Two-Stage Revision




Articulating Spacer in Two-Stage Revisions



Jacob T. Munro, MBChB, FRACS
Bassam A. Masri, MD, FRCSC
Donald S. Garbuz, MD, MHSc, FRCSC
Nelson V. Greidanus, MD, MPH, FRCSC



CASE STUDY


A 65-year-old man presented to our office 2 years after a right total knee arthroplasty (TKA). He has had increasing right knee pain and swelling during the past 2 months, particularly in the past week. His primary procedure was complicated by a superficial wound infection, which was treated with oral antibiotics. He has hypertension and is a type 2 diabetic who uses oral hypoglycemic drugs for control.


Examination confirmed a limp and a temperature of 37.5° C. His knee was warm and had a large effusion, but there was no deformity. The midline wound had healed. A complete blood cell count found a white blood cell (WBC) count of 14,000 cells/μL with a C-reactive protein (CRP) level of 120 mg/L and erythrocyte sedimentation rate (ESR) of 80 mm/hr. Anteroposterior and lateral radiographs of the knee showed progressive osteolysis around both components compared with views obtained 3 months postoperatively. Analysis of the aspirated fluid revealed a WBC count of 30,000 cells/μL with 80% polymorphonucleocytes. The Gram stain result was negative. Culture grew methicillin-sensitive Staphylococcus epidermidis . After a discussion with the patient about treatment options, we performed a two-stage revision to address the chronic, deep infection.



Chapter Preview


Chapter Synopsis


Articulating, antibiotic-impregnated spacers are useful devices in the treatment of infected total knee arthroplasties (TKAs). By reducing the articular surface friction and providing improved knee stability, the prosthesis with antibiotic-loaded acrylic cement (PROSTALAC) facilitates early mobility and discharge from the hospital. It promotes a higher level of activity for the patient, preserves motion through the knee, and improves quadriceps strength before definitive reimplantation. Local delivery of antibiotics can be tailored to the infecting organism by adding customized doses of antibiotics to the bone cement before the spacer is manufactured in the operating room.


Important Points





  • Use of heat-stable antibiotic powder is based on the known or suspected organisms, and it is applied at a minimum dose of 4 g of antibiotic per 40 g of cement. For most organisms, we use vancomycin (1.5 g per 40 g of cement) with tobramycin (3.6 g per 40 g of cement).



  • Bone stock is preserved whenever possible, but all infected and devitalized tissue must be débrided. Bone loss is preferred to leaving infected tissue in situ.



  • The size of the PROSTALAC system is customized to create a stable, well-aligned knee that aids rehabilitation and subsequent reconstruction. Asymmetric defects are augmented with extra antibiotic-loaded cement.



  • Patients should be encouraged to move and to work on quadriceps strengthening. We encourage range-of-motion exercises and allow weight bearing as tolerated depending on the quality and quantity of remaining host bone.



Clinical/Surgical Pearls





  • Increasing the amount of antibiotic in the cement increases the curing time.



  • Combining two antibiotics in the cement increases the overall elution rate.



  • After the implants have been sized, manufacturing the spacer on the back table while débridement and irrigation are being completed increases surgical efficiency.



Clinical/Surgical Pitfalls





  • The PROSTALAC components should not be removed from the molds until they are completely set. They will not release properly and may be damaged if removed too early.



  • The molds are lubricated with sterile mineral oil to prevent the cement sticking to them.



Introduction


Antibiotic-eluting spacers are a well-established means of delivering local antibiotic therapy. They most commonly are used during two-stage revision of an infected total knee arthroplasty (TKA) but also can be used to treat septic arthritis with joint degeneration when reconstructive arthroplasty will eventually be required. Static spacers deliver local antibiotic therapy but do not allow knee motion and limit the patient’s function. Articulating (mobile) spacers are designed to allow mobility and weight bearing while maintaining soft tissue tension and joint stability.


