 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
The Journal of Bone and Joint Surgery (American). 2007;89:871-882. doi:10.2106/JBJS.E.01070 © 2007 The Journal of Bone and Joint Surgery, Inc.
Antibiotic-Impregnated Cement Spacers for the Treatment of Infection Associated with Total Hip or Knee Arthroplasty
1 Department of Orthopaedic Surgery, University of Virginia, Box 800159,
Charlottesville, VA 22908
Disclosure: In support of their research for or preparation of this
work, one or more of the authors received, in any one year, fellowship funding
in excess of $10,000 from Smith and Nephew Inc. In addition, one or more of
the authors or a member of his or her immediate family received, in any one
year, payments or other benefits in excess of $10,000 or a commitment or
agreement to provide such benefits from a commercial entity (Stryker
Orthopaedics). Also, a commercial entity (Smith and Nephew Inc.) paid or
directed in any one year, or agreed to pay or direct, benefits in excess of
$10,000 to a research fund, foundation, division, center, clinical practice,
or other charitable or nonprofit organization with which the authors, or a
member of their immediate families, are affiliated or associated.
 Infection at the site
of a total joint arthroplasty can be classified into four basic categories:
Type I (early postoperative), Type II (late chronic), Type III (acute
hematogenous), and Type IV (positive intraoperative cultures with clinically
unapparent infection).
Total joint replacement is one of the most frequent and successful types of operations in orthopaedics. Infection is a rare yet devastating complication of the procedure, with a reported prevalence of 0.5% to 3% and with a higher reported prevalence after total knee arthroplasty than after total hip arthroplasty1-4. There is also a higher rate of infection after revision hip and knee arthroplasties than after primary hip and knee arthroplasties1-8. Two-stage revision surgery was first described in 1983 by Insall et al., who demonstrated the necessity of removing the implants as well as the cement and of introducing antibiotic therapy for definitive treatment9. This procedure has emerged as the standard of care for a late chronic infection at the site of a total joint replacement4,5,10-17. Garvin and Hanssen reviewed twenty-nine studies and found that two-stage procedures without antibiotic-loaded cement had a better success rate (82% of 158 joints) than one-stage exchange arthroplasties (58% of sixty joints), although systemic antibiotics were used for both procedures18. With the addition of antibiotic cement, the rates of successful eradication of the infection increased to 91% (385 of 423 joints) for the two-stage technique and 82% (976 of 1189 joints) for the one-stage revision18. Two-stage revision arthroplasty without the use of spacers allows complete removal of foreign materials, with later reimplantation after eradication of the infection. However, this procedure has several disadvantages as soft-tissue contractures and joint instability may develop and the patient will have difficulty with mobility. From a technical perspective, the disadvantage of the procedure is that it makes reimplantation during the second-stage operation more difficult as a result of arthrofibrosis and the loss of tissue planes6,9,12. Compared with revision total hip arthroplasty for treatment of aseptic loosening and compared with primary total hip arthroplasty, two-stage revision for the treatment of infection is associated with a lengthier hospital stay, an increased number of hospitalizations, and increased perioperative morbidity19-21. Use of antibiotic-impregnated polymethylmethacrylate bone-cement spacers is now considered to be the standard of care for patients with a chronic infection at the site of a total joint replacement. These spacers provide direct local delivery of antibiotics while preserving patient mobility and facilitating reimplantation surgery1-11. This operative treatment decreases cost and improves patient outcomes as well as addresses some of the disadvantages of two-stage revision procedures in which spacers are not used1-12,14-17,19-42. We will systematically review the various types of spacers and their uses in two-stage revision arthroplasty for treatment of infection at the site of a total joint replacement.
Infection at the site of a total joint arthroplasty can be classified into four basic categories: Type I (early postoperative), Type II (late chronic), Type III (acute hematogenous), and Type IV (positive intraoperative cultures with clinically unapparent infection) (Table I)11,16,43-47. Specific operative modalities are recommended for eradication of each type of infection, although these recommendations are not universally followed.
