Total Ankle Arthroplasty

Development of Total Ankle Arthroplasty Systems

Since the first report of TAA in the 1970s, many TAA systems have been introduced. The first-generation, cemented, constrained designs were very stable but required extensive bony resection for implantation and frequently failed because of loosening, subsidence, and extensive osteolysis. Second-generation, less constrained implants required less bone resection and did not require cement fixation; because shear forces and torsion at the bone-prosthesis were reduced, loosening was less frequent. However, increased polyethylene wear and failure compromised the stability of the components, often leading to painful impingement and subluxation or complete dislocation of the components. Contemporary, third-generation, semi-constrained total ankle systems consist of three components: a metallic baseplate that is fixed to the tibia, a domed or condylar-shaped metallic component that resurfaces the talus, and an ultrahigh-molecular-weight polyethylene bearing surface interposed between the tibial and talar components. Systems in which the polyethylene component is locked into the baseplate are often referred to as “two-piece” or “fixed-bearing” designs, whereas those with the polyethylene component not attached to the baseplate are called “three-piece” or mobile or meniscal bearing systems. Currently, there are over 50 different TAA systems in use worldwide, most of which have not been approved for use by the US Food and Drug Administration (FDA). There is no convincing evidence of the clear superiority of one design over another among the currently available systems; choice of a prosthesis depends on individual patient factors, institutional factors, and the surgeon’s training and experience.

Design Rationale

The development of an implant system that mimics the normal anatomy and biomechanics of the ankle and achieves success rates similar to those of hip and knee arthroplasty has been hampered by several anatomic features of the ankle joint: (1) the ankle has significantly less contact area between joint surfaces than the hip or knee; (2) the ankle experiences 5.5 times body weight with normal ambulation, compared with three times body weight at the knee; and (3) the articular cartilage surface of the ankle is uniformly thinner than that of the knee.

The biomechanical concepts that have resulted in the most recent generation of ankle arthroplasties are somewhat beyond the scope of a surgical-oriented textbook; however, the number and variety of implants on the market demand a familiarity of basic principles of component design.

The design of TAA systems continues to evolve as we learn more about the modes of failure from longer-term studies. First-generation implants had high rates of osteolysis, implant loosening, tibial and talar bone loss, and wound complications. Second-generation designs used porous metal-backed surfaces to improve osseous integration; replaced the tibiotalar, talofibular, and medial-malleolar-talar articulations; and/or improved stability by fusing the syndesmosis. Complications related to nonunion of the syndesmosis, polyethylene wear, and migration and impingement of the implants led to a number of modifications in third- and fourth-generation designs: less bone resection, more bony ingrowth, retention of ligamentous support, and anatomic balancing.

Fixed-Bearing VersusMobile-Bearing Designs

Most modern implants fall into two basic groups: those with a mobile polyethylene component that has the ability, at least in theory, to move under the tibial component to adapt to changes in joint forces ( Fig. 10.1A ) and those with the polyethylene component fixed rigidly to the tibial component ( Fig. 10.1B ). Mobile-bearing designs are used most commonly in Europe and have a long history of outcomes from which they can be evaluated. Another theoretical advantage of these designs is the “forgiveness” of the implant, which allows small variances in alignment to be compensated for by a reorientation of the prosthesis to accommodate the joint forces. The ability of the polyethylene component to move should, in theory, keep the articulation between the talar component and the polyethylene component more congruent and less likely to lead to edge load and advanced wear. In experienced hands, however, there is a question of how much the polyethylene component actually moves under the tibia. Barg et al. found very little anteroposterior movement of the talar component under the tibia in follow-up radiographs of a three-component, mobile-bearing design. They noted that the prosthesis functioned largely like a fixed-bearing design, but with a possible advantage of allowing an individualized position of the polyethylene insert in response to individual soft-tissue loads produced by different ankle joint configurations. Aside from the STAR ankle implants ( Fig. 10.2 ), implants approved for use in the United States are fixed-bearing designs.

Proponents of fixed-bearing designs suggest that the normal ankle joint, as opposed to the knee, has a more stable central axis of motion and less need for an additional degree of freedom of motion. Backside polyethylene wear against the tibial component is a major concern with mobile-bearing designs and less with fixed designs. Attention to detail in the proper alignment of the prosthesis along the mechanical axis of the limb has been suggested to prevent excessive wear.

