Archived copy

Page 2

Epidemiology

The current importance of CoNS infections is emphasised by the frequency with which these organisms are isolated from the blood of hospital patients. In 1978, around 250 episodes of CoNS infection in the British Isles (excluding Scotland) were reported to the Communicable Disease Surveillance Centre of the Public Health Laboratory Service. This figure rose to more than 3000 in 1993. In contrast with S. aureus infections, which arise equally between cases in the community and cases in hospital,³ CoNS infections are almost entirely associated with hospital care.

Among the many recent advances in medicine, the ability to deal with degenerative or traumatic joint disease, vascular heart disease, malignant conditions and kidney failure, has been aided by the use of biomedical material implants. Furthermore, the ability to monitor and treat patients with acute medical conditions effectively has been assisted by the availability of pressure monitoring systems, arterial grafts and intravenous catheter devices. However, each of these devices carry a 0.1-50% risk of becoming infected (see Table 2 and Figure 2).

Fig 2Arterial pressure monitoring catheter showing infection at the penetration site.

Infection can arise for a variety of reasons. With intravascular devices, for example, there is direct contact between the blood and the external skin flora. Other prosthetic devices which are inserted deep within the body, such as heart valves or joints, run the risk of becoming infected during the operation. Clinical recognition of such infection can be delayed for weeks or months after surgery, owing to the indolent nature of many CoNS infections. The results can be particularly devastating with prosthetic joints, as these are often inserted into the elderly, who are less able to cope with the further major surgery that is usually needed to eradicate persistent infection.

Similarly, infection of a prosthetic heart valve presents a major clinical dilemma. The infection is usually present as small crumbly vegetations attached to the valve and associated structures. These can break loose and settle in other parts of the body, causing serious problems, such as a stroke or gangrene of a finger or limb. Furthermore, antibiotic treatment is often unsuccessful and replacement of the infected implant is required. Another serious problem arises when a spinal fluid shunting system inserted into patients with hydrocephalus (water on the brain) becomes infected. The patient (often a child) is at risk of persistent meningitis and valve malfunction, both of which have serious consequences.

Pathogenesis of infection

Attachment and evasion To be successful, a pathogen must enter a suitable host and grow within body tissues and fluids, while avoiding being detected and killed by the host's immune system. Here we will focus on some of the events leading to CoNS infection of medical devices.

Virtually all medical prostheses placed into normally sterile body sites are at risk from infection. For example, where catheters are used, the CoNS can gain entry through the skin during insertion of the tip, or by tracking along the insertion site, or as a consequence of blood-borne colonisation of the catheter surface. Once the skin barrier has been breached, the first step to infection involves the bacterium adhering to the surface of the implanted medical device. Despite the variety of polymers, plastics and metals used to make medical devices, CoNS have the unusual property of being able to stick to and colonise any such implants.⁴ How this is mediated and why CoNS, rather than their more virulent relative S. aureus or other pathogenic bacteria, reign supreme in this context has been the subject of much debate and experimental investigation.

What has emerged is that CoNS stick avidly not only to a variety of human cells and tissues (see Figure 3), but also to polymers such as polystyrene. Although many investigators have suggested that the major factor influencing the initial attachment is bacterial surface hydrophobicity, an adhesive surface macromolecule (called PS/A, for capsular polysaccharide/ adhesin) which helps S. epidermidis stick to silicon rubber has been described by Gerald Pier's group at the Brigham and Women's Hospital in Boston. Loss of PS/A by mutation reduces S. epidermidis' ability to cause disease in experimental animal models, implying a direct link between PS/A production and pathogenicity.⁵ A 220kDa surface-associated protein in S. epidermidis also promotes attachment to polystyrene,⁶ suggesting that there may be a number of factors which modulate initial CoNS-adhesion to polymer surfaces.

Fig 3False colour image of a monolayer of epithelial cells, which form the outer skin and the lining of many human body organs, attached to a plastic surface. The Staphylococci stick to the surface of these cells. The yellow areas show the gaps between the cells

Although this kind of interaction may explain the initial attachment of CoNS to medical devices in the laboratory, in the patient additional complexities arise. This is because, once they're implanted, medical devices soon become coated with components of the body fluids in which they are bathed. This is known as 'surface conditioning', and initially involves the deposition of blood proteins and platelets, then connective tissue and extracellular matrix components such as fibronectin, fibrinogen and collagen.⁴ Such conditioning does not deter the staphylococci; on the contrary, these components can act as 'receptors' for bacterial attachment.

Indeed, in the case of S. aureus, several research groups have characterised, at the molecular and genetic level, specific bacterial cell wall proteins that help binding to fibronectin and fibrinogen.⁷ At the University of Lund in Sweden, Torkel Wadstrom and co-workers have clearly demonstrated that while CoNS bind to a similar spectrum of host proteins, they use cell wall proteins that are quite distinct from those of S. aureus.⁸ These findings are beginning to explain the propensity of CoNS to colonise medical devices constructed from a variety of different materials.

Following the initial attachment phase, the CoNS adopt a highly characteristic mode of growth in which they begin to produce a mucoid substance or 'slime'. This growth method was first implicated in the development of infection back in 1972 by Roger Bayston, then at the Children's Hospital in Sheffield.⁹ However, his observations were not fully appreciated until the 1980s when CoNS infections of intravascular catheters became a highly significant clinical problem. It is now widely accepted that the second stage of CoNS infection involves slime production and biofilm formation (see Figure 4).^4,10,11

Fig 4 Scanning electron micrograph of a microcolony layer of coagulase-negative staphylococci (round cells) embedded in slime and forming a biofilm on the top of the internal surface of a catheter

The slime mainly consists of polysaccharides, with about 10-20% proteins, and probably stabilises the biofilms by promoting bacterial cell-to-cell and cell-to-surface associations so that multi-layered cell clusters accumulate on the implant surface. Life within this sticky matrix seems to help the CoNS to survive, allowing them to feed and interfering with host cellular defences (such as phagocytosis and T-cell proliferation) and antibody production.¹² However, the data directly supporting slime as a virulence factor remain equivocal, largely because of the lack of precise chemical definition and because the genes responsible for directing slime production have yet to be located.

[Page 1] [Page 3]

20 March 1995