Atomic force microscopy and scanning electron microscopy as alternative methods of early identification of pathogens causing catheter-related bloodstream infections of patients in ICU

According to various sources, the incidence of catheter-related bloodstream infections (CRBSIs) ranges from 0.1 to 22.7 per 1,000 catheter days [1, 2, 3, 4, 5, 6, 7]. Both Gram-positive (40–68%) and Gram-negative (19–77.8%) bacteria may cause catheter-related bloodstream infections in intensive care units. The following pathogens are predominant in CRBSIs: Staphylococcus aureus, Staphylococcus epidermidis, and other coagulase-negative staphylococci. The most commonly cultured Gram-negative rods are: Pseudomonas spp., Acinetobacter spp., E. coli, Klebsiella spp. and Enterobacter spp., whose epidemiological significance is growing, as they are responsible for increasing morbidity and mortality of patients [1, 3, 5, 6, 7, 8].

Currently central venous catheter tip culture (CTC) is the method used to confirm CRBSIs. An infection is diagnosed if the culture value is equal to or greater than 15 CFU (colony forming units) and if the same pathogen is cultured from peripheral blood. Another method is the ratio of quantitative culture (RQC), which compares the growth of pathogens from blood collected simultaneously from a potentially infected central venous catheter and peripheral blood. Differential time to positivity (DTP) is a comparison of the growth of pathogens from blood collected simultaneously from a potentially infected central venous catheter and peripheral blood. A catheter-related infection is diagnosed if the same pathogen is cultured from both biological materials and if the growth of pathogens from the central venous catheter precedes the growth of one pathogen from peripheral blood by at least 2 hours. It is necessary to collect exactly the same volume of blood into all test tubes [2, 9, 10, 11, 12, 13]. These methods are used less frequently due to technical difficulties. Central venous catheter tip culture (CTC) forces the removal of the catheter tip from the patient’s vascular bed. If a multi-lumen catheter is used, the Infectious Diseases Society of America (IDSA) recommends collecting blood from all catheter channels and a culture exceeding 100 CFU/ml should be regarded as a positive result. However, all these methods are difficult to interpret. The identification of the pathogen responsible for CRBSI is time-consuming. There is a positive relation between rapid identification of the pathogen and the implementation of targeted therapy. This relation is of key importance for the patient, as it is necessary to control the clinical symptoms of sepsis and septic shock, which are associated with a high mortality rate [1, 12, 14, 15].

The aim of the study was to determine whether it is possible to reliably identify the presence and type of pathogens on the surface of catheters collected from patients with suspected catheter-related bloodstream infection by means of an atomic force microscope (AFM) and scanning electron microscope (SEM), and whether they could comparably replace the traditional microbiological methods of identification of pathogens. Another aim of the study related to the standardization of the method was to determine the usefulness of this procedure in microbiological diagnostics and the wait time for the result confirming the presence of the pathogen, and to compare this method with the microbiological identification method recognized as a diagnostic model.

The colonization of catheters with pathogens was assessed with an atomic force microscope (AFM) and a scanning electron microscope (SEM). The AFM enables not only assessment of the test surface but also measurement of the size of structures observed. Images are magnified 10⁸ X. The theoretical resolution in relation to the X- and Y-axes is up to 0.1 nm, and to the Z-axis even up to 0.01 nm. The SEM uses secondary electrons and thus enables observation of solid samples, without the need to prepare thin films to observe the surface layer of the sample, that is, its topography and structure. SEM measurements were made with a FEI Quanta 250 FEG microscope. During the measurements, accelerating voltages from 5 to 10kV were used. It was imaged in SE mode with LFD or GSED detectors depending on the pressure in the chamber. Low vacuum mode (60–70 Pa) or environmental mode (100–200 Pa) was used. The samples were not covered with any conductive layer prior to measurement. AFM measurements were taken with an Agilent 5500 microscope. The microscope was operated in an atmosphere of air at room temperature. All measurements were made in intermittent contact mode. Cantilevers Budget Sensors Allinone with a resonance frequency of approx. 150 kHz and spring constant about 7.4 N/m was used for imaging. The typical radius of curvature of the tip was less than 10nm. Scanning speed was approx. 1 line per second. In both measurements (SEM and AFM) the outer surfaces of the sampled fragments of the central catheters were examined. The data were processed with the WSxM software [17].

