Fusion for Degenerative Arthritis of the Ankle

Published on 10/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 1907 times

CHAPTER 17 Fusion for Degenerative Arthritis of the Ankle

Troy M. Gorman, Florian Nickisch, Timothy C. Beals, Charles L. Saltzman

Post-traumatic degeneration of the ankle is the principal cause of end-stage arthritis, and unlike arthritis of the hip or knee, it often affects younger patients. In addition to post-traumatic arthritis, other conditions that lead to degeneration of the ankle joint include rheumatoid arthritis, primary osteoarthritis, osteochondritis and osseous necrosis of the talus, postinfectious arthritis, crystalline arthropathies, hemachromatosis, and neuropathic degenerative disease.

Unlike hip and knee arthritis, primary ankle osteoarthritis is rare, and most degenerative ankle arthritis is post-traumatic in nature. Autopsy studies demonstrate that degenerative changes are about three times more prevalent in the knee than the ankle and that degenerative changes in both joints increase with increasing age.¹^,² Radiographic studies that aim to quantify the prevalence of ankle osteoarthritis are of limited value because of the low correlation between the identification of osteophytes on plain films and the development of symptomatic osteoarthritis.³ Clinical studies suggest that knee and hip osteoarthritis is 8 to 10 times more common than ankle osteoarthritis.¹^,⁴

Because ankle arthritis is primarily post-traumatic, it affects a younger and more active patient population than hip and knee osteoarthritis. This fact demands a durable surgical option when nonoperative interventions have failed. The use of total ankle replacement is evolving, and new designs are being developed, but the gold standard for most cases of end-stage ankle arthritis has been open ankle arthrodesis. During the past 2 decades, arthroscopic ankle arthrodesis has become a viable alternative to the open procedure and has shown encouraging results.⁵^–⁸ Proposed advantages of arthroscopic techniques are less postoperative pain and morbidity, decreased blood loss, and shorter hospital stay. One important advantage of arthroscopic ankle arthrodesis is that it can be performed in patients with a poor soft tissue envelope (Fig. 17-1).

FIGURE 17-1 A patient with a poor soft tissue envelope after a burn injury went on to have a successful arthroscopic ankle arthrodesis.

ANATOMY AND PATHOGENESIS

The ankle joint is formed by the interaction of the tibia, the talus, and the fibula. The distal tibia and the medial and lateral malleoli form the ankle mortise, which contains the talus. The ankle mortise offers inherent bony stability due to its congruency, and it is further stabilized by soft tissue structures. These structures include the ligaments of the syndesmosis, the ankle capsule, the anterior talofibular ligament, the calcaneofibular ligament, the posterior talofibular ligament, the intermalleolar ligament, and medially, the deltoid ligament complex.

The ankle is relatively resistant to primary degenerative osteoarthritis, possibly because of the properties of ankle cartilage, including relatively better retention of tensile fracture stress and tensile stiffness with age, as described by Kempson and colleagues.⁹ The cartilage in the ankle is metabolically different from that in the knee. Ankle cartilage appears to be less affected by the catabolic cytokine interleukin 1 (IL-1) and deleterious collagenases that are produced in response to IL-1.¹⁰

Secondary osteoarthritis of the ankle can develop after fracture or ligamentous injury. Rotational ankle fractures and ligamentous injury with recurrent instability are the most common causes.¹¹^–¹⁵ In Saltzman’s practice during 13 years, 445 (70%) of 639 patients with Kellgren-Lawrence grade 3 or 4 ankle arthritis were post-traumatic cases, and only 46 patients (7.2%) had primary oseteoarthritis.¹³ Other recorded causes in this study of ankle arthritis included neuropathic disease (e.g., Charcot neuroarthropathy), inflammatory arthropathies (e.g., rheumatoid arthritis), crystalline arthropathies (e.g., pseudogout), osteochondritis, osseous necrosis, and postinfectious arthropathy.

The relative resistance of the ankle joint to primary osteoarthritis is likely a combination of its congruency, which results in inherent stability and restrained motion; unique tensile properties; and distinctive metabolic characteristics. Unfortunately, the ankle is quite susceptible to post-traumatic arthritis, and this may be related to its thinner and stiffer articular cartilage not being able to accommodate articular step-offs or the stresses of improperly constrained motion. Step-offs lead to increased local contact stresses that the thin cartilage of the ankle may not be able to accommodate as well as the thicker cartilage in the hip and knee.¹⁶^,¹⁷ These increased localized contact stresses likely contribute to the degeneration of articular cartilage that is seen after trauma. Other disease processes, such as Charcot arthropathy and osteochondritis, with large osteochondral defects can lead to step-offs or incongruity in the articular surface and result in increased contact stresses (Figs. 17-2 and 17-3).

