Total Ankle Arthroplasty:
Indications, Results, and Biomechanical Rationale
Thomas H. Lee, M.D.
.
orthopedic Foot and Ankle Center, 6200 Cleveland Ave., Columbus, Ohio 43231. Telephone: 614‑895‑8747.
Abstract:
Total ankle replacement was developed in the 1970s, after the success of total hip and knee arthroplasty. The goal of total ankle arthroplasty is to decrease pain and improve function in the lower limb. Ideally, in order to be superior to an arthrodesis, the ankle replacement should provide the patient good patterns of joint motion, the ability to walk and run, and have low complication rates. Unfortunately, total ankle arthroplasty has not been as successful as replacement of other joints. Published studies of early series with greater follow-up show that ankle arthroplasties did not provide lasting pain relief or improve function, and most ultimately failed. During the 1980s many authors concluded that total ankle arthroplasty was not warranted because of the generally poor long-term results and the high rate of complications. However, newer second-generation design techniques, innovative operative procedures, and dissatisfaction with the results of ankle arthrodesis have renewed interest in total ankle arthroplasty. This review describes the numerous types of ankle joint replacements, critically reviews the results, and reports on newer prostheses that incorporate more anatomical designs
INTRODUCTION:
The options that are available for the treatment of severe disabling arthritis of the ankle are conservative nonoperative treatments, soft-tissue and bony debridement, ligamentous reconstruction, corrective osteotomy, arthrodesis, and arthroplasty.1 Conservative, nonoperative treatments include activity modification, oral non-steroidal anti-inflammatory medications, analgesics, bracing, orthotic devices and, occasionally, intra-articular steroid injections. The most common and probably the most successful operative treatment of diffuse symmetric ankle arthrosis is an arthrodesis. Numerous authors and reports have advocated ankle arthrodesis and believe that it can consistently lead to a painless, stable foot in 59% to 75% of patients.2-4 It is considered by many authors to be the "gold standard" for the surgical treatment of post-traumatic arthritis, especially in young, active patients. In many surgeons’ hands and for many patients, arthrodesis will relieve pain and correct malalignment of the ankle joint, subsequently restoring the patient to a higher level of activity.1,5
Unfortunately, the various forms of ankle arthrodesis can be fraught with complications, such as infection, skin slough, nerve injury or entrapment, and even nonunion and malunion.3,6 In addition, sacrificing ankle motion to relieve pain puts greater stress on the surrounding joints. Specifically, gait analyses and cadaveric studies have shown that a tibiotalar fusion places increased stress on knee, subtalar, and midfoot regions. With time, the increased stresses can lead to hindfoot and midfoot arthritis and the possibility of surgical fusion of these joints, as well. Bauer et al, reported that up to 80% of patients with an ankle arthrodesis developed arthritic changes in the adjacent joints within 12 years.7 Other disadvantages included a 10% - 22% rate of pseudoarthrosis, a long period of cast immobilization, and poor patient acceptance of permanent ankle stiffening.8-10 Other studies have shown a 15% incidence of subtalar arthritis after a tibiotalar arthrodesis and an overall unsatisfactory result in 20% of patients.11
Ankle arthrodesis has been shown to produce abnormal gait patterns, specifically abnormal knee flexion at stance phase and slight flexion in the mid- and forefoot.12 Loss of ankle motion, specifically tibiopedal motion, results in definite functional deficits such as difficulty with inclined surfaces, pain with rigorous activities, and difficulty running.2,11,13 Although velocity of gait has been shown to be slowed and the length of stride shortened in patients with an ankle arthrodesis, the lost motion can be compensated for by motion of the small joints in the foot, altered motion of the ankle in the contralateral limb, and appropriate footwear.2 Patients with pre-existing arthritis in the subtalar and midtarsal joints, such as rheumatoid patients, have loss of the compensatory movements of the midfoot that are required for comfortable gait after ankle fusion.5,14 Furthermore, there is speculation that joints at the knee and hip may be compromised after ankle fusion as the body attempts to compensate for the lost range of motion at the ankle.15 The diminished overall motion and increased stresses on the remaining joints may lead to a poor result in an ankle arthrodesis. After a pantalar arthrodesis, dorsiflexion is diminished approximately 63% and plantarflexion about 82%.3 A triple arthrodesis diminishes dorsiflexion 13% and plantarflexion 16%.3 Other disadvantages include a long period of immobilization in a cast or brace, which may cause some patients to be non-ambulatory. Hansen believes that an ankle fusion is only helpful initially and that eventual failure due to the subtalar or talocalcaneal joints becoming over-stressed is inevitable.16 Therefore, a reliable ankle replacement system would be a welcome addition to an orthopedic practice.