The results of microscopic examinations were compared with the results of microbiological analysis (central venous catheter tip culture and blood collected through the catheter). The central venous catheter tip culture samples were collected from 24 patients from three intensive care units (ICUs): two adult units and one pediatric unit. The material for the study was collected between February and October 2014 and between June 2018 and February 2019. The patients exhibited clinical symptoms of catheter-related bloodstream infections (CRBSIs): that is, an increase in selected inflammatory parameters (CRP or C-reactive protein; PCT or procalcitonin; WBC or white blood cells). CRBSI was suspected due to the time the central venous catheter remained in the vascular bed (minimum 48 hours). The patients with suspected CRBSI who were transferred from another hospital were excluded from the study. After the central venous catheter was removed from the patient with suspected CRBSI, a three-centimeter distal part of the catheter was cut off under aseptic conditions and divided into three parts. One of them was sent for microbiological analysis, and the other two were sent to a microscopy laboratory (one was fixed in 95% ethanol and the other was not fixed). Because there were no significant differences in the visualization of microorganisms between the catheter (fixed/not fixed in 95% ethanol) samples tested, we used a sample fixed in alcohol for testing, which was safer (it prevented further multiplication). The results of microscopic examinations were obtained not later than 6 hours after the collection of material. They were compared with the results of microbiological analysis of the central venous catheter tip and blood collected from the catheter. Blood cultures were prepared automatically with a Bactec apparatus (Becton Dickinson). The central venous catheter tips were cultured onto a solid microbial agar with 5% sheep blood. The growth of 15 or more colonies indicated high probability of catheter-related sepsis. Additionally, when the tip was cultured onto a solid medium and the count of microorganisms was low, it was transferred to a broth medium for microorganisms to proliferate. After one day of incubation the broth was sieved onto solid mediums. Currently our microbiological laboratory is capable of identifying the species of bacteria and fungi by means of mass spectrometry (matrix-assisted laser desorption/ionization – time of flight, or MALDI–TOF MS). The method is based on the measurement of specific proteins. The spectra of proteins are compared with the current library (microbial database). Thus, microorganisms are identified and classified to a specific species. At present it is the best and most reliable testing method.

Catheter-related bloodstream infection was defined according to the CDC guidelines [12, 16]: positive blood cultures collected from separate catheters and related positive culture of the CVC tip with the same organism and exclusion of another source of bacteremia.

The presence of biofilm was considered possible in the SEM analysis if the biofilm-like matrix and the shape of bacterial cell structures were visualized. The presence of bacterial cells was confirmed if there were visible cocci (0.5–1 µm) or rod-shaped cells (2–5 µm) and the image corresponded to at least one of the following traits: cell division or characteristic spatial arrangements, cell clusters compatible with a given microcolony (staphylococci, bundles, streptococci) (Fig. 4, 5, 6, 7, 8).

The study group consisted of patients who stayed in the ICU for an average of 33.83 days. 5 patients died within 30 days of the suspicion of catheter-related bloodstream infection. The most common risk factors observed in the patients were: severe infection – 23 patients, steroid therapy – 12 patients, active neoplastic disease – 2 patients, and bone marrow suppression – 3 patients. 7 patients were aged over 65 and 3 patients suffered from an autoimmune disease. During hospitalization there were 7 cases of sepsis and 6 patients were in shock, including 1 patient in cardiogenic shock and 5 patients in septic shock. 21 patients received blood products. 8 patients underwent nutritional therapy via central venous catheter.