Early postoperative infections (Type I), both superficial and deep, are defined as wound infections that occur less than four weeks after the primary operation. Superficial Type-I infections are typically treated with débridement and a course of antibiotic therapy, and deep Type-I infections are usually treated with replacement of the polyethylene insert, retention of the metal prosthetic components, and intravenous administration of antibiotics11,22,46,47. Occasionally, antibiotic-loaded polymethylmethacrylate beads are also inserted2,11,47. Late chronic infections (Type II) are defined by their occurrence more than four weeks after the operation. They typically present with worsening pain and loosening of the prosthesis and are usually treated with a two-stage reconstruction. Treatment includes removal of all prosthetic components and bone cement, débridement of necrotic and granulation tissue, placement of an antibiotic-impregnated cement spacer, administration of a course of intravenous antibiotics, and delayed reimplantation arthroplasty when there is no longer evidence of infection11,12,16,22,23,28,44. It is important to note that one-stage exchange arthroplasty also has been used, more commonly in Europe than in the United States, but strict patient selection and use of antibiotic-loaded cement for fixation of the prosthesis are strongly recommended1,2,48,49. Acute hematogenous infections (Type III) are defined by bacteremia and are typically managed with débridement, replacement of the polyethylene insert, and retention of the prosthesis if there is no implant loosening, followed by a course of intravenous antibiotics11,43-47. Patients with positive intraoperative cultures (Type IV) within days after the performance of a revision arthroplasty for the treatment of aseptic loosening are typically managed with a course of intravenous antibiotics with retention of the prosthesis11,16,43-47.
There are two types of antibiotic-impregnated cement spacers that are typically used in two-stage revisions of total hip and knee arthroplasties: nonarticulating (block or static; Figs. 1-A and 1-B) and articulating (mobile; Figs. 2-A, 2-B and 2-C, 3).
Nonarticulating spacers allow local delivery of a high concentration of antibiotics and at the same time function to maintain joint space for future revision procedures. Their disadvantages include a limited range of motion of the joint after the operation, resulting in quadriceps or abductor shortening, scar formation, and bone loss. Cohen et al.50 and Wilde and Ruth51 reported the use of nonarticulating spacers that consisted of polymethylmethacrylate cement mixed with antibiotics and were shaped to fit the defect that had been left after the removal of a total joint prosthesis associated with infection. In contrast, articulating spacers permit more joint motion and can improve function prior to the second-stage reimplantation. From a technical perspective, improved joint function and decreased scar formation after resection arthroplasty can facilitate exposure during reimplantation4,17,52. Although the distinction between articulating and non-articulating spacers is somewhat controversial, use of a well-molded, well-fitted articulating spacer that restores soft-tissue tension and allows a greater degree of joint motion has been reported to have a better outcome than use of a nonarticulating spacer, which may limit joint freedom4,5,26. As stated earlier, antibiotic-impregnated cement spacers can maintain limb length, minimize soft-tissue contracture, facilitate reimplantation, and provide local antibiotic therapy. However, there is considerable variation in their form and function12,16,17,23,33,53. A spacer may be commercially made, or it may be custom-made in the operating room. It may be made entirely of polymethylmethacrylate cement, or it may be a cement-coated metal composite or a sterile prosthesis partially coated with antibiotic-impregnated cement. Favorable results have been reported with each of these types of spacers2-8,10,17,27,29,33,36,38,42,52,54,55.
The first nonarticulating knee spacers (Figs. 1-A and 1-B) were made with antibiotic-loaded cement in the operating room and were formed to fill the bone defect left after removal of a prosthesis associated with infection. Use of these types of spacers has yielded inferior results, in terms of postoperative range of motion and pain, compared with those following single-stage aseptic revisions performed because of aseptic loosening6,7,26,28,56. A number of different types of articulating spacers can be employed in the two-stage revision of a total knee arthroplasty that was complicated by infection. One commercially available system is the prosthesis of antibiotic-loaded acrylic cement (PROSTALAC) (DePuy, Warsaw, Indiana), which was originally developed for the hip and was subsequently used in the knee6,57. The PROSTALAC knee system includes femoral and tibial components made of antibiotic-loaded Palacos bone cement with a small metal-on-polyethylene articular surface (Fig. 3). In a study of forty-five patients followed for an average of forty-eight months (range, twenty to 112 months) after treatment with the PROSTALAC prosthesis, Haddad et al.6 reported eradication of the infection in forty-one patients (91%) despite the use of the metal and polyethylene components during the first stage of the two-stage procedure. The PROSTALAC system provided an increased range of motion, minimized pain, improved function, and facilitated the second-stage procedure by maintaining soft-tissue planes6,17,41. The cost of a knee revision due to infection is twice that of an aseptic revision and three to four times that of a primary total knee replacement, with most of the increased cost due to the prolonged and repeated hospitalization6. While articulating spacers reduce the amount of time spent in the hospital, the cost of the PROSTALAC system, which requires an entire system of molds and instrumentation, may prohibit its use. Also, the PROSTALAC system offers a limited choice of sizes, and development of cheaper and more customizable alternatives may be beneficial. Hofmann et al.12 used high-dose antibiotic-impregnated polymethylmethacrylate bone cement (4.8 g of tobramycin per 40 g of cement) in an articulating spacer in fifty patients during the first stage of a two-staged procedure. The spacer was created intraoperatively by cleaning, autoclaving, and reinserting the removed femoral component, which articulated with a new thin tibial polyethylene insert with a large amount of antibiotic-loaded cement placed between the tibial polyethylene and the bone; occasionally, a new patellar polyethylene insert was used as well. Care was taken to apply the cement early to the components and late to the bone to allow molding to the defects without adherence and interdigitation. At an average of seventy-three months (range, twenty-four to 150 months) after the reimplantation procedure (the second stage), the rate of good to excellent results was 90%, with only six of the fifty patients having a reinfection. Emerson et al.26 achieved similar results using the model described by Hofmann et al. They compared twenty-six patients treated with a nonarticulating spacer (average duration of follow-up, 7.5 years) with twenty-two treated with an articulating spacer (average duration of follow-up, 3.8 years) and found a 9% reinfection rate in the patients with the articulating spacer compared with an 8% rate in those with a nonarticulating spacer. However, the range of motion of the joints with an articulating spacer was an average of 14° greater.
