Total Ankle Arthroplasty

Scope note. This page is the dedicated total ankle arthroplasty (TAR) procedure guide — implants, patient selection, technique and outcomes. For the full spectrum of ankle arthritis management (conservative care, arthrodesis, distraction arthroplasty, and the TAR-versus-fusion decision in context), see ankle arthritis.

What it is. Total ankle arthroplasty has evolved from a rarely performed procedure with poor outcomes to a viable motion-preserving alternative to ankle arthrodesis for end-stage ankle osteoarthritis with disability despite conservative management. It preserves ankle motion and improves objectively measured gait, and it reduces the need for later subtalar fusion, though by about two percentage points at five years rather than dramatically. Patient selection and surgical precision decide whether it succeeds.

Survivorship, stated precisely. The registries on this page do not support parity with hip and knee replacement. The National Joint Registry linkage study for England and Wales (PMID 37345846) reports 90.2% survival at five years; the Norwegian registry (PMID 17966015), the early registry-era benchmark, reports 89% at five years falling to 76% at ten, and its authors describe the revision rate as acceptable against other ankle series but high compared with total knee and hip arthroplasty. Every major registry finds the revision burden higher than for hip and knee replacement, so quote five-year figures around 90% and a ten-year figure closer to three-quarters, and do not claim the ankle has caught up. Registry data (AOANJRR, Norwegian, NJR) confirm the improvement over earlier designs in real-world practice.

Who has it. Mean age at surgery is 60 years, with 40-75 the optimal range. Post-traumatic arthritis accounts for 70% of cases, primary osteoarthritis 20% and inflammatory arthropathy 10%; the sexes are equally represented in post-traumatic disease, while females predominate in inflammatory arthritis. The intended patient is sedentary to moderately active, not the high-impact athlete or heavy labourer.

The primary indication is end-stage ankle arthritis in an appropriate candidate who has failed extensive conservative management. The best candidate profile: age 55, BMI 26, unilateral post-traumatic OA, sedentary-to-moderate activity, neutral alignment, intact ligaments, good bone stock, a non-smoker, compliant, with realistic expectations.

Relative indications. Favourable but not ideal: these patients need counselling about increased complication risk and the possible need for staged procedures. Age over 60 brings lower activity demand and reduced implant stress, but the bone quality must still be good despite age. Bilateral ankle arthritis, where motion preservation is critical for gait, is staged 6 months apart if both ankles are replaced. Ipsilateral hindfoot or midfoot arthritis may need simultaneous or staged fusions. A correctable coronal deformity of 10-15° may need a staged supramalleolar osteotomy or ligament reconstruction first. A BMI of 30-35 raises complication risk without being an absolute bar: optimise weight and counsel about the increased failure risk.

Absolute contraindications. Do not attempt TAA in their presence: the failure rate is high and the complications can be catastrophic, and each should prompt strong consideration of arthrodesis as the safer alternative. A patient unable to follow the strict postoperative weight-bearing protocol is in the same category.

Relative contraindications. Proceed with extreme caution, because each increases the failure risk significantly; multiple risk factors together warrant counselling about a high probability of failure and an offer of fusion instead.

The joint. The talocrural joint is a highly congruent, constrained hinge that transmits 3-5 times body weight during gait. The tibial plafond is a concave articular surface, wider anteriorly, and requires a flat tibial component with stable fixation; the talar dome is convex with a trochlear contour, and an anatomic talar component preserves its bone stock.

Motion. Normal ankle motion is roughly 20° dorsiflexion and 50° plantarflexion, but gait itself requires only about 10° of dorsiflexion and 20° of plantarflexion. A replaced ankle achieving a 30° arc can therefore walk near-normally without approaching native range.

The talus. It has no muscular attachments and relies entirely on the surrounding arteries: the artery of the tarsal canal, the artery of the tarsal sinus and the deltoid branches. That dependence makes it vulnerable to avascular necrosis, and extensive dissection compromises the supply.

The ligaments. The deltoid complex medially and the lateral collateral complex (ATFL, CFL, PTFL) provide the coronal-plane stability that TAA success depends on, and incompetence of either edge-loads the polyethylene.

• Deltoid incompetence (medial stability, resists valgus and external rotation) lets the talus tilt into valgus: the medial clear space opens and the polyethylene edge-loads laterally
• Lateral laxity (resists varus and inversion) lets the talus tilt into varus: medial edge loading and medial gutter impingement

View larger

Implants are classified by generation, and the generations tell the story of why modern designs look the way they do.