A brand-new catheter was first analyzed microscopically to obtain a pattern of its inner and outer surface and reveal its structure in both the AFM (Fig. 1a) and SEM microscopes (Fig. 1b). The catheter was examined immediately after opening the package under aseptic conditions. Both photos show longitudinal depressions, which are 50–100 nm wide and probably were formed in the manufacturing process. The scratch depth measured with the AFM was about 200 nm. This type of surface topography may favor microbial colonization [17].

Fig. 1

The microscopic examination and microbiological analysis of both the blood and central venous catheter samples confirmed the presence of microorganisms in 16 cases (double positive result). The microbiological and microscopic examinations did not confirm the presence of pathogens (double negative result) in 7 cases. The microscopic and microbiological examinations of the central venous catheter did not confirm the presence of pathogens in the sample collected from one patient, but the result of the blood culture was positive (Staphylococcus hominis). This means that the patient’s blood was infected from another source (sepsis was not related to the central venous catheter).

The microscopic images of central venous catheters confirmed 16 cases of colonization with Gram-positive and Gram-negative bacteria. Several slides revealed a highly developed structure of the biofilm connecting bacterial colonies through an exoskeleton (Fig. 2a, b, c, 3a, 6, 7, 8). The rounded structures were connected with other long fibers. The diameter of the bacteria was usually about 1 µm. The fiber diameter was about 200–300 nm and the average length was 10 µm. The biofilm acted like a diffusion barrier protecting microorganisms from antibiotics. In Figure 3b the bacteria were completely covered with a continuous biofilm structure without fenestrations. The microbiological analysis confirmed the presence of the Staphylococcus epidermidis pathogen both in the blood and in the catheter tip culture. The S. epidermidis biofilm contained more than 50 cell clusters embedded in an extracellular matrix, which was up to 160 µm thick.

Fig. 2

Fig. 3

The microscopic image of the catheter from patient No. 20 revealed typical bacterial colonies arranged in clusters (Staphylococcus spp.) (Fig. 4). The diameter of the bacteria was about 1 µm. Each colony shown in the image contained 5 to 50 bacteria. The structure of the biofilm is not clearly visible, or it is a very thin layer (Figure 4b). The microbiological analysis of the catheter collected from the patient confirmed its colonization with the Staphylococcus epidermidis and Pseudomonas aeruginosa strains. Only the Staphylococcus epidermidis strain was cultured from the patient’s blood. However, the Pseudomonas aeruginosa bacteria may have formed an intact biofilm and therefore they were not identified with conventional microbiological methods.

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

The microscopic examination of the catheter tip collected from patient No. 21 showed elongated structures (length: 2–5 µm, diameter: 100–300 nm) on the catheter surface. No biofilm was found. The microbiological analysis revealed the colonization of the catheter with the Enterobacter cloacae complex and Staphylococcus epidermidis strains, but only the Enterobacter cloacae complex strain was cultured from the patient’s blood. The S. epidermidis bacteria may not have entered the bloodstream from the catheter. There were 6 cases where the presence of bacteria on the surface of the catheter was confirmed by microbiological analysis, but these bacteria were not cultured from the patients’ blood.