FIGURE 17-2 Magnetic resonance imaging of a patient’s ankle with osseous necrosis of the talus shows collapse of subchondral bone.

FIGURE 17-3 An intraoperative photograph demonstrates fragmentation of the talus due to osseous necrosis.

PATIENT EVALUATION

History and Physical Examination

The first step in evaluating the patient with ankle pain is to obtain a proper history, particularly any history of trauma. One major ankle sprain can result in a significant injury to ankle joint cartilage or persistent instability, with the eventual development of degenerative changes.¹³ Inquiring about other joint pain is important because multiple joint involvement may indicate systemic causes, such as inflammatory arthropathy. Ankle arthritis is not usually the first manifestation of a systemic disease process. Ascertaining a diabetic history is important, because diabetes is a risk factor for Charcot neuroarthropathy and related to perioperative morbidity in those with end-stage organ disease.

It is helpful to identify what activities exacerbate or alleviate the pain. Pain with uphill walking can point to anterior impingement. Alternatively, pain at the back of the ankle with downhill walking may suggest a symptomatic os trigonum, posterior osteochondral defect, or posterior impingement. Discomfort resulting from walking on uneven ground suggests subtalar pathology or ankle instability. Lateral hindfoot pain can be caused by subfibular impingement, peroneal tendon problems (i.e., subluxating peroneals or tendinopathy), and fractures of the lateral process of the talus. Posteromedial pain is often associated with posterior tibial tendon pathology. A general sense of instability with recurrent swelling can suggest medial, lateral, and combined chronic ligament insufficiency. Because all of these pathologies may coexist, careful evaluation is essential.

Examination begins with observation, especially when the examiner has the opportunity to watch the patient walk into the examination room. Having the patient walk as part of the examination is informative, and observing overall lower extremity alignment and gait pattern is critical. Restricted ankle motion leads to early heel rise and a bent knee gait. The posture of the forefoot when it strikes the ground should be observed, because excessive forefoot varus or valgus is important to consider in surgical planning. On standing, the position of the hindfoot should be recorded. External rotation of the lower extremity is a common feature of patients with ankle arthritis.

The seated examination includes evaluation of range of motion in the ankle, hindfoot, midfoot, and forefoot. Ankle stability should be assessed by drawer testing, with the foot in plantar flexion and neutral position, which investigates the competence of the anterior talofibular ligament and calcaneofibular ligament, respectively. Talar tilt should be assessed. Foot alignment is important because deformity in the foot may cause secondary ankle disease. For example, pes planus with medial column instability may be associated with secondary ankle valgus and eventual degenerative change. Conversely, realignment of a deformed ankle can alter the foot position and adversely or positively affect function of other joints, particularly the subtalar joint. If compensatory foot deformities are identified on examination, their passive, manual correctability has to be assessed.

Tendons should be palpated to identify potential sources of pain. Finding the point of maximal tenderness during the examination may help in diagnosis. A vascular examination should be performed with palpation of pulses and assessment of the distal capillary refill, and a neurologic assessment looking for motor or sensory deficits rounds out the examination. It is important to identify patients with impaired balance.

Diagnostic Imaging

Radiographs should be done weight bearing if possible. The four radiographic views we use in our clinic to evaluate ankle pain include the anteroposterior, lateral, mortise, and hindfoot alignment views. Radiographs of the degenerative ankle can show joint space narrowing, osteophyte formation, subchondral sclerosis, and subchondral cysts. The hindfoot alignment view is important in the evaluation hindfoot varus or valgus and ankle deformity in the coronal plane.¹⁸ It is taken with the patient standing on a platform facing a collector that angles away from the platform at 20 degrees. The x-ray tube is posterior to the ankle, with the beam perpendicular to the plane of the film at the level of the ankle. The source-to-collector distance is 40 inches. On average, the most inferior aspect of the calcaneus lies just medial to the longitudinal midaxis of the tibia (Fig. 17-4). Magnetic resonance imaging (MRI) has limited usefulness unless the examiner suspects an osteochondral lesion of the talus, osseous necrosis, or ligamentous abnormality that will alter patient care. In such cases, MR arthrography may be advantageous. Computed tomography (CT) is the better choice for three-dimensional bony imaging, and it allows visualization near hardware, unlike MRI. Noninvasive joint distraction plus air-contrast arthrography enhances visualization of the ankle’s articular features.¹⁹

FIGURE 17-4 Hindfoot alignment view demonstrates no significant coronal malalignment.

Selective fluoroscopically guided injections can be helpful in diagnosing patients who have clinical or radiographic findings that suggest more than one source of pain. It is reasonable to expect 75% pain relief in an area that is injected.²⁰ It is important to identify the patients’ ankle pain as global (i.e., affecting most of the joint) or focal (i.e., affecting a specific region), because this distinction may guide the treatment options. It is essential to identify patients with coexisting subtalar pain, because that population needs to be counseled more intensely about the risks of residual pain and progression of adjacent joint arthritis.

