COMPLICATIONS OF TOTAL ANKLE REPLACEMENT
January 1st, 2003
Steven M. Raikin, MD and Mark S. Myerson, MD
ABSTRACT
Intermediate clinical and radiographic results of the new generation of total ankle arthroplasties have been encouraging. Many of the devastating problems of earlier designs have been avoided, yet these newer implants are not without complications. Many of these can be avoided with careful preoperative planning and by paying meticulous attention to technique during implantation. Common pitfalls with the implantation process of the Agility (DuPuy, Warsaw, IN) total ankle replacement are discussed. The diagnosis and treatment of complications occurring intraoperatively and in the intermediate- and longer-term postoperative period are outlined.
INTRODUCTION
Earlier generations of total ankle arthroplasties were plagued with high complication rates and poor results. Constrained systems such as the Mayo total ankle and the Beck-Steffe Conaxial ankle resulted in loosening rates of 6019 and 90%,37 respectively, and survival rates of 42% at 10-year followup.19 The unconstrained systems had similarly dismal results, in these cases because of instability of the construct. The Smith multiaxial ankle had 52% poor results as early as 2 years after surgery,9 whereas the Newton ankle had a 50% failure rate at just 3 years.31 The first generation of semiconstrained ankle replacement was the Imperial College, London Hospital system. However, this had a domed convex polyethylene component that resulted in high wear rates and early failure. The 5-year results of this system had only 25% good and excellent results. The high rate of failure and poor results led to the opinion that arthrodesis should be considered as the treatment of choice for painful arthritis.6 Total ankle arthroplasty was considered to be contraindicated in patients with severe rheumatoid arthritis31 and, because of these complications, it was stated that they should no longer be implanted.19,37
Although arthrodesis remains a commonly accepted treatment of painful ankle arthritis, this treatment, too, continues to be associated with high rates of complications and functional impairment. McGuire et al29 reported a 30% higher complication rate for ankle arthrodesis than for than ankle arthroplasty, whereas Kofoed and Sturup23 found that patients treated with ankle replacement had better overall function and pain relief than those treated with arthrodesis. Many different techniques of arthrodesis have been described, yet the reported nonunion rate has been as high as 41%10 and the infection rate, as high as 28%.29 Patients with arthrodeses of the ankle often have difficulty climbing stairs and walking on uneven surfaces. Physiologically, these patients have a decreased gait velocity of 16%, an increased oxygen consumption of 3%, and an overall 10% decrease in gait efficiency.35 This increased physiologic demand is not well tolerated by elderly patients who, because of other medical conditions often have a diminished ability to compensate for this higher demand. Ankle arthrodesis also results in increased compensatory motion at adjacent joints as well as on the contralateral limb,28 resulting in increased degenerative changes in the subtalar and the transverse tarsal joints at long-term followup.1,17,25,30 Although for some patients, arthritis of the subtalar and transverse tarsal joints may predate the ankle arthrodesis, the incidence of arthritis in these joints increases after arthrodesis, which may necessitate more extensive or pantalar arthrodesis.32
A new generation of uncemented semiconstrained ankle arthroplasties has been designed and implanted since the 1980s. The TNK Ankle (designed by Dr. Takakura, Kyocera, Japan) is a ceramic hydroxyapatite-coated prosthesis with good early results. Two different meniscal bearing prostheses have also had good clinical success: The Scandinavian Total Ankle Replacement (STAR) (Waldemar Link GmbH & Co, Hamburg, Germany) has had promising results in patients both under and over 50 years of age at up to 15 years followup,21 and the Buechel-Pappas Ultra Total Ankle Replacement (Endotec, South Orange, NJ) has produced 86% good and excellent results at an average of 5.5 years (San Giovanni TP, Thomas WH, Wilson MG, Theodore G: Long-term follow-up with a second generation, cementless, total ankle replacement. Presented at the 15th Annual Summer Meeting of the American Orthopaedic Foot and Ankle Society, Fajardo (PR), July 11, 1999). Another unique design, the Agility Total Ankle Replacement (DuPuy, Warsaw, IN), was developed with the aid of CAD-CAM computer technology and aimed to specifically address the modes of failure of its predecessors.3 This semiconstrained, uncemented device has a highly congruent talar-polyethylene articulation. The prosthesis has been designed to simulate the anatomical alignment of the talar dome and to achieve increased component stability by the addition of an arthrodesis of the tibiofibular syndesmosis. The latter theoretically distributes the weightbearing load between the distal tibia and the fused distal tibiofibular articulation.