The prosthesis with antibiotic-loaded acrylic cement (PROSTALAC) was the first articulating spacer described in the English literature. The prototype, which was first used in 1987, consisted of a handcrafted copy of a conventional prosthesis made of highly antibiotic-loaded cement. In 1991, the device was improved by the use of flexible polyethylene molds, which created smooth articular surfaces on the femoral and tibial components. Although it allowed a greater range of relatively pain-free motion and improved mobility, this design had several problems. The cement-on-cement articulation was not a low-friction design and interfered with the rhythm of a normal gait. It was difficult to maintain a consistent thickness of the cement, making restoration of the joint line problematic. This spacer did not have a means of substituting for the posterior cruciate ligament (PCL) and was prone to instability.


The current PROSTALAC spacer was introduced in 1994 and had modifications to improve performance and stability. Its femoral and tibial components are made with antibiotic-loaded Palacos acrylic cement (Zimmer, Warsaw, Ind.). Each component is cast in a size-specific mold. The tibial mold allows adjustment of the thickness of the spacer to assist in restoration of bone loss. It has a post and cam mechanism formed from cement between two inlay polyethylene plateaus. The femoral component incorporates small metal runners linked by a posterior crossbar, creating a metal-on-polyethylene bearing surface with a posterior-stabilized system ( Fig. 33A.1 ).




FIGURE 33A.1


Anteroposterior ( A ) and lateral ( B ) images of the prosthesis with antibiotic-loaded acrylic cement (PROSTALAC) spacer after implantation.


Indications and Contraindications


The primary indication for the use of the PROSTALAC system is for treatment of an infected TKA as the first-stage device in a two-stage reimplantation. It can be used in the treatment of infected unicompartmental and revision TKAs. The PROSTALAC device may also be used to treat severe intraarticular infections when secondary reconstruction with joint replacement is appropriate.


Use of the PROSTALAC system is contraindicated when there is severe bone or soft tissue loss that prevents future reconstruction. Because the PROSTALAC design substitutes for the PCL, it requires intact collateral ligaments for optimal joint stability. Other means of management should be considered for patients who have severe life-threatening infections in multiple sites, those with severe immunocompromise, those with severe vascular deficiency, and in those with lifestyle issues that preclude major reconstructive surgery.


Equipment


Placement of the PROSTALAC system requires the following equipment:




  • General or regional anesthesia as appropriate



  • Standard supine positioning and an available tourniquet



  • Instruments to facilitate implant removal, including specialized knee extraction osteotomes, a Gigli saw, and an angled punch or femoral and tibial component extractors



  • Soft tissue débridement and pulse lavage systems



  • Palacos acrylic bone cement with gentamicin (three 40-g batches: one for the femoral component, one for the tibial component, and one to secure the spacer and fill minor defects)



  • Antibiotic powder: vancomycin (1.5g per 40 g of cement) with tobramycin (3.6 g per 40 g of cement)



  • Sterile mineral oil



  • Metal femoral condylar runners and polyethylene tibial skids ( Figs. 33A.2 and 33A.3 )




    FIGURE 33A.2


    Metal runners for the femoral component of the articulating joint prosthesis.



    FIGURE 33A.3


    Polyethylene tibial skids for the tibial component of the articulating joint prosthesis.



  • Femoral and tibial molds ( Figs. 33A.4 to 33A.6 )




    FIGURE 33A.4


    Molds for the femoral component. The base mold (left) supports the articular surface of the implant, and the bone surface mold (right) contours the bone-implant surface and compresses the cement into the runners. Both have been coated with mineral oil.



    FIGURE 33A.5


    The mold housing supports the implant while it is setting, allowing the surgeon to compress the bone surface mold into the base mold.



    FIGURE 33A.6


    Molds for the tibial component. The base mold (top) supports the polyethelene skids and forms the post for the posterior stabilizing mechanism. The housing (left) controls the thickness of the component by engaging a piston (right) that compresses the cement into the base mold and can be set to a required depth depending on bone loss. All components have been coated with mineral oil.



Examination and Imaging


Patients should be assessed for deformity and range of motion (ROM). An effusion or boggy swelling suggests synovitis. Careful attention should be paid to previous incisions and scars, especially recently healed lesions or actively draining sinuses. Vascular and neurologic assessments should be completed.