Total ankle arthroplasty was developed in the 1970s after success with total hip and knee arthroplasties. It has been indicated as an alternative to ankle arthrodesis that has the potential to provide pain relief while preserving joint motion and stability. To date, however, there is no ankle replacement system that shows results comparable to those achieved in knee or hip replacements. Although early results of total ankle arthroplasty were encouraging, further experience and longer follow-up periods revealed poor long-term results. For example, in an early report, Lachiewicz, et al reported on 15 patients with an average follow-up of 3.3 years, with the results considered excellent.17 However, when Unger, et al reported on the same 15 patients with a longer follow-up (average 6.2 years), deterioration in their clinical scores was apparent.17,18 In the 1970s good results and implant survival were reported in 80% - 85% of the patients (Table I). These studies consistently showed better results in patients with rheumatoid arthritis or in those with low physical demands and restricted activity levels than in patients with post-traumatic arthritis.19 In addition, previous ipsilateral foot/ankle operative procedures and younger age have been reported to be associated with a higher risk of failure.20
Based on these early results, the indications for total ankle arthroplasty were expanded from rheumatoid and severe degenerative ankles to include younger patients and patients with isolated traumatic arthritis. Unfortunately, in many later reports published in the 1980s with follow-up periods averaging more than 4 to 5 years, failure occurred in a high percentage of the total ankle arthroplasties (Table II). In fact, Bolton-Maggs, et al reported such poor results with total ankle arthroplasties that in 1985 they recommended arthrodesis as the treatment of choice for any arthritic ankle.4 However, more recent reports using newer ankle replacement systems appear to be promising. This review describes the numerous types of ankle joint replacements, critically reviews the results, and reports on newer prostheses that incorporate more anatomical designs. It should be noted that it is difficult to compare series of total ankle arthroplasties, mostly due to variabilities in diagnosis, patient age, length of follow-up, prosthesis design, and scoring systems.
The three main types of ankle replacement systems are constrained, nonconstrained, and semiconstrained. They have either 2 or 3 components. Unfortunately, most reports have shown that the long-term results of all designs have been disappointing.4,7,11,18,21,22 In addition, the different ankle arthroplasties have various types of fixations: pegs, long or short stems, and cylindrical or rectangular bars. Earlier designs had tibial components made of polyethylene, while newer ones are metal-backed.7
Constrained designs incorrectly treat the ankle as a hinge joint and transfer all of the normal ankle stresses to the bone-cement interface. The advantage of constrained systems is their inherent stability but this is believed to lead to a high rate of loosening, especially of the tibial components.13 In addition, they are less prone to subluxate or cause impingement The TPR (Thompson, Parkridge & Richards),23 ICLH (Imperial College London Hospital),24 Conaxial(Beck-Steffee),13 Oregon,25 and Mayo Clinic Ankle(1976)20,21 prostheses are designs of this type. After reviewing the 9-year follow-up results of one of the more widely used implants, the constrained Mayo prosthesis, Kitaoka and Patzer no longer recommend it for any ankle arthritic condition.21 In their series, 41% of the ankles needed further surgery, including debridements for infection, revisions for pain, decompression for malleolar impingement, and arthrodesis for failures.21 In another study of constrained implants, 90% of the Conaxial prostheses were loose after10 years and a there was a complication rate of 60%.13
On the other hand, nonconstrained implants such as the multi-axial design, have been reported to be unstable and associated with malleolar impingment. However, they may allow better stress distribution than the more constrained prostheses, resulting in less loosening. They allow for motion in multiple planes and rely entirely on the ligaments without any certainty of maintaining the normal ankle axis.20,26 The Mayo Clinic Ankle (1989),21 St. George-Buckholz,27 Smith,28,29 Newton,30 and the Irvine31are designs of this type.
Several reasons for the long-term failure of these early prostheses have been suggested. First, many original designs required excessive bone resections and relied on cement fixation onto soft cancellous bone. Constrained prostheses placed excessive stress on the cement – cancellous bone interface. Subsequently, the main reason for failure was aseptic loosening. Nonconstrained prostheses, as previously mentioned, resulted in a lower incidence of loosening but failure occurred due to malleolar and soft-tissue impingement. In addition, many talar components were often placed on top of the talar dome, thus changing the normal rotational axis of the ankle. Kofoed eloquently summarized the failure of early designs as caused by the lack of respect for the anatomy, kinematics, alignment, and stability of the ankle joint.32
The challenge of designing successful ankle prostheses is to understand the biomechanical factors relating to the failure of prior implants. These factors include the anatomy of the ankle joint along with its variability in axis of rotation and subsequent motions, forces generated across the joint, stability of the ankle joint, and amount of bone resection required for implantation.
Anatomical Considerations
The ankle joint is a complex system composed of the articulation between the talus and the mortise. It includes the distal end of the tibia and the lateral and medial malleoli. There are three sets of articular surfaces: the tibia and the superior talus, the medial malleolus and the medial talus and, the lateral malleolus and lateral talus. This bony geometry and orientation of the ligamentous and muscular systems make it a very stable joint. The axis of the ankle passes just distal to the tip of each malleolus and is directed laterally and posteriorly.
Although it is often regarded as a single-axis, dorsiflexion-plantarflexion hinge, the ankle joint axis orientation may vary slightly, especially when it is loaded and horizontal-plane rotation is allowed to occur.32 The ankle axis has been described as a changing axis or changing instant center of rotation. This is caused mainly by the shape of the talar trochlea and action of the soft tissues.33 In cadaveric and gait studies the rotation has been shown to be between 10° – 12°.34 It is this variable center of rotation which allows the talus to glide and slide within the ankle mortise during flexion and extension.8,35 In addition, the curvature of the talus and the distal tibia show varying radii7,31 that allow horizontal rotations to occur in the foot or leg with movements of the ankle. Authors have described the axial rotation of the tibia relative to the talus to be between 6° and 12°.34,36 Specifically, on dorsiflexion of the ankle, the foot deviates outward and on plantarflexion the foot deviates inward. This external and internal rotation varies with the obliquity of its axis.