As is the case for total knee arthroplasties complicated by infection, two-stage revision arthroplasty is the standard of care when total hip arthroplasties are complicated by infection4,5,11,14,27,33,38,42. A PROSTALAC hip system is also commercially available. It consists of a snap-fit all-polyethylene, loosely cemented acetabular component with a metal endoskeleton, a femoral head, and a centralizer that are inserted into a mold and filled with antibiotic-impregnated cement to create the implant (Fig. 4)17,34,41,58,59.
Wentworth et al.41 reported success in 83% of 116 patients treated with the PROSTALAC prosthesis. Successful treatment was defined as no growth of a microorganism on culture of any specimen obtained from the operative site at the time of the second-stage operation. Younger et al.8 had better results with the PROSTALAC system, with the infection eradicated in 94% (forty-five) of forty-eight patients at an average of forty-three months (range, twenty-four to sixty-three months) postoperatively. The disadvantages of this system include cost, the need for special equipment, and the limited choices with regard to the sizes of the components. Etienne et al.27 described a method to construct a prosthesis similar to the PROSTALAC system but at a lower cost. They used a spacer consisting of either the removed, autoclaved femoral component or an inexpensive modular femoral component coated with a mantle of antibiotic-loaded cement. A polyethylene acetabular liner was cemented in place with the same antibiotic-impregnated cement, so that the system temporarily functioned like a conventional total hip prosthesis. Etienne et al. reported a reinfection in only three of thirty-two patients at a mean of 1.7 years (range, one to three years) postoperatively, and they noted quick component assembly, low cost, a good range of motion, and the possibility for partial weight-bearing, which facilitated revision surgery. Specially designed reusable silicone or metal molds have been fabricated as temporary articulating spacer endoprostheses, with a metal endoskeleton for support4,10,42. Durbhakula et al.4 used a Rush pin as the endoskeleton and reported no reinfections in twenty patients followed for an average of thirty-eight months (range, twenty-six to sixty-seven months); eighteen patients had a successful two-stage revision. Yamamoto et al.42 used two 2.0-mm bent Kirschner wires as the endoskeleton and also reported good results, with eradication of the infection in all of seventeen patients followed for a mean of three years and two months (range, fourteen to sixty-two months); however, there was one periprosthetic fracture and one dislocation.
Implantation of antibiotic-loaded polymethylmethacrylate bone cement has become a useful technique in the treatment of infections at the sites of total joint arthroplasties (see Appendix). Spacers formed with use of antibiotic-loaded cement deliver high doses of antibiotics at the site of the infection and can achieve local concentrations higher than those achieved with systemic antibiotics alone, with little effect on serum or urine levels12,34,39,60,61. Because they achieve high concentrations of antibiotics at the site of an infection, spacers can be used to treat infected avascular bone that is isolated from systemic antibiotics while avoiding the potential systemic toxicity that can result from intravenous use2,11,16,39,43,61.
Choice of Antibiotics
Fungal infections are extremely rare at the sites of total joint arthroplasties and are difficult to treat. In most reported cases, the offending organism has been a member of the Candida species28,63,64. An in vitro analysis of antifungal-impregnated polymethylmethacrylate bone cement showed that amphotericin B and fluconazole remained active with zones of inhibition, while 5-flucytosine did not65. Fungal infection at the site of a total joint arthroplasty has been treated successfully with antibiotic-impregnated polymethylmethacrylate bone-cement spacers and staged reimplantation28,63,66,67.
