Endotracheal tube biofilm in critically ill patients during the COVID-19 pandemic : description of an underestimated microbiological compartment

Study characteristics, patients and data collection

We collected ETTs for routine microbiological analysis from all patients who were admitted to the two participating ICUs and who required invasive MV during the third wave of the COVID-19 pandemic in France (March 1 to May 31, 2021). Two ICUs participated in the study, namely one ICU from a general (non-academic) hospital (William Morey Hospital, Chalon-sur-Saône, France) and one medical ICU from a university teaching hospital (University Hospital of Dijon, Dijon, France). Patients over 18 years of age, admitted to the ICU and requiring MV for more than 48 h were included in the study. Patients whose ETT could not be properly collected and dispatched for microbiological analysis and confocal microscopy within 24–48 h were excluded from the study.

At extubation, each ETT was cleaned three times with a sterile saline solution to remove any excess of tracheal secretion (inner and outer surfaces), and was then immediately placed in a sterile bottle, and dispatched to the microbiological laboratory for further analysis, according to a local biofilm protocol (see “Microbiological characterization” section below). All ETTs were cut into four sterile sections, approximately 0.5 cm thick (two below-the-cuff sections and two above-the-cuff sections) for further mesostructural and microbiological analysis.

Data were retrospectively collected using a dedicated electronic Case Report Form (eCRF) (CleanWeb®), completed by the investigators in each study center. For all eligible patients, we recorded the following demographic and clinical information regarding the ICU stay: age, gender, type of admission, COVID-19 disease status (confirmed by PCR), number of organ dysfunction and severity of the disease, medical history, anthropometric reference data, duration of MV, length of ICU stay, occurrence of reintubation, ICU mortality and occurrence of VAP. The eCRF also recorded the structural characteristics (shape, thickness) and microbiological data related to the characterization of ETT-deposited BF, as well as a global microbiological colonization cartography for each patient (respiratory samples from tracheobronchial aspirate (TBA) or bronchoalveolar lavage (BAL), blood cultures from a central venous catheter (CVC) or arterial catheter (AC), urine analysis, and other colonization/infection sites).

Study definitions

Both William Morey Hospital and University Hospital of Dijon adapted their prevention policies to be in line with that described in the guidelines of the French Society of Anesthesia and Intensive Care Medicine and the French Society of Intensive Care on healthcare-associated pneumonia in the ICU21. VAP diagnosis was confirmed by the physician directly in charge of the patient, after a minimum duration of MV of 48 h, according to the same guidelines.

Immunosuppression was defined as immunosuppressive therapy, long-term corticosteroid therapy (> 3 months), Human Immunodeficiency Virus (HIV) infection, solid organ and bone marrow transplants.

Prior antimicrobial treatment referred to any antimicrobial treatment from among amoxicillin/clavulanic acid, quinolone, second and third generation cephalosporins, during the 3 months prior to ICU admission.

Commensal oropharyngeal flora (COF) were considered for each sample, retrieving the association of more than two bacterial species among classical anaerobic (like Peptostreptococcus species) and aerobic species (Viridans streptococci, Staphylococcus spp., Streptococcus spp., Haemophilus spp., Neisseria spp., Corynebacterium spp., …) from normal oropharyngeal flora.

Mesostructural characterization

Adapted from a recent method developed by Thorarinsdottir et al. for the processing of ETT before electron microscopy analysis22, half of the ETT sections (below- and above-the-cuff sections dedicated to confocal microscopy) were fixed in a solution of 4% formaldehyde diluted with phosphate-buffered saline (PBS) for 30 min at room temperature, then washed in normal saline solution to remove the excess formaldehyde with gentle shaking for 10 min. All ETT sections were then allowed to dry at room temperature for a minimum of 72 h before confocal microscopy acquisitions.

Confocal imaging was achieved with a VivaScope®3000 system from Caliber Imaging and Diagnostics (Rochester, NY, USA). Briefly, ETT below- and above-the-cuff sections were placed under the handheld optical lens of the VivaScope®3000 system to provide up to ten [750 µm × 750 µm]-wide image z-stacks covering the whole inner surface of the ETT. Each image z-stack included between ten and forty separate acquisitions by an optical pace of 5 µm. Given its large field of view, the VivaScope®3000 gives access to high resolution imaging of the biofilm structure at the mesoscopic scale (approximately within 1 mm), as defined in the work by Wagner et al.23. Table S1 introduces the main characteristics and differences between both the VivaScope®3000 and a classical Leica Confocal Laser Scanning Microscopy (CLSM) system.

Quantitative confocal image analysis and ETT-deposited BF thickness were accessible with a contrast-based ImageJ 1.52o protocol (Plot Profile Function) and calculated as the mean of three representative measurements throughout each ETT section. ETT-deposited BF mesostructural acquisitions and analysis were performed by a physician who was blinded to all patient information based on three qualitative morphological criteria: surface aspect (smooth or rugged), attachment to the inner surface of the ETT (detached or tied) and inner covering of the ETT (continuous or intermittent). Ribbon-shaped BF were described as smooth, detached, and continuous. Mushroom-shaped BF were both rugged and intermittently tied to the inner surface of the ETT.

Microbiological characterization

The remaining half of the ETT sections (below- and above-the-cuff sections) were gently homogenized in a 20% Digest-EUR® solution (Eurobio Scientific, Les Ulis, France) for 15 min at 37 °C according to a previously described protocol16,24. After a few seconds of gentle shaking, an aliquot of the solution was then spread on different culture plates (trypticase soy agar, blood agar under aerobic and anaerobic atmospheres, Drigalski agar, chocolate agar with the addition of 5% CO2, and Sabouraud agar) incubated for 48 h at 37 °C, either pure or at 1/10 dilution. Subsequent BF bacterial counts were determined according to laboratory protocols used for respiratory samples. Bacterial identification was done using matrix-assisted laser desorption-ionization-time of flight mass spectrometry (MALDI-TOF MS, Bruker Daltonics, Germany) and susceptibility testing was performed using standard clinical routine methods according to the latest CASFM / EUCAST guidelines25.

For each patient, microbiological BF characterization was compared to all respiratory samples (TBA-BAL) and global colonization sites (TBA-BAL, CVC-AC blood cultures, urine analysis as well as other colonization / infection sites) available during the whole ICU stay.

Statistical analysis

GraphPad Prism version 5 was used for statistical analysis. Results are displayed as number (percentage, %) for categorical variables and as median (interquartile range, IQR) for quantitative variables. The Chi-square test was used to compare qualitative variables, and the Mann–Whitney test for quantitative variables after verification of the normality of distribution. Pearson’s correlation coefficient (r) was calculated to evaluate the strength of the linear correlation between ETT-deposited BF thickness and MV duration. All p values were two-tailed and differences were considered as significant for p values < 0.05 (* for p values between 0.01 and 0.05 ; ** for p values between 0.001 and 0.01 ; *** for p values < 0.001 ; ns stands for non significant).

Ethics approval and consent to participate

In line with current French legislation governing clinical research, this study falls outside the scope of Jardé’s law20, and as such, the Ethics Committee CPP (Comité de Protection des Personnes) Est I confirmed its approval for all experimental methods and informed consent was not required. The study protocol was conducted in accordance with the relevant guidelines and regulations. Patients were nonetheless informed in the admission booklet and via posters in the ICU that the data from their medical files could be used retrospectively for research purposes.