TREATMENT

Nonoperative and operative treatments can help reduce symptoms and improve the function of painful ankle arthritis. There are no well-designed retrospective or prospective clinical trials reporting on nonoperative treatment of diffuse ankle arthritis. Our experience is that nonsteroidal anti-inflammatory drugs have variable efficacy in addressing the pain of ankle arthritis. The judicious use of corticosteroid injections may provide temporary relief and be beneficial in acute exacerbations in someone who has tolerable steady-state pain. Clinical trials suggest that hyaluronate-based injectables may diminish ankle pain and improve function in patients with ankle osteoarthritis, and additional trials exploring this treatment modality are being conducted.²¹^,²²

Nonoperative interventions primarily focus on mechanical unloading and immobilization. Devices that address this goal include the cane, ankle foot orthoses (AFOs), and the leather ankle lacer with an embedded polypropylene shell.²³ To fully unload the ankle joint requires a rigid linkage between the acceptance of force on the sole of the foot and the leg above the ankle. The prototypic device, a weight-bearing patellar brace (PTB), is poorly tolerated because load transfer irritates the soft tissues in front of the knee. Adding a rocker-bottom sole to a shoe or the use of a solid ankle-cushioned heel (SACH) may also provide relief by reducing ankle excursion with gait.

Operative intervention should be considered only after failure of nonoperative treatment methods. The surgeon should identify the cause of the patient’s problem and to what extent ligamentous instability, malalignment, or other foot deformity contributes to the perceived pain from ankle arthritis. When planning surgical interventions, recreating normal foot alignment can encourage improved foot function, regardless of the chosen surgical technique. Surgical options for end-stage degenerative ankle arthritis include osteotomies about the ankle, débridement, distraction, ankle arthroplasty, and ankle arthrodesis. Ankle arthrodesis can be performed in a variety of ways, including open, mini-open, and arthroscopic techniques.

If arthrodesis is the surgical treatment of choice, the next step is ensuring appropriate alignment at the fusion site, which takes into account the alignment of the entire limb. Malalignment or angulation of the tibia may require special consideration. Adjustment may need to be made in the position of the fusion to ensure a plantigrade foot if the foot has significant forefoot varus or valgus or other foot malalignment. For example, if a patient has significant fixed forefoot varus, the ankle joint needs to be positioned in a little more valgus so a plantigrade foot can be created. Concomitant knee and ankle arthritis and deformity should be fully assessed, and realignment at the knee is a priority before ankle arthrodesis.

General indications for ankle arthrodesis include degenerative arthritis with significant pain unresponsive to nonoperative interventions, end-stage arthritis due to other causes (e.g., rheumatoid arthritis, pseudogout, postinfectious arthritis, hemachromatosis), large osteochondral defects not amendable to other interventions, osseous necrosis of the talus, failed total ankle replacement, and malalignment or instability from a paralytic deformity. Absolute contraindications to ankle arthrodesis include active infection and active Charcot arthropathy. However, after appropriate treatment of an infection and resolution of the metabolic issues associated with Charcot arthropathy, arthrodesis is an acceptable treatment for these problems. Some surgeons may consider active smoking by the patient a relative contraindication.

In terms of arthroscopic ankle arthrodesis, the indications remain the same, with the exception of a failed total ankle replacement. Well-aligned ankles and those that are easily realigned are excellent candidates for arthroscopic fusion. Patients with soft tissue compromise (e.g., those with prior trauma, burn victims, patients with skin grafts) or vasculopathy are strongly considered for an arthroscopic approach. In the past, it was thought that ankle varus or valgus greater than 5 degrees was an absolute contraindication to arthroscopic arthrodesis. However, later reports suggested that substantial ankle varus or valgus is a relative contraindication rather than an absolute one.²⁴^,²⁵ The investigators consider any ankle that can be re-aligned properly after arthroscopic débridement appropriate but acknowledge that patients should be counseled that conversion to an open approach is prudent if an extensive capsulotomy is required to achieve correct alignment. Additional contraindications for the arthroscopic procedure are significant focal bone loss and deformity and extremely rigid ankles. In general, the desired position of the arthrodesis is neutral dorsiflexion, 5 degrees of ankle valgus, equal or slightly greater external rotation compared with the contralateral leg, and placement of the anterior aspect of the talar dome at the level of the anterior aspect of the tibia.

