Work by Kempson ⁵⁴ shows that the tensile properties of ankle and hip articular cartilage differ and that these differences increase with age (Figs. 21-2 and 21-3). In particular, the tensile fracture stress and tensile stiffness of ankle articular cartilage deteriorate less rapidly with age than those of the hip.⁵⁴ The tensile fracture stress of hip femoral articular cartilage is initially greater than that of talar articular cartilage. However, with age it declines exponentially in the hip but linearly in the ankle (see Fig. 21-2). As a result of these aging differences, ankle articular cartilage can withstand greater tensile loads than hip articular cartilage, beginning in middle age, and this difference increases with increasing age. Age-related changes in hip and ankle articular cartilage tensile stiffness follow a similar pattern (see Fig. 21-3).

Figure 21-2 Femoral head and talus articular cartilage superficial layer tensile fracture stress versus age. Notice that the tensile fracture stress of ankle articular cartilage is greater beginning in middle age than the tensile fracture stress of femoral head articular cartilage and that the difference increases with increasing age. This illustration was developed from data reported by Kempson.⁵⁴ (Graph courtesy Joseph A. Buckwalter, MD.)

Figure 21-3 Femoral head and talus articular cartilage superficial layer tensile stiffness versus age. Notice that the tensile stiffness of ankle articular cartilage is greater beginning in middle age than the tensile fracture stress of femoral head articular cartilage and that the difference increases with increasing age. This illustration was developed from data reported by Kempson.⁵⁴ MPa, megapascals. (Graph courtesy Joseph A. Buckwalter, MD.)

Presumably, age-related declines in articular cartilage tensile properties result from progressive weakening of the collagen fibril network in articular cartilage. The cause of the age-related weakening of the articular cartilage matrix has not been explained, but age-related changes in collagen fibril structure and collagen cross-linking have been identified that might contribute to changes in matrix tensile properties.11,13 Kempson⁵⁴ has suggested that these differences in tensile properties might explain the apparent vulnerability of the hip and knee to degenerative changes with increasing age and the relative resistance of the ankle to development of primary osteoarthritis.

Articular Cartilage Metabolism

Ankle articular cartilage can differ from that of other joints in the expression of an enzyme that can degrade articular cartilage and in response to the catabolic cytokine interleukin-1 (IL-1). Chubinskaya and colleagues ¹⁵ detected messenger ribonucleic acid (RNA) for neutrophil collagenase (matrix metalloproteinase-8 [MMP-8]) in chondrocytes of human knee articular cartilage but not in those of ankle articular cartilage. IL-1 inhibited proteoglycan synthesis by chondrocytes in knee articular cartilage more effectively than in those of ankle articular cartilage.⁴⁷ The difference in the response to IL-1 between chondrocytes in knee and ankle articular cartilage appears to be due to a greater number of IL-1 receptors in the chondrocytes of knee articular cartilage. These observations need further study, but they suggest that metabolic differences exist between knee and ankle articular cartilage, which might help to explain the relative rarity of primary ankle osteoarthritis.

Prevalence of Ankle Osteoarthritis

Determining the prevalence of ankle osteoarthritis is more difficult than it might seem at first. As in other joints, the correlation between degenerative changes in the joint and the clinical syndrome of osteoarthritis is not consistent.28,106 In addition, it is extremely expensive and difficult to obtain and study unbiased samples of populations to determine the prevalence of osteoarthritis. For these reasons, studies of the prevalence of osteoarthritis by examination of autopsy specimens, evaluation of radiographs of populations of patients, and evaluation of patients presenting with symptomatic osteoarthritis have significant limitations.

Autopsy Studies

Despite their limitations, including relatively small numbers of joints examined and lack of random or systematic sampling of populations, autopsy studies can provide useful information concerning differences in prevalence of degeneration among joints.

Meachim et al 66–69 examined knee, shoulder, and ankle joints at autopsies performed on adults. They found full-thickness chondral defects in 1 of 20 ankle joints from people older than 70 years.⁶⁶ Cartilage fibrillation was much more frequent than full-thickness defects in all joints.

