Central venous access device (CVAD)-related sheaths are a phenomenon that may lead to minor functional problems of CVADs or significant sequelae for patients, such as the development of extravasation (Mayo, 1998; Kumwenda et al, 2018a) and are associated with venous occlusion (Krausz et al, 2014). CVAD functionality may be adversely affected—there may be failure to withdraw blood (persistent withdrawal occlusion), or infusion rates can be reduced (Kuter, 2004). Both of these lead to delayed or missed intravenous therapy. Protocols within the authors’ institution for persistent withdrawal occlusion related to CVAD sheaths may include chest X-ray, administration of thrombolytic agents and saline infusions to ensure the device is safe to use. This process has a negative impact on the patient experience and clinical and financial resources.
The incidence of CVAD sheaths has been reported to be 78-100% (Kuter, 2004). Seifert et al (1997) explained that virtually all foreign-body material soon becomes coated with glycoproteins containing fibrinogen, fibronectin, collagen and other proteins.
Composition and pathophysiology
The development of catheter-related sheaths was first documented by Motin and colleagues (1964). In early animal studies clinicians sought to identify the composition of the sheath phenomenon. Xiang et al (1998) studied 123 animal subjects with silicone catheters placed that were then removed at different time intervals, including immediate withdrawal of the catheter and dwell times of days, weeks and up to 10 months. The pathogenic effects were then studied. For the catheters that were immediately removed, at day one 100% of catheters were found to have thrombus sheaths.
| Specific factor | Species | Possible mode of action |
|---|---|---|
| Fibronectin | S aureus | Adherence, promotion |
| Fibrinogen | S aureus | Adherence, promotion |
| Thrombospondin | S aureus | Adherence, promotion |
| Platelets | S aureus | Adherence, promotion |
Source: Seifert et al, 1997
Reddy et al (2007) supported Xiang et al's findings and suggested that the sheath is formed from endothelial injury at the insertion site, but added that the catheter tip's interaction with the wall of the superior vena cava also encourages sheath growth. Xiang et al (1998) identified in all subjects that a translucent sheath was still visible 2 weeks, 2 months and 10 months after catheter withdrawal and suggested the sleeve is considered a predictable consequence of exposing an artificial surface to blood.
The incidence of retained sheaths after catheter removal was investigated by Krausz et al (2014) who found 14% of sheaths were retained; however, the CT protocols were not optimised for fibrin remnant detection, which may have led to a lower incidence being reported. Niehaus (2015) confirmed a high-resolution CT approach is required and not readily available for detection of biofilm.
Nycz et al (2018) examined dialysis catheters that had been indwelling from 4 months to 1 year with a scanning electron microscope, atomic force microscope and a goniometer and described the composition of the film that had formed on the catheter as fibrinogen, globulins, albumins and other proteins, and blood cells as well as coagulation factors (Nycz et al, 2018). They found that the proximal portion of the catheter layer contained precipitate crystals of sodium heparin, due to the proximity of the injection site of the anticoagulant.
Marcia Ryder summarised the process of sheath formation:
‘Upon arrival of the catheter into the venous system, circulating plasma proteins instantly collide and rapidly bind with the biomaterial. In the next few minutes, the coagulation cascade and the complement system are activated, attracting platelets and polymorphonuclear leukocytes to the foreign material. The matrix of absorbed protein, adherent leukocytes, aggregated platelets, and accumulated fibrin composes a “conditioning layer” that envelops the catheter as a fibrinous sheath. A partial or occlusive pericatheter thrombus may develop over the fibrin sheath within the next few days.’
Jordan et al (2012) explained that fibrinogen is most likely cross-linked transversely in its molecular design, accounting for its stretchiness and ability to retract after being stretched. Wallace et al (2017) made the point that ‘a key component in thrombosis is fibrinogen and it is likely one of the first components to adsorb to artificial surfaces’.