First generation (1970s-1980s): cemented and constrained. Two components, tibia and talus, with no polyethylene insert, a flat tibial surface and no translation. Cement concentrated stress at the bone-cement interface, the metal-on-plastic articulation wore and produced osteolysis, over-constraint caused early mechanical failure, and non-anatomic loading caused subsidence and talar collapse. Aseptic loosening exceeded 50% at five years. The examples are the Mayo (1974), St Georg (1978) and TPR (1974) ankles, and the unacceptably high failure rates led to a recommendation against TAA in the 1980s-1990s that nearly ended the field.

Second generation (1990s-2000s): uncemented and semi-constrained. Porous coating gave biological fixation, a separate UHMWPE insert reduced wear, and a flat or sulcus tibial surface improved conformity; the anterior approach became standard, with less soft-tissue disruption but a required violation of extensor hallucis longus. The designs still needed extensive bone resection, the fixed bearing still constrained motion, and malalignment produced edge loading. The examples are the Agility (1984) and Buechel-Pappas (1988) ankles. Five-year survival improved to 80%, but ten-year survival was only 60-70% and revision rates stayed high: uncemented fixation proved the concept, and the fixed bearing limited longevity.

Third generation (2000s-present): mobile-bearing and anatomic. Three components: a flat or sulcus tibial component, uncemented with porous coating or pegs; an anatomic talar component, uncemented with a central peg or dual pegs; and a mobile polyethylene insert between them that translates and rotates, reducing constraint and edge loading. The sulcus in the tibial surface gives anterior-posterior stability with mobility, allowing 10° of anterior-posterior translation. The anatomic talar component preserves bone stock and contour, which is what makes later revision or conversion to fusion possible; porous coating achieves 90% osseointegration at two years without cement stress; and a minimally invasive approach preserving the soft-tissue envelope reduced wound complications from 20% to 10%. Current implants include the mobile-bearing STAR (Stryker) and the fixed-bearing Infinity (Stryker), Cadence (Integra), INBONE (Wright Medical) and Salto Talaris (Integra), supported by the TARVA RCT, the STAR pivotal trial and multiple national registry studies. This is mature technology with reproducible good outcomes in appropriately selected patients.

Fourth generation (2020s-future): patient-specific and 3D-printed. CT-based planning maps the patient's anatomy for a precise fit, titanium lattice structures are printed for bone ingrowth, custom cutting guides improve alignment and reduce operative time, and modular components allow intraoperative adjustment; robotics-assisted implantation, biologic coatings to enhance osseointegration and convertible designs are the stated future directions. Against that, most studies have under five years of follow-up, the implants cost 2-3 times a standard implant, not all designs are FDA approved, the planning software and technique need specific training, and no superiority over modern third-generation designs has been shown. Long-term outcome data are needed before widespread adoption can be recommended.

A well-executed TAA in a poor candidate will fail, and the wrong patient is the most common cause of preventable failure. The assessment establishes whether the ankle, the patient and their expectations suit a replacement, whether staged procedures are needed first, or whether fusion is the safer offer.

History. Characterise the pain (rest versus activity-related, night pain) and its cost in function (walking distance, stairs, activities of daily living), and record what has already been tried: injections, bracing, physiotherapy, surgery. Establish the aetiology (trauma, inflammatory, primary osteoarthritis) and the medical comorbidities that threaten the wound and the implant: diabetes, peripheral vascular disease, neuropathy. Ask about occupation and activity goals, because a heavy-labour occupation contraindicates TAA and high-impact sport is unrealistic after it, and record smoking status, because active tobacco use is a contraindication until the patient has stopped.

Examination. Watch the gait (antalgic pattern, foot progression angle, hindfoot alignment) and measure standing hindfoot varus or valgus, clinically and on the hindfoot-alignment radiograph. Record active and passive dorsiflexion and plantarflexion, and decide whether any deformity is fixed or corrects with manual stress. Test the ligaments with varus and valgus stress and an anterior drawer, check the dorsalis pedis and posterior tibial pulses and sensation, inspect the skin for scars, coverage and previous approaches, and examine the subtalar, talonavicular and midfoot joints, because a stiff or painful hindfoot complicates the operation and a painful adjacent joint may be the real pain source.

Outcome scores. Baseline scores are essential for outcome assessment. The validated measures are:

• AOFAS Ankle-Hindfoot Score - pain, function and alignment on a 100-point scale; the minimal clinically important difference is 10 points
• FAAM - Foot and Ankle Ability Measure, with ADL and sports subscales
• VAS pain - 0-10
• SF-36 - general health quality of life
• EQ-5D - health utility for cost-effectiveness

What to promise. The realistic goals are pain reduction, better walking and a return to low-impact activity: a normal walking pattern without a limp, stairs without a rail, and normal shoes for most patients, with an expected arc of about 30° (on average 10° dorsiflexion and 20° plantarflexion). The patient must understand that a replaced ankle is not a normal ankle: no high-impact sport, heavy labour or running, and the modification is lifelong.