Huch et al ⁴⁷ resected 36 knees and 78 ankles from both limbs of 39 organ donors to evaluate the prevalence of ankle osteoarthritis. The joints were evaluated using a scale described by Collins.¹⁹ Grade 0 is normal gross appearance of a joint, grade 1 is fraying or fibrillation of the articular cartilage, grade 2 is fibrillation and fissuring of the cartilage and osteophytes, grade 3 is extensive fibrillation and fissuring with frequent osteophytes and 30% or less full-thickness chondral defects, and grade 4 is frequent osteophytes and greater than 30% full-thickness chondral defects. In these studies, grades 3 and 4 were defined as osteoarthritis, and grade 2 was defined as early osteoarthritis.⁴⁷ However, the authors did not have information concerning possible symptoms associated with the joints studied, so it is not certain if the degenerative changes they identified were associated with clinical osteoarthritis. Using the Collins grading scale, Huch’s group found grade 3 and 4 degenerative changes in 5 of 78 (6%) ankle joints and in 9 of 36 (25%) knee joints (Fig. 21-4). Degenerative changes were most commonly found on the medial aspect of the ankle.

Figure 21-4 Histogram showing the prevalence of ankle joint degeneration in autopsy studies reported by Huch and colleagues ⁴⁷ and Muehlman and colleagues.⁵⁴ In these studies, the criteria for joint degeneration (osteoarthritis) were extensive articular cartilage fibrillation, osteophytes, and regions of full-thickness cartilage loss (Collins grades 3 and 4). Notice that joint degeneration was more than three times as common in the knee as in the ankle and that the prevalence of joint degeneration in the knee and the ankle increased with age. (Histogram courtesy Joseph A. Buckwalter, MD.)

In another series of investigations, Muehlman et al ⁷⁴ examined seven joints, including the knee and ankle of both lower legs in 50 cadavers. The cadavers studied ranged in age from 36 to 94 years, with a mean age of 76 years. Sixty-six percent of the knee joints had grade 3 and 4 degenerative changes compared with 18% of the ankle joints (see Fig. 21-4). Ninety-five percent of the knees had grade 2, 3, or 4 degenerative changes compared with 76% of the ankles. The authors also observed that the medial compartments of both the knees and the ankles were more commonly involved than the lateral compartments. Radiographs often showed no evidence of degenerative changes, although direct examination of the joints showed regions of full-thickness cartilage erosion. Overall, the autopsy studies demonstrate that advanced degenerative changes are at least three times more prevalent in the knee than in the ankle and that the prevalence of degenerative changes in both joints increases with increasing age (see Fig. 21-4).

Radiographic Evaluations

Although epidemiologic studies based on radiographic evaluations document a striking increase with increasing age in the prevalence of degenerative changes of all joints, including those of the foot and ankle, the reported studies have not focused on ankle osteoarthritis. Radiographic studies of ankle joint degeneration have important limitations because there is no strong correlation between formation of osteophytes and development of clinical osteoarthritis ¹⁰⁶ and because it is difficult to evaluate the thickness of ankle articular cartilage, particularly on radiographs that were not performed in a standardized fashion. Furthermore, ankle radiographs often do not show signs of joint degeneration even when the ankle joint has regions of full-thickness erosion of articular cartilage.⁷⁴ Attempts to evaluate the prevalence of ankle degeneration and osteoarthritis by plain radiographs alone therefore have limited value.

Clinical Studies

Very few studies of the prevalence of osteoarthritis have included patients with ankle osteoarthritis. The available information suggests that knee osteoarthritis is 8 to 10 times more common than ankle osteoarthritis.22,47 Yet the best currently available estimates suggest that knee replacements are performed more often than ankle replacements and ankle fusions combined. These observations, combined with the data from autopsy studies showing that advanced knee joint degeneration is about three to five times more common than advanced ankle joint degeneration, suggest that surgical procedures are performed less often for patients with advanced ankle osteoarthritis than for those with advanced osteoarthritis of the knee.

The reasons for this are unclear. It is possible that joint degeneration and osteoarthritis cause less severe pain and functional limitation in the ankle than in the knee. Lack of understanding of the evaluation and treatment of ankle osteoarthritis among physicians, the efficacy of nonsurgical treatments for ankle osteoarthritis, and the lack of effective and widely accepted surgical treatments for ankle osteoarthritis can also explain the apparent difference in the frequency of surgical treatment of ankle and knee osteoarthritis.