Anteroposterior, lateral, and skyline plain radiographs should be obtained, and old radiographs are useful for comparison. They are assessed for lysis, implant migration, and likely bone defects. The extent of cement within the femoral and tibial canals should be accounted for in the preoperative plan for removal.


Anatomy and Approaches


When presented with multiple previous surgical scars, the surgeon should try to use the most lateral incision. Transverse scars should be crossed with a perpendicular incision. Incisions should be planned to débride areas where an open wound or sinus exists. If there are concerns about soft tissue viability or about obtaining closure, particularly over the proximal tibia, the opinion of a plastic surgeon should be sought preoperatively. It is always best to achieve stable soft tissues at the first stage rather than wait until the second stage. Having healthy soft tissues at the time of reimplantation reduces the risk of recurrent infection.


A medial parapatellar approach is most commonly used, although lateral approaches have been described. Débriding the synovium from the suprapatellar pouch and then reestablishing the medial and lateral gutters can facilitate exposure and removal of potentially infected tissue. Eversion of the patella is neither required nor recommended. Remnants of the lateral patellofemoral ligament or scar in that area should be released, and the patella should be subluxated laterally instead of everted. After the liner is removed, a medial release is performed, and the tibia is externally rotated to assist in the subluxation of the patella. This protects the extensor mechanism, and extensile maneuvers are rarely required unless the medullary canal of the tibia needs exposure to remove hardware or cement. In that case, a tibial tubercle osteotomy (TTO) or other extensile maneuver (e.g., rectus snip, quadriceps turndown) is performed in keeping with the indications for these techniques.


Surgical Procedure


Step 1. Removal of Insert


The modular polyethylene insert can be removed using an osteotome or implant-specific instruments. In the setting of all-polyethylene or cementless trabecular metal prostheses, the implant can be carefully cut with a saw across the fixation lugs or post, which can then be removed separately. The lugs can be removed using a pencil-tipped, high-speed bur.


Step 2. Femoral Component Removal


The femoral component is debonded at the cement–implant or bone–implant interface using thin, flexible osteotomes. A Gigli saw may be used under the anterior flange. Small osteotomes must be used around the derotation lugs because they can pull condylar bone away when the component is removed. The posterior condylar portion of the implant and the implant around the intercondylar notch need to be debonded with angled osteotomes. Ideally, the component should be lifted free with a punch or a femoral extractor without excessive tapping or undue force.


Step 3. Tibial Component Removal


The tibial component is debonded initially with an oscillating saw using a variety of wide and narrow blades and subsequently debonded with osteotomes. Care should be taken to free the component anteriorly and posteriorly to avoid a plateau fracture. A series of wide osteotomes can be used to lift the component with gentle pressure. The stacked osteotome technique can be used to gently and slowly lift the tibial component, allowing it to be removed with a punch. The stacked osteotome technique allows the surgeon to avoid excessive leverage against soft tibial bone, which can make bone defects worse.


Step 4. Patellar Component Removal


The patellar component can be removed with a saw at the cement–implant or bone–implant interface. The remaining cement and lugs can be removed with a high-speed bur.


Step 5. Débridement and Preservation


The remaining bone can then be débrided, preserving bone where possible. The posterior soft tissue compartment can be easily débrided with protection of the neurovascular structures.


Step 6. Assessment of Bone Loss and Component Size


Bone loss should be assessed. The prior implants can give an indication of appropriate component size. Because a temporary spacer is being used, the exact location of the joint line is not as critical as it is for definitive reimplantation. There are three sizes of femoral and tibial components, and the most appropriate size is chosen for both sides. Trial implants that closely match the definitive PROSTALAC components are inserted, and the appropriate tibial liner thickness is determined. Bone loss can be accommodated by adding more cement on one side or the other to correct varus or valgus deformities when the definitive PROSTALAC components are cemented.


Step 7. Tibial Component Thickness


The thickness of the tibial component is established from the trial tibial insert. The tibial mold also allows alteration of the thickness of the tibial component in 2-mm increments based on the size of the trial tibial liner. For example, a 12-mm trial liner requires a 12-mm tibial PROSTALAC implant ( Fig. 33A.7 ; see Fig. 33A.6 ).