Normally, ankle motion ranges from 20° - 36° with approximately 30° needed for walking, 37° for ascending stairs, and 56° for descending stairs.8,34,37,38 These rotations coupled with loads across the ankle joint generate axial as well as substantial rotational forces. Stauffer, et al reported a mean maximum ankle joint reaction force of about 4 – 7 times body weight for normal people during normal level walking. In addition, the shear forces across the ankle joint reach a peak of about 80% of body weight. This is greater than forces stated for the hip and knee joints.39 Normal ankle ligaments, tendon and tendon sheaths, muscles, compression at the talomalleolar surfaces, and conformity of the tibiotalar surface under load control these rotational forces.40
Therefore, the biomechanics of an ankle arthroplasty must take into account these important forces and motions. These very large forces are distributed over a large weight-bearing surface area, which is actually slightly larger than either the hip or knee joint (10 cm2).8 With regard to force distribution characteristics, one can think of ankle replacements as either being congruent designs, allowing single-plane motion, or incongruent designs, allowing multiaxial motion. This is based on the geometry of their articulating surfaces. Trochlear, bispherical, concave-convex, and convex-convex are variants of the incongruent type, while spherical, spheroidal, conical, cylindrical and sliding-cylindrical are types of the congruent design.41
Early incongruent ankle arthroplasties suffered poor wear and deformation resistance due to high local stresses at their incongruent surfaces. Studies have shown that incongruent contact between polyethylene and metal results in excessive wear and is unsuitable for long-term use.34 Congruent arthroplasties, on the other hand, allow good pressure distribution and better surface deformation resistance; however these designs produce undesirable constraining forces.34,41
Biomechanical as well as clinical studies have shown that an ankle prosthetic design cannot treat the joint as just a simple hinge joint as did the original constrained prostheses, which have had the highest reported rates of loosening. Prostheses such as the Irvine,31 Smith,29 and Newton22 designs, which failed to provide inversion-eversion or internal-external rotatory stability, led to unstable ankles, as well.40,42
In constrained prostheses, especially cemented implants, the forces are transmitted directly to the bone-cement interface. These high torsional or shearing loads may ultimately lead to the failure of the implant. For example, the Mayo,19 the Oregon,25 and the Takakura43 suffered high constraint forces. Furthermore, the TPR23 and Newton22 replacements suffered from high contact stresses which may have led to their failure.42 Therefore, an implant should minimize tensile or shear loads and transmit the forces by compression. The LCS sliding cylindrical design attempts to achieve this goal by using broad, congruent surfaces on both the tibial and talar components and allowing axial rotation, mediolateral and anteroposterior sliding.34
Fully non-constrained prostheses rely on the ankle ligaments, capsule, and malleoli restoration for stability. Therefore, it is imporant to maintain the normal ankle anatomy and its subsequent distribution of stresses. Burge, et al showed that even an error of 1-2 mm in tensioning of the soft tissues after arthroplasty can have a major effect on prosthetic joint laxity.40 Three-part ankle replacements with variably sized meniscal implants offer a theoretical advantage in that ligament tension can be adjusted after the components have been placed. Thicker meniscal implants can provide increased stability. One cadaveric study, however, showed that a 3-part replacement failed to provide anteroposterior stability and that this may put stress on the ligaments leading to their failure.40 To date, no clinical studies of these types of prostheses have reported this problem.
A design must consider the amount of joint resection necessary for implantation. Excessive resection leaves the prosthesis supported only by soft cancellous bone; cases in which prostheses were placed in cancellous bone, even with the use of cement, had poor results.43 For example, migration and sinkage of the talar component was seen in almost every ICLH implant on follow-up. This was attributed to the soft cancellous bone supporting the implant.44 It has been theorized that some uncemented designs do not have cortical bone support anteriorly and posteriorly and may lead to the subsequent sinking and loosening seen in some of the reported cases.45 Takakura, et al believe that an arthroplasty should be performed with the prosthesis protruding from the anterior and posterior surfaces of cortex for additional support.43 Additionally, it is important to maintain sufficient bone stock after prosthetic removal to allow fusion or secondary reimplantation. Furthermore, minimizing removal of bone from the talus is important to prevent avascular necrosis.46
Newer second-generation designs are are considered semi-constrained. These designs preserve bone stock and retain the normal rotational axis and tibiopedal alignment. The New Jersey LCS, STARR, and Agility prostheses are designs of this type. The LCS and STARR are 3-component designs that incorporate a free-gliding meniscus. The upper articulation between the tibia and meniscus allows rotation and gliding, while the articulation between the talar component allows flexion and extension. The Agility is a 3-component design that has a polyethylene component secured within the tibial component. The design allows some axial rotation as well as some medial-lateral translation of the talar component within the polyethylene component.