Factors Affecting Elution of Antibiotics from Polymethylmethacrylate Spacers Compared with commercially available antibiotic-loaded cement, hand-mixed cement is associated with a decreased release of antibiotics, whereas vacuum-mixing has been shown to result in only a minor reduction in antibiotic release70-74. Unlike antibiotics that are commercially mixed in cement, hand-mixed antibiotics do not have a homogeneous distribution in the cement, which decreases their rate of elution from a given surface area70,71,73. Vacuum-mixing decreases the porosity of the cement, which also decreases the rate of elution of the antibiotics71. The elution of antibiotics from polymethylmethacrylate bone cement is determined by a combination of surface area and porosity60,71,73. One study showed that increasing the surface area of polymethylmethacrylate bone cement by 40% resulted in a 20% higher rate of elution of vancomycin68. Dextran has been added to cement to enhance porosity and increase antibiotic elution rates. Kuechle et al. found that the addition of 25% dextran to cement increased the release of antibiotics in the first forty-eight hours approximately four times compared with that associated with routine preparation and increased the duration of the elution to up to ten days compared with only six days with routine preparation75. For decades, hand-made polymethylmethacrylate bone-cement spacers containing high doses of antibiotics have been used successfully in the treatment of established infections at the sites of prosthetic joints, and they can release high levels of antibiotics locally1-11. It is important to note that the Food and Drug Administration has approved commercial production of only low-dose antibiotic-loaded bone cements. These include Simplex P (Stryker Howmedica Osteonics, Mahwah, New Jersey), which contains 1 g of tobramycin; SmartSet GHV and MHV (DePuy Orthopaedics, Warsaw, Indiana), which contain 1 g of gentamicin; and Palacos G (Biomet, Warsaw, Indiana), which contains 0.85 g of gentamicin28,60,61. These low-dose antibiotic-loaded cements are not suitable for treatment of established infections at the sites of prosthetic joints, although they may be used prophylactically for high-risk patients undergoing a total knee or hip arthroplasty with cement or in the second stage of a two-stage revision total joint arthroplasty after the initial infection has been eradicated11,31,60,61. Therefore, surgeons need to add antibiotics to the cement in order to achieve the high doses suggested for the treatment of periprosthetic joint infection2,18,34,61.
Method for Mixing Antibiotic-Impregnated Cement
Changes in Biomechanical Properties
Antibiotic Doses for Polymethylmethacrylate Spacers In one study, Fehring et al.22 observed effective results with the use of 1.2 g of tobramycin per 40 g of bone cement alone. The average duration of follow-up was thirty-six months (range, twenty-four to seventy-two months) for the patients who received a static spacer and twenty-seven months (range, twenty-four to thirty-six months) for those treated with an articulating spacer. Three patients who had received a static spacer had a reinfection, and the final infection eradication rate was 88% (twenty-two of twenty-five). One patient who had received an articulating spacer had persistent drainage after the implant was removed and required an arthrodesis, leading to an infection eradication rate of 93% (fourteen of fifteen) in this group. As stated previously, fungal infection at the site of total joint replacement represents a very difficult challenge. However, Evans successfully treated fungal infections with polymethylmethacrylate bone-cement spacers containing 500 mg of amphotericin B followed by second-stage reimplantation in six patients28. Similar techniques have been used by other authors66,67. Phelan et al.63 used staged reimplantation arthroplasty and systemic administration of antifungal agents to treat four Candida infections at the sites of total joint arthroplasties. They also identified an additional six cases in the literature that had been treated with the same regimen. In addition to resection arthroplasty, eight patients received amphotericin B, either alone or in combination with other antifungal therapy, and one patient received fluconazole alone. Eight of the patients did not have a recurrence of the infection at a median of 50.7 months (range, two to seventy-three months) following reimplantation arthroplasty.
Safety Issues
The use of antibiotic-loaded polymethylmethacrylate bone cement in spacers offers not only a more effective treatment for periprosthetic infection, with eradication rates in the literature ranging from 90% to 100%, but also improved function, decreased pain, increased patient satisfaction, shorter hospital stays, and decreased cost4-8,12,16,17,28,31,61,85.