Despite current advances and satisfactory intermediate results,3,33 total ankle replacement remains technically demanding. Because the Agility ankle currently is being used by the authors and is the only total ankle of the current generation of device systems that fully is approved by the FDA., the purpose of the current study was to define some of the complications associated with this device intraoperatively and at intermediate- and long-term postoperative followup, and to document their diagnosis and treatment.
PREOPERATIVE PLANNING
Patient selection and preoperative planning is very important in influencing the results of total ankle replacement. Although the preferred patient is older, has a relatively sedentary life style, and weighs less than 240 pounds, none of these variables have, as yet, been clinically studied with the Agility system. Clinical experience suggests that the patient weight should be offset by the expected size of implant to be used. That is, a muscular 240-pound patient with a large bony architecture who is templated as requiring a size six implant (the largest size) would be expected to have a longer survivability of that total ankle replacement than would an obese patient of the same weight with a small bony architecture, templated as requiring a small (size two or three) implant. A recent study on the STAR system found no difference in revision rates between patients older than 55 years compared with those younger than 55 years.21 Patients with ankle arthritis because of rheumatoid arthritis have been shown to have better outcomes with the STAR ankle replacement than do patients with posttraumatic arthritis or osteoarthritis;22 however, such improved outcomes have not been seen thus far with the Agility system.33
Contraindications to ankle replacement are neuropathic arthropathy of the ankle (or known peripheral neuropathy that may progress to a neuroarthropathy) and infection of the ankle joint, distal tibia, or talus. Although no specific guidelines are given for total ankle arthroplasty, evidence of an active infection within the previous 12 months remains a contraindication in other joint arthroplasties13 and should be used as a guideline in ankle replacements. Assessment of an intraoperative frozen section specimen showing less than 10 polymorphoneucleocytes per high-power field is the accepted criteria for exclusion of ongoing infection.26 This frozen section assessment should be routinely obtained in suspected cases and those with known previous infection. Osteonecrosis of the talus or distal tibia should be considered a relative contraindication to this procedure, depending on the extent of the necrosis. The extent of osteonecrosis varies, but rarely is the entire talar body affected. Preoperative magnetic resonance imaging is useful in determining the amount of bone involved and the probability of subsequent subsidence of the components. If the magnetic resonance image indicates that most of the avascular bone needs to be removed when the appropriate bony cuts are made in the talus or the distal tibia, one can assume that the prosthesis will be supported by healthy bone and that, therefore, arthroplasty is not contraindicated. However, if after the appropriate bony resection intraoperatively, the remaining bone is inadequately vascularized, the arthroplasty should be aborted and an arthrodesis, usually requiring bulk bone graft, should be performed. Alternatively, impaction bone grafting a small osteonecrotic section may facilitate bone ingrowth and stability of the prosthesis, but this possibility has yet to be studied. Bone vascularity may be difficult to assess intraoperatively and may require temporary release of the tourniquet.
Hindfoot alignment needs to be carefully assessed before ankle replacement. Alignment is assessed with a careful physical examination and weightbearing radiographs of the ankle and foot. Occasionally, stress radiographs in varus and valgus may be required to confirm hindfoot malalignment or instability and to differentiate instability from ankle ligament insufficiency. Hindfoot deformity should be corrected before replacing the ankle. Failure to adequately address the hindfoot problem will result in altered biomechanical stress on the replaced ankle joint that will lead to instability of the ankle, abnormal loads on the joint, and premature loosening. Hindfoot valgus can result in increased stress on the lateral side of the replaced ankle articulation and increased wear of the polyethylene. As the valgus progresses, stress fractures of the fibula or midfoot collapse may occur. Valgus alignment of the tibia can be corrected with a supramalleolar osteotomy or a triple arthrodesis. Failure of this correction may necessitate reconstruction of the deltoid ligament to balance the ankle. Similarly, varus alignment of the hindfoot may result in increased medial polyethylene wear and, with progressive deformity, lateral ankle ligament insufficiency and instability. The current authors correct the alignment of the hindfoot as a first-stage procedure, generally with an osteotomy or an arthrodesis, and reform the ankle replacement 6 to 12 weeks later.