FIGURE 33A.7


The base mold has been removed, and the component can be ejected after it is set. A series of pinholes on the piston (bottom) allow control of height of the implant as compression is applied.


Step 8. Loading Cement in Molds


Antibiotic-loaded cement is then packed into the appropriate molds that have been lubricated with sterile mineral oil ( Fig. 33A.8 ; see Figs. 33A.4 and 33A.5 ). Curing time is substantially lengthened by the addition of antibiotic.




FIGURE 33A.8


The completed femoral component is allowed to completely harden.


Step 9. Débridement Completion and Lavage


While the molded PROSTALAC components are setting, débridement is completed, and the knee is thoroughly lavaged with no less than 9 L of normal saline. If the medullary canals of the tibia and femur are opened, they must be thoroughly débrided and irrigated. During the lavage phase, a small amount of antibiotic-loaded cement is inserted into the post defect in the tibia so that it is molded within the defect. It is then removed without disturbing its shape and is allowed to set on the back table for later insertion into the post defect after the cement has hardened. It is called the tibial plug .


Step 10. Inserting Implants


After they have set, the implants are removed from the molds ( Fig. 33A.9 ; see Fig. 33A.7 ). Large cement flashes created by the edges of the mold can be easily removed with a rongeur. The definitive implants are then inserted in a press-fit manner to determine appropriate fit and to determine whether the alignment needs to be adjusted. The surgeon determines how the alignment needs to be adjusted and whether additional cement on one side of the knee is required to correct alignment; this serves as a trial step. The tibial plug is then inserted in a press-fit manner without cement into the tibial post defect. The implants are cemented using Palacos acrylic bone cement impregnated with gentamicin but without additional antibiotic-loaded cement because the large surface area of the implant elutes sufficient amounts of antibiotics and the handling characteristics of highly antibiotic-loaded cement are too poor for cementations. The surgeon should not be afraid to cement the mold for fear of thorough interdigitation, because this never happens with the relatively poor quality of bone.




FIGURE 33A.9


The femoral component is released from the mold, and cement flashes are removed with a rongeur (not shown).


After the tibial component is cemented onto the tibia, the cement bonds to the tibial component and to the tibial plug, converting the tibial component into a stemmed component to increase its stability and avoid dislocation. It is important to insert the femoral component first because the post on the tibia interferes with placement of the femoral component.


Step 11. Unresurfaced Patella


The patella is not resurfaced.


Step 12. Closure


The tourniquet should be released, and bleeding should be controlled. The knee is closed in layers. If a TTO is performed, it is stabilized by conventional means using wires. A drain is not used to avoid the removal of antibiotic in the hematoma.


Controversies


The use of static spacers is still advocated by some as a means of tissue rest in the treatment of infection. Although some cases of severe bone loss are not suitable for articulating spacers, the advantages of improved function and easier second-stage reconstruction are desirable, especially when infection clearance rates are comparable.


Several systems of articulating spacers exist, including cement-on-cement, cement-on-polyethylene, and metal-on-polyethylene. Although no design has shown inferior infection clearance rates, a low-friction, stabilized design remains important. Issues with pain and poor motion linked to the high-friction, cement-on-cement bearing are improved markedly with metal-on-polyethelene designs.


The duration between the first and second stage was 6 weeks when a description of two-stage exchange arthroplasty was published in 1983. Intravenous antibiotics are administered during the 6 weeks. Although there is no direct evidence that earlier reconstruction results in inferior outcomes, this period is appropriate to allow soft tissue healing, muscle strengthening, and improvement in the patient’s general condition. Early implantation is possible when no spacer is used. However, when a spacer is used, rapid reimplantation is unnecessary and undesirable in our opinion.


After the 6-week course of intravenous antibiotics is finished, a drug holiday for an additional 6 weeks allows the surgeon to further monitor the patient and inflammatory markers for signs of resolution of the infection. We monitor the CRP level every 1 to 2 weeks after surgery, rather than the ESR, because the CRP measurement returns to normal earlier than the ESR. After the CRP level has normalized (<10 mg/L) or has consistently shown a downward trend, reimplantation is safe.