Pathogenesis of Ankle Osteoarthritis

Clinical experience and published reports of the treatment of ankle osteoarthritis indicate that primary ankle osteoarthritis is rare and that posttraumatic arthritis, which develops after ankle fractures or ligamentous injury, is the most common cause of ankle osteoarthritis.26,44,91,94,110 Over a 13-year period in the senior author’s practice, 445 of 639 patients (70%) with Kellgren-Lawrence grade 3 and 4 ankle arthritis were posttraumatic cases, and only 46 (7.2%) had primary arthritis (Table 21-1).⁹¹ The most common causes of posttraumatic arthritis were rotational ankle fractures (37%) and recurrent ankle instability (15%) (Table 21-2). Curiously, 61 patients in this group gave a history of a single major ankle sprain that never healed completely. Of the 46 patients who were classified as having primary ankle osteoarthritis, 23 (59%) had clinically significant hindfoot malalignment, which emphasizes the intrinsic resistance of the ankle joint to primary articular degeneration and the relative rarity of primary osteoarthritis of the ankle (Table 21-3).

Table 21-1

All Ankle Arthritis Patients * in the Senior Author’s Practice over a 13-Year Period

SD, standard deviation.

^*Total of 639 patients.

Table 21-2

Posttraumatic Ankle Arthritis Patients in the Senior Author’s Practice over a 13-Year Period

SD, standard deviation.

Table 21-3

Primary Ankle Arthritis Patients in the Senior Author’s Practice over a 13-Year Period

In a similar study, Valderrabano et al ¹⁰⁵ reviewed the etiology of symptomatic end-stage ankle arthritis for patients presenting to their clinic over a 10-year period. The results of their study were similar to ours in that they also showed that posttraumatic arthritis was the most common cause of end-stage ankle arthritis (78%). These authors also found that malleolar ankle fractures were the most common cause of posttraumatic arthritis (39%), followed by ligamentous injuries (16%) and pilon fractures (14%). Arthritis secondary to other causes was present in 13% of patients, whereas primary osteoarthritis occurred in only 9% of patients.¹⁰⁵

Patients with neuropathic degenerative disease of the ankle and degenerative disease after necrosis of the talus with collapse of the articular surface make up a small portion of the patients with degenerative disease of the ankle. Primary osteoarthritis is the most common diagnosis for patients treated with hip and knee replacements. In contrast, posttraumatic osteoarthritis is the most common diagnosis in patients treated with ankle arthrodesis or replacement. This observation raises the possibility that the ankle may be at least as vulnerable as, and perhaps more vulnerable than, the hip and knee to development of severe posttraumatic osteoarthritis.

The relative rarity of primary osteoarthritis of the ankle might be the result of the congruency, stability, and restrained motion of the ankle joint, tensile properties and metabolic characteristics of ankle articular cartilage, or a combination of these factors. The thinness of ankle articular cartilage and the small contact area, leading to high peak contact stresses, can make the joint more susceptible to posttraumatic osteoarthritis. In particular, the thinner, stiffer articular cartilage of the ankle may be less able to adapt to articular surface incongruity and increased contact stresses than the thicker articular cartilage of the hip and knee, and the contact stresses may be higher in the ankle.

Joint injuries can cause damage of articular cartilage and subchondral bone that, if not repaired, creates articular surface incongruencies and decreases joint stability. Long-term incongruence or instability can increase localized contact stress. The ankle osteoarthritis that occurs after injuries appears to follow a pattern that is consistent with the hypothesis that posttraumatic ankle osteoarthritis results from elevated contact stress that exceeds the capacity of the joint to repair itself or adapt. According to this hypothesis, the development of posttraumatic ankle osteoarthritis progresses through three overlapping stages: articular cartilage injury, chondrocyte response to tissue injury, and decline in the chondrocyte response.

Neuropathies and necrosis of the talus that cause incongruity of the articular surface also lead to secondary ankle osteoarthritis. Patients with neuropathies can develop rapidly progressive joint degeneration after minimal injury or in the absence of a history of an injury. This can occur because the loss of positional sense leads to undetected ligamentous or articular surface injuries that create localized regions of increased contact stress. Articular surface incongruence resulting from necrosis of the talus can have the same effect.

Consistent with the hypothesis that excessive contact stress causes degeneration of ankle articular cartilage, the significant residual joint incongruity and severe disruption of the ankle joint articular surface predictably lead to joint degeneration, commonly within 2 years of injury. Advanced joint degeneration can also develop within 2 years after injuries that cause relatively little apparent damage to the articular surface. In some of these latter cases, the joint surface might have sustained damage that is not apparent by radiographic evaluation. In others, joint instability resulting from alterations of the anatomy of the mortise, such as spreading of the distal tibiofibular syndesmosis, shortening and rotation of the fibula, or capsular and ligamentous laxity, can cause degeneration of the joint. However, some patients develop progressive joint degeneration after ankle injuries without apparent articular surface damage, alteration of the joint anatomy, or joint instability. On the other hand, some patients with articular surface incongruity or joint instability do not develop progressive joint degeneration. The pathogenesis of posttraumatic ankle osteoarthritis is therefore more complex than it appears and needs extensive further study.