Arthroscopic ankle arthrodesis was first developed in the mid-1980s. Myerson and Quill reported the first comparative series, with 17 arthroscopic procedures compared with 16 open procedures done with a medial malleolar osteotomy.⁷ They found an average fusion time of 8.7 weeks in the arthroscopic group and 14.5 weeks in the open group, with similar fusion rates for both groups. Their criteria for clinical union included an absence of tibiotalar motion, crepitus, or pain with ambulation or on examination. Radiographic union was defined as the appearance of osseous trabeculae across the tibiotalar arthrodesis site and incorporation of bone graft into the fusion mass when bone graft was used.

One of the first long-term series was published in 1996 by Glick and colleagues.⁶ They reported 34 cases with an average follow-up of 8 years. Fusion rates were 97%, and good or excellent results were reported for 86% of patients. Three ankles in this series were rated poor because of subtalar pain and a nonunion, and a malunion accounted for the other poor results.

O’Brien and colleagues added support to arthroscopic arthrodesis by showing comparable fusion rates between 19 arthroscopically treated patients and 17 patients undergoing open arthrodesis using flat cuts.⁸ The arthroscopic group had decreased operative times, decreased tourniquet times, less blood loss, and decreased hospital stays.

Since the publication of these three studies, the popularity of arthroscopic ankle arthrodesis has grown and the technique has evolved. Our current techniques are described subsequently in the context of individual cases.

Case 1

The patient is a 49-year-old woman with a history of rheumatoid arthritis and multiple left ankle sprains as a young adult. She has had increasing left ankle pain over the past few years. She underwent arthroscopic débridement of the ankle 3 years earlier, but the procedure did not improve her pain. An over-the-counter lace-up ankle brace did not ease the pain. She did get some relief with a corticosteroid injection in the left ankle, but the pain returned. Her pain was aggravated with walking and ascending and descending stairs.

On physical examination she had mild hindfoot varus and an ankle arc of motion of 15 degrees. She had a small ankle effusion, and palpation elicited tenderness along the medial, anterior, and lateral joint lines. The subtalar joint was not tender to palpation and had normal motion. The result of the anterior drawer test was negative. She was neurovascularly intact. A hindfoot alignment radiograph showed minimal varus (see Fig. 17-4). Anteroposterior, lateral, and mortise views showed loss of tibiotalar joint space and osteophyte formation (Fig. 17-5). The subtalar joint was preserved (Fig. 17-6).

FIGURE 17-5 A mortise view shows significant narrowing of the tibiotalar joint space and involvement of the medial and lateral gutters.

FIGURE 17-6 A lateral view shows significant tibiotalar degeneration but preservation of the subtalar joint.

Treatment options discussed with the patient included ankle arthrodesis and total ankle replacement. Given her young age and relatively poor motion, she was not considered to be a good candidate for ankle arthroplasty. Arthrodesis appeared to be the more reliable option, and the patient and surgeon decided to perform the procedure arthroscopically from an anterior approach.

Anterior Arthroscopic Ankle Arthrodesis

Room Setup

We recommend general anesthesia to relax the gastrocnemius-soleus complex, and it is augmented with a regional block to aid in postoperative pain control. At our center, a popliteal-level indwelling catheter and a single-shot saphenous block are placed with ultrasound guidance.

The bed is placed in a beach chair position and then in slight Trendelenburg so the operative leg is nearly parallel to the floor. The heel is positioned just off the edge of the table. The break in the bed should be positioned at the knee, so that when the end of the table is dropped, traction can be applied to the limb without moving the patient. This reflexed position provides sufficient resistance against the thigh to allow for distraction when traction is applied (Fig. 17-7). Preparing above the knee allows for assessment of rotational alignment. An alternative is to use a well-leg holder against the posterior thigh of the operative limb. We mark landmarks, including malleoli, branches of the superficial peroneal nerve, and the expected level of the tibiotalar joint space.

FIGURE 17-7 The position of the bed allows traction against the bed without using a well-leg holder. The fluoroscopic scanner can be positioned to enable easy visualization of the ankle joint.

Instruments include a pump, a 4.5-mm arthroscope, a 4.5-mm aggressive shaver, and 4.0-mm burr and curettes. Distraction, noninvasive or invasive, may or may not be used. We find distraction helpful during the preparation of the articular surfaces because it allows easier movement of the instruments. We use a thin wire through the calcaneus attached to a custom distractor and apply 70 to 90 pounds of force to allow for better visualization and navigation around the joint. If external strapping techniques are used to distract the joint, the force (usually less than 25 pounds) should be decreased during preparation of the anterior talar dome, because this takes tension off the anterior capsule, allowing easier access to the talar dome.

Joint Preparation

Use of thigh tourniquet is optional. Saltzman inflates a tourniquet only when needed for better visualization. An anteromedial approach is used to instill 15 to 20 mL of saline. A fluoroscope may be used to verify that the needle is correctly placed in the joint. With fluid injection, the ankle joint usually plantar flexes.