The New Jersey LCS total ankle arthroplasty is an uncemented 3-component prosthesis. It consists of a noncemented metal inlay tibial component, a polyethylene mobile bearing, and a noncemented metal onlay talar component. The prosthesis is designed to allow sliding in the mediolateral or anteroposterior directions without constraint from the implant. Initial results are encouraging with no clinical or radiologic failures and 87% (20 of 23) reporting excellent pain relief at an average of 35.3 months. The major complications reported were poor wound healing, reflex sympathetic dystrophy, and deep infection.34 In a later study using implant revision as an end point, Buechel and Pappas reported a survival rate of 94.75% at 10 years with 85% good to excellent clinical results.47 Mechanical problems related to meniscal-bearing subluxation occurred in 4 of 40 patients and required revision in 2 cases. Doets48 reported on 30 LCS implants (28 pts); 25 were placed for rheumatoid arthritis, 1 for osteoarthritis, 1 for juvenile arthritis, and 1 for psoriatic arthritis. Complications included a deep infection, 3 delayed wound healings, and 5 malleolar fractures. Four failures (13%), secondary to varus instability, were converted to arthrodesis. One case was lost to follow-up and 4 died of unrelated causes. The remaining 21 implants (70%) were functioning well at an average of 6 years. Doets concluded that the LCS ankle yielded better results than earlier ankle replacements. The original LCS developers redesigned the implant to include a deepened talar trochlear groove and additional fixation fins with the goal of improving resistance to meniscal bearing subluxation and eliminating talar component subsidence. Initial 3-year results seem encouraging with 100% implant survival, 92.9% good-to-excellent clinical results, and no talar subsidence.47
The Agility Ankle is an uncemented 3-component, semi-constrained prosthesis that incorporates an arthrodesis of the tibiofibular syndesmosis. This is believed to provide increased surface area for ingrowth and added stability. It is porous coated chrome-cobalt with a polyethylene spacer secured within the tibial component essentially forming a 2-component implant. Circumferential cortical loading is achieved by resecting the articular surfaces of the malleoli as well as extending the posterior lip of the metal tray to contact the posterior cortex of the tibia. It permits approximately 60° of flexion/extension along with axial rotation. Early reports appear to be encouraging, however, the follow-up is limited with the longest report only 4.8 years. Of these first 100 ankle arthroplasties, only 5 revisions were performed: 1 for component fracture, 2 for loosening, 1 for malposition, and 1 was revised to an arthrodesis because of pain. There were no deep infections, 2 superficial infections, and 6 superficial peroneal nerve injuries. Ninety-seven percent of patients experienced some degree of pain relief; 55% reported no pain, 28% reported mild pain, 16% moderate pain. Non-union and delayed union of the syndesmosis were associated with poor results.26,49
The STAR (scandinavian total ankle replacment) prosthesis is an uncemented 3-component hydroxyapatite-coated implant with a polyethylene free-gliding disc between a flat tibial glide plate and a talar cap. Tibiopedal motion acts on a cylindrical basis between the talar component and the disc. Slight rotation is possible, limited by the malleoli. In 20 STAR implants (18 for osteoarthritis, 2 for haemochromatosis) with a 1- 4 year follow-up, there was only 1 case of subsidence which occurred in the first 3 months, and excellent or good results were reported in 18 of 20 ankles.45 A multicenter study of 131 STAR replacements (68 for rheumatoid arthritis, 63 for osteoarthritis) followed for at least 1 year had only 8 failures (6.1%; 5 were revised and 3 underwent an arthrodesis). Of the 71 patients with a 2-year follow-up there were only 5 failures (7%); 1 underwent arthrodesis and 4 were revised. It is encouraging that were no failures among patients followed for more than 2 years, including 5 patients followed for 7 years.50
Cement or No Cement?
Uncemented implants have better overall results than those inserted with cement.7,11,34,43,51 This may be explained by the fact that above 1-1.5 cm from the distal tibia, only soft fatty marrow is present which is not suited for cement.51 Compared to the talus, the softer bone structure of the distal tibia makes it particularly prone to loosening.52 In addition, pressurization of cement in the distal tibia is difficult due to the anatomy of the ankle joint. Cement may intrude into the back of the joint causing interference with motion.
Cementless or press-fit prostheses include the Scholz,11 Takakura,43 Alvine,49 LCS,34 and STAR45 prostheses. Tillmann, et al found that the uncemented prostheses performed more physiologically than the cemented implants.32 In one study, 7 of 30 (23%) uncemented ceramic ankle prostheses showed loosening at 4.1 years with 67% satisfactory results.43 Another report described 10 2-component uncemented ankle arthroplasties with no failures at 2 years and excellent clinical results.11 Buechel, et al reported a series of 23 uncemented 3-component ankle arthroplasties with only 1 revision of the polyethelene meniscus and excellent results at 2.9 years.34 Initial reports of the uncemented, 3-component STAR prosthesis show only 1 failure due to loosening.45
On the other hand, most reports concerning cemented prostheses have had higher aseptic loosening rates and poorer results. Demottaz, et al reported on 21 cemented arthroplasties (Mayo, TPR, Buchholz, Oregon, Waugh).14 With an average follow-up of 14.7 months, 88% of the prostheses exhibited progressive radiolucent lines and 10% showed loosening. Only 19% of the patients experienced complete pain relief and functional improvement was disappointing.14 Bolton-Maggs, et al reported that 33 of 62 prostheses loosened by 5.5 years and that only 13 results could be described as good.4 Takakura, et al reported that 85% of the 33 cemented ankle replacements showed loosening and sinking which began to appear as soon as 5 years postoperatively. They concluded that patients without osteoporosis undergoing ankle arthroplasty should receive cementless implants.43 Jensen and Kroner reported a 52% rate of loosening in a series of 23 cemented TPR prostheses and concluded that ankle arthroplasty should not be performed at all.52
In contrast to these discouraging reports on cemented prostheses, a recent report by Kofoed and Sorrensen showed a 75% survival of 52 cemented ankle arthroplasties for both osteoarthritis and rheumatoid arthritis followed for 14 years.53 The prostheses were either a congruent 2-component device or a 3-component, meniscal bearing device. In addition, 88.5% of their patients were able to perform activities of daily living without ankle pain.53 These results are comparable with survival rates of uncemented prostheses.