Impact of Ankle Arthritis

The significant impact of knee and hip arthritis on patient function is well documented, and recent attempts have been made to better quantify the impact ankle arthritis has on the lives of patients. Several studies have established a baseline for the functional limitations on patients with end-stage arthritis of the ankle.2,39,98 Using different measures, all of these studies have shown that ankle arthritis severely impacts the lives of patients. Glazebrook et al³⁹ used the Short Form-36 (SF-36) to prospectively compare patients with end-stage ankle to patients with end-stage hip arthritis. These authors showed that patients with ankle arthritis had SF-36 scores equivalent in severity to patients with end-stage hip arthritis and were two standard deviations below normal patients. Similarly, using the Musculoskeletal Functional Assessment (MFA), Agel et al² showed that patients with ankle arthritis scored three times worse than normal patients. In a more recent study, Segal et al⁹⁸ correlated the SF-36 and MFA scores with gait kinematics and step count. These authors found that patients with ankle arthritis had reduced function based on SF-36 and MFA scores. In addition, patients showed reduced ankle motion, ankle plantar flexion moment, peak ankle power absorbed, and peak ankle power generated in the affected limb when compared with the normal contralateral limb.

Taken together, the results of these studies show that patients are severely affected by ankle arthritis. The severity of the impact on function has been shown with different measures of functional outcomes, as well as quantified with gait analysis.

Approach to the Patient with Ankle Arthritis

History and Physical Examination

Taking a good history and performing a careful physical examination are essential. First, determine if there is a clear history of trauma contributing to the development of ankle arthritis. Although a past fracture is the most common cause of ankle degeneration, recurrent sprains (or even one major sprain without resolution) can also be responsible. Ankle arthritis is usually not the first manifestation of generalized inflammatory arthritis, but certainly it is relatively common in patients with severe multiarticular disease. Hemophilia, gout, talar avascular necrosis (AVN), or infection can all contribute to the development of end-stage arthritis.

Next, determine which activities cause ankle pain or limit function. Walking uphill causes bony impingement in the anterior ankle or the talonavicular joints. Pain caused by downhill walking suggests a problem at the back of the ankle and can include posterior soft tissue impingement, trigonal problems, or synovial chondromatosis. Pain that is primarily caused by walking on uneven ground and that is experienced in the back or lateral aspect of the ankle can indicate subtalar joint disease. Subfibular pain might not be from the ankle or subtalar joints but might be due to malalignment and secondary bony impingement of the calcaneus on the lateral process of the talus, peroneal tendons, or fibula. Posteromedial pain typically indicates a tendon problem rather than ankle arthritis.

The examination should be done with the patient sitting and standing. In the seated position, a careful vascular and neurologic assessment can be made, joint motion estimated, and points of maximal tenderness identified. The ligaments around the ankle should be tested for stability. All major extrinsic tendons need to be palpated to determine if there are associated tendinopathies. Alignment of the foot should be evaluated. Patients with recurrent instability often have a declinated first ray, whereas patients with severe flatfeet and secondary ankle disease often have clinical instability of the medial column. In patients with rheumatoid arthritis, careful attention should be given to the skin and nails to rule out punctate infarcts suggestive of ongoing vasculitis.

The standing and walking examinations complement the seated examination. Alignment of the hindfoot is assessed from behind. Excessive varus or valgus angulation of the heel should be noted. Restriction of ankle motion can lead to early heel rise or back-knee gait. The posture of the forefoot upon striking the ground should be noted. Patients who load the lateral part of the foot might have fixed varus deformity of the ankle or transverse tarsal region.

Ankle Joint Imaging

Radiography

Plain radiographs should be taken with the patient standing whenever possible. At our center, we have a standard minimum series of radiographs taken for patients with ankle problems. These include standing ankle lateral, anteroposterior, mortise, and hindfoot alignment views.⁸⁹ The hindfoot alignment view is particularly important in situations where the heel is in varus or valgus and the ankle has coronal plane tilting (Fig. 21-5). If we are considering any surgery distal to the tibiotalar joint, we also obtain standing views of the entire foot.