Differential diagnosis. Several conditions mimic or coexist with end-stage ankle osteoarthritis, and each changes the plan.

Standing radiographs. The first-line study is a standing series: AP ankle, mortise, lateral ankle and a hindfoot alignment view. Read them for joint space narrowing and subchondral sclerosis, coronal deformity (tibiotalar angle, talar tilt), sagittal deformity (anterior tibial translation, posterior talar subluxation), the quality and quantity of tibial and talar bone stock, and osteophytes and bone cysts. The measurements that matter:

• Tibiotalar angle - normal 90° ± 3°
• Talar tilt - normal under 5°
• Anterior translation - normal under 3mm
• Hindfoot alignment angle on the hindfoot view

CT. A 3D CT is highly recommended for preoperative planning. It gives an accurate assessment of bone stock and cysts, detects talar body AVN, measures deformity precisely in all planes, templates component size and position, assesses the subtalar and talonavicular joints and plans the bone cuts. It is particularly useful in revision cases and complex deformity.

MRI. Selective, not routine: suspected talar AVN, soft-tissue assessment of ligaments and tendons, osteochondral lesions and cyst characterisation. A straightforward primary TAA does not need one.

SPECT-CT. Rarely, for a diagnostic dilemma when conventional imaging is inconclusive: the painful TAA (identifying component loosening), an uncertain pain source in polyarticular disease, and distinguishing subtalar from ankle arthritis.

Algorithm

View larger

The decision. The fundamental choice in end-stage ankle arthritis is between a motion-preserving arthroplasty and a motion-sacrificing arthrodesis, and it is not either/or: it is individualised to the patient's age, activity demands, bone quality, deformity, ligament stability and expectations. There is no universal right answer. Both procedures relieve pain effectively with different trade-offs, and TARVA's comparable one-year outcomes put the choice squarely on patient factors.

Supramalleolar Osteotomy: Realignment and the Joint-Preserving Alternative

The principle. A tibial (with or without fibular) osteotomy just proximal to the plafond redistributes load away from the overloaded side of the joint, shifting the mechanical axis toward the better-preserved compartment: conceptually the ankle equivalent of a high tibial osteotomy at the knee. For varus ankle OA (medial overload) the configuration is a medial opening-wedge (or lateral closing-wedge) osteotomy; for valgus OA (lateral overload), a medial closing-wedge (or lateral opening-wedge), chosen to neutralise the deformity and any tibial plafond obliquity.

Two roles. As a joint-preserving alternative in its own right, it suits the younger or higher-demand patient with mid-stage, asymmetric OA, reasonable residual joint space, preserved motion and a congruent or correctable joint: it keeps the native joint and defers arthroplasty or fusion. As a staging step before TAA, it corrects supramalleolar deformity so the prosthesis sits on a neutral mechanical axis, since uncorrected deformity drives edge loading and early failure. It is not a treatment for end-stage concentric arthritis.

Caveats. It requires adequate residual joint space and motion; nonunion, hardware problems and incomplete correction are the main complications; and it is frequently combined with hindfoot or soft-tissue procedures to balance alignment fully.

Alignment is everything. A neutral mechanical axis (tibiotalar angle 90° ± 3°) is mandatory: varus or valgus malalignment over 5° doubles the failure risk, so navigation or intraoperative fluoroscopy confirms the axis before final implantation.

Balance and bone. Ligament balancing is essential: if the ankle is unstable to varus stress, reconstruct the lateral ligaments; if unstable to valgus, address deltoid competence. Preserve bone, because a later revision or conversion to fusion depends on what is left.

The scale of the problem. Complication rates after TAA are significantly higher than after hip or knee replacement: an overall complication rate of 30-40%, a reoperation rate of 20-25% and a revision rate of 15% at 10 years. Wound complications are the most common, followed by aseptic loosening, subsidence, periprosthetic fracture and infection, and across the registries aseptic loosening, subsidence and instability are the leading reasons for revision, with conversion to fusion the commonest salvage. Patient selection and technical precision are what keep these figures down.

Early, 0-6 weeks. Wound dehiscence declares itself at 2-3 weeks; meticulous closure, elevation and minimal tension prevent it, and local wound care, VAC therapy and delayed closure or a flap if the defect is large treat it; deep wound complications requiring flap coverage occur in 3%. Deep infection presents with fever, drainage, pain and wound erythema; prophylactic antibiotics and sterile technique reduce it, and it is treated by debridement and intravenous antibiotics, with component removal if the organism is virulent. Venous thromboembolism occurs in 2-3% despite mechanical and pharmacological prophylaxis, presents as leg swelling, chest pain or dyspnoea, and is treated with anticoagulation per protocol, with an IVC filter if recurrent. Compartment syndrome is rare (severe pain, swelling, neurological deficit); avoid tight dressings, monitor perfusion, and keep a high index of suspicion, because the treatment is emergency fasciotomy. Early complications need prompt recognition and aggressive management to salvage the reconstruction.