A multicenter study by Gaudot et al. comparing fixed-bearing with mobile-bearing implants found no significant differences in accuracy of positioning, clinical and radiographic mobility, or morbidity. In a more recent randomized trial, Queen et al. also found no clinically meaningful differences in outcomes between the two implant types when examining gait mechanics and pain at 1-year follow-up. The most recent American Orthopaedic Foot and Ankle Society (AOFAS) scores were higher for patients with fixed-bearing implants than for those with mobile-bearing implants, and radiolucent lines and subchondral cysts were less frequent. In contrast, Currier et al. analyzed 70 retrieved ankle implants, most commonly for loosening and polyethylene fracture. Loosening occurred more frequently in fixed-bearing designs than in mobile-bearing designs. Lundeen et al. in a study of patients with a third-generation mobile-bearing TAA concluded that the multiplanar articulation in mobile-bearing TAA may reduce excessively high peak pressures during tibial and talar motion, which may have a positive effect on gait pattern, polyethylene wear, and implant longevity.

Alignment

Currently available implant systems are designed to be placed along the mechanical axis of the limb and depend on satisfactory alignment above and below the ankle joint. The most common method of obtaining correct alignment is an external alignment jig, using intraoperative fluoroscopy to judge the alignment. At least one system uses an intramedullary alignment rod ( Fig. 10.3 ). One innovation is a tomography-produced customized cutting jig, such as those available for knee arthroplasty. Patient-specific instrumentation has been successfully used in arthroplasty of other joints (e.g., knee, shoulder, hip), but has rarely been described in TAA. Cadaver and clinical studies have shown this instrumentation to provide accuracy and reproducibility among multiple surgeons, implants, and facilities. Hamid et al. did not find a difference in postoperative alignment with patient-specific instrumentation or standard referencing; however, use of patient-specific instrumentation significantly decreased operative time and, thus, costs. Another fairly recent system uses a lateral approach to more accurately reproduce the center of rotation of the ankle and minimize bone resection ( Fig. 10.4 ).

Preoperative Evaluation

A thorough understanding of the patient’s medical history and review of systems are important in the decision-making process and the consideration of the patient for TAA. Systemic diseases such as diabetes, inflammatory arthritis, chronic obstructive pulmonary disease, and peripheral vascular or heart disease may adversely affect the outcome and healing of the incision. Conditions such as sleep apnea, malnutrition, vitamin D deficiency, and depression are associated with decreased functional outcomes and poor results. We do not perform elective TAA in active smokers. It must be clear that the ankle joint is indeed the cause of the patient’s primary complaint. Many patients have adjacent joint disease that might also need to be treated before or at the time of surgery. Selective injections of lidocaine are helpful in accurately identifying the painful pathologic process. A complete assessment of the limb is important. A lumbar spine pathologic process with sciatica and radicular lower extremity pain or degenerative disease of the hip or knee may cause a change in the management plan. Patients with combined knee and ankle arthritis and deformity often are best managed by correction of the knee deformity first, followed by the ankle replacement.

A thorough evaluation of the neurovascular status of the limb is essential, and any concerns should prompt a formal vascular evaluation. The patient’s gait should be evaluated for limp, and any alterations of knee or hip motion to compensate for the arthritic ankle and limb-length difference should be assessed. The standing evaluation is important for clinical assessment of ankle and hindfoot alignment. Is there a supramalleolar deformity that must be corrected? Is the hindfoot well aligned, or is there a component of varus or valgus? Clinical assessment of the gastrocsoleus complex and the Achilles tendon is important. The Silfverskiöld test for selective gastrocnemius tightness might reveal a contracture that is independent of ankle range of motion and that must be released intraoperatively. Coetzee and Castro demonstrated the inability to distinguish true range of motion of the tibiotalar joint on clinical examination and proposed a radiographic evaluation of the range of motion preoperatively. Nonetheless, an idea of sagittal plane range of motion is important. Overall hindfoot motion is important as well. A stiff, arthritic hindfoot might be the difference between choosing arthroplasty or arthrodesis. Strength testing of the leg motor groups should not reveal major deficits that would impair the outcome. The anterior skin should be stable and without lesions that would impair the healing of the surgical incision.