Figure 21-5 A, The hindfoot alignment view is taken with the patient standing on a platform, with the toes pointed straight ahead at the film plate. B, The x-ray beam is directed toward the ankle in a posterior to anterior direction, tilted 20 degrees caudad. The film is placed near the toes, also tilted 20 degrees, oriented exactly perpendicular to the beam.⁸⁹ A line passing down the central longitudinal axis of the tibia should bisect the calcaneus.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) is a very useful adjunct for imaging the ankle. It is excellent in delineating abnormalities of soft tissues around the ankle joint. However, in assessing ankle arthritis, the value of standard MRI is often limited. First, any hardware near the ankle joint generates major artifacts that obscure visualization of articular features. Second, the articular surfaces of the ankle are naturally close packed and congruous. Unlike the knee, where joint surfaces are not congruent, there is no clear separation of articular surfaces of the tibiotalar joint. Third, the now standard, hospital-based MRI, using a 1.5-tesla (T) magnet, only captures three to four pixels across a healthy ankle articular surface. Instruments with smaller magnets (i.e., 0.50 or 0.75 T) cannot capture any articular features of normal ankles. In cases of degeneration, current standard MRI magnets cannot easily distinguish focal articular features, even with distraction.

Computed Tomography

Advances in computed tomography (CT) have revolutionized ankle imaging. The total time for scanning has been reduced to 2 to 5 minutes compared with an average capture time of 20 minutes for MRI. As a result, CT images are much less susceptible to motion artifacts. The 4- and 16-slice helical units now capture anisotropic, 1-mm ⁵⁷ voxel-based data sets that can be postprocessed to show two- or three-dimensional renderings of any feature of interest (e.g., bone, tendons, cartilage). Intraarticular injection of contrast material before scanning can be used to enhance the accurate visualization of ankle articular features (Fig. 21-6).³² Computed tomography also has the large advantage over MRI of being able to work in an environment near retained hardware. CT arthrograms to delineate if a patient is suffering from global or focal tibiotalar arthritis or evaluate the condition of the subtalar joint have become a common adjunct to plain radiography at our center.

Figure 21-6 Injection of contrast media can facilitate the visualization of articular cartilage during computed tomography (CT) evaluation of a joint. If articular cartilage is present, it appears black compared with the lighter contrast material or adjacent bone. Preoperative sagittal (A) and coronal (B) CT images for a patient being evaluated for ankle arthrodesis show the ability of the technique to delineate articular cartilage. For this patient, the status of the articular cartilage for the tibiotalar and subtalar joints were evaluated.

Selective Injections

Selective injections are used to help identify the source of pain for patients who have clinical or radiographic findings that suggest more than one focal source of pain.⁵⁵ Before the injection, the patient is asked to perform activities that cause pain in the ankle (e.g., walk on uneven ground, walk up or down stairs, run). Diagnostic injections are done under fluoroscopic control. Contrast dye is first instilled to confirm the exact location of the injection. This is followed by injecting a local anesthetic. If the pain is not reduced by at least 75%, we look for a second source. In a study of foot and ankle fusion patients, Khoury et al⁵⁵ reported that the reduction in pain after an intraarticular injection correlated to the response from surgery.

Global Ankle Arthritis Versus Focal Ankle Arthritis

When evaluating the patient with ankle pain and apparent arthritis, one of the first tasks of the clinician is to determine whether the problem is global (affects the majority of the joint), or focal (affects a specific region of the joint). Inflammatory arthritides such as rheumatoid disease and seronegative spondyloarthropathies are, by definition, global processes. Similarly, hemophiliac, gouty, crystalline deposition, and septic arthropathies are diffuse joint processes. Intraarticular pilon fractures and neuroarthropathic fractures that cause ankle arthritis generally induce a global arthritic response, especially with multiple fracture lines involving the tibiotalar joint. Conversely, tibial shaft malunions, ankle instability, and foot malalignment problems that lead to ankle cartilage loss initially often cause well-localized focal problems.

Treatment of Global Ankle Arthritis

Conservative Treatment

Little has been written about the nonoperative treatment of diffuse ankle arthritis. Indeed, no retrospective or prospective clinical trials have been reported. Nonoperative treatment is based on experience and patient preferences. In our experience, the efficacy of nonsteroidal antiinflammatory drugs (NSAIDs) varies. Care must be exercised when prescribing these medications because of their sometimes substantial side effects.