The PROSTALAC implant is durable enough to allow ongoing use if the second stage is delayed for any reason. In some cases in which second-stage reconstruction was not possible, the PROSTALAC system has been used indefinitely. Because function is inferior compared with a definitive knee implant, we do not recommend it routinely.


Complications


To ensure the PROSTALAC components release freely from the molds, they should be well lubricated with mineral oil, and the cement should be allowed to cure completely. Aggressive impaction should be avoided during insertion because it can fracture the spacer. If this occurs, it may be possible to cement it back together. The spacer can break at any time after implantation, especially if it is damaged or left in place for a longer duration than intended. Although this does not necessarily compromise treatment unless the spacer is fragmented, it may restrict knee mobility until reimplantation can proceed. In 17 years of using the current-generation PROSTALAC system, we have not seen a single case of a fragmented spacer.


In some cases, use of articulating spacers has been complicated by catastrophic failure. This is related to the failure of the surgeon to cement the spacer. When a spacer is not cemented and gaps are left between the spacer and the bone, the spacer may not be strong enough to withstand the forces applied to it. We strongly advocate that the spacers be cemented to avoid fracture, dislocation, and catastrophic failure of the spacer with potentially disastrous consequences. Articulating spacers have been criticized because of these complications, but they can be avoided by paying careful attention to detail and following the steps of the procedure outlined earlier.


In cases of severe bone or soft tissue loss, it may be difficult to create a stable spacer. The PROSTALAC system incorporates a posterior-stabilized design, but it does not compensate for the loss of collateral integrity. Failure to adequately balance soft tissues may result in dislocation. Care should be taken to appropriately rotate the implants to facilitate patellar tracking. Poor tracking may restrict motion and, along with thinning of the patella, lead to dislocation or fracture.


Outcomes


The concept of a staged revision to treat infected TKAs was introduced in the 1970s. In 1983, Insall and colleagues published their results of successfully treating eleven patients with removal of the prosthesis, débridement, antibiotic therapy, and delayed reimplantation. The use of antibiotic beads and spacers was introduced to increase local antibiotic delivery and to control soft tissue tension. In 1988, Wilde and Ruth published a series showing a 90% success rate for controlling infection with spacer use and an 80% rate without spacer use. Although the findings were not statistically significant, advantages included improved patient satisfaction and easier reconstruction.


The effort to overcome the functional difficulties of static systems led to the development of several types of articulating spacer, which produced infection clearance rates ranging from 85% to 100% and acceptable functional outcomes ( Table 33A.1 ). Although no randomized trial of static versus articulating spacers exists, several comparative studies have been published. Park and co-workers compared two groups treated with static or all-cement articulating spacers. They found a small but significantly better ROM and functional outcome when using articulating spacers, but they observed no difference in reinfection rates. Emerson and associates and Fehring and colleagues reported outcomes of static versus articulating spacers using metal-on-polyethelene and cement-on-cement designs, respectively. Jämsen and co-workers compared all-cement spacers with metal-on-polyethelene spacers. There were no differences in reinfection rates or functional outcomes in any of these studies.