The problem of catheter-related bloodstream infections in patients in intensive care units (ICUs) is complex. The increasing possibilities of supporting critically ill patients in the complex treatment of failures of individual systems and organs creates the need to use different and numerous intravascular catheters at the same time. On the one hand, this involves invasive treatment of the patient’s tissues and vascular bed, which may be temporarily damaged by the catheter and/or the catheterization method. On the other hand, this involves a wide range of infectious complications. Due to the disease profile, often combined with the patient’s weakened immunity, it is usually necessary to apply long-term, sometimes combined, antibiotic therapy. Microorganisms form a biofilm to protect themselves from the antibiotic. It is a three-dimensional, complex multicellular structure, surrounded by a layer of various components such as exopolysaccharides, cellulose, proteins, and DNA material, which are produced by these bacteria. Microorganisms form a living membrane connected to the surface of the catheter. It has a complex system of channels, whose function resembles the structure of the circulatory system. A biofilm can be found in as many as 70% of nosocomial infections [11, 12]. Foreign bodies (catheter, implants, drains) are excellent materials on which microorganisms can form a biofilm, which is a defense mechanism enabling their survival in the host’s body. So far research has shown that the bacteria which can create a biofilm have higher colonization capacity. In consequence, their resistance to the body’s mechanical, metabolic, and immunological defense mechanisms increases. It is difficult for phagocytes to access bacterial pathogens and the supply of reactive oxygen species is inhibited. The biofilm membrane increases pathogens’ resistance to antibiotics by weakening the ability of drugs to penetrate through it and access bacteria or fungi. This mechanism involves complex biochemical mechanisms based on systems of biological pumps, including ion pumps, which reduce the effect of drugs by releasing substances neutralizing the effectiveness of antibiotics. Planktonic cells – that is, free-floating cells – are released from the biofilm network. The first report on the role of biofilm was published in 1982. It described catheter-related bloodstream infections with Staphylococcus epidermidis. Since that time the role of biofilm in the pathogenesis of catheter-related bloodstream infections has been confirmed in numerous studies [3, 9, 12, 17, 18, 19].

The following factors increase the risk of CRBSI: failure to apply aseptic technique during catheterization and while using catheters; long-term catheterization; wrong site of catheterization and the number of central venous catheters applied to a patient; diabetes, neutropenia, immunosuppression; prior antibiotic therapy; parenteral nutrition via central venous catheter; duration of stay in the ICU; and cause of admission to the ward [3, 5, 11, 20, 21, 22]. All the central venous catheters examined in our study were installed under sterile conditions through the internal jugular vein or subclavian vein into the superior vena cava. Five of our patients did not survive 30 days after the suspicion of CRBSI. All of these patients had more than two CRBSI risk factors. Microorganisms may colonize the central venous catheter within 48 hours after its insertion. During this time microorganisms may form a biofilm both on the inner and outer surface of the catheter, which may cause bacteremia or even catheter-related sepsis [3, 9, 22]. The susceptibility of the catheter to adhesion, colonization, and the resulting catheter-related bloodstream infection are directly related to the type of catheter surface and micro-clots on it. AFM images of a brand-new catheter, which is ready to be installed in a patient, show its uneven surface, which facilitates microbial colonization (Fig. 1a and b). The catheters examined in our study were used for 6–14 days. All of them were consecutively installed in individual patients. 23 patients suspected of CRBSI received an antibiotic while the catheter was being replaced.

So far there have been some studies on the use of a scanning electron microscope (SEM) for diagnosing catheter-related bloodstream infections [23]. Researchers have stressed the advantage of this method: three-dimensional images with the distribution of biofilm on the inner and outer surface of the catheter. However, the high cost of the method is an obstacle to its everyday use. The results of studies showed the high correlation between the sonication method (the use of ultrasound to disintegrate particles in a suspension) and the SEM analysis. The authors of the cited article used the sonication method, which revealed the microorganisms that had not been detected with other methods [23]. The interest in using SEM and AFM in microbiology and many other areas of medicine is constantly growing and it offers more and more opportunities for cooperation [24, 25, 26, 27, 28].

This method is a useful tool in basic research and in diagnostics. SEM has also been used in peridontology to assess implant infections, tissue morphology, and the interaction of bacteria and tissues [24, 29]. Moreover, it facilitated the knowledge and understanding of the processes of biofilm formation by microorganisms on the surfaces of materials used in medicine [30]. By the use of this diagnostic method, the relationship between chronic Chlamydia infection and diseases such as atherosclerosis and pulmonary emphysema was confirmed [31]. This method proved to be helpful also in the observation of the interaction of the Ebola virus with the host cells, showing the morphological changes occurring during this process [32].

Is it useful to remove central venous catheters?