Outcome Studies of Knee Cement Spacers
In a retrospective review, Calton et al.56 found scarring and capsular contractions with bone loss in 60% of twenty-four patients in whom an infection at the site of a total knee replacement had been treated with a nonarticulating spacer. On the average, 6.2 mm of bone loss was noted in the tibia, while 12.8 mm was noted in the femur, frequently with invagination and migration of the spacer. The authors recommended intramedullary extension of the spacer to prevent migration, adequate thickness of the spacer to tense the collateral ligaments to prevent contracture, and a wide-enough block to rest on the cortical rim and prevent invagination into the cancellous bone. There was no difference in the infection rate, operating time, or functional outcome between the patients treated with the nonarticulating spacer and those treated with an articulating spacer. A study of twenty-five static and thirty mobile spacers demonstrated that the articulating spacers facilitated reimplantation and were not associated with bone loss22. Emerson et al.26 reported a greater range of motion with articulating spacers than with static knee spacers, with knee flexion averaging 107.8° and 93.7°, respectively, and no evidence of higher complication rates. Durbhakula et al.7 reported on twenty-four patients treated with a two-stage revision involving use of an antibiotic-loaded articulating cement spacer formed with vacuum-injected silicone molds that had been designed to fabricate articulating femoral and tibial components. With such a system, there is no need for a metal-on-polyethylene articulating surface and the cost is reduced by the employment of reusable molds that cost approximately $300 each. Durbhakula et al. reported no problems with dislocation, fracture, or fragmentation of the spacer, and the infection eradication rate was 92% at an average of thirty-three months. Goldstein et al.29 designed a low-cost, all-cement system that could be formed with instruments and supplies that are available at most hospitals. They described a technique in which heavy aluminum foil was used to form the osseous anatomy, hand-molded cement was applied around the foil to prevent interdigitation, and a layer of sterile lubricant was used in between for easy removal of the foil. The femoral condyles were molded with the trial tibial insert, and the tibial insert was used to approximate the size and thickness of the cemented tibial component. The authors reported early success in five patients.
MacAvoy and Ries52 described an inexpensive molding method to fabricate a "ball-and-socket" articulating spacer. The relatively constrained articulation between the spacer components may be particularly useful for patients with large amounts of bone loss and instability (Figs. 5-A, 5-B, and 5-C). The technique was used in twelve patients who had an infection at the site of a total knee arthroplasty complicated by severe comorbidities. At an average of twenty-eight months postoperatively, the infection was eradicated in nine of thirteen knees. All patients were able to walk with the spacer in place with minimal assistance. The average knee flexion was 98° (range, 45° to 135°) at the time of follow-up. With the spacer in place at the time of reimplantation, the average range of motion of the knee was 79° (range, 40° to 100°).
Outcome Studies of Hip Cement Spacers In order to fill bone defects resulting from the removal of components associated with infection, Leunig et al.33 created customized, inexpensive, hand-molded implants with gentamicin-loaded cement and placed them in the area of the femoral neck or medullary canal. The implants were reinforced by inserting plates and/or screws into the cement before polymerization. While the authors reported complete eradication of infection in twelve patients followed for an average of 2.2 years, they also reported five dislocations and one fracture, suggesting a weakness in the design. Barrack5 used a Rush pin to reinforce hand-molded, antibiotic-loaded cement to make a temporary prosthesis. Twelve patients so treated had no fractures or dislocations and were free of infection at a minimum of two years postoperatively. The advantage of using Rush pins is that they are available in numerous lengths and diameters, which allows the surgeon to produce hand-made prostheses with a wide range of lengths and offsets. This technique is a cost-effective alternative to the use of a commercially available hip spacer.
An infection at the site of a total hip or knee arthroplasty is a challenging clinical problem. The gold standard of treatment for late chronic infections is two-stage revision arthroplasty, which includes placement of an antibiotic-impregnated cement spacer after removal of the prosthesis and thorough débridement, followed by a course of intravenous antibiotics and delayed second-stage revision total joint arthroplasty. The choice of the spacer is based on many factors, including the amount of bone loss, the condition of the soft tissues, the need for joint motion, the availability of prefabricated spacers or molding methods, and the selection of the antibiotics. Thermostable, water-soluble, susceptibility-directed antibiotics should be used. Hand-mixing of additional antibiotics into antibiotic-impregnated bone cement is preferred to increase the antibiotic dosage. The cement should be mixed first, and the antibiotics should then be added. Articulating spacers should be the first option since they appear to provide a better functional outcome.
Tables summarizing the results of the use of antibiotic-loaded cement spacers in the hip and knee as reported in the literature are available with the electronic versions of this article, on our web site at jbjs.org (go to the article citation and click on "Supplementary Material") and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM). 
|
Top
Introduction
Classification of Infection at...