Radiographic lucency may be an important factor in predicting long-term survival of the ankle prosthesis. Some believe it to be a sign of impending failure and others are of the opinion that it is a benign finding. The incidence of lucent lines in hip and knee arthroplasty is much higher than the incidence of loosening, and many other factors have been implicated in the formation of these radiolucent lines.54-56 There are few studies regarding lucent lines in ankle replacements; however, significant radiolucencies adjacent to the tibial and talar components have been reported in up to 90% of cases.4,14,17,44 Although many ankle arthroplasties show evidence of radiographic loosening, other studies have shown that radiolucent lines are not invariably associated with clinical failures. Kitaoka and Patzer studied 204 Mayo total ankle arthroplasties and found radiographic evidence of loosening in 8% of talar components and 57% of tibial components; however, this did not correlate with their 35% clinical failure rate.21 Helm and Stevens studied 19 ICHL ankle replacements and found no correlation between the radiologic and clinical results.54 Almost all ankles had radiolucent lines and of the 8 patients with migration of the prosthesis, 6 had good or excellent results with no significant pain.54 Newton reviewed 50 arthroplasties and found no correlation between the presence of radiolucent zones and symptoms.22 Seventeen of the ankles had complete radiolucent zones but the patients did not experience pain.22 Pyevich, et al found that radiolucent lines were always apparent around the ankle implants within 2 years of implantation but were usually not progressive.26 In addition, after a few years, most areas of lysis were not progressive and were believed to indicate a lack of implant fixation.26
Complications other than loosening have been reported in total ankle arthroplasties.4,7,13,19,21,34 Delayed wound healing, nerve injuries, superficial and deep infections, wound dehiscence, fractures of the medial or lateral malleolus, and painful fibular impingement have been reported. Delayed wound healing was more often associated with an approach lateral to the extensor hallucis longus. Subsequent cases in which a more medial approach between the extensor hallucis longus and tibialis anterior was taken showed better primary wound healing.11,21 Painful fibular impingement may be caused by progressive hindfoot valgus and may require revision to an arthrodesis. Fracture of the medial or lateral malleolus may result in complete displacement of the ankle prosthesis -- a disastrous complication. In addition, muscle weakness about the ankle postoperatively, especially of the plantar flexors, has been suggested to cause abnormal gait patterns and ankle motion after total ankle replacement. This may result in the lack of pain relief and functional improvement reported in some studies.14
Little has been written on salvage procedures for failed ankle replacements. After removal of the prosthetic components, many times the bone stock is insufficient for re-implantation of a new prosthesis and arthrodesis is the only available option. Prior reports of Charnley’s compression technique, variations of the technique, intramedullary rod fixation, screw fixation and external fixation have been reported to lead to successful unions. Furthermore, revision using autograft and allograft to fill the voids left by the prosthesis and to aid in fusion has been successful.8,57-59
CONTRAINDICATIONS
Contraindications for implantation of a total ankle arthroplasty include talus avascular necrosis, Charcot joint, neurological problems (insensate foot), absence of muscular function in the distal leg, previous ankle arthrodesis with removal of the malleoli, severe tibiotalar malposition, acute or chronic infection, long-term steroid therapy or a fused ankle.7,8,22,23,54 Relative contraindications include a youthful, active individual with degenerative ankle joint disease, prior infection in the ankle, and vasculitic ulcers on the leg or foot. In addition, Lachiewicz believes that rheumatoid patients who are taking methotrexate should discontinue the drug at least 1 month prior to the surgery to minimize the risk of infection and delayed wound healing.5
INDICATIONS
A review of the literature finds that many authors advocate ankle replacement in patients with rheumatoid arthritis or other inflammatory arthritic disorders such as hemachromatosis, who have multi-joint involvement. Post-traumatic or primary osteoarthritis in older inactive patients is an indication, as well. Newton limited the replacement to those patients exhibiting osteoarthritis who had good ligament stability, reasonably normal anatomy, and a lack of significant varus or valgus deformity (less than 20°).22 Arthrodesis of the ankle in rheumatoid patients has been shown to accelerate degeneration of joints distal to the tibiotalar joint and for the aforementioned reasons, an arthroplasty may be a better alternative. Some authors additionally believe that an ankle arthroplasty is indicated in ankles with marked deformation and destruction, especially where subtalar or midtarsal joints are involved. These patients tend to lead lives of decreased activity. Most studies on younger patients with post-traumatic arthritic ankles report poor results, most likely due to the increased stress placed on the implant. An active older patient may be a poorer indication than an inactive younger patient.8,15,21,43 In addition, a patient should be considered a candidate for a total ankle arthroplasty if he or she cannot tolerate prolonged periods of immobilization as would be necessary after an ankle arthrodesis.