At a minimum, standing radiographs of the ankle in anteroposterior, lateral, and mortise views should be obtained. Any suspicion of proximal limb malalignment should be evaluated with standing lower extremity films. Because an accurate assessment of the alignment of the hindfoot is not possible with standing films of the ankle, Frigg et al. described a hindfoot alignment view ( Fig. 10.5 ) that gives a better appreciation of overall alignment and helps to determine if an adjunctive procedure is needed to improve the alignment of the foot distal to the ankle joint.

Although coronal plane deformities usually are the focus of radiographic evaluation before TAA, sagittal plane deformities have been shown to alter the mechanics and joint reaction forces more than coronal plane deformities. Several methods for evaluating the lateral position of the talus on radiographs have been described. Veljkovic et al. described a sagittal talar position measurement that they named the lateral tibial station (LTS); the LTS defines the center of rotation of the talus as related to the anatomic tibial axis. In a study of 82 ankles, they showed that this measurement can be reliably obtained on preoperative weight-bearing lateral radiographs; the mean LTS measurement was 1.7 mm (normal range, 0.8076 to 3.1496 mm). The value of the LTS in evaluating ankle pathology remains to be established.

Radiographic evaluation should include assessment of the quality of the bone stock, coronal plane alignment of the ankle with supramalleolar deformities or joint incongruencies, the presence of osteophytes requiring removal, adjacent joint arthritis or malalignment that requires correction, calcaneal pitch angle as a predictor of gastrocsoleus contracture, and the presence of major cysts or defects that will need grafting.

The effect of bone mineral density (BMD) on outcomes of TAA has not been clearly delineated. Dual-energy x-ray absorptiometry (DEXA) scanning generally is not obtained to evaluate BMD before TAA because of the additional time and costs involved. CT scans are, however, often obtained before TAA, and Hounsfield units measured on CT have been shown to correlate well with DEXA T-scores. In a study of 198 ankles, BMD (Hounsfield units) was measured on preoperative CT scans and compared to outcomes at 2.4-year follow-up. After controlling for age, gender, and BMI, only tibial Hounsfield units of less than 200 were significantly associated with periprosthetic fracture. Prophylactic internal fixation of the medial malleolus may be considered for patients with tibial Hounsfield units of less than 200 as measured on CT scans. Severe osteoporosis often is listed as a contraindication to TAA, but newer implant designs and surgical techniques (e.g., noncemented fixation and patient-specific implants) have overcome some of the problems of TAA in osteoporotic bone, and outcomes have improved, especially in older, more sedentary patients with osteoporosis. DEXA scanning or BMD evaluation on CT scans is indicated in patients at risk for osteoporosis.

Indications

Although degenerative, inflammatory, and posttraumatic arthritic conditions of the ankle are the primary indications for TAA, there is little clinical evidence on which to base more specific indications and contraindications. The ideal candidate for ankle arthroplasty has been described as an older, thin, low-demand individual with minimal deformity and retained ankle range of motion. These descriptions, however, are vague and controversial. Some have defined “young” as younger than 50 years of age and “thin” as weighing less than 200 lb, but there is no clinical evidence to support these classifications. Commonly cited contraindications to TAA include age younger than 50 years, history of poor patient compliance, heavy industrial labor, heavy smoking, uncontrolled diabetes with neuropathy, significant ankle instability, angular deformity of more than 10 to 15 degrees, vascular insufficiency, obesity (over 250 lb), significant bone loss, osteonecrosis, and active or previous infection. More recently, however, a number of these contraindications have been questioned. Demetracopoulos et al. reviewed outcomes in 395 consecutive patients according to age (younger than 55, 55 to 70, and older than 70 years of age) and found no differences in pain relief or physical outcomes or in the incidences of wound complications, reoperations, or revisions. Tenebaum et al. compared clinical and gait outcomes in patients older than 70 years to those in patients between 50 and 60 years of age; improvements were equivalent in the two groups. Good results have been reported in both obese patients and diabetic patients (see sections on Obesity and Diabetes), as well as those with angular deformities of more than 20 degrees (see section on Deformity Correction).