The medial portal is established just medial to the tibialis anterior tendon sheath and slightly distal to the ankle joint by incising the skin with a no. 15 blade and bluntly dissecting down to the capsule with a small, straight hemostat. The hemostat is then punched into joint, and the tips are used to spread the tissue to create a path for the arthroscope. The arthroscope is then introduced into the anterior ankle. The ankle is distended by turning on the fluid pump, and the entrance for the lateral portal is visualized with the camera. To avoid distending the soft tissues, the pump should always be run at the lowest possible flow rate and pressure to maintain adequate visualization. The lateral portal is made just lateral to the long extensor tendons and the peroneus tertius and just distal to the ankle joint. Placing portals too proximal, particularly if anterior osteophytes are present, should be avoided. The trajectory of the portal should be tested with an 18-gauge needle. We like to have a distance of 5 to 10 mm from the palpable branches of the superficial peroneal nerve when creating this portal. The joint is entered bluntly with the small hemostat, and a thin periosteal elevator is placed into the joint. If necessary, a shaver is used first and an anterior synovectomy is performed to improve visualization.

Any residual cartilage on the tibia and talus are removed with the shaver or curettes (Video 1). Having a set of thin-handled, angled curettes is helpful. If the lateral gutter does not show significant wear, the cartilage can be left in this location, and the fibula is excluded from the fusion. However, if there are degen erative changes in this region, we remove all residual cartilage and incorporate the fibula into the fusion.

After all the cartilage is removed, a 3- to 4-mm burr is used to pockmark the subchondral surface so it looks like the surface of a golf ball (Fig. 17-8). It is crucial to do this across the entire undersurface of the tibia and the anterior two thirds of the talar dome. Distraction helps us to preserve the natural contours of the ankle joint. If the joint is unable to be débrided because it is too tight, the incision can be extended and a laminar spreader placed into the joint to facilitate cartilage removal to perform a mini-open procedure.²⁶ Adequacy of débridement is then determined by deflating the tourniquet, turning off the pump, and inspecting the joint for sufficient punctate bleeding.

FIGURE 17-8 An intra-articular arthroscopic view reveals the pock-marked surface of the tibia and talus and shows that the overall congruency of the tibiotalar joint is preserved.

Alignment

Aligning the foot correctly under the tibia is the critical step. Because most arthroscopic patients have a mild deformity, the goal is getting the foot plantigrade and neutral in the sagittal plane. This can sometimes be very challenging. We recommend placing the ankle in what is the best position clinically and temporarily fixing it with two large Kirschner wires or cannulated pins. Mechanically compressing the bones while the pins are placed can help to achieve proper compression.

During the next step, the assistant holds the leg without grabbing the foot and goes to the side of the patient to assess the sagittal position of the ankle with the foot gently dorsiflexed. The position must be perfect before moving onto fixation. The leg should be straightened to ensure proper coronal plane positioning and rotational alignment of the ankle.

Fixation

For the tibiotalar fusion, cannulated or solid screws can be used. We typically use two large, cannulated screws and check the position and length of our guide pins with a fluoroscope before placing the screws.

The first pin is placed through a 2-cm incision placed on the anterior leg about 5 cm above the joint. This incision needs to be long enough to ensure that neither the superficial nor deep peroneal structures are injured by screw placement. The pin is placed on the anterolateral part of the tibia starting 3 cm above the joint, just superior and medial to Chaput’s tubercle, and it is directed 10 to 20 degrees posterior from vertical into the posterior half of the talar dome. Lateral and anteroposterior views are needed to confirm pin placement before a large, cannulated, partially threaded compression screw (usually 6.5 mm) is used to pull the talus up into the mortise (Fig. 17-9).

FIGURE 17-9 Anterolateral screw placement is confirmed by fluoroscopy. The screw is placed on the anterolateral part of the tibia starting 3 cm above the joint, just superior and medial to Chaput’s tubercle, and it is directed 10 to 20 degrees posterior from vertical into the posterior half of the talar dome.

A second screw is placed from a posteromedial position in the tibia to an anterocentral position in the talus (Fig. 17-10). The leg is placed in a figure-of-four position, and a 1-cm incision is made along the posteromedial edge of the tibia 3 cm above the joint line. Blunt dissection and retraction get the pin down to the bone. The pin should start at least 4 mm anterior to the sheath of the posterior tibial tendon and traverse the posteromedial corner of the ankle, coursing into the center of the neck of the talus (Fig. 17-11). Direct visualization of the posterior tibial tendon sheath is recommended to prevent injury to the tendon. After the guide pin is overdrilled, a partially threaded, large, cannulated screw is placed, which may slightly compress the construct back into the posteromedial corner.

FIGURE 17-10 The guide pin for the posteromedial screw and the location of the anterolateral screw are seen on anteroposterior fluoroscopy.

