Originally published November 1997
Advances in medical devices such as catheters, guidewires, stents, pacemakers, and other invasive products have enormously improved diagnostic and therapeutic practices in medical care. However, the benefits of catheters and other invasive devices are often limited by the occurrence of infections associated with the devices, even when the best aseptic techniques are practiced. Each year, as many as 2 million hospital patients in the United States develop nosocomial infections; approximately 80% of the 80,000 annual deaths in this country from nosocomial infections are device related.1
Antiinfective coatings can be applied to many indwelling medical devices. Inset photo: Zone-of-inhibition assay showing effective activity against bacterial organisms. Photo: STS Biopolymers, INC.
Urinary catheters and central venous catheters are notorious examples of infection-prone devices. Urinary-tract infections occur in about 20% of patients with Foley catheters in place for more than 10 days, and in more than 40% of patients with Foley catheters in place for more than 25 days. There are approximately 500,000 cases of these infections in U.S. hospitals each year, and most are associated with catheters.2
Nosocomial bloodstream infections in the United States number more than 100,000 per year, with annual mortality ranging from 10,000 to 20,000 and cost of treatment estimated at $1 billion. At least 50,000 cases of these infections are associated with central venous catheters.3 Other IV devices, such as midline catheters and peripherally inserted central catheters (PICCs), are also cause for significant concern in the medical community.
Device-related infection results from the introduction of organisms, primarily bacteria, during the device insertion or implantation procedure, or from attachment of bloodborne organisms to the newly inserted device and their subsequent propagation on its surface. Good clinical practicesuch as thoroughly cleaning and disinfecting the area prior to insertion, proper prepping by the clinical staff, and care in handling the device to maintain sterility prior to insertionwill reduce but not eliminate the occurrence of infection. Infection also can occur postprocedurally, either from bacteria in the blood or urine attaching to the device or, in the case of externally communicating devices, from bacteria that use the device as a pathway into the body, in some cases long after the device has been inserted.
BACTERIA AND BIOFILM FORMATION
Most bacteria that cause device-related infections enter the body when the device is initially placed. The organisms first attach to the device surface through the secretion of polymers (polysaccharides) or the extension of fibrils, which anchor the bacteria to the surface. After attachment, cell division of the bacteria produces sister cells that form microcolonies and create a protective barrier commonly known as biofilm or bioslime (see Figure 1). Once this barrier is formed, the bacteria can propagate within the biofilm and release substantial amounts of bacterial cells into the surrounding fluids and tissues. The infections that ensue can be difficult to treat, because the body's macrophages and antibiotics are unable to reach the primary source of the infecting bacteria (see Figure 2). Often, effective treatment requires removal of the offending device.4
Figure 1. Biofilm formation on device surface.
Figure 2. Biofilm as a reservoir and shelter for bacteria.
BENEFITS OF DRUGS ON DEVICE SURFACES
In recent years, there have been numerous efforts to sequester antimicrobials and antibiotics on the surface of or within devices that are then placed in the vasculature or urinary tract as a means of reducing the incidence of device-related infections. In this article, we will review the basic approaches and some recent results in this area. Our focus is on the addition of drugs to the device to counteract conditions induced by the placement or presence of the device in the body. It must be remembered that the presence of active antiinfective agents in or on the device is secondary to the device's primary therapeutic or diagnostic function. The use of devices as a means of drug delivery to treat preexisting conditions is therefore outside the scope of this article.
Locating active agents or drugs only at the surface of or in the vicinity of the device to reduce the incidence of device-related infections and thromboses is often preferable to administering the same drugs systemically. Systemic administration requires maintaining dose levels throughout the body, whereas local administration from the device surface concentrates the drug at the precise site where it is needed.
Avoiding systemic administration of high levels of drug is especially important for patients with certain conditions. For example, diabetes, HIV disease, kidney or liver dysfunction, poor circulation, pregnancy, and advanced age are conditions for which systemic administration is not recommended unless necessary. The increasing concerns about bacterial resistance are another argument for surface as opposed to systemic administration of antibiotics. Finally, the active agent or drug can be expected to be more effective when administered directly at the site.
