with MRSA pneumonia, we found that those treated with LNZ achieved better pharmacokinetic and pharmacodynamic indices in serum and lung tissue [
11], very high levels of LNZ within ETT biofilm, and a lower ETT biofilm MRSA burden in comparison with untreated controls; however, similar rates were not found in the VAN group [
12,
13]. What is more, in a study in piglets, Luna et al. found that LNZ was associated with a lower pathology score, better survival, and a trend towards better clearance of MRSA in comparison with glycopeptides [
14]. Since it is well known that findings in pigs are not always reproducible in humans [
15], we designed a clinical observational study in ICU patients to assess this issue.
Our study aimed to determine the effect of systemic treatment with LNZ vs VAN on ETT biofilm from ICU patients with respiratory MRSA infection, including LNZ and VAN concentration measurements within plasma and endotracheal aspirate (ETA) 72 h after treatment initiation and within ETT biofilm upon extubation.
Materials and methods
Patients
The study was conducted at the medical and surgical ICUs of four university hospitals in southern Europe, three in Spain and one in Italy. The following hospitals enrolled patients: Hospital Clinic, Barcelona, Spain (including the following ICUs: Respiratory, Medical, Surgical, Cardiovascular and Hepatic), Hospital del Mar (Critical Care Dept), also in Barcelona, Spain, Hospital Universitario Central de Asturias, in Oviedo, Spain (Intensive Medicine Service), and the Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, in Milan, Italy (Adult Intensive Care).
Data were prospectively collected from September 2013 to December 2016. The investigators made daily rounds in all ICUs. Patients were included consecutively, and only the first episode was analyzed. All patients were over 18 and had respiratory infection due to S. aureus (confirmed microbiologically) with ≥ 48 h of orotracheal intubation and ≥ 48 h of treatment with either LNZ or VAN. Patients with severe immunosuppression (neutropenia after chemotherapy or hematopoietic stem cell transplantation, drug-induced immunosuppression in solid–organ transplant or cytotoxic therapy, and HIV infection-related disorders) were not registered.
The study was carried out in compliance with the Declaration of Helsinki (current version, Fortaleza, Brazil, October 2013) and was conducted in accordance with the requirements of the 2007 Spanish Biomedical Research Act. The study was approved by the institution’s Internal Review Board (registry number 2012/7927). Written informed consent was obtained from patients or their next of kin.
Definitions
The clinical suspicion of pneumonia was based on clinical criteria. We considered VAP in patients with previous invasive mechanical ventilation for 48 h or more. Patients were classified as VAP or non-ventilator ICU-acquired pneumonia (i.e., cases that do not meet the VAP criteria) [
16]. Early-onset VAP was defined as occurring within the first 4 days of invasive mechanical ventilation. The respiratory infection was considered ventilator-associated tracheobronchitis when at least two of the aforementioned criteria for pneumonia were found in the absence of radiographic signs of new pneumonia [
17]. Severe community-acquired pneumonia (SCAP) was defined according to the 2007 IDSA/ATS guidelines [
18] and as previously defined [
19]. All SCAP patients included required invasive mechanical ventilation.
Microbiology and antimicrobial treatment
The microbiological evaluation has been extensively addressed in previous reports [
20]. Microbial identification and susceptibility testing were performed by standard methods [
21,
22].
The initial empiric antimicrobial treatment was administered according to local adaptations of international guidelines [
5] and subsequently revised according to the microbiology results.
Data collection and severity assessment
All relevant data were collected at admission and at the onset of pneumonia from the medical records and bedside flow charts, including clinical, laboratory, radiological, and microbiological information. Patients were followed until the end of mechanical ventilation.
The severity assessment included the APACHE-II [
23] and the Sequential Organ Failure Assessment (SOFA) [
24] score on ICU admission at microbial diagnosis and at orotracheal extubation.
Endotracheal tube preparation and microbiology analysis
All ETTs from the patients included were collected and stored at − 80 °C until analysis. All ETTs were number coded so that the investigators would be blind to treatment group allocation during the analysis. For the first time, we also included the microbiological culture of the ETT cuff. The ETT cuff was dissected and microbiologically processed before rinsing the outer surface and slicing the ETT, following our methodology published elsewhere [
12]. Both ETT cuff and ETT were sonicated before microbiological cultures. Bacterial growth was quantified and reported as logarithmic scale of colony-forming units per milliliter (log
10 CFU/mL). Susceptibility to oxacillin, linezolid, and vancomycin was assessed for all
S. aureus strains isolated from ETT through E-test strips (Biomerieux, France) using ATCC25923 strain as standard laboratory testing control strain, following the manufacturer’s recommendations.