Medium-term, 6 weeks to 2 years. These relate to the implant-bone interface and the healing process. Subsidence presents as pain, loss of motion and radiographic settling of the component, from osteoporosis, an undersized component or premature weight-bearing. Periprosthetic fracture presents as pain, inability to bear weight and deformity, from trauma, a stress riser or osteoporosis. Persistent pain despite healing is investigated with CT, SPECT-CT and diagnostic injections for malalignment, component malposition or adjacent joint arthritis, and revised if the component is the problem. Stiffness from heterotopic ossification, adhesions or malposition is treated with physiotherapy; manipulation under anaesthesia is rarely indicated, and severe heterotopic bone is excised.

Late, over 2 years. TAA is not a lifetime solution for every patient, and late complications often need revision surgery. Aseptic loosening presents with progressive pain, swelling and instability, from micromotion, malalignment and wear-debris osteolysis. Periprosthetic cysts are often asymptomatic and found on routine imaging, from wear debris, stress shielding and osteolysis; they are revised if expanding beyond 15mm or symptomatic. Polyethylene wear, from high activity, malalignment or a thin insert, presents as instability, osteolysis and component tilt: exchange the insert, and revise the metal components if worn or loose. Instability, from ligament attenuation, wear or malalignment, presents as recurrent swelling, giving way and pain with activity, and is treated by ligament reconstruction or component revision if malpositioned. Adjacent joint degeneration (subtalar, talonavicular, midfoot pain) is less common than after fusion but does occur, and is treated by injection, bracing or fusion of the symptomatic joint if severe.

Salvage of the Failed TAA: Conversion to Arthrodesis

The core problem is bone loss. When a TAA fails, commonly from aseptic loosening, subsidence, instability or deep infection, the principal definitive salvage is conversion to arthrodesis, the leading endpoint after a failed or revised TAR in the registry data. Removing the tibial and talar components and any osteolytic cysts leaves a large segmental gap and a shortened, often poorly vascularised talus, so a simple in-situ ankle fusion is rarely possible.

Rebuilding the gap. The defect is usually reconstructed with bulk structural allograft (typically femoral-head allograft) or autograft to restore height and provide a fusion bed; without it, conversion causes unacceptable limb shortening. When the talar body is deficient or the subtalar joint is involved, the fusion is extended across the subtalar joint as a tibiotalocalcaneal (TTC) fusion stabilised with a retrograde intramedullary hindfoot nail or a blade plate, spanning tibia, talus and calcaneus: it sacrifices subtalar motion but achieves a stable construct across poor bone.

What to expect. Nonunion and complication rates are higher than after a primary ankle fusion (poor vascularity, bone loss, prior surgery, possible occult low-grade infection); union is slower and residual shortening is common. Conversion is broadly successful, roughly 80%, but it is a salvage with a real nonunion risk, not a routine fusion. Always exclude infection (aspiration, CRP/ESR) before a single-stage conversion, because an unrecognised infected TAA mandates a staged protocol. (See the dedicated ankle arthrodesis topic for primary fusion technique.)

What predicts success. The selection criteria predict the outcome for the reasons they were chosen: age 50-65 brings lower activity demands, adequate bone quality and a reasonable lifespan expectation; a BMI under 28 reduces implant stress and wound complications; primary or post-traumatic OA gives predictable bone quality with less systemic involvement; neutral alignment or under 5° of deformity gives balanced loading without edge wear; intact or reconstructible ligaments prevent instability and component tilt; and good tibial and talar bone stock gives fixation and osseointegration. Two factors belong to the surgeon rather than the patient: experience of over 20 cases annually, which improves alignment and reduces technical error, and a third-generation mobile-bearing implant with proven longevity and a low complication rate.

What predicts failure. The mirror image. Age under 40 brings activity demands and a lifespan requirement that exceed implant durability; a BMI over 35 increases loading, wound complications and failure; varus or valgus malalignment over 10° causes edge loading, accelerated wear and loosening; uncorrectable medial or lateral laxity causes component tilt, edge loading and early loosening; osteoporosis, AVN or large cysts leave inadequate fixation and a subsidence risk; heavy labour or high-impact sport accelerates wear and loosening; smoking at surgery causes wound complications, nonunion and infection; and prior ankle or hindfoot sepsis risks reactivation around the implant.

Global Epidemiology

Guidelines & Registry Evidence, Side by Side

Practice Variation