Because TAA is generally considered a “motion-sparing” procedure rather than a “motion-producing” procedure, its use in individuals with ankle stiffness has been limited; however, Brodsky et al. compared outcomes to preoperative range of motion and found that, although a low preoperative range of motion was predictive of overall lower physical function, patients with stiff ankles had clinically greater improvements in function. These findings suggest that TAA can offer clinically meaningful improvements in gait function and should be considered for patients with end-stage tibiotalar arthritis, even with limited sagittal range of motion.

Reports are variable concerning the outcomes of TAA in arthritis of different etiologies; most studies, however, have not found etiology to significantly affect implant survival. Bennett et al. reviewed the outcomes of 173 TAAs of differing etiologies (osteoarthritis, rheumatoid arthritis, pilon fracture, ankle fracture, and posttraumatic arthritis without previous fracture) and found no major differences in any of the reported outcomes at 2-year follow-up. Whatever the etiology of the arthritis, severe involvement does not necessarily portend a poor result. In a group of 124 patients, Chambers et al. found no differences in Short-Form 36 (SF-36) scores regardless of Kellgren-Lawrence grade of arthritis severity. Those with the most severe arthritis (Kellgren-Lawrence grade 4) had the most improvement in all domains of the Foot and Ankle Outcome Score and were more satisfied (91%) with their outcomes than all other groups (50%); 94% of patients with severe arthritis thought their quality of life had been improved compared to 47% in those with severity grades of less than 4. Although radiographic severity is an important factor that should be considered, it does not appear to contradict TAA.

Total Ankle Arthroplasty or Ankle Arthrodesis for Ankle Arthritis

Ankle arthrodesis (see Chapter 11 ) has long been the gold standard for the surgical treatment of moderate to severe ankle arthritis. It is, therefore, reasonable to ask if there is a compelling reason to pursue TAA as a treatment option for patients with ankle arthritis. Although the patient satisfaction rate after ankle arthrodesis is fairly high, there are certainly circumstances in which arthrodesis might not be the best procedure, including preexisting subtalar or other hindfoot arthritis, contralateral hindfoot or ankle arthritis, and hip or knee impairment such that motion through the ankle joint may be beneficial to the overall limb and patient function.

No level I studies have directly compared the two procedures, and reports in the literature are contradictory ( Table 10.1 ). The most recent reports seem to favor TAA with the latest-generation implants over arthrodesis, citing better functional outcomes, fewer complications, and better patient satisfaction. Some gait studies have noted no difference in gait patterns after arthroplasty and arthrodesis, whereas others report more nearly normal gait and better walking on uneven surfaces after arthroplasty; gait appears to be improved by either procedure. Daniels et al. compared intermediate outcomes (mean 5.5-year follow-up) of arthrodesis (107 patients) and arthroplasty (281 patients) in a diverse cohort of patients and found comparable clinical outcomes; however, rates of reoperation and major complications were higher after ankle arthroplasty. Norvell et al. found that ankle-specific adverse events were infrequent and only weakly associated with operative procedure. Careful patient selection is mandatory for the success of either of these procedures in the treatment of ankle arthritis.

Outpatient Total Ankle Arthroplasty

The success of outpatient total hip, knee, and shoulder arthroplasty, with no compromise of safety, satisfaction, or results, has prompted many surgeons to move TAA to an ambulatory surgical center. Although there are relatively few reports of the outcomes of this procedure, available reports cite good outcomes with few complications, in addition to cost savings and patient satisfaction. Two reviews of complications in outpatient TAA reported overall complication rates of 5% and 15%; most complications were minor, and readmissions and reoperation were infrequent. In a comparison of outpatient and inpatient TAA, Gonzales et al. found a 13% cost savings in the outpatient group, with a low complication rate and high patient satisfaction. The most important aspect of outpatient TAA, as with any outpatient arthroplasty, is careful patient selection. An algorithm developed by our surgeons for selecting patients for outpatient total joint arthroplasty is a helpful guide ( Fig. 10.6 ). There is limited evidence to determine specific patient factors that may preclude outpatient TAA surgery; however, experience in arthroplasty of the foot and ankle and other joints has identified various patient factors that are associated with perioperative complications and a longer length of hospital stay, such as elevated HbA1c level, obesity, hypoalbuminemia, age greater than 64, increased operating room time, American Society of Anesthesiologists (ASA) score of 2 or higher, and presence of comorbidities. Taylor and Parekh formulated a detailed algorithm specifically for outpatient TAA based on a number of patient and procedure characteristics ( Fig. 10.7 ).