SURGICAL TECHNIQUES
Most surgeons use an anterior approach to the ankle joint.5,11,13,30,31,34 Kofoed recommends curving the incision in order to stay medial to the anterior tibial tendon and extensor hallus longus tendons. REFERENCE NEEDED All other structures, including the vessels and nerves remain lateral. Scholz described a similar anterior “c-type” incision between the extensor hallucis longus and anterior tibial tendons.11 Samuelson, et al used a posterior approach with excision of an os-calcis bone block including the Achilles tendon insertion.24 Buchholz, et al described a lateral approach with osteotomy of the fibula and division of the lateral ligaments.27
Elevating the leg and using a tourniquet may facilitate surgical exposure. Care should be taken to avoid the cutaneous branch of the superficial peroneal nerve which frequently crosses the field of dissection. Distraction of the ankle joint can be obtained manually with a small lamina spreader or with the aid of an external fixation apparatus. Synovial tissue and marginal osteophytes should be removed. Depending on the system and the implant used, talar and tibial resection guides are used to remove just enough of the bone to allow fit of the prosthesis. Trial reduction should be obtained, and once the final prosthesis is implanted, the anterior capsule and extensor retinaculum should be closed over a drain to prevent wound complications.
CONCLUSIONS
Patients with symptomatic ankle arthritis have many options available for treatment. To achieve good and lasting results, an ankle replacement must provide good alignment and stability. Review of the original literature on ankle arthroplasties reveals that the only indications for ankle replacement were for very disabled, rheumatoid arthritis patients. The success of early ankle arthroplasty may be largely attributable to the patients’ decreased level of activity and associated low stress on the arthroplasty. More recently, however, the reasonably active arthritic patient, with good bone stock and a foot that is plantigrade with little valgus or angulation within the ankle mortise, may be an ideal patient for one of the second-generation total joint arthroplasties. It is critical that the ankle implant bond solidly to bone, maintain the ankle’s anatomic alignment and shape, and allow easy implantation by orthopedic surgeons. Unfortunately, most attempts to design and implement a total ankle replacement in the past have been unsuccessful. They have failed to incorporate the biomechanical characteristics of the ankle joint, including the soft cancellous bone and the varying radii of curvature that allows complex motions to occur. The design of the implant should permit effective transfer of joint loads, be inherently stable, allow ease of surgical implantation/removal with minimum bone loss, have resistance to wear, creep, fatigue failure, and compressive and shear loading. In addition, it should be easy to manufacture at an affordable price.42 The newer second-generation uncemented, semi-constrained, porous- coated designs appear promising. Although acceptable long-term results are anticipated with these systems, further follow-up will be necessary to determined their effectiveness..
References
1. Demetriades L, Strauss E, Gallina J. Osteoarthritis of the ankle. Clin Orthop, 1998;349:28-42.
2. Mazur J, Schwartz E, Simon S. Ankle arthrodesis: long-term follow-up with gait analysis. J Bone Joint Surg 1979; 61-A:976-88.
3. Mann RA. Arthrodesis of the foot and ankle. In: Mann RA, Coughlin MJ, editors. Surgery of the foot and ankle. St.Louis:Mosby, 1993: 673-713.
4. Bolton-Maggs BG, Sudlow RA, Freeman MA. Total ankle arthroplasty. A long-term review of the London Hospital experience. J Bone Joint Surg 1985; 67-B(5):785-90.
5. Lachiewicz PF. Total ankle arthroplasty. Indications, techniques, and results. Orthop Rev 1994;23(4):315-20.
6. Hagen R. Ankle arthrodesis problems and pitfals. Clin Orthop 1986;202:152-162.
7. Bauer G, et al. Total ankle replacement. Review and critical analysis of the current status. Foot Ankle Surg 1996;2:119-126.
8. Stauffer RN. Total joint arthroplasty. The ankle. Mayo Clin Proc 1979;54(9):570-5.
9. Lance E, et al. Arthrodesis of the ankle, a follow-up study. J Bone Joint Surg 1971;53-A:1030.-
10. Johnson EW Jr, Boseker EH. Arthrodesis of the ankle. Arch Surg 1968;97:766-73.
11. Scholz KC. Total ankle arthroplasty using biological fixation components compared to ankle arthrodesis. Orthopedics 1987;10(1):125-31.
12. Nelissen RGHH, Doets HC, Meskers C. The value of ankle prostheses - a gait analysis approach. In: Kofoed H, editor. Current status of ankle arthroplasty. Berlin:Springer, 1998:72-77.
13. Wynn AH, Wilde AH. Long-term follow-up of the Conaxial (Beck-Steffee) total ankle arthroplasty. Foot Ankle 1992; 13(6):303-6.
14. Demottaz JD, Mazur JM, Thomas WH, Sledge CB, Simon SR. Clinical study of total ankle replacement with gait analysis. A Preliminary Report. J Bone Joint Surg 1979; 61-A:976-988.
15. McGuire MR, et al. Comparative analysis of ankle arthroplasty versus ankle arthrodesis. Clin Orthop 1988;226:74-81.
16. Hansen ST. Long-term complications of ankle arthrodesis - a survey. In:Kofoed E, editor. Current status of ankle arthroplasty. Beriln:Springer, 1998:129-131.
17. Lachiewicz PF, Inglis AE, Ranawat CS. Total ankle replacement in rheumatoid arthritis. J Bone Joint Surg 1984;66-A(3):340-3.
18. Unger AS,et al. Total ankle arthroplasty in rheumatoid arthritis: a long-term follow-up study. Foot Ankle 1988;8(4):173-9.