Malalignment of the ankle should be addressed intraoperatively during ankle replacement. Postoperative varus alignment of the ankle should not be accepted. Persistent heel varus can be corrected with an osteotomy. If the varus is because of the presence of a tilt in the tibiotalar joint itself, a partial release or lengthening of the deltoid ligament is required. Inserting a lamina spreader into each side of the joint and distracting the ankle is useful in assessing the balance of the ligaments. The ligaments then are released, creating an equal medial and lateral joint space opening. In cases of varus because of lateral ankle instability, correction is obtained by performing a lateral ankle ligament reconstruction procedure. Persistent varus after such corrective procedures have been performed can be addressed with a lateral sliding and closing wedge osteotomy of the calcaneus. If ankle malalignment is not corrected, the replaced ankle has an increased chance of failing. Failure can occur because of subsidence as a result of eccentric loading of the tibial tray, progressive instability of the ankle, or increased wear of the polyethylene from eccentric loading. Progression of the angular deformity will eventually result in stress fractures because of the increased load on the adjacent bone and potential catastrophic failure of the arthroplasty (Fig 1).
Before embarking on the procedure, the surgeon should be familiar with the system and its instrumentation. Preoperative templating to plan the required bony cuts and to size the needed component should be routinely performed.
INTRAOPERATIVE COMPLICATIONS
Surgery can be performed through a single anterior incision or through two incisions, one anterior and one lateral. The double-incision technique was developed in conjunction with syndesmotic arthrodesis for easier access to the fibula and the distal tibiofibular articulation. However, the second posterolateral incision is usually required only in cases where the fibula lies more posterior to the tibia than normal, as is seen in patients with posttraumatic arthritis with a fibula malunion and in patients with a congenital equinovarus deformity where the fibula is always seen posteriorly. Concern about whether excessive retraction of the skin margins using a single incision can cause later wound complications is not warranted. Full visualization of the fibula for the tibiofibular fusion is quite possible through the single anterior incision. During surgery, excessive skin tension is best prevented by retracting either in a medial or lateral direction and by avoiding concurrent retraction in both directions.
Because the anterior incision is positioned directly over the superficial peroneal nerve, this nerve should be identified, carefully retracted laterally, and protected throughout the procedure. Injury to the nerve will not cause any motor dysfunction, but a resulting neuroma can be debilitating. If the nerve is accidentally transected, it should be buried into a drill hole in the bone, away from its superficial location where irritation might occur. The deep peroneal nerve, which crosses the midline of the ankle deep to the anterior compartment muscles in association with the dorsalis pedis artery, should be identified and retracted. The neurovascular bundle can be retracted medially or laterally, depending on the level and angle at which it crosses the ankle joint.
The bone cuts for implantation of the tibial component require resection of the inner one-third of the medial and the lateral malleolus. Before making the bone cuts, careful fluoroscopic evaluation of the cutting block position is required to ensure optimal placement. The ankle should be rotated to ensure a true anteroposterior view of each malleolus and the cutting block. This step cannot be emphasized enough, because resecting too much bone may result in intraoperative or potential postoperative fractures of the malleoli. It is occasionally necessary to perform part of the bone resection without the use of the cutting blocks. Each cut is made once its position has been marked, allowing the malleoli to be visualized during the procedure. Particular care should be taken in patients with soft osteopenic bone, such as those with rheumatoid arthritis, who have an increased risk of fracture should the structural integrity of the malleoli be compromised. If a fracture does occur intraoperatively, it should be corrected with internal fixation because the malleoli are required for the stability of the prosthesis. Screw fixation of a malleolar fracture is difficult because of the location of the tibial component and the size of the remaining portion of the malleolus. For this reason, a medial malleolar fracture is stabilized with Kirschner wires, with or without a tension band construct. Fractures of the lateral malleolus may be stabilized using either a tension band or a one-third tubular plate and screw fixation. If excessive bone is inadvertently resected without resulting in a fracture, bone graft from resected distal tibia should be packed into the defect to help strengthen the malleolus and prevent future fractures. The fibula is particularly prone to stress fractures in patients with planovalgus deformity; if a fracture occurs postoperatively in such a patient, it should be addressed in the same manner as an intraoperative fracture.