Catheter sheath formation is a complex multifaceted process. In addition to the mechanics of the insertion site, endothelial injury and interaction between the catheter tip and vessel, invisible molecular interactions are at work. Namely, van der Waals forces, electrostatic and hydrophobic interactions are also a factor in sheath formation (Seifert et al, 1997). Van der Waals forces are intermolecular forces, the higher the molecular weight the higher the van der Waals force, where the nucleus of an atom attracts the electrons from another. These are relatively weak forces but together with other biological processes they play a role in adhesion. Authors investigating the effect of van der Waals forces on cell adhesion found that the adhesion force measured between macroscopic polymer spheres was found to be strongest when the surfaces were absolutely smooth and clean (Kendall and Roberts, 2015). Kendall and Roberts (2015) surmised that van der Waals force is a ubiquitous attraction across all molecules and thereby must contribute to cell adhesion. Pascual (2002) made a similar suggestion and explained that van der Waals forces play a role in protein adhesion to biomaterial and thus the formation of catheter sheaths.
Infection and CVAD sheath
Microbial adhesion to sheaths is affected by van der Waals forces, including electrostatic and hydrophobic interactions, which lead to strong bonds being created between the microbes and the host proteins, namely the catheter sheaths (Pascual, 2002). Bouza et al (2002) supported the relationship between sheaths and microbes, suggesting that bacteria including Staphylococcus aureus adhere to the host-derived proteins within ‘thrombin sheaths’ covering the internal and external surfaces of catheters.
As explained by Pascual (2002), many bacteria have hydrophobic components on their surface and when two hydrophobic entities interact in an aqueous environment they tend to become attached. Plastic biomaterials commonly used to make intravascular devices are hydrophobic, facilitating attachment of bacteria (Pascual, 2002). The correlation between bacterial adhesion to hydrophobic biomaterials has been demonstrated for Staphylococcus epidermidis (Pascual, 2002).
The relationship between catheter sheaths and infection has been the focus of a number of studies. Mehall et al (2002) showed in an animal study that catheters that were actively challenged with bacterial injections had a significantly higher incidence of infection if a ‘fibrin sheath’ was present versus catheters without sheaths. Positive blood cultures were obtained in 55 of 68 animals and of those infected 85% (n=47) were associated with catheter sheaths versus 15% (n=8) without catheter-related sheaths.
Ryder (2001) suggested that certain constituents of sheaths act as foundation for infections. Fibrinogen and fibronectin enhance the binding of S. aureus. Mehall et al (2002) reaffirmed this point as in their study a fibrin sheath was shown to promote and mediate bacterial attachment and infection. S. aureus is known to bind to plastic surfaces via fibronectin as a mediator of adherence; blood and platelets also play a role in S. aureus adhesion (Seifert et al, 1997).
Catheter sheaths are not often completely removed at the time of CVAD removal. Figure 1 shows the rare occurrence of partial sheath still present on the polyurethane catheter of an implanted port, following removal. The ghostly remains of the catheter sheaths may continue to host microbes that pose a significant threat to their hosts long after the catheter has been removed. Sheikh et al (2017) investigated a patient with S. epidermidis bacteraemia. Transoesophageal echocardiography was undertaken and a right-sided heart endovascular infection was identified related to a fibrin sheath, still present from a catheter that had been removed 2 months earlier. Tang et al (2015) studied 11 patients with confirmed infective sheath vegetations that were still present even though the catheter had been removed. Mortality was 36% with death ranging from 44 to 251 days after diagnosis and hospital readmission rate was 55%; 8 of the 11 patients were septic and required intensive care (Tang et al, 2015).
Biofilm
During polymer colonisation many staphylococci exude slime that embeds and covers the cells and protects them against host defences. This complex extracellular matrix, known as biofilm, forms a protective barricade around bacteria, making it difficult for the body's defences and impossible for some antibiotics to penetrate (Clayton and Mah, 2017). Bacteria are social creatures that interact and co-ordinate behaviours, cell-to-cell communication or quorum sensing facilitates processes such as proliferation and biofilm formation (Pascual, 2002). Bacteria have the ability to sense bacterial density via cell-to-cell signalling then form aggregates between cells via a self-produced extracellular matrix, and illustrate an interconnected synergy of communication and interaction (Passos De Silva et al, 2017). Quorum sensing is known to affect biofilm formation on both Gram-negative and Gram-positive species (Passos De Silva et al, 2017). Niehaus (2015) found the presence of biofilm on 90% of adult intravascular catheters examined and 70% of paediatric catheters.