In order for local administration to be effective, however, there must be sufficient amounts of the agent released from the device, and the duration of release must be appropriate for the condition. If there is good elution of drug from the device, drug concentration will be high at and near its surface, but will diminish with distance (see Figure 3). This is the desired effecthigh concentration near the device surface and very low systemic concentrationand is easily seen in relatively static environments, such as those of urinary catheters. In more dynamic physiological environments like the circulatory system, the laws of fluid mechanics can bring about the same effect: the flow of blood in a vessel may be quite high but will slow at the approach to the surface of an implanted device. At the surface itself, there is practically no flow at all. The net result is that the drug eluting from the device will be highly concentrated at and near the surface.5
Figure 3. Antiinfective coatings can provide a high concentration of drug near the device surface and low systemic concentration.5
SELECTING A SURFACE-TREATMENT METHOD FOR DRUG ATTACHMENT
To be effective, device-based drugs must be available at and near the surface in sufficiently high concentration to preclude bacterial attachment and propagation. In other words, the device surface must serve as a reservoir for a large amount of drug and be capable of releasing it over time in appropriate quantities.
It is also critical that the drug remain potent after sterilization. For this reason, devices incorporating heat-, radiation-, or ethylene oxidesensitive antibiotics need to be tested carefully for efficacy after sterilization. Some methods of sequestering drugs on devices have been shown to protect antibiotics from the deleterious effects of sterilization.
For how many hours or days must high concentrations of drug be released from the device? How long should the drug or agent elute at effective concentrations? To answer these questions, one must consider both short- and long-term effects. There is a period of increased susceptibility to infection that occurs during and immediately following the insertion of a device. If the device is to be in place only during a procedure or perhaps for a few days, then the drug only needs to be available for a few hours or a day or two at most.
The longer the device is in place, the greater the probability of infection. Two longer-term routes of infection are possible: the attachment of already internally present bacteria (existing independent of the device-based procedure) onto the surface of the device, and transmission of bacteria from external sources via externally communicating devices such as central venous catheters, PICC lines, midline catheters, drainage tubes, and urinary catheters, among others. In the latter case, the proximal end of the device allows for introduction of bacteria, and the device itself acts as a pathway for their migration into the body.
The occurrence of device-centered infections when there is no external exposure together with the high incidence of infection with urinary catheters and with externally communicating vascular and drainage devices suggests that both of these longer-term routes of infection are commonly at work. Thus, the ideal antiinfective device would be one that elutes an antimicrobial or antibiotic for the duration of its placement or, failing that, for as long as possible. In addition, in the case of catheters, both the interior and exterior surfaces can be pathways for infection. Therefore, providing drug on both inside lumens and exterior surfaces may further reduce catheter-related infections.
A number of important characteristics must be taken into account in assessing the choice of an antiinfective surface treatment, among them
BiocompatibilityThe full complement of biocompatibility tests should be considered for all devices that contact body fluids and tissues. (For general testing requirements, see the ISO 10993/EN 30993 standard and the FDA Blue Book Memorandum G95-1.)
Drug AvailabilityThe amount of drug available is obviously critical. Any surface-modification system that cannot provide drug in sufficient quantities over the needed time period allows for unnecessary exposure to infection.
AdhesionThe selected surface treatment cannot shed or peel. Loss of large particles from the surface could create emboli or distribute the drug to nontargeted areas of the body.
DurabilityThe surface treatment must be able to withstand the rigors of the insertion process and any subsequent device manipulation after placement.
FlexibilityAny surface treatment that measurably adds to the diameter of the device can be expected to add some stiffness. Minimizing this added stiffness can be crucial for devices such as small-diameter catheters and guidewires that rely on very flexible tips to minimize the risk of perforation.
CoverageThe selected treatment should entirely cover whatever surfaces of the device are exposed to body fluids, so as to reduce the risk of exposure to bacteria.
SterilizabilityThe device must be presented sterile. For commercial products, this means that it must be packaged and sterilized without diminishing the efficacy of the antibiotic, antimicrobial, or antithrombogenic agent.
StabilityThe surface treatment and drug must remain stable under normal storage and use conditions and must have a reasonable shelf life. Radiation sterilization and some types of surface treatmentsfor example, exposure to UVmay cause cross-linking. In many polymers, cross-linking reactions will continue even after the exposure to radiation or UV has been terminated. Products that rely on cross-linking as a surface treatment or that are radiation sterilized should be tested for this continuation of the cross-linking process, which can cause embrittlement.
Ease of UseTo be clinically viable, the treated device must be relatively easy to use. This presents a drawback for devices to which the drug must be added during the clinical procedure.
CostAn obvious consideration in all product development decisions. Do the benefits generated justify the costs?
METHODS OF DRUG ATTACHMENT AND ENTRAPMENT
For the most part, the development of technologies for placing drugs in and on the surfaces of devices can be viewed as a progression from early studies in surface adsorption and covalent attachment through incorporation of the drug into the device material and ultimately to the attachment or entrapment of drugs in surface coatings.5 Surface coatings that contain drugs or drug/polymer layers represent a significant advance because they permit greater control of drug release and allow larger amounts of drug to be available at the surface, at the same time providing additional properties such as lubricity and resistance to adhesion.