While making both the talar and the tibial bone cuts, the surgeon also needs to keep in mind the anatomy of the posteromedial soft-tissue structures to avoid potential injury. The posterior tibial tendon, flexor hallucis longus tendon, and the neurovascular bundle run directly behind the posterior cortex of the tibia and can be accidentally transected during bone resection. In cases of posttraumatic arthritis, these structures may be scarred down to the underlying bone, and an additional small posteromedial incision may be useful to help visualize, elevate, and retract these structures. If transection of the posterior tibial tendon or the neurovascular bundle occurs, it should be repaired immediately through a posteromedial incision. It is probably acceptable to leave a transected flexor hallucis longus tendon unrepaired because marked loss of function of the hallux does not result because of the multiple communicating bands between the flexor hallucis longus and the flexor digitorum longus tendons just distal to the Master Knot of Henry in the foot.
Great care and careful planning are required in selecting the level of the bone cuts. The ankle should be distracted approximately 1 cm while the cuts are made, either with a lamina spreader internally or with an external fixator. One of these methods must also be used to balance the soft tissue on the medial and lateral side of the ankle as discussed earlier. Studies on the quality of the periarticular bone around the ankle joint have seen a statistically significant decrease in bone resistance to compression as one moves further from the subchondral plate.2,16 Resecting more than 4 to 5 mm of bone from either side of the joint may result in an increased risk of component subsidence after surgery (Fig 2). Conversely, removing too little bone can result in “overstuffing” the joint with components, which will result in a decreased postoperative arc of motion and ankle pain. If limited range of motion is found with the trial components, overstuffing should be addressed by removing the prosthesis and resecting additional bone. This bone is usually harvested from the talus, but intraoperative fluoroscopic assessment of the bone ends should be used as a guide to determine from where bone may be safely resected. Although insufficient bone resection may be responsible for limited range of motion, dorsiflexion loss usually should be corrected with lengthening of the Achilles tendon. Generally, on the operating table, 20° of dorsiflexion and 30° of plantar flexion of the prosthetic joint is desirable, but dorsiflexion is more commonly restricted.
Alignment of the inserted components will influence the outcome of the replacement. For example, placing the tibial component in more that 4° of valgus is associated with a significantly higher rate of pain.33 Malalignment is prevented by ensuring acceptable alignment in all planes intraoperatively via ankle rotation and fluoroscopic assessment. It is also important to drape the limb for the procedure with the knee exposed to ensure optimal positioning and alignment of the jig and cutting guides.
INTERMEDIATE TERM COMPLICATIONS
Wound breakdown, occasionally seen postoperatively, usually responds to dressing changes and local care. Superficial dehiscence is treated with daily wet-to-dry saline dressings, and deeper wound breakdowns are managed with silver sulfadiazine (Silvadine) dressings. If an eschar develops over an area of the incision, the eschar is best managed with observation and twice daily dressing changes, but without debridement. The current authors have not experienced a failure of this treatment, and although minor wound dehiscence occasionally occurs, in the current authors’ experience it has not been associated with infection.
Little long-term information is available for the Agility Total Ankle Replacement system. In a 2- to 12-year follow up study (average, 4.8 years), Pyevich et al33 found a 6% revision rate and a number of other complications. The most common postoperative complication associated with the Agility seems to be a plantar flexion contracture of the ankle, which, in the series by Pyevich et al,33 averaged 7°. Desirable ankle range of motion is an arc between 30 and 40°, with 10 to 15° being in dorsiflexion. Currently, it is not clear whether this plantar flexion contracture influences the patient’s reported pain or outcome, but functions such as stair climbing are affected adversely. Biomechanically, a plantar flexion contracture of the ankle has been shown to increase the peak forces across adjacent joints,36 which may lead to increased loading and therefore future arthritis. Thus, it is recommended that all patients undergoing ankle arthroplasty who do not achieve at least 15° of dorsiflexion intraoperatively undergo percutaneous Achilles tendon lengthening or gastrocnemius recession at the time of ankle replacement surgery, followed by appropriate immobilization of the ankle postoperatively to prevent recurrent equinus. The current authors have observed no problems with initial cast immobilization in patients who undergo lengthening of their Achilles tendons.