Like clinicians, science and industry strive to reduce CVAD infective complications, coated or impregnated catheters have been developed. Better outcomes have been seen with catheters impregnated with chlorhexidine and silver sulfadiazine and some other antimicrobial catheters (Wang et al, 2018); however, great efforts are being made to reduce the use of antimicrobials in the attempt to reduce antibiotic resistance, limiting the use of antimicrobial catheters. Nyzc et al (2018) explained that the conditioning layer that forms shortly after implantation can mask the function of some antimicrobial coatings of impregnated catheters.
In terms of preventive measures against catheter sheaths some laboratory studies have shown promise. In one study catheters were inserted into animal specimens, who then received S. epidermis injections. A portion of the animals received enoxaparin, and were compared with those who did not receive the anticoagulant. The authors showed that the amount of fibrin sheath surrounding the catheters decreased in the enoxaparin group and the incidence of catheter colonisation decreases when the amount of fibrin decreases (Keller et al, 2006). Sylvia et al (2018) placed four different coated peripherally inserted central catheters (PICCs) into sheep for 14 and 30 days. Histology analysis was performed on serial catheter sections, which showed fibroblastic sleeve at both 14 and 30 days in all catheters. However, chlorhexidine-coated PICCs showed a 64% lower mean fibroblastic sleeve weight when compared with untreated catheters at day 14, and 81% lower sleeve weight at day 30. A multi-site study examining the efficacy of Syner-Kinase on persistent withdrawal occlusion, a known consequence of catheter sheaths, showed catheter occlusion and 89% success rate for intervention (Kumwenda et al, 2018b).
The study
A nuclear medicine scan can image bones, organs and other parts of the body by using a small dose of a radioactive tracer. A number of radionuclides are routinely employed, Tc99m (technetium) being the most common, but F18 and Ga68 find use in positron emission tomography-CT imaging. The choice of radiopharmaceutical is determined by which organ or part of the body is to be scanned, and for what reason. The preferred route of administration of the radiopharmaceutical is via a cannula in a peripheral vein, but for patients with poor venous access a CVAD offers a convenient alternative.
During some routine radioisotope scans it was noticed that a number of the CVAD catheters exhibited uptake or retention of the radioisotope tracer (Tc99m), when the CVAD was used to administer the tracer. A discussion about these anomalies took place between the nuclear medicine and the vascular access teams, and in view of the potential relationship between CVAD sheaths, thrombosis and infection further detailed investigation was planned.
Tests within the radioisotopes department were undertaken to assess if the CVADs retained isotopes when they were not placed within the patient. The experiments using CVADs showed that after flushing with saline there is no intrinsic tendency of the catheter polymer materials to retain any significant amount of radioactivity, for the radiopharmaceuticals Tc99m and 18FDG (unpublished data). More recently, the authors have also shown the same to be true for 223RaCl2 (radium chloride), m-131IBG (iodobenzoic guanidine), 68Ga-dotatoc and 177Lu-dotatate. Tc99m products and 18FDG are by far the most commonly used tracers in the radioisotopes department; it may be likely that other tracers show comparable behaviour, but the department has simply not yet seen sufficient numbers of cases to substantiate this. Only once the catheters have entered the bloodstream do they have the potential to absorb isotope tracer due to the biological process of sheath development around the catheter.
It would seem reasonable to assume a radioisotope can illustrate the extent of CVAD sheath present, due to the tracer's propensity to bind to protein molecules and an affinity for fibrin (Oliveira and Caravan, 2017). Then those catheters with a greater amount of sheath formation should display a higher degree of radioisotope residual uptake. In other words, the amount of catheter uptake of isotope displayed on the scans delineates the extent of the sheath formation on the catheter of the CVAD.
Figures 2-5 show variable uptake of radioisotope on indwelling vascular access devices.