FIGURE 17-11 A lateral view shows the anteromedial pin starting 3 cm above joint line and coursing in the center of the neck of the talus. The pin should start at least 4 mm anterior to the sheath of the posterior tibial tendon.

If the lateral gutter (fibulotalar articulation) has no significant arthritis, it does not need to be débrided or fixed. However, if the fibulotalar joint is arthritic, a third screw is placed from a posterolateral position in the fibula to an inferocentral position in the talar neck and body. We often downsize the width of this screw to avoid cracking the fibula and add accessory screws directed from the fibula to the talar dome (Figs. 17-12 and 17-13). Another consideration is to perform a transverse osteotomy of the fibula a few centimeters proximal to the joint line to decouple forces from the proximal fibula to the distal fibular fragment. This may enhance fusion of the syndesmotic region, and the distal fibular fragment can be secured with the same screw arrangement described earlier.

FIGURE 17-12 An anteroposterior view of the final construct demonstrates a downsized screw from the posterolateral fibula into the anterocentral neck of talus. An accessory screw was added from the fibula into the talar dome.

FIGURE 17-13 A lateral view of the final construct demonstrates incorporation of the lateral gutter into the fusion with a downsized posterolateral screw and accessory lateral screw.

Small gaps usually exist between the bony surfaces after fixation. If the lateral malleolus is included in the fusion, there is often a small gap laterally because the medial screw may pull the talus medially. We inject about 5 to 10 cc of demineralized bone matrix into these small gaps through our portals. Alternatively, the demineralized bone matrix can be placed after the guide pins are placed, but it should be done before final screw placement.

Postoperative Care

The ankle is placed in a well-padded posterior splint. The splint and sutures are removed after 10 to 14 days, and a below-knee cast is placed. We allow patients 5 to 10 pounds of heel weight bearing so they can maintain their balance. Others allow full weight bearing at this point.²⁷ Our group has no experience with early weight bearing, and we remain concerned about increasing the rate of nonunion.

At 6 to 8 weeks, radiographs are taken. At this point, we hope to see early bridging, which has a ground-glass appearance in the joint space. If there is doubt about the adequacy of the fusion, CT can be used to better visualize the joint. If fusion is apparent and the patient has minimal pain in the joint, the leg is placed into a removable boot. It should be worn at all times except for sleeping, bathing, and sitting. The patient then begins progressive weight bearing as tolerated in the boot, as long as there is no significant pain.

At 10 weeks, a second set of radiographs is obtained. If the radiographs show increased bridging and the patient has no pain with standing, he or she is weaned from the boot. Formal physical therapy is not mandated, but a 4- to 8-week course of balance and gait training after removal of the boot seems to speed recovery, particularly in elderly patients.

Case 2

A 33-year-old man had an injury to the left ankle caused by automobile-pedestrian accident 2 years before presentation. The initial injury was suspected to be a large talar dome fracture, which was addressed by internal fixation. Two years after the injury, the patient underwent partial removal of hardware and subsequently became infected. He was treated with irrigation and débridement and with antibiotics. He was finishing antibiotic therapy at the time of presentation.

The patient had much ankle pain and hindfoot pain. Examination revealed that he had three healed incisions about the ankle. It was diffusely tender to palpation, and the ankle and subtalar range of motion was limited. Radiographs and CT scans demonstrated osseous necrosis of the medial half of the talus, including the dome and the body, and arthrosis of the tibiotalar and subtalar joints (Figs. 17-14 and 17-15). The patient’s overall alignment was well preserved (Fig. 17-16). Because the patient had numerous prior incisions, there was concern that his soft tissue envelope might be compromised. He was considered for a combined prone arthroscopic ankle and subtalar arthrodesis.

FIGURE 17-14 A mortise view shows significant tibiotalar degenerative disease and partial osseous necrosis of the talus.

FIGURE 17-15 A lateral view shows additional degenerative disease of the subtalar joint.

FIGURE 17-16 A hindfoot alignment view demonstrates preservation of the coronal alignment.

Combined Posterior Arthroscopic and Subtalar Arthrodesis

Room Setup

The patient is positioned in the prone position, with the involved ankle placed approximately 10 cm distal to end of the table. We sling the safety strap posterior and inferior to the buttocks to resist the force of distraction. General anesthesia and a regional nerve block are used. Instruments include a custom distraction device; a 2.7-mm, 30-degree angled arthroscope; a larger, 4.0-mm arthroscope; thin periosteal elevators; long and thin-handled curettes; aggressive shavers; and a 4.0 mm burr. Before standard preparation and draping, the nonoperative limb is flexed to about 90 degrees at the knee and held in this position for the duration of the surgery (Figs. 17-17 and 17-18). This allows full access to the operative limb. On the operative side, the Achilles tendon, the medial and lateral malleoli, the sural nerve, the posteromedial neurovascular bundle, and the estimated levels of the tibiotalar and subtalar joints are marked with a surgical pen.