One of the more concerning problems with the Agility ankle is the rate of arthrodesis of the tibiofibular joint. Fusion of the syndesmosis correlates with stability of the tibial component and has the greatest influence on the outcome of the replacement.24,33 Conversely, nonunion or delayed union results in a statistically significant increase in tibial subsidence rates and ballooning lysis (Fig 3) at the interface between the bone and the tibial component.33 A solid fusion mass between the distal tibia and fibula should be radiographically evident by the fourth postoperative month. Patients with persistent ankle pain and swelling but inadequate radiographic evidence of union by this time require additional intervention. An initial period of immobilization in a cast or a weightbearing boot may be sufficient to allow the fusion to mature adequately, but an external bone stimulator should be used if union does not occur by the sixth postoperative month. Persistence of the nonunion beyond 6 to 9 months necessitates a revision of the arthrodesis with placement of supplemental cancellous bone graft.
A high incidence of radiographic circumferential lucent lines (less than 2 mm in width) around the tibial component commonly is seen. This is significantly more common in patients with nonunion or delayed union of the syndesmosis, but it is not exclusive to this group. These lines, however, rarely are progressive after 2 years and possibly represent failure of ingrowth of bone into the coated tibial base-plate, rather than loosening of the components.27,33 The presence of these lucent lines is of unknown clinical significance, and no correlation has been found between these lines and clinical failure in other ankle prostheses.34
A patient who has loosening of the tibial tray because of failure of bony ingrowth into the porous coating will often experience pain on initiation of walking (“start-up” pain). This pain will usually subside with continued ambulation but will recur with more prolonged walking. The stable “welding” of the tibial tray to the overlying bone can be seen on standard radiographs and computed tomography scans. This is recognized by the formation of trabeculae radiating from the region of ingrowth of the prosthesis. The absence of this pattern may indicate early loosening, which should be treated with immobilization and the addition of an external bone stimulator as described for problems of syndesmotic arthrodesis.
Tibial and talar subsidence will occur if the components are inserted too far from the subchondral plate. The more bone that is resected, the greater is the chance and degree of subsidence likely to be seen. Subsidence, which may be very subtle, may also indicate loosening of the component,27 which may result in eventual failure of the arthroplasty and require revision. The overall subsidence rates in the Agility ankle at an average 4.8 years after implantation is 13% for the tibial component (Fig 2) and 8% for the talar component.33 If subsidence is detected and noticed to be progressive on sequential radiographs, it should be addressed before catastrophic failure and increased bone loss occur. It is not clear what constitutes marked progressive subsidence necessitating revision, and each case must be monitored and treated accordingly. If it is thought that marked subsidence is present and likely to be progressive, then it can be corrected with revision of the components using a larger sized prosthesis and perhaps supplemental bone graft. Inability to revise the replacement is addressed with an arthrodesis of the ankle, usually with a bulk allograft interposition.(6057)
Occasional tibial base-plate fractures were seen with the initial design of the Agility system, but since the base-plate was made thicker in 1989, no case of fracture has been reported. Similarly, early cases were complicated by a high rate of talar component loosening, which also has not been a problem since the talar component was changed from titanium to cobalt chrome. The polyethylene insert is between 3.7 and 4.7 mm thick. Experience from arthroplasties of other joints suggests that a thickness of less than 8 mm leads to accelerated polyethylene wear15 and increased osteolysis. Whether such increased wear and osteolysis proves to be the case in the ankle remains to be seen. Substantial osteolysis secondary to polyethylene wear has not been reported to date, although followup remains relatively short. Arthroscopic evaluation of the ankle replaced with the Agility system has not shown macroscopic evidence of wear (F. G. Alvine, personal communication, 1999). Polyethylene wear depends on the bearing surface and the stability of the prosthesis. Although the talar component conforms highly to the dome of the polyethylene, and somewhat is constrained, malalignment of the components can lead to loss of this conformity and increased polyethylene wear.5 Highly congruent arthroplasties, such as the arthroplastic hip joint, produce smaller sized particles of ultrahigh density polyethylene during wear. Particles of 0.3 to 10 microns are phagocytosable by macrophages,11 resulting in apoptotic cell death.8 The released cytokines result in periprosthetic osteolysis that may lead to failure of the prosthesis.4 Longer-term followup and a greater number of replacements are required before it is known whether this mode of failure will be seen in the replaced ankle
If catastrophic failure of the ankle arthroplasty occurs, arthrodesis with bone grafting is the treatment of choice. Revision arthroplasty with larger components may occasionally be indicated as described above, but this historically has a success rate of only 20%.12 In the two largest series of arthrodesis for failed ankle replacement,7,20 a compression external fixator was used as the method of fixation. These resulted in fusion rates of 89% and 81%, respectively, although 19% of these required more than one surgical attempt. In both studies, patients had good or excellent results in 80% of cases. To maintain limb length, intercalated bone graft using iliac crest12 or femoral head allograft is recommended. In the event of a nonunion, salvage arthrodesis can be performed (with 78% union rates18) but residual symptoms are common. Clinical results in these cases only have a 40% good or excellent results.18
A final concern is overgrowth of bone around the medial and lateral margins of the prosthesis, which may result in painful impingement. Adequate resection of preexisting osteophytes must be completed and all bone fragments must be cleared from the joint margins during the replacement surgery to prevent this problem. Symptomatic impingement occurring postoperatively may be managed by local injection of hydrocortisone and physical therapy. If this modality fails, open or arthroscopic resection can be performed to remove the excess bone.