Much of the early work in the field focused on surface adsorption. The simplest surface-adsorption technique is the immersion of the device in a solution of the drug. This approach is limited by the short time the drug remains on the surface of the device: because it is not bound to the surface or sequestered in any way, it washes away from the surface very quickly, generally in less than an hour. In addition, only a thin film is deposited on the surface, typically yielding, at best, only moderate release levels of drug. Better results from adsorption techniques have been achieved by using drugs with limited solubility. For example, the silver salts of anionic antibiotics are less soluble than the sodium salts, and devices coated with the former can exhibit effects lasting from days to weeks.68
Most of the adsorption methods that have been developed require that the clinician soak the device in a drug solution or suspension just prior to the procedure. This can be a limiting factor because of the increased potential for contamination and the additional steps required for clinical staff to prepare the device and the solution/suspension. Alternatively, it is possible for a manufacturer to prepare the device with a surface-adsorbed drug, and then dry, package, and sterilize it for later use. However, because the drug is not bound or encapsulated, sterilization is problematic, and commercial processes may degrade or erode the drug. The most significant drawbackwhether the device is prepared just prior to use or prepackaged and sterilizedis the loss of activity that normally occurs after only a few hours or, at most, a few days.
It has long been recognized that many antibiotics have negative charges analogous to that of heparin.9 This finding has led to a method of binding antibiotic molecules to the surface of prosthetic materials through the adsorption of positively charged surfactantssuch as benzalkonium or tridodecylmethylammonium chloride (TDMAC).10,11 The bound surfactant acts as an anchor for subsequent binding of negatively charged antibiotic molecules, which include, for example, the penicillin and cephalosporin families of drugs. The pharmacological agents are not irreversibly bound to the prosthesis, however, and after exposure to blood or body fluid are slowly released, resulting in a local environment of high drug concentration at the surface of the prosthesis, far in excess of what could be achieved by systemic administration. This high concentration of antibiotic causes localized inhibition of bacterial growth.5
If the antimicrobial material disappears from the surface in a matter of hours or days, the device once again becomes a potential site for the attachment and propagation of bacteria. In response to the rapid loss of activity of surface-adsorbed drugs, there has been some interesting work centered on covalent attachment.
Covalent attachment can keep a drug present at the surface for as long as the device is in place. There are, however, some restrictions to this technique. The choice of drugs that can be covalently bound is limited, and the covalent attachment mechanism often occupies the therapeutic sites on the drug, rendering it ineffective. The technique is therefore limited to drugs that can be covalently attached or modified for covalent attachment without losing their therapeutic qualities. Because the amount of available drug is limited and there is no release of drug from the device into the adjacent fluid or tissue, the antimicrobial properties of covalently attached agents are limited to the contact areas of the device. Complete coverage and tight packing of the drug become critical, since any gaps in coverage, even very small ones, will be vulnerable to biofilm formation.
A third method of making a drug available at the device surface is incorporating it into the polymer from which the device is fabricated. The concept is that the device substrate can be a reservoir that allows the drug to elute, providing antimicrobial activity at and adjacent to the surface. A significant amount of drug can be entrapped within the device substrate by compounding the agent into the plastic prior to injection molding or extrusion, in the same manner that pigments, stabilizers, and strengtheners are added to the resin.
There have been reports of good experimental results using this technique, with antimicrobial activity demonstrated up to 3 or 4 weeks.5 The major drawback is the extensive R&D effort necessary to determine commercial viability and the potential cost of the final product. Significant experimental work is required to qualify polymers appropriate for devicesresins whose molding integrity is not compromised by the addition of the drug. One must also determine appropriate drug-plastic combinations that will allow for controlled release at a sufficient level over the requisite time period. Because of the iterative nature of the testing and the complexity of the molding setups, the time and costs required to achieve adequate results may be daunting.
The fourth and final approach involves sequestering drugs into device coatings. This approach avoids some of the problems encountered when trying to compound the drug into the bulk polymer: the surface treatments can be applied without changing the basic properties of the device, and a sufficient quantity of drug can be incorporated. The several commercially available systems are generally prepared by one of two methods: surface treating devices by cross-linking polymers that contain drugs, or coating devices with polymer solutions that contain antimicrobial agents.
An example of the first approach uses exposure of prepolymer derivatives to UV light to form a polymer that is covalently bonded to the device surface. Drugs can be incorporated into this polymer layer either before or after the polymerization is carried out.