In addition to selecting appropriate patients, appropriate use of a multimodal perioperative and postoperative pain management protocol is essential. The choice of pain management protocols has shifted away from oral and intravenous administration of narcotics to peripheral nerve blocks and indwelling catheters as better options. Continuous popliteal sciatic nerve block is commonly used for major foot and ankle reconstructions. Liposomal bupivacaine (Exparel; Pacira Pharmaceuticals, Parsippany, NJ) administered as a periarticular injection has been shown to have marked benefits in postoperative pain control. Mulligan et al. compared liposomal bupivacaine to popliteal sciatic nerve block for TAA and found no significant differences between the groups with regard to complications, emergency department visits, readmissions, reoperations, visual analog scale (VAS) pain score, and narcotic use.

The management of perioperative blood loss, especially postoperative hemarthrosis, is another modifiable factor affecting patient recovery, complication rates, and costs. Tranexamic acid (TXA) has been shown to be effective for reducing perioperative blood loss with arthroplasty of other joints, but little information is available for determining its effectiveness in TAA. Nodzo et al. described 50 patients with uncemented TAA, 25 of whom received TXA and 25 who did not. Drain output and change in preoperative to postoperative hemoglobin levels were significantly less and the overall wound complication rate was lower in patients who received TXA. We do not routinely use TXA for TAA.

Considerations for Adjunctive Procedures

Deformity Correction

Osteoarthritic ankles considered for arthroplasty should have minimal periarticular deformity, or this deformity should be correctable with osteotomy or arthrodesis.

Determination of the site of the deformity is mandatory. Bonasia et al. characterized deformities as varus or valgus, incongruent or congruent ( Table 10.2 ). In the valgus ankle and hindfoot, the following procedures should be considered: medial displacement osteotomy of the calcaneus (see Technique 83.7 ), Cotton osteotomy of the medial cuneiform or selective arthrodesis of the medial midfoot (see Techniques 83.8 and 85.5 ), subtalar arthrodesis with or without talonavicular arthrodesis (see Technique 85.6 ), posterior tibial tendon reconstruction with tendon transfer (see Technique 83.2 ), and closing wedge osteotomy of the distal tibia (see Technique 58.10 ). Demetracopoulos et al. evaluated 80 patients with preoperative valgus deformities of at least 10 degrees (average of 15 degrees). After TAA, the average postoperative deformity was 1.2 degrees, with significant improvements in VAS, SF-36, American Orthopaedic Foot and Ankle Society (AOFAS), and Short Musculoskeletal Function Assessment (SMFA) scores. The authors concluded that correction of coronal alignment could be obtained and maintained in patients with moderate-to-severe preoperative valgus malalignment. Lee et al. compared intermediate and long-term outcomes of TAA in 144 ankles with preoperative varus, valgus, or neutral alignment. Outcomes similar to those in ankles with neutral alignment were obtained in ankles with varus or valgus malalignment of up to 20 degrees when neutral alignment was achieved with TAA.

TABLE 10.2

Ankle Joint Pathologies That Include Distal Tibial Articular Surface Malalignment, Talar Tilt due to Ligamentous Instability, or Both

Modified from Bonasia DE, Dettoni F, Femino JE, et al: Total ankle replacement: When, why, and how?, Iowa Orthop J 30:119–130, 2010.

Deformity Type	Abnormal Angles
Varus tibial deformity-congruent joint	Increased LDTA, CORA at the level of tibial articular surface, normal tibial-talar angle
Valgus tibial deformity-congruent joint	Increased LDTA, CORA at the level of tibial articular surface, normal tibial-talar angle
Varus tibial deformity-incongruent joint	Decreased LDTA, CORA at the level of tibial articular surface, tibial-talar angle >10 degrees
Valgus tibial deformity-incongruent joint	Increased LDTA, CORA at the level of tibial articular surface, tibial-talar angle >10 degrees
Incongruent joint	Normal LDTA, tibial-talar angle >10 degrees

ADTA , Anterior distal tibial angle, sagittal plane—increased ADTA represents recurvatum deformity; CORA , center of rotation of angulation, at or proximal to joint line; LDTA , lateral distal tibial angle, coronal plane—decreased LDTA represents varus deformity; T-T angle—angle formed by tibial and talar articular surfaces: >10 degrees = incongruent joint.