PAIN AFTER ANKLE REPLACEMENT
The source of pain after an ankle replacement can be difficult to isolate and subsequently treat. The current authors have arbitrarily divided the presentation of pain into three categories based on time of presentation after replacement [early (less than 3 months), intermediate (3 to 6 months), and late (more than 6 months)] and have devised an algorithm that may aid in diagnosis and treatment (Fig 4). Pain because of wound complications and neuromas have been discussed above. It may be difficult to differentiate a chronic regional pain syndrome (previously known as reflex sympathetic dystrophy) from an infection in both the early and intermediate time periods. Both present with a diffusely painful ankle, and a careful history and examination are required to differentiate the two. Patients with severe disproportionate pain with respect to their clinical findings should be evaluated for chronic regional pain syndrome. Such patients often complain of hypersensitivity of the overlying skin if it is touched or stimulated; the limb is cool and clammy, and may have sympathetically induced temperature and color changes. Chronic regional pain syndrome is managed with a course of lumbar sympathetic blocks (that act as a definitive diagnostic and therapeutic tool), physical therapy (which should include continued active ankle motion and a desensitization program), and medical treatment. It is recommended that a pain center be involved in the treatment of these patients.
Patients with an infection usually have a history of a fever and chills before presentation, the ankle is painful if moved, and although the skin is erythematous and warm to the touch, it is not hypersensitive. If the diagnosis cannot be made clinically, an erythrocyte sedimentation rate, a c-reactive protein level, a full blood count (including white cell differential count), and blood cultures should be obtained. Antibiotics should be administered only after the joint has been aspirated and fluid has been sent for culture and sensitivity. If infection is present, an incision and debridement should be performed promptly.
Based on the literature on hip and knee replacements, infections presenting within 3 months of the initial surgery can usually be treated with component salvage.14 Patients presenting more than 3 months after replacement surgery, those with persistent deep infection, or those with infections caused by virulent Gram-negative organism should be treated with component removal, intravenous antibiotics, and a second-stage revision or arthrodesis.
Patients presenting with pain that commenced more than 3 months after replacement commonly have problems related to the surgery itself. Varus or valgus malalignment of the hindfoot or the replaced ankle components places an increased stress on the ankle and adjacent joints. Malalignment can be diagnosed with clinical and radiographic evaluation and managed with hindfoot correction or component revision, as discussed earlier. Failure of bone ingrowth into the components coating may occur. Patients with this condition usually complain of “start-up” pain on initiation of ambulation, which subsides as the components settle with continued walking. The pain may be related to delayed or nonunion of the distal tibiofibular arthrodesis, and it should be managed initially with immobilization in a cast, with the addition of a bone stimulator if union has not occurred after 6 months. Revision with bone grafting is performed if the pain because of the lack of ingrowth is persistent, if the tibiofibular joint does not go on to unite, or if the components begin to subside into the tibia or talus.
Ankle instability is another cause of pain presenting during the intermediate period. The clinical diagnoses is typical, with a positive anterior drawer test of the ankle, or marked valgus or varus instability. This diagnosis is confirmed with stress radiographs and, if present, it should be treated as described earlier.