Methods
Consecutive data were collected on 184 patients, who had CVADs in place and underwent isotope scans, where isotope tracer was infused via the CVAD. The amount of uptake that was retained on the CVAD was then categorised into three groups: no uptake, moderate uptake or significant uptake, representing the relative extent of catheter-related sheath present.
All patients’ medical records were reviewed including microbiology results, radiology reports and electronic/paper medical records. The medical notes were examined from the date the CVAD was inserted to 12 months after the date of the isotope scan, or until the device was removed or to the date the patient died. Data collection included:
Clinicians gathering the data were blinded to whether there was any isotope uptake observed on the scan.
Results
The amount of uptake on the scans, which illustrates the extent of the sheath formation on the catheter, was categorised into three groups: no uptake, moderate uptake or significant uptake. Of the 184 patients’ medical notes reviewed, 179 patient reviews were completed; some had devices placed in alternate institutions and so had insufficient records available to access. From the 179 data sets reviewed 130 patients had no uptake, 30 had moderate uptake and 19 had significant uptake.
Persistent withdrawal occlusion
The rates of persistent withdrawal occlusion were reviewed for patients with each classification of uptake. The figures were not too dissimilar and were between 5% and 7% for all levels of uptake. However, the incidence of persistent withdrawal occlusion is less reliably recorded in the medical records, when compared with thrombosis and infection where there are multiple data sources to obtain the information.
Infection
The numbers of bloodstream infections were measured for each patient from the time of CVAD insertion until 12 months after the scan, or until the devices were removed, or until the patient died. Bloodstream infection incidence for the no uptake group was 7% (n=9), moderate uptake 10% (n=3), and for patients with significant uptake 16% (n=3). The rate of infection was found to be more than double for patients with significant catheter sheaths (significant uptake) compared with those without (no uptake).
Thrombosis
The number of symptomatic thromboses related to the vascular device were recorded, confirmed by ultrasound Doppler or other radiological imaging. The incidence of thrombosis for the no uptake group was 0.8% (n=1), moderate uptake 6.7% (n=2), and the significant uptake group had no incidence of thrombosis. No thrombosis identified in the significant uptake group may be due to the smaller number of patients within this group, pointing towards a need for larger sample sizes in future studies. However, there is a marked difference between patients with no observable sheath and those with catheter sheaths. If significant sheath and partial sheath patients are grouped together, patients with catheter sheaths were five times as likely to develop a thrombosis (incidence of 4.4% vs 0.7%). That the presence of a catheter sheath can act as a precursor to thrombosis has been suggested by Sheikh et al (2017), which appears to be supported by this study.
Total complications
Infective, thrombotic and persistent withdrawal occlusion episodes were combined when looking at total complications (TC). Each level of uptake was compared. The no uptake group experienced 15% TC, moderate uptake 23% TC and significant uptake 21% TC. Alternatively, when comparing no uptake with uptake groups (combined moderate and significant), no uptake TC was 15% versus uptake 22%.
Conclusion
The discussion of the pathological process of catheter sheath formation offered in the literature suggests sheath development is a result of the injury of the endothelium at insertion site and subsequent proliferation of host cells over the catheter. A sheath develops at the tip of the catheter as it comes in contact with the vessel wall, intravascular biomaterial and plasma proteins collide and bind to the catheter, and the coagulation system is activated as the surface becomes covered with protein (Xiang et al, 1998). The adhesion of the sheath to the catheter and subsequent microbial adhesion are as a result of intermolecular forces – Van der Waals, electrostatic and hydrophobic interactions. The presence of fibrin sheaths may remain following removal of a CVAD, and these ‘ghost remnants’ of the catheter may significantly affect patient morbidity and mortality. Following the removal of CVADs patients may still remain at risk, clinical vigilance is necessary in the weeks and months following, particularly in immunocompromised patients. Patients may present with an infection of unknown source related to bacteria burden on the sheath remnants.
This single-centre study showed that oncology patients known to have moderate to significant catheter sheaths were at higher risk of thrombotic and infective complications and experienced a higher total complication rate.
In view of these findings the authors recommend that when patients are identified with higher uptake of isotopes around catheters, signifying the presence of catheter sheaths, the patients should be highlighted to their respective medical teams and infection control team for monitoring.