FIGURE 17-17 The setup for posterior ankle arthroscopy is shown. The nonoperative limb is flexed to 90 degrees at the knee.

FIGURE 17-18 Initial draping is shown for posterior ankle arthroscopy. The Mayo stand cover is placed over the nonoperative limb.

Joint Preparation

A transosseous, 1.6- to 1.8-mm wire is placed medial to lateral through the calcaneus and attached to a foot plate of a ring fixator or a modified traction bale to allow distraction with a hand-driven winch (Video 2). Posterolateral and posteromedial portal sites immediately adjacent to the Achilles tendon and equidistant between the ankle and subtalar joints are established.

The posterolateral portal is established first. An 18-gauge needle is used to inject 10 mL of normal saline into the subtalar joint and 10 mL into the tibiotalar joint under fluoroscopic guidance. The joints are observed to distract with fluid injection. A small, longitudinal incision is then made through the skin (previously marked) just lateral to the Achilles. A short, straight hemostat is the used to dissect to the lateral aspect of the subtalar joint. The hemostat is used to penetrate the joint capsule, and then the 2.7-mm, 30-degree angled arthroscope is inserted. An arthroscopic pump controls the fluid flow and is held to a minimum flow rate to allow visualization. Next, an 18-gauge needle is inserted at the site marked for the posteromedial portal and directed toward the central aspect of the subtalar joint. The needle is visualized, making sure that it is lateral to the flexor hallucis longus. All work is performed lateral to the flexor hallucis longus to avoid injuring the posteromedial neurovascular structures. Visualization and clear delineation of the flexor hallucis longus is the first step in this procedure. The needle is then removed, and the posteromedial portal is established, directing all instruments centrally. At this point, the 4.0-mm arthroscope can be used to improve fluid flow and visualization during the procedure.

If necessary, a third portal is established approximately 1 cm proximal and 1 cm posterior to the tip of the lateral malleolus. The sural nerve is at risk, and caution is required when inserting instruments. This portal gives greater access to the posterior facet of the subtalar joint. Most of the procedure is done with the arthroscope in the posterolateral portal and the instruments in the posteromedial portal, although both portals are used in an alternating fashion for viewing and for instrumentation. Arthroscopic cannulas may be used.

Synovectomy is performed first with the shaver to improve visualization. The residual articular cartilage of the entire posterior facet is then removed with a thin, curved periosteal elevator, curettes, and an aggressive shaver (Fig. 17-19). Visualization of the interosseous ligament signifies the anterior extent of the débridement, and by staying posterior to this structure, the surgeon avoids the sinus tarsi vasculature. The bone surfaces are then pockmarked with the burr, as described for the ankle arthrodesis. It is best to preserve the overall congruency of the joint during the débridement and pockmarking to get good opposition of the joint during fixation.

FIGURE 17-19 A periosteal elevator is used for débridement of residual cartilage from the subtalar joint.

Instruments are removed from the subtalar joint, and the tibiotalar joint is entered. A hemostat is used to bluntly enter the ankle from the lateral side, and all work is performed lateral to the flexor hallucis longus tendon. Using the same order of instruments described previously, the ankle is débrided of residual cartilage, and the bony surfaces are prepared. Depending on the angle of entry to the tibiotalar joint, an additional proximal portal that is in line with the other portals may be required for adequate instrument access. Demineralized bone matrix can be inserted by means of a small cannula through the posterolateral portal into the ankle and subtalar joints, and traction is released.

Fixation

Three choices are available for fixation: screws, intramedullary rod, and ring fixation. We typically use only screws when the bone is of good quality. Fixation of the subtalar joint can be achieved using two large, cannulated, 6.5- or 7.3-mm, partially threaded cancellous screws placed under fluoroscopic guidance. The initial guide pin is started at the posterolateral calcaneus and is angled anterosuperior to the talus neck or body. It is placed proximal to the weight-bearing surface of the heel and distal to the initial attachment of the Achilles. A second guide pin is placed medially in a similar direction to the first. The pins are then overdrilled and the cannulated screws placed. Alternatively, the second screw can be directed from the inferolateral portion of the calcaneus into the head and neck region of the talus. We use a 4.5-mm screw for this purpose.

The tibiotalar joint is fixed first with a compression screw in a posterior to anterior direction. Having the patient prone makes placement of the first pin easier. The first pin is placed from a posterolateral position in the tibia to an anterocentral position in the talus. A 1-cm incision is made along the posterolateral edge of the Achilles 5 cm above the joint line. Blunt dissection and retraction are necessary to get the pin down to the bone and avoid injury to the sural nerve. The pin should traverse the posterior centrolateral distal tibia and end in the center of the neck of the talus. After the guide pin is overdrilled, a partially threaded, large, cannulated screw is placed, which usually compresses the construct back into the posterior malleolus. A second posteromedially based screw is placed from just anterior to the posterior tibial tendon, as described for the anterior ankle arthrodesis.