For patients presenting with pain more than 6 months after replacement, one must be suspicious of joint or bone failure, which may present as a stress fracture of one of the malleoli, loosening, or subsidence of the components. Fractures can be managed with open reduction and internal fixation, but the underlying problem causing malalignment must be corrected as described above.
Polyethylene wear and osteolysis have not been a major problem with the Agility system to date, but as with all other arthroplasties, these complications may occur with time. Osteolysis, too, may result in loosening with subsidence of the components, progressive bone loss, and subsequent failure. Should wear occur without substantial osteolysis or loosening of the components, the polyethylene insert can be replaced without removal of the tibial or talar components. Loosening, bone loss, and subsidence should be managed with revision of the arthroplasty with a larger component and bone grafting, with arthrodesis being the ultimate salvage if arthroplasty is not possible.
In conclusion, the Agility Total Ankle Replacement System is a good alternative to arthrodesis for painful arthritis of the ankle. The results are better than those seen with the first generation of total ankle replacements in the 1970s and 1980s. Although the radiographic complication rate is high, it does not seem to relate to the overall clinical outcome in the intermediate term.
Insertion of the prosthesis is technically very demanding and has a steep learning curve. Great care has to be taken in preoperative planning and meticulous attention to detail is required during insertion of the components. Doing so can diminish the risk of developing most of the known complications, or can increase the possibility of managing them effectively and in a timely fashion.
The long-term results remain unknown, and an adequate revision system is not as yet available. Salvage at this points remains revision with a larger component or conversion to an ankle fusion. Salvage arthrodesis usually requires bone block interposition grafting (usually with allograft) to prevent excessive shortening of the limb (Fig 5). In the past, this modality has had good long-term results with salvage by other total ankle arthroplasty systems.19
REFERENCES
1. Abdo RV, Wasilewski SA: Ankle arthrodesis: a long-term study. Foot Ankle 13:307-312, 1992.
2. Aitken GK, Bourne RB, Finlay JB, Rorabeck CH, Andreae PR: Indentation stiffness of the cancellous bone in the distal human tibia. Clin Orthop 201:264-270, 1985.
3. Alvine FG: Total ankle arthroplasty: new concepts and approach. Contemp Orthop 22:397-403, 1991.
4. Anthony PP, Gie GA, Howie CR, Ling RS: Localised endosteal bone lysis in relation to the femoral components of cemented total hip arthroplasties. J Bone Joint Surg 72B:971-979, 1990.
5. Blunn GW, Joshi AB, Minns RJ, et al: Wear in retrieved condylar knee arthroplasties. A comparison of wear in different designs of 280 retrieved condylar knee prostheses. J Arthroplasty 12:281-290, 1997.
6. Bolton-Maggs BG, Sudlow RA, Freeman MA: Total ankle arthroplasty. A long-term review of the London Hospital experience. J Bone Joint Surg 67B:785-790, 1985.
7. Carlsson AS, Montgomery F, Besjakov J: Arthrodesis of the ankle secondary to replacement. Foot Ankle Int 19:240-245, 1998.
8. Catelas I, Petit A, Zukor DJ, et al: Induction of macrophage apoptosis by ceramic and polyethylene particles in vitro. Biomaterials 20:625-630, 1999.
9. Dini AA, Bassett FH, III: Evaluation of the early result of Smith total ankle replacement. Clin Orthop 146:228-230, 1980.
10. Frey C, Halikus NM, Vu-Rose T, Ebramzadeh E: A review of ankle arthrodesis: Predisposing factors to nonunion. Foot Ankle Int 15:581-584, 1994.
11. Green TR, Fisher J, Stone M, Wroblewski BM, Ingham E: Polyethylene particles of a 'critical size' are necessary for the induction of cytokines by macrophages in vitro. Biomaterials 19:2297-2302, 1998.
12. Groth HE, Fitch HF: Salvage procedures for complications of total ankle arthroplasty. Clin Orthop 224:244-250, 1987.
13. Hanssen AD, Osmon DR, Nelson CL: Prevention of deep periprosthetic joint infection. Instr Course Lect 46:555-567, 1997.