A judiciously timed injection of the joint with steroids can help a patient enjoy an important life event (wedding, vacation). We do not like to give repeated injections of steroids because of the catabolic risks to soft tissues. Recently, there has been an increased interest in the use of injectable viscosupplementation for the treatment of ankle arthritis.* Review of the studies related to this subject show a large incidence of industry support and need to be interpreted with an understanding that this might introduce bias. Studies have shown an improvement in Ankle Osteoarthritis and American Orthopaedic Foot and Ankle Society (AOFAS) scores with hyaluronic acid injection; however, this improvement was no greater than control saline injection.^18,24,88 Although the benefit of viscosupplementation beyond placebo effect is still controversial, there does not appear to be any significant adverse events associated with the injections.^99,100,109 The authors do not routinely use hyaluronic acid injections for the conservative treatment of ankle arthritis unless there is a contraindication or need to delay surgery. In these situations, viscosupplementation is considered as an alternative to repetitive corticosteroid injections.

The standard nonoperative treatment for end-stage ankle arthritis is mechanical unloading. A cane can be very helpful. If patients accept the cosmetic and functional limitations of an ankle–foot orthosis (AFO), they often obtain partial pain relief. We like to use an AFO that can be molded to the contour of the posterior calf muscles. This permits some unloading of the ankle. The two designs that appear to work best are a leather ankle lacer with an imbedded polypropylene shell for structural support or a calf lacer AFO.⁹² The former fits in a shoe; the latter requires metal drop locks fixed to a single shoe (Fig. 21-7). Adding a solid ankle cushion heel (SACH) and a rocker sole can help by further limiting ankle motion.

Figure 21-7 Brace treatment of ankle arthritis is directed at limiting motion and reducing axial loading. The two braces we use are a custom-made leather-lined polypropylene ankle–foot orthosis (A) and a calf lacer with metal drop locks fitted to a rocker-bottom shoe (B). These usually are made with lace-up straps, but they can be fitted with hook-and-loop (Velcro) straps, as shown in A, for patients with profound loss of grip strength from end-stage rheumatoid disease.

Operative Treatment

The decision to operate on ankle arthritis requires a clear assessment of the patient’s functional needs and a complete understanding of the cause of the patient’s problem. Isolated, primary global ankle osteoarthritis is relatively rare. More commonly, malalignment secondary to trauma, ligamentous instability, or foot deformity is present with painful and arthritic ankles. Regardless of treatment strategy, reestablishing normal foot alignment encourages improved foot function. With total ankle replacement, perfect coronal and anteroposterior alignment of the ankle holds the greatest promise of giving long-lasting function and low wear rates. With ankle fusion, correct alignment helps to maintain residual natural adjacent joint motion, especially in the subtalar joint, and can delay the development of secondary hindfoot arthritis.

The indications for surgery continue to evolve as techniques change and evidence for effectiveness accumulates. Since the mid-1990s, we have witnessed the emergence of several alternative strategies to treat end-stage ankle arthritis. Periankle osteotomies hold promise of prolonging ankle function by redistributing forces across joints. Ankle joint distraction with tensioned wires can similarly prolong ankle function. Ankle replacement is emerging as a viable alternative for selected patients. Of all the surgical techniques, though, ankle fusion remains the workhorse of surgical reconstruction.

Primary Ankle Fusion

Surgical Considerations

In a general sense, ankle fusion has a few clear advantages over other techniques. When the pain originates within the ankle joint, a successful arthrodesis usually eliminates it. Short-term results and complication rates have been markedly improved by modern techniques of limited periosteal stripping, rigid internal fixation, and meticulous attention to alignment and position. Pain relief is more reliable with fusion than with most other techniques. Secondary operations, other than occasional hardware removals, are relatively rare.

However, ankle fusion is not entirely without problems. First, tibiotalar bone-bridging after attempted fusion surgery is not completely reliable, and reported rates of initial fusion range from 60% to 100%.*

Second, initial pain relief can be elusive if other causes of pain are present or revealed by ankle fusion. Third, functional limitations are common, even with clinically successful fusions (Table 21-4).⁷⁵ Fourth, shoe modifications may be needed to improve the transition from heel strike and toe-off, including use of a SACH or a rocker-bottom sole. Finally, accelerated degeneration of other foot joints, especially the subtalar and talonavicular joints, can occur after ankle fusion.¹⁷ This degeneration, in turn, can lead to further bracing or fusion surgery and must be considered when contemplating an ankle arthrodesis for a young patient.