If the lateral gutter has no significant arthritis, it does not need to be débrided or stabilized. However, if the fibulotalar joint is arthritic, a third screw is placed across the tibiotalar joint from the fibula to inferocentral in the talar neck and body. Addition of a fully threaded screw from the plantar lateral calcaneus into the tibia allows enhanced stabilization.

Another option for fixation is a tibiotalocalcaneal fusion nail. After the joints are débrided and alignment verified, the nail can be placed with fluoroscopic guidance. We prefer a transverse incision in the heel, and care is taken to avoid the plantar nerves. Typically, the incision is placed mostly on the lateral half of the heel fat pad at the junction between the middle and distal thirds. Fluoroscopic confirmation is essential. The dissection down to the calcaneus is done bluntly, and retractors are used to prevent injury to the local neurovascular structures. The nail is inserted with reaming. The construct is compressed before final interlocking screws are placed (Figs. 17-20 and 17-21). Postoperative instructions are the same as for the ankle fusion described previously.

FIGURE 17-20 Postoperative, anteroposterior radiograph demonstrates fusion using a tibiotalocalcaneal fusion nail.

FIGURE 17-21 Postoperative, lateral radiograph demonstrates the neutral position of the ankle and fusion across the ankle and subtalar joints using a tibiotalocalcaneal fusion nail.

CLINICAL RESULTS

Many clinical series published since 1990 have reported the outcomes of arthroscopic ankle arthrodesis. All the studies are retrospective in nature, and only a few have compared the outcomes of arthrodesis with those of open treatment. Myerson and Quill published the first comparison of arthroscopic with open techniques.⁷ Seventeen patients were treated with arthroscopic ankle arthrodesis, and 16 patients were treated with an open technique (i.e., medial malleolar osteotomy). The investigators found that the arthroscopic group had a shorter average time to fusion (8.7 vs. 14.5 weeks) and a much shorter hospital stay (1.5 vs. 4 days). The arthroscopic group had one pseudoarthrosis, for a fusion rate of 94.1%,compared with 100% in the open group. However, this study was not controlled, and the two groups were dissimilar, with the open group having patients with greater deformity and bone loss.

O’Brien and colleagues attempted to devise a better study by having two comparable groups in regard to the amount of deformity.⁸ Nineteen patients were treated arthroscopically, and 17 were treated with open arthrodesis with flat cuts. The investigators reported an 84% fusion rate for the arthroscopic group and an 82% rate for the open group. They found the arthroscopic group had a shorter operative time by 12 minutes (166 vs. 184 minutes) and a significantly shorter hospital stay (1.6 vs. 3.4 days). The complication rate was similar for both groups.

The following results are compiled from an additional 13 clinical series (Table 17-1). Reported fusion rates range from 70% to 100%.^{5,6,24,25,27-36} Five studies reported 100% fusion rates, and the other eight studies had rates ranging from 89% to 97%, with only one study reporting a fusion rate below 89%.³⁶ A common definition for fusion is an ankle that is clinically stable on examination and pain free on weight bearing, and it has radiographic signs of bridging trabeculae. Five studies reported mean fusion times of less than 10.5 weeks, with the fastest mean time to fusion reported at 8.9 weeks.⁶^,²⁹^,³⁴^–³⁶ The other studies reported mean fusion times ranging from 11 to 16 weeks.

TABLE 17-1 Published Series of Arthroscopic Ankle Arthrodesis

Five of the series reported the clinical outcomes of patients, with good or excellent results for 80% to 95% of patients and with follow-up ranging from 14 months to 18 years.^5,6,25,33,36 In the largest of the studies, Winson and colleagues reported good or excellent results for 83 (80%) of 104 patients at an average follow-up of 5.4 years.²⁵ Other studies have reported satisfaction rates of 95% to 100%.³⁴^,³⁵

No studies have reported on arthroscopic tibiotalocalcaneal fusions. However, Amendola and colleagues described the technique and early results of posterior arthroscopic subtalar arthrodesis (PASTA).³⁷ They reported 11 feet in 10 patients that underwent PASTA. Ten of the 11 joints fused by 10 weeks, with one patient going on to nonunion. Eight patients were very satisfied with the results, one was satisfied, and one patient was not satisfied with the results of subtalar fusion. The patients demonstrated a 50-point improvement in their modified American Orthopaedic Foot and Ankle Society (AOFAS) score (i.e., 36 to 86 points). Other than the one nonunion, there were no postoperative complications.