Table 33A.1

Outcomes for Two-Stage Treatment of Infected Total Knee Arthroplasties




























































































































Spacer Study No. of Knees
(Avg. Months of Follow-up)
Knee Reinfections Reported Outcomes After Reimplantation
None Insall et al, 1983 11 (34) 1 with different organism Avg. flexion: 95 degrees (2 patients with lag of >20 degrees)
HSS scores: 5 excellent, 4 good, 2 fair
None Rosenberg et al, 1988 26 (29) 0 ROM: 21 had >90 degrees
HSS scores: 12 excellent, 6 good, 2 fair, 4 poor
None Windsor et al, 1990 38 (48) 1 with same organism
3 with different organisms
Avg. ROM: 87 degrees
HSS scores: 11 excellent, 13 good, 6 fair, 7 poor
Beads Whiteside, 1994 33 (19) 4 required one or two more operations Avg. flexion: 100 degrees
Static Booth and Lotke, 1989 25 (25) 1 with same organism Avg. flexion: 100 degrees
Avg. HSS scores: 81.5; function, 64
Static vs C/C Park et al, 2010 Static: 20 (36)
C/C: 16 (29)
Static: 3
C/C: 1
Avg. flexion: static, 92 degrees; C/C, 108 degrees (significant, P = .04)
Avg. HSS scores: static, 80; C/C, 87 (significant, P = .04)
Avg. flexion during treatment: static spacer, 9 degrees; C/C, 80 degrees
HSS scores during treatment: static, 48.2; C/C, 57.2
C/C Durbhakula et al, 2004 24 (33) 2 with same organism Avg. flexion: 104 degrees
Avg. HSS score: 82
C/C Pitto et al, 2005 21 (24) 2 never reimplanted Avg. flexion: 94 degrees
Avg. KSS: 81; function, 84
Avg. flexion during treatment: 77 degrees
Avg. KSS during treatment: 74; function, 75
C/C Ha, 2006 12 (range, 24-42) 0 Avg. flexion: 102 degrees
Avg. KSS: 87; function, 80
Avg. flexion during treatment: 85 degrees
C/C Van Thiel et al, 2011 60 (35) 3 with same organism
4 with different organisms
Avg. flexion: 101.3 degrees
Avg. KSS: 79
C/PE Evans, 2004 31 (min. of 24) 2 with same organism Avg. ROM: 0 to 130 degrees
Avg. ROM during treatment: −10 to 80 degrees
M/PE Cuckler, 2005 44 (65) 1 with same organism Avg. flexion: 110 degrees
Avg. KSS: 84
Avg. flexion during treatment: 110 degrees
M/PE Hofman et al, 2005 50 (73) 3 with same organism
3 with different organisms
Avg. ROM: 4 to 104 degrees
Avg. HSS score: 89
Avg. ROM during treatment: −6 to 91 degrees
M/PE (PROS) Haddad et al, 2000 45 (48) 1 with same organism
3 with different organisms
Avg. flexion: 94.5 degrees
Avg. HSS score: 71.5
M/PE (PROS) Meek et al, 2003 47 (41) 2 with same organism (MRSA) Avg. flexion: 87 degrees
Avg. scores: WOMAC pain, 68.9; WOMAC function, 77.1; Oxford, 67.3; SF-12 mental, 53.7; SF-12 physical, 41.2
M/PE (PROS) Gooding, 2011 115 14 reinfected; all repeat two-stage procedures, of which 2 failed Avg. flexion: 93.2 degrees at 2 years
Avg. scores at 2 years: WOMAC pain, 68.6; WOMAC function, 63.2; Oxford, 60.8; SF-12 mental, 55.4; SF-12 physical, 35.5
Static vs C/C Fehring et al, 2000 Static: 25 (36)
C/C: 30 (27)
Static: 3
C/C: 1 (NS)
Avg. flexion: static, 98 degrees; C/C, 105 degrees (NS)
Avg. HSS score: static, 83; C/C, 84 (NS)
Static vs M/PE Emerson et al, 2002 Static: 26 (90)
M/PE: (46)
2 in each group; all with different organisms Avg. flexion: static, 93.7 degrees; M/PE, 107.8 degrees (significant, P = .01)
C/C vs M/PE Jämsen et al, 2006 C/C: 10 (48)
M/PE: 24 (25)
C/C: 3 repeat two-stage procedures; 1 failed; 1 AKA after first stage
M/PE: 6 repeat two-stage procedures; 2 failed; 2 never reimplanted after first stage
Avg. flexion: C/C, 92 degrees; M/PE, 103 degrees (NS)
Avg. C/C KSS: pain, 46; function, 53; overall knee score, 79
Avg. M/PE KSS: pain, 46.8; function, 58, overall knee score, 81
Avg. flexion during treatment: C/C, 44.3 degrees; M/PE, 87.3 degrees
Avg. C/C KSS during treatment: pain, 15, function, 13, overall knee score, 34
Avg. M/PE KSS during treatment: pain, 19.1; function, 17.9; overall knee score, 40

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

May 29, 2019 | Posted by in ORTHOPEDIC | Comments Off on Two-Stage Revision

Full access? Get Clinical Tree

Get Clinical Tree app for offline access