Determination of MLST from Sanger data and phylogenetic analyses
The Sanger sequences were used to obtain the allelic profile of seven
S. aureus housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, yqiL). The genes were concatenated by the
MLST.net database. The MLST results were compared to references in NCBI and the
S. aureus MLST database in order to assign sequence type (ST). The MLST results were compared against the MLST database (
https://pubmlst.org/saureus/) using comparative eBURST V3 software employing the BURST algorithm [
25]. Accessory gene regulator (agr) type I, II, III, IV, or V was confirmed by conventional polymerase chain reaction (PCR) using previously described primers and reaction conditions [
26].
Antibiotic concentration in biological matrixes
LNZ or VAN concentrations in biological matrixes (i.e., plasma, ETA, and ETT biofilm) were determined using high-performance liquid chromatography (HPLC) as previously described [
11]. To release antibiotics from ETA and biofilm and to perform the HPLC, we applied our methodology previously described elsewhere [
12]. The lower limit of detection of HPLC was 2.5 μg mL
− 1 for both antibiotics. When the sample was below detection limit (BDL), the value assigned was 1.25 μg/mL.
A bioassay was alternatively performed for the detection of vancomycin, as previously reported [
12].
Bacillus subtillis (ATCC 6633) in Mueller-Hinton Agar was used for the analysis. The lower limit of detection of the vancomycin bioassay was 0.70 μg mL
− 1. When the sample was BDL, the value assigned was 0.35 μg/mL.
Scanning electron microscopy
Biofilm was imaged and thickness measured via scanning electron microscopy (SEM) [
12]. Briefly, a 1-cm-long hemisection of the ETT distal dependent parts were fixed, dehydrated in graded alcohol series, dried using a polaron critical point drying apparatus, and mounted on commercial SEM stubs (Ted Pella, Inc. Spain). To avoid charge artifacts, the section was sputter-coated with a gold thin layer (sc 510, Fisons Instrument, East Sussex, UK) and carefully silver painted. Samples were imaged via a scanning electron microscope (JEOL JSM 7001F FEG, Japan), and micrographs were recorded on a personal computer. We measured minimal, maximal, and mean biofilm thickness using dedicated software (ImageJ, Wayne Rasband, NIH, USA).
Statistical analysis
Categorical variables were reported as number (%), while continuous variables were reported as mean SD or median (interquartile range, IQR), if the distribution was normal or non-normal respectively. Continuous variables between groups were compared using the one-way analysis of variance (ANOVA) or the Kruskal-Wallis test as appropriate. Post hoc pairwise comparisons were carried out via Tukey’s honestly significant difference (HSD) test. Paired samples were compared with the paired t-test or non-parametric Wilcoxon signed-rank test when appropriate. Spearman’s correlation analyses were performed to determine associations between continuous variables. A two-sided p value ≤ 0.05 was considered statistically significant. Data were processed with IBM SPSS Statistic for Windows, version 22.0 (IBM Corporation, Armonk, NY, USA).
Discussion
Systemic treatment of MRSA respiratory infection with LNZ in mechanically ventilated ICU patients resulted in lower ETT biofilm and ETT cuff MRSA burdens than in patients who received VAN. Indeed, MRSA eradication was superior in ETT (50 vs 30%) and ETT cuff (75 vs 20%) in the LNZ than in the VAN group, although the difference was statistically significant only with regard to ETT cuff. Accordingly, the concentration of LNZ was higher than VAN in ETA and also in ETT biofilm, even though both drugs achieved therapeutic plasma levels at 72 h after treatment initiation.
This is the first report in the literature comparing the effects of LNZ and VAN in ETT from mechanically ventilated humans. Our findings indicate that LNZ is more effective in ETT cuff than VAN, since MRSA presence and loads were significantly lower in the LNZ group. Why is this important?
On the one hand, ETT cuff microfolds, formed in contact with the tracheal wall, are considered a common route of microbial access to the lower respiratory airways [
27]. For this reason, attempts have been made in order to minimize ETT cuff aspiration of subglottic secretions. Although these systems lose efficacy over time [
28‐
32], their efficacy can be complemented by systemic antibiotics with ETT biofilm and cuff effect like LNZ, but not VAN.