19. Stauffer RN, Segal NM. Total ankle arthroplasty: four years' experience. Clin Orthop 1981;160:217-21.
20. Kitaoka HB, Patzer HL, Ilstrup DM, Wallrichs SL. Survivorship analysis of the Mayo total ankle arthroplasty. J Bone Joint Surg 1994;.76(7):974-9.
21. Kitaoka HB, Patzer GL. Clinical results of the Mayo total ankle arthroplasty. J Bone Joint Surg 1996;78(11):1658-64.
22. Newton SED. Total ankle arthroplasty. Clinical study of fifty cases. J Bone Joint Surg 1982; 64(1):104-11.
23. Zych G, Mnaymneh W. Total ankle replacement. A new approach to an old problem. J Florida Med Assoc 1979.66(1):96-100.
24. Samuelson KM., Freeman MA, Tuke MA. Development and evolution of the ICLH ankle replacement. Foot Ankle 1982;3(1):32-6.
25. Spector E. Ankle Implants. Clin Pod 1984;1:225-235.
26. Pyevich M T, Saltzman CL, Callaghan JJ, Alvine FG. Total ankle arthroplasty: A unique design. Two to twelve-year follow-up. J Bone Joint Surg 1998;80-A(10):1410-1420.
27. Buchholz H, Engelbrecht E, Siegel A. Totale Sprunggelenkendoprothese Modell "St. Georg". Chirurg 1973;44:241-244.
28. Kirkup J. Richard Smith ankle arthroplasty. J R Soc Med 1985;78(4):301-4.
29. Dini AA, Bassett FH III. Evaluation of the early result of Smith total ankle replacment. Clin Orthop 1980;146:228-230.
30. Newton SE. An artificial ankle joint. Clin Orthop 1979;142:141-5.
31. Waugh TR, Evanski PM, McMaster WC. Irvine ankle arthroplasty. Prosthetic design and surgical technique. Clin Orthop 1976;114:180-4.
32. Kofoed H. Current status of ankle arthroplasty. Berlin: Springer, 1998:157-.
33. Barnett CH, Napier J. The axis of rotation at the ankle joint in man, its influence upon the talus and mobility of the fibula. J Anat 1952;86: -.
34. Buechel FF, Pappas MJ, Iorio LJ. New Jersey low contact stress total ankle replacement: biomechanical rationale and review of 23 cementless cases. Foot Ankle 1988;8(6):279-90.
35. Lundberg A., et al. The axis of rotation of the ankle joint. J Bone Joint Surg 1989;71-B:94-99.
36. Close J. Some applications of functional anatomy of the ankle joint. J Bone Joint Surg 1956;A-38:761-781.
37. Murray MP, Drought AB, Kory RC. Walking patterns of normal men. J Bone Joint Surg 1964;46-A:335-360.
38. Mann RA. Biomechanics of the foot and ankle. In: Mann RA, Coughlin MJ, editors. Surgery of the foot and ankle. St. Louis:Mosby, 1993:3-45.
39. Stauffer R, Chao E, Brester R. Force and motion analysis of the normal, diseased and prosthetic ankle joint. Clin Orthop 1977; 127:189-196.
40. Burge P. Evans M. Effect of surface replacement arthroplasty on stability of the ankle. Foot Ankle 1986;7(1):10-7.
41. Pappas M, Buechel F, DePalma AF. Cylindrical total ankle joint replacement: surgical and biomechanical rationale. Clin Orthop 1976;118: 82-92.
42. Lewis G. The ankle joint prosthetic replacement: Clinical performance and research challenges. Foot Ankle Int 1994;15(9):471-476.
43. Takakura Y, Sugimoto K, Tamai S, Masuhara K. Ankle arthroplasty. A comparative study of cemented metal and uncemented ceramic prostheses. Clin Orthop 1990;252:209-16.
44. Herberts P, et al. Endoprosthetic arthroplasty of the ankle joint. A clinical and radiological follow-up. Acta Orthop Scand 1982; 53(4):687-96.
45. Kofoed H, Danborg L. Biological fixation of ankle arthroplasty. Foot 1995; 5(1):27-31.
46. Neufeld S, Lee TH. Current Management of Talar Neck Fractures. Am J Orthop - Supplement, 1999(Jan).
47. Buechel FF, Pappas MJ. Survivorship and clinical evaluation of cementless, meniscal-bearing total ankle replacments. Seminars in Arthroplasty 1992;3(1):43-50.
48. Doets HC. The low contact stress/Buechel-Pappas total ankle prosthesis. In: Kofoed H, editor. Current status of ankle arthroplasty. Berlin:Springer, 1998:29-33.
49. Alvine F. Total ankle arthroplasty: new concepts and approach. Contemp Orthop 1991;22:397-403.
50. Schernberg E. Current results of ankle arthroplasty - European multi-center study of cementless ankle arthroplasty, In:Kofoed H, editor. Current status of ankle arthroplasty. Berlin:Springer, 1998:41-46.
51. Kofoed H. Cylindrical cemented ankle arthroplasty: a prospective series with long- term follow-up. Foot Ankle Int 1995;16(8):474-9.
52. Jensen N, Kroner K. Total ankle joint replacement: A clinical follow up. Orthopedics 1992;15:236-239.
53. Kofoed H, Sorensen TS. Ankle arthroplasty for rheumatoid arthritis and osteoarthritis: prospective long-term study of cemented replacements. J Bone Joint Surg 1998;80-B(2):328-32.