For the varus ankle, procedures to consider include deltoid ligament release or sliding osteotomy of the medial malleolus, opening wedge osteotomy of the distal tibia (see Technique 11.1 ), Dwyer closing wedge osteotomy of the calcaneus (see Technique 87-11 ), dorsiflexion osteotomy of the first metatarsal (see Technique 84-19 ), and subtalar, double, or triple arthrodesis (see Chapter 85 ).

Varus deformity of the distal tibia above the level of the joint is best treated with supramalleolar osteotomy. Varus deformity of the tibial plafond at the joint from erosion of the medial malleolus or medial subchondral bone can be corrected by accurate placement of the tibial cut. Joo and Lee reported satisfactory clinical and radiographic outcomes in patients with moderate and severe varus deformities similar to those in patients with neutral alignment when postoperative neutral alignment was obtained, and special care was taken to correct causes of the varus malalignment with additional procedures.

For the varus unstable ankle with deformity below the level of the joint, sometimes an osteotomy of the hindfoot is required ( Fig. 10.8 ). If instability persists intraoperatively, a lateral ligament reconstruction should be done. Judicious release of the deltoid ligament, especially the deep deltoid ligament, may be wise in this setting. To avoid devascularization of the talus by injury to the deltoid branch of the posterior tibial artery, a sliding osteotomy of the medial malleolus has been described, with or without fixation. Reddy et al. reported correction of coronal plane deformity without osteotomy in ankles with an average of 18 degrees of varus. Deltoid release was necessary for all ankles with more than 18 degrees of varus deformity, and all ankles with more than 25 degrees of varus developed recurrent deformity. Hobson et al. suggested that TAA could be safely done with up to 30 degrees of coronal plane deformity. In their short-term follow-up of 103 patients with severe varus deformities, Sung et al. found that those with more than 20 degrees of varus deformity had outcomes similar to those with varus deformities of less than 20 degrees, with no significant differences in postoperative complications or implant failures. Adjunctive procedures, such as osteotomy, ligament release or lengthening, and tendon transfers, were done as needed. In the comparison study of Lee et al., adjunctive procedures were required in 71% of ankles with varus deformities, in 56% of those with valgus deformities, and in 39% of those with neutral alignment. Percutaneous Achilles tendon lengthening and release of the medial deltoid ligament were the most frequently done concomitant procedures; calcaneal osteotomy was done in five ankles (three in the varus group and two in the valgus group).

Tan and Myerson divided varus ankle deformities into anatomic levels and described procedures for correction at each level. For extraarticular deformity above the ankle joint, they recommended a medial opening wedge osteotomy or, for severe ankle arthritis, a dome osteotomy. With a medial opening wedge osteotomy, they recommended a staged procedure in which total ankle replacement is done later. The dome osteotomy is useful for multiplanar supramalleolar deformity and can usually be done simultaneously with replacement ( Fig. 10.9 ). For deformity at the level of the ankle joint and a congruent joint, a “neutralizing” distal tibial cut may be all that is needed for realignment. A wedge of the distal tibia is removed with minimal bone resection at the eroded medial plafond and a larger resection at the lateral plafond. For a severely tilted talus, additional procedures are required, including the removal of osteophytes from the lateral gutter and a lateral ankle stabilization procedure. Medial-side releases of the deltoid and posterior tibial tendon have been described, but Tan and Myerson recommended a lengthening medial malleolar osteotomy, as described by Doets et al. ( Fig. 10.10 ), rather than soft-tissue releases, because it allows controlled lengthening of the medial side of the ankle and provides reliable bony healing. With more severe varus tilt of the talus with a markedly dysplastic medial malleolus and incongruent joint, a useful alternative osteotomy is the medial tibial plafondplasty, which is done as a separate, staged procedure before ankle replacement. Residual heel varus that remains after component implantation can be corrected with a lateralizing calcaneal osteotomy. Combined deformities are generally best treated with correction of the deformities, followed by a staged ankle arthroplasty. Supramalleolar deformities are corrected first, followed by correction of hindfoot and forefoot varus and any ligamentous reconstruction needed.