Table 21-4

Functional Limitations after Ankle Fusion for 28 Highly Satisfied Patients ⁷⁶

After an ankle fusion, approximately 50% of patients demonstrate arthroses distal or proximal to the fusion site within 7 years. Although most of these changes are seen on radiographs, their presence at 7 years does not bode well for what will develop at these joints over the next 20 to 30 years. Many factors probably affect the onset of this arthrosis besides the increased stress. One factor is probably related to the overall stiffness or laxity of the surrounding joints. The stiffer the surrounding joints, the less the patient is able to dissipate the increased stress created by the fusion compared with a patient with more joint laxity. Because an arthrodesis is often performed on a traumatized limb, the adjacent joints, although not demonstrating arthrosis, might have sustained tissue damage at the time of the initial injury that makes them more prone to develop arthrosis when subjected to increased stress. Patients with generalized flexibility and ample residual foot flexibility after the development of ankle arthritis tend to have improved outcomes after fusion than those who have intrinsic stiffness.

Once it has been decided to perform an arthrodesis, the next most critical factor is to establish the proper alignment of the fusion site. To do this, the surgeon must consider the entire lower leg and not just the foot. The position of the knee or the bow of the tibia, which can occur either naturally or because of prior trauma, must be carefully examined when planning the arthrodesis. The alignment of the extremity distal to the fusion site is also important to create a plantigrade foot. Sometimes when an ankle arthrodesis is being contemplated, an underlying postural deformity of the foot precludes placing the ankle in the optimal position. If a patient has a significant forefoot varus deformity, the ankle joint needs to be aligned in a little more valgus position to create a compromised plantigrade foot. Placing the ankle into its normal alignment of 5 degrees of valgus can keep the patient from placing the forefoot flat on the ground. Conversely, if the patient has a cavus foot with a fixed forefoot valgus, the ankle fusion might be placed into neutral or slight varus to create a foot that is as plantigrade as possible. Sometimes a selective osteotomy or arthrodesis of the forefoot is necessary to create a plantigrade foot after an ankle fusion if it cannot be created by the fusion alone.

The biomechanics of the foot dictates its optimal alignment. When the subtalar joint is placed into an everted position, it provides instability to the joint, creates flexibility of the transverse tarsal joint, and results in a supple forefoot. When the subtalar joint is in an inverted position, it provides stability to the joint, locks the transverse tarsal joint, and creates a rigid forefoot. It is therefore important to place the subtalar joint in 5 to 7 degrees of valgus when a fusion is done to maintain flexibility of the forefoot. A fusion in varus position creates a rigid forefoot and increased stress under the lateral aspect of the foot, which is poorly tolerated by the patient.

Soft tissue considerations are paramount to a good surgical result. Avoid making incisions through contracted, scarred skin because wound sloughs and soft tissue coverage are challenging in this anatomic region. When making an incision, the surgeon must always be cognizant of the location of the cutaneous nerves around the ankle. Although cutaneous nerves tend to lie in certain defined anatomic areas, great variation exists. It is therefore important to be always on the lookout for an aberrant cutaneous nerve as the incision is continued through the subcutaneous tissues. The most common nerves injured during ankle surgery are the medial branch of the superficial peroneal nerve, encountered in anterior longitudinal approaches, and the anterior branch of the sural nerve, encountered beneath the lateral malleolus during transfibular approaches. The cutaneous nerves can be quite superficial and easily transected or can become adherent within scar tissue. If this occurs, a painful scar or dysesthesias distal to the injury can cause the patient to be dissatisfied with an otherwise a successful fusion.

Basic Surgical Principles

Respect the soft tissues. Retract carefully. Meticulously avoid local cutaneous nerve injury or entrapment during all stages of the procedure, including closing of the wound. Remove all cartilage, and feather and penetrate into subchondral bone. Create congruent cancellous surfaces that can be placed into apposition to permit a fusion to occur. Use bone graft or bone graft substitutes only to fill defects. Stabilize the arthrodesis site with rigid fixation, if possible. Align the hindfoot to the lower extremity and the forefoot to the hindfoot to create a plantigrade foot. For really stiff feet, slight (less than 5 degrees) dorsiflexion may be preferred. Immobilize the ankle and limit weight-bearing activity until bone-bridging is certain.