54. Helm R, Stevens J. Long-term results of total ankle replacement. J Arthroplasty 1986;1(4):271-7.
55. Silverton C., Rosenberg A. Long-term results in total knee arthroplasty. In: Callaghan JJ, et al, editors. orthopedic knowledge update hip and knee reconstruction. Rosemont:American Academy of orthopedic Surgeons, 1995:303-317.
56. Whirlow J, Rubash H. Aseptic loosening in total hip arthroplasty. In: Callaghan JJ, et al, editors. orthopedic knowledge update hip and knee reconstruction. Rosemont,American Academy of orthopedic Surgeons, 1995:147-157.
57. Carlsson AS, Montgomery F, Besjakov J. Arthrodesis of the ankle secondary to replacement. Foot Ankle Int 1998;19(4):240-5.
58. Groth H, Fitch H. Salvage procedures for complications of total ankle arthroplasty. Clin Orthop 1987;224:244-250.
59. Charnley J. Compression arthrodesis of the ankle and shoulder. J Bone Joint Surg 1951;33-B:180-191.
60. Evanski PH, Waugh TR. Management of arthritis of the ankle. An alternative of arthrodesis. Clin Orthop 1977;122:110-5.
TABLE I: Short Term Follow-up
|
Author |
Prostheis design |
Diagnosis/ Number |
Average f/u |
Survival rate |
| |||
|
|
SA(56) |
|
|
| ||||
|
Stauffer & Segal19 |
Mayo |
OA(3) |
1.9 yrs |
93% |
| |||
|
|
|
RA(43) |
|
|
| |||
|
|
|
|
|
|
| |||
|
Lachiewicz, et al17 |
Mayo |
RA(15) |
3 yrs |
100% |
| |||
|
|
|
|
|
|
| |||
|
Herberts, et al44 |
ICLH |
RA(14) |
3 yrs |
86% |
| |||
|
|
|
OA(7) |
|
|
| |||
|
|
|
|
|
|
| |||
|
|
|
RA(10) |
|
60% |
| |||
|
Newton22 |
Newton |
OA(34) |
3 yrs |
38% |
| |||
|
|
|
SA(6) |
|
0% |
| |||
|
|
|
|
|
|
| |||
|
Kofoed45 |
STARR |
OA(4) |
2.5 yrs |
90% |
| |||
|
|
|
SA(16) |
|
|
| |||
|
|
|
|
|
|
| |||
|
Dini & Bassett29 |
Smith |
RA(5) SA(16) |
2.5 yrs 2.1 yrs |
80% 75% |
| |||
|
|
|
|
|
|
| |||
|
Evanski & Waugh60 |
Irvine |
RA(5) OA(23) |
9 months |
93% |
| |||
|
|
|
|
|
|
| |||
|
Buechel et al34 |
LCS |
RA(6) OA(4) SA(13)
|
2.9 yrs |
100 |
| |||
(SA=2nd arthrosis, OA=osteoarthritis, RA=Rheumatoid)
TABLE II: Longer term reports
|
Author |
Prosthesis Design |
Diagnosis/nnumber |
Average Follow-up |
Survival Rate |
|
Jensen & Kroner52 |
TPR |
RA(21) OA(2) |
4.9 yrs |
48% |
|
|
|
RA(125) |
5 yrs |
79% |
|
Kitaoka, et al20 |
Mayo |
SA(65) |
10 yrs |
65% |
|
|
|
OA(14) |
15 yrs |
61% |
|
|
|
|
|
|
|
Kitaoka & Patzer21 |
Mayo |
RA(96) |
9 yrs |
64% |
|
|
|
SA(64) |
|
|
|
|
|
OA(8) |
|
|
|
|
|
|
|
|
|
Wynn, et al13 |
Beck-Steffee |
RA(18) |
2 yrs |
73% |
|
|
|
SA(12) |
5 yrs |
40% |
|
|
|
|
10 yrs |
10% |
|
|
|
|
|
|
|
Helm, et al54 |
ICLH |
RA(19) |
4.5 yrs |
83%
|
|
Bolton-Maggs, et al4 |
ICLH |
RA(34) OA(13) SA(15)
|
5.5yrs |
47%
|
|
Pyevich,et al26 |
Agility |
RA(19) |
4.8 yrs |
94% |
|
|
|
OA(19) SA(47) |
|
|
|
|
|
|
|
|
|
Unger, et al18 |
Mayo |
RA(23) |
5.6 yrs
|
65% |
|
Takura43 |
Takakura uncemented- ceramic |
RA(9) OA(18) SA(3) |
4.1 yrs |
77% |
|
|
|
|
|
|
|
|
Takakura cemented |
OA(20) RA(11) SA(2)
|
8.8yrs (metal) 6.7yrs(ceramic) |
15% |
|
Kofoed51 |
Cylindrical 2- piece cemented |
RA(13) OA(15) |
12yrs |
70% |
|
|
|
|
|
|
|
Buechal & Pappas47 |
LCS |
OA(8) RA(8) SA(22) |
10yr |
94.75% |
|
|
|
|
|
|
|
Kofoed & Sorensen53 |
2 piece (early) 3 piece (later) |
OA(25) RA(27) |
14yrs |
72.7% (oa) 75.5% (ra)
|
(SA=2nd arthrosis, OA=osteoarthritis, RA=Rheumatoid)