A prominent example of the second approach involves the use of nonreactive hydrophilic-hydrophobic polymer matrix coatings that contain active agents and are bonded to the device surface. The hydrophilic component of the mix hydrates very quickly and enables the drug to diffuse to the surface for elution to the adjacent body fluids. By altering the proportions of the constituent polymers, fabricators can vary the thickness of the coatings, the amount of drug sequestered, and the rate of elution. Applicable to any of the metals and polymers commonly used for device construction, the coatings have been demonstrated to elute drug in active quantities over periods from a few hours to 10 or more weeks. Their nonreactive nature helps avoid problems of inadequate drug diffusion that can occur in systems in which the drug is tightly bound by reactive coatings or ionic/covalent bonding.
Figures 4 and 5 show the results of zone-of-inhibition testing for several coatings based on this nonreactive technology. In this test, segments of the coated device are extracted in serum or urine and then incubated in agar that has been seeded with a bacterium. If the material effectively elutes its antimicrobial or antibiotic, there will be a zone around the segment in which the bacterial growth has been inhibited. The larger the zone of inhibition, the greater the effective elution. By repeating the test at daily intervals, one can determine how long the coated device will elute effective amounts of the drug. Zone-of-inhibition testing has been shown to be a good surrogate for in vivo results.5
Figure 4. Results of zone-of-inhibition testing of antiinfective coating (STS Biopolymers MEDI-COAT, active agent: Norfloxin; organism: E. coli).
Figure 5. Results of zone-of-inhibition testing of antiinfective coating (STS Biopolymers MEDI-COAT, active agents: Rifamacin and BKC; organism: S. epidermidis).
The surfaces of indwelling medical devices are an excellent platform for the formation of life-threatening infections. Although aseptic techniques can reduce the incidence of these infections, a significant risk remains. Incorporation of antimicrobials in the bulk material that constitutes a device can be effective but costly, and various surface treatments are emerging as important and less-expensive alternatives. Early efforts that concentrated on adsorption of antibiotics to device surfaces achieved limited results. Sequestering drugs more permanently at the device surface is preferable as long as the drug is not too tightly bound to the surface. One promising approach features nonreactive hydrophilic-hydrophobic polymer matrices that entrap but do not bind the drug, allowing for immediate and extended elution.
We wish to thank Richard Harvey, PhD, of the Robert Wood Johnson Medical School, University of Medicine and Dentistry of New Jersey, for his generous assistance with this and other projects.
1. "Infection Control and Biosafety: Trends, Products and Opportunities," Medical Data International report, Irvine, CA, MDI, 1:3344, 1993.
2. Plott R, Polk B, Murdock B, et al., "Mortality Associated with Nosocomial Urinary-Tract Infection," New Engl J Med, 30(11):637642, 1982.
3. Adal KR, and Farr BM, "Central Venous CatheterRelated Infections: A Review," Nutrition, 12:208213, 1996.
4. Bisno AL, and Waldvogel FA (eds), Infections Associated with Indwelling Medical Devices, Washington, DC, American Society of Microbiology, p 161, 1989.
5. Harvey R, "Utilizing Prostheses for Drug Delivery," in Implantation Biology, Greco RS (ed), Boca Raton, FL, CRC Press, ch 19, 1994.
6. White JB, Benvenisty AI, Reemtsma K, et al., "Simple Methods for Direct Antibiotic Protection of Synthetic Vascular Grafts," J Vasc Surg, 1:372, 1984.
7. Modak SM, Sampath L, Fox CL, et al., "A New Method for the Direct Incorporation of Antibiotic in Prosthetic Vascular Grafts," Surg Gynecol Obstet, 164:143, 1987.
8. Benvenisty AI, Tannenbaum G, Ahlborn TN, et al., "Control of Prosthetic Bacterial Infection: Evaluation of an Easily Incorporated, Tightly Bound Silver Antibiotic PTFE Graft," J Surg Res, 44:1, 1988.
9. Jagpal R, and Greco RS, "Studies of a Graphite-Benzalkonium-Oxacillin Surface," Am Surg, 45:774, 1979.
10. Greco RS, Harvey RA, Henry R, et al., "Prevention of Graft Infection by Antibiotic Bonding," Surg Forum, XXXI: 29, 1980.
11. Kamal GD, Pfaller MS, Rempe LE, et al., "Reduced Intravascular Catheter Infection by Antibiotic Bonding," JAMA, 265(18):2364, 1991.
Xianping Zhang, PhD, is a research chemist at STS Biopolymers, Inc. (Henrietta, NY), where she specializes in antimicrobial and antibiotic coatings for medical devices.
Richard Whitbourne is chairman and Richard D. Richmond is president of STS Biopolymers, which develops and manufactures biocompatible, lubricious, antimicrobial, antibiotic, antithrombogenic, echogenic, and other specialty coatings for medical device manufacturers.