Open Ankle Arthrodesis

Indications

The main indication for ankle arthrodesis is arthrosis that causes pain or deformity (or both) of the ankle joints. The arthrosis may be primary or secondary. The secondary form is caused by posttraumatic changes, postsepsis condition, avascular necrosis of the talus, failed total ankle replacement, or malalignment from a paralytic deformity. In some cases, a subtalar or transverse tarsal joint fusion (or both) is included, which results in a pantalar arthrodesis.

Position of Arthrodesis

The desired position of the arthrodesis is as follows:

Dorsiflexion/plantar flexion: Neutral

Varus/valgus: 5 degrees of valgus

Rotation: Equal or slightly more externally rotated than the opposite extremity

Posterior displacement: Anterior aspect of the talar dome is brought to anterior aspect of tibia

Under some circumstances, such as weakness of the quadriceps muscle, the arthrodesis can be placed with the foot in 10 degrees of equinus, which will help to stabilize the knee joint. Conversely, for really stiff feet, slight (<5 degrees) dorsiflexion may be preferred. In general, however, both equinus and calcaneus positioning should be minimized to prevent a back-knee thrust with the equinus and excessive weight bearing by the calcaneus. The foot is externally rotated slightly (5 degrees) compared with the other side to allow normal knee motion. This, in turn, avoids the problem of requiring the knee to externally rotate in stance phase, which can result in gradual laxity of the medial collateral knee ligament.

Mann Technique

Preparation

1. Alignment of the normal and abnormal limb is assessed to evaluate the degree of deformity. This is done by aligning the patella parallel to the operating table and observing the degree of rotation of the foot in relation to a perpendicular line from the patella. In this way, the surgeon can determine the degree of internal or external rotation of both the normal and the abnormal extremity. The advantage of this measurement approach is that it is relatively easy to check the alignment of the arthrodesis site at surgery.

2. A thigh tourniquet is applied, and a sandbag is placed under the ipsilateral hip to enhance the visibility of the lateral side of the foot and ankle.

Lateral Approach

3. The skin incision begins approximately 10 cm proximal to the tip of the fibula, is carried down over the shaft of the fibula, and then swings gently distally another 10 cm toward the base of the fourth metatarsal. Although this incision extends between nerves, with the sural nerve passing posteriorly and the superficial peroneal nerve passing anteriorly, the surgeon must be aware of an anterior branch of the sural nerve that might pass through the plane of the incision (Fig. 21-8A and Video Clip 21).

Figure 21-8 A, Line of skin incision. B, Diagram demonstrates cuts made in the fibula, distal end of tibia, and talus. Incision is between the superficial peroneal nerve and sural nerve. C, The cut in the tibia is made perpendicular to the long axis of the tibia. D, After the cut has been made in the talus, absolute bone apposition should exist without tension when the foot is in neutral position with regard to dorsiflexion/plantar flexion. E, Model demonstrates sites for placement of two screws across arthrodesis site: one on the plantar aspect at the junction between the neck of the talus and body and the other in the lateral process. F, The arthrodesis site is stabilized with two K-wires. Two 3.2-mm drill bits are placed from distal to proximal across the arthrodesis site, marking the site of screw placement. G, After placement of 6.5-mm screws across arthrodesis site. H, Placement of screws in lateral and anteroposterior projections. I, Preoperative and postoperative radiographs demonstrate ankle arthrodesis performed with technique described. J, Ankle arthrodesis, preoperative and postoperative radiographs. A washer was used in the screw placed in the lateral process because the bone was soft. K, Ankle arthrodesis, preoperative and postoperative radiographs. The medial malleolus has been removed to displace the talus medially to gain better alignment of the lower extremity. L, Postoperative radiographs after ankle arthrodesis using third screw from side to gain increased fixation.

4. The skin flaps are developed to create a full-thickness flap along the skeletal plane. The periosteum is stripped from the fibula anteriorly and posteriorly, and the incision is carried on distally to expose the posterior facet of the subtalar joint and the sinus tarsi.

5. The dissection is carried across the anterior aspect of the tibia and ankle joint. With a periosteal elevator, the surgeon strips soft tissue from the distal end of the tibia, ankle joint, and proximal talar neck and then medially to the medial malleolus. Care is taken not to dissect distally over the neck of the talus to protect the blood supply into the talus.

6. The fibula is osteotomized approximately 2 cm proximal to the level of the ankle joint and beveled to relieve the sharp prominence (Fig. 21-8B). The distal portion of the fibula is removed by sharp and blunt dissection to expose the lateral aspect of the tibia and ankle joint as well as the posterior facet of the subtalar joint. As this is carried out, the peroneal tendons are reflected posteriorly.