On the other hand, we demonstrated the superiority of LNZ over VAN in ETA and ETT biofilm drug concentration. The efficacy of LNZ penetrating into respiratory secretions is emphasized by the therapeutic levels achieved by both drugs in plasma compared with their concentrations in ETA 72 h after the first drug administration, in which LNZ remained several folds above the MIC but VAN levels remained subtherapeutic in most of the samples. Notably, the concentrations of LNZ and VAN in ETA (72 h) are indicative of their concentration in ETT biofilm after extubation. Nevertheless, the usefulness of ETA (72 h) for predicting other drug concentrations in ETT biofilm after extubation needs to be investigated further.
Although biofilms exhibit intrinsic tolerance to antibiotics [
10,
33], in the ETT, the presence of antibiotics and the development of the biofilm are concomitant. This increases the ability of systemic LNZ treatment to limit biofilm development, as its ETT MRSA eradication rate is 67% higher than that of VAN. However, in critically ill patients, the distribution of LNZ within the ETT, mainly driven by respiratory secretions, is not homogeneous; therefore, its efficacy for eradicating MRSA is not always guaranteed.
All the ETT MRSA were susceptible to LNZ and VAN MIC after long periods of intubation. This clearly highlights that the emergence of resistant strains associated with biofilms is less likely in intubated patients, a finding that is at odds with previous findings in other respiratory diseases [
34]. All VAN MIC but two were below 1.5 μg/mL, a threshold MIC that has been previously associated with lower clinical response, higher relapse [
35], and increased mortality.
In contrast, we did not find differences in biofilm thickness between LNZ and VAN groups. This may be due to differences between secretion production and microbiota in pigs and in human patients. Thus, the proposal that thickness might be a good indicator of treatment efficacy in a highly controlled experiment [
12,
13] may not apply to human patients, where the underlying conditions and other concomitant issues may influence the biofilm and secretions accumulated within the ETT, in addition to length of stay and treatment efficacy.
The results of our study corroborate those of many previous randomized clinical trials. The Zephyr study [
36] observed higher rates of clinical cure in nosocomial
S. aureus pneumonia (both MRSA and MSSA) when comparing LNZ to VAN. Surprisingly, the IDSA/ATS guidelines still place VAN and LNZ at the same level [
4], even though there are enough clinical and animal data to change this recommendation in favor of LNZ [
5]. Our study is also in line with a previous study published by our group in a pig model of MRSA pneumonia in animals ventilated for 72 h. However, findings in animals require replication in humans.
The strengths of our study are the following: (1) this is the first comparison of linezolid and vancomycin in ETT biofilms obtained from humans on long-term mechanical ventilation and (2) ETT (including cuff) biofilms and bacterial burden are studied in depth.
A few potential limitations of this study deserve further clarification. This was not a randomized study, and so there is no possibility of comparing the outcomes. In addition, the fact that we had to recruit patients from different hospitals increased the heterogeneity of the ST types involved. Nevertheless, we did not find any differences in patients’ characteristics or in the length of orotracheal intubation between VAN and LNZ groups, and so the heterogeneity of the MRSA ST collected emphasizes the validity of our results and provides realistic epidemiologic data. Secondly, the use of VAN is becoming less and less frequent in Europe, and for this reason, the number of ETT within this group of study was lower than in the LNZ group. Ultimately, these patients received concomitant antimicrobials that may have combined effects with vancomycin or linezolid. However, concomitant antimicrobials were homogeneously distributed between the two treatment groups.
The main clinical implication of our results is that LNZ, which acts effectively in ETT biofilms and cuffs, performs much better than VAN in MRSA eradication and may be important in preventing relapses in MRSA VAP pneumonia.
Acknowledgements
A.M. is the recipient of a Long-Term Research Fellowship (LTRF 2017-01-00073) from the European Respiratory Society (ERS) and the Spanish Society of Pulmonology and Thoracic Surgery (SEPAR). G.L.B. is the recipient of a postdoctoral grant from the Strategic Plan for Research and Innovation in Health (PERIS) 2017–2021. AT was awarded with an ICREA Academy Grant (2014-2018). The authors gratefully acknowledge the support of Josep Mª Nicolas, Elizabeth Zavala, Javier Fernandez, Miquel Ferrer, and Irene Rovira for obtaining endotracheal tubes at their respective ICUs, Jordi Vila for his critical review of the manuscript, and the technical staff at Scientific and Technological Centres of the University of Barcelona (CCiTUB) for their technical support.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.