The duration of use of a central venous catheter in the patient’s body is a confirmed factor of catheter colonization (cc) and catheter-related bloodstream infections (CRBSI) [3]. The removal of a central venous catheter should be a justified procedure if CRBSI is suspected. The circumstances of catheter-related bloodstream infections (skin and subcutaneous tissue infection, suppurative phlebitis) should be taken into account in clinical evaluation. Clinical symptoms such as pulmonary embolism, endocarditis, persistent bacteremia, recurrent infections despite treatment, abscesses in the liver or spleen, and hemodynamic instability, which cannot be explained by another cause, should always be considered as symptoms potentially dependent on the presence of a central venous catheter. If there are difficulties replacing the central venous catheter, such as in patients with long-term catheters after numerous CRBSIs, antimicrobial lock therapy (ALT) should be considered: that is, filling the central venous catheter with a concentrated solution of an antibiotic (dissolved in a small volume) in order to act locally on the biofilm. This therapy lasts 10–14 days. It is recommended if catheter colonization is suspected. Systemic administration of antibiotics is recommended only if the patient meets the sepsis criteria. There have also been scientific reports on ethanol lock technique (ELT), which consists in a two-hour exposure of the catheter to 70% ethanol [11, 12, 19].

The central venous catheter should always be removed whenever inflammatory parameters increase and clinical symptoms occur (elevated CRP, PCT, WBC, body temperature), provided that other sources of infection have been eliminated. A study was conducted on patients with suspected CRBSI in 18 intensive care units in Spain. The patients were divided into three groups (1 – catheter removed immediately, 2 – catheter removed at a later stage, and 3 – catheter left). There was no statistical difference in thirty-day mortality between the groups. The thirty-day mortality of the patients with CRBSI was lower than in the group of patients where clinical symptoms of the infection were related to another source. Moreover, the regression model showed that the risk of death within 30 days was higher among the patients who had the central venous catheter replaced immediately than in the group where the catheter replacement was delayed (p = 0.02). It may have happened because empirical antibiotic therapy was not initiated after immediate removal of the catheter, and the detachment of the biofilm from the catheter resulted in sepsis. The authors suggested that immediate replacement of the central venous catheter, which may be a source of infection, should be considered only in certain groups of patients, such as in cases of immunosuppression, hemodynamic instability, or organ transplantations. Simultaneously, it is necessary to consider the risk of mechanical complications which may occur during replacement of the catheter [1, 14].

1. The pilot study showed the possibility of using the atomic force microscope (AFM) and scanning electron microscope (SEM) to confirm the presence of pathogens on the surface of central venous catheters collected from patients with suspected catheter-related bloodstream infections. These advanced microscopic technologies may support traditional microbiological analysis. Due to the shorter time of initial diagnosis, they are beneficial for ICU patients.

2. In order to use these methods, it would be advisable to start bilateral and formal cooperation with a center enabling quick microscopic diagnosis of removed catheters. This would help to make quick and more effective therapeutic decisions.

3. In order to prepare a diagnostic procedure protocol based on microscope 1 and microscope 2, it is necessary to create as a next step a diagnostic model corresponding to repetitive images deposited in a digital catalogue and to determine what area of the technological material should be sent for diagnosis. This method of microbiological analysis seems to be the most appropriate for the aforementioned conditions.

4. It cannot be ruled out that the advanced technologies presented in our study may become decisive in the treatment process if it is impossible to confirm the presence of a pathogen in a patient with sepsis and septic shock by means of microbiological analysis. Unfortunately, traditional microbiological diagnostics may fail to identify pathogens for various reasons.

AK, HT, MN - Research concept and design

AK, HT, KK, ABŚ, MG - Supervising the project, carrying out the experiments

AK, TK, HT, MN - Acquisition of data

AK, HT, MN - Data analysis and interpretation, writing – original draft preparation

AK, HT, KK, MG – Writing – review and editing

MN, AK - Visualization

AK - Literature review

KK, AK, HT, MG - Final proofreading and approval of the version for publication

KK - Funding acquisition