14. Hanssen AD, Rand JA: Evaluation and treatment of infection at the site of a total hip or knee arthroplasty. Instr Course Lect 48:111-122, 1999.
15. Heim CS, Postak PD, Greenwald AS: Factors influencing the longevity of UHMWPE tibial components. Instr Course Lect 45:303-312, 1996.
16. Hvid I, Rasmussen O, Jensen NC, Nielsen S: Trabecular bone strength profiles at the ankle joint. Clin Orthop 199:306-312, 1985.
17. Jackson A, Glasgow M: Tarsal hypermobility after ankle fusion--fact or fiction? J Bone Joint Surg 61B:470-473, 1979.
18. Kitaoka HB: Salvage of nonunion following ankle arthrodesis for failed total ankle arthroplasty. Clin Orthop 268:37-43, 1991.
19. Kitaoka HB, Patzer GL: Clinical results of the Mayo total ankle arthroplasty. J Bone Joint Surg 78A:1658-1664, 1996.
20. Kitaoka HB, Romness DW: Arthrodesis for failed ankle arthroplasty. J Arthroplasty 7:277-284, 1992.
21. Kofoed H, Lundberg-Jensen A: Ankle arthroplasty in patients younger and older than 50 years: A prospective series with long-term follow-up. Foot Ankle Int 20:501-506, 1999.
22. Kofoed H, Sorensen TS: Ankle arthroplasty for rheumatoid arthritis and osteoarthritis: Prospective long-term study of cemented replacements. J Bone Joint Surg 80B:328-332, 1998.
23. Kofoed H, Sturup J: Comparison of ankle arthroplasty and arthrodesis. A prospective series with long-term follow-up. Foot 4:6-9, 1994.
24. Kronemyer B: Results of first 100 Agility total ankle arthroplasties encouraging. Orthop Today 17:16-17, 1997.
25. Lance EM, Paval A, Fries I, Larsen I, Patterson RL, Jr.: Arthrodesis of the ankle joint. A follow-up study. Clin Orthop 142:146-158, 1979.
26. Lonner JH, Desai P, DiCesare PE, Steiner G, Zuckerman JD: The reliability of analysis of intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J Bone Joint Surg 78A:1553-1558, 1996.
27. Manaster BJ: From the RSNA refresher courses. Total hip arthroplasty: Radiographic evaluation. Radiographics 16:645-660, 1996.
28. Mazur JM, Schwartz E, Simon SR: Ankle arthrodesis. Long-term follow-up with gait analysis. J Bone Joint Surg 61A:964-975, 1979.
29. McGuire MR, Kyle RF, Gustilo RB, Premer RF: Comparative analysis of ankle arthroplasty versus ankle arthrodesis. Clin Orthop 226:174-181, 1988.
30. Morgan CD, Henke JA, Bailey RW, Kaufer H: Long-term results of tibiotalar arthrodesis. J Bone Joint Surg 67A:546-550, 1985.
31. Newton SE, III: Total ankle arthroplasty. Clinical study of fifty cases. J Bone Joint Surg 64A:104-111, 1982.
32. Papa JA, Myerson MS: Pantalar and tibiotalocalcaneal arthrodesis for post-traumatic osteoarthrosis of the ankle and hindfoot. J Bone Joint Surg 74A:1042-1049, 1992.
33. Pyevich MT, Saltzman CL, Callaghan JJ, Alvine FG: Total ankle arthroplasty: A unique design. Two- to twelve-year follow-up. J Bone Joint Surg 80A:1410-1420, 1998.
34. Unger AS, Inglis AE, Mow CS, Figgie HE, III: Total ankle arthroplasty in rheumatoid arthritis: A long-term follow-up study. Foot Ankle 8:173-179, 1988.
35. Waters RL, Barnes G, Husserl T, Silver L, Liss R: Comparable energy expenditure after arthrodesis of the hip and ankle. J Bone Joint Surg 70A:1032-1037, 1988.
36. Wayne JS, Lawhorn KW, Davis KE, Prakash K, Adelaar RS: The effect of tibiotalar fixation on foot biomechanics. Foot Ankle Int 18:792-797, 1997.
37. Wynn AH, Wilde AH: Long-term follow-up of the Conaxial (Beck-Steffee) total ankle arthroplasty. Foot Ankle 13:303-306, 1992.
|
 |
