Tarsal Coalition

The normal motion of the subtalar joint during walking involves both rotation and gliding. The axis of motion of the subtalar joint is defined as a line deviated 42º from the horizontal surface and 16º medially or internally rotated from a line extending from the center of the calcaneus to a point between the first and second metatarsals.

During stance phase, the subtalar joint rotates from a position of 4º external-valgus to 6º internal-varus. This motion compensates for the horizontal external rotation of the tibia during this phase. When the internal rotation is restricted by the coalition, the tarsal joints have to compensate, with flattening of the foot and loss of longitudinal arch and a valgus appearance in the horizontal plane. This leads to an adaptive shortening of the peroneal tendons, reactive peroneal spasm, and so-called peroneal spastic flatfoot. Prolonged restriction of motion eventually may lead to posterior facet arthrosis of the subtalar joint.

Another motion of the subtalar joint lost is the gliding motion demonstrated during foot dorsiflexion. Cineradiographic studies have demonstrated a hinge motion instead, with widening at the plantar aspect of the midtarsal joints, narrowing at the dorsal surfaces, and overriding of the navicular on the talar head at maximum dorsiflexion. The traction effect of the later motion on the ligaments and capsule of the talonavicular joint is thought to be the genesis of the talar beaking seen in many radiographs of tarsal coalition.

The pain symptoms associated with tarsal coalitions may therefore be attributed to the following:

• Ligament sprain
• Peroneal muscle spasm
• Sinus tarsi irritation
• Subtalar joint irritation
• Arthritic changes

The variability of the symptoms in different patients may be due to the variability of subtalar restriction of the different coalitions. The specific but different ages of presentation of different coalitions would then be due to the different ages at which coalitions ossify and thus restrict subtalar motion.

Pfitzner first proposed that the coalitions were the result of incorporation of accessory ossicles into the adjacent normal tarsal bones. This theory received some support from various authors, including Slomann, Badgley,[7] and Harris and Beath.[8] Leboucq first proposed in 1890 that the failure of segmentation of primitive mesenchyme was the cause of tarsal coalitions. His theory was supported by Solger (1890), Dwight (1907), Trolle (1948), and Jack (1954).[9]

In 1955, Harris found evidence of mesenchymal coalitions in fetal cadavers, conclusively proving Leboucq's theory. Through the work of Leonard, Wray, Herndon, and others, it is believed today that tarsal coalition is inherited in an autosomal dominant fashion with high penetrance.

The incidence of talocalcaneal coalitions in the general population is thought by most authors in the literature to be approximately 1%. Various authors have credited Pfitzner with the first incidence study but quote different percentages, ranging from 1.5-6% for the overall incidence of tarsal coalitions to 2.9% for calcaneonavicular coalitions.

Harris and Beath found 74 cases in 3600 consecutive prospective army recruits examined.[8] Rankin and Baker found 24 cases in approximately 60,000 basic trainees at Fort Dix from 1971 to 1972 (0.4%),[10] whereas Vaughn and Segal found 28 coalitions in 2000 cases of painful feet.[11] Shands and Wentz found an incidence of 0.9% in cases of painful flat feet in a pediatric clinic.[12] Leonard found that 39% of relatives of patients with symptomatic coalitions had asymptomatic coalitions.[13] Snyder et al found that 91 of 223 (63%) young patients with ankle sprains had radiographic findings suggestive of tarsal coalitions.[14]

In an evaluation of 2982 complete skeletons collected from 1910 through 1940, Cooperman et al found that 26 of them had calcaneonavicular coalitions (incidence of 1.2%).[15] Leonard found that all of the 76% of the first-degree relatives who had radiographic evidence of tarsal coalitions were asymptomatic.[13] This would suggest that the actual incidence in the population may be higher than was previously thought.

In terms of the distribution of types, Stormont and Peterson found in a review of the literature that 48.1% of the tarsal coalitions presented were talocalcaneal and 43.6% were calcaneonavicular. In their own series, the distribution was 53% calcaneonavicular and 37% talocalcaneal. The literature overall mentions a bilaterality rate of 22-60% in talocalcaneal coalitions and 40-68% in calcaneonavicular coalitions, with most being just over 50%.[16]

Sex ratios vary, with male-to-female ratios ranging from 1:1 to 4:1.

Most studies have found no statistical difference in racial distributions. However, a study by Park found a prevalence of 6% in 839 East Asian (Korean) patients, significantly higher than the figures reported for white patients.[17]

Although some authors have found that nonoperative treatment methods have the potential to achieve pain relief and prevent or delay surgery for symptomatic tarsal coalitions,[18] such treatment approaches have not been found to be uniformly successful. Accordingly, most long-term studies have focused on outcomes following surgical treatment. Proper patient selection is a prerequisite for optimal results.

Patients with extensive or multiple coalitions typically undergo fusion procedures, and those with less extensive or isolated coalitions undergo resection and soft-tissue interposition of their coalitions. Talar beaking must be critically evaluated but is not necessarily a contraindication for tarsal coalition resection. The vast majority of calcaneonavicular coalitions can be resected with the expectation of successful long-term results. Resection of symptomatic talocalcaneal coalitions yields optimal results when the coalition involves no more than about one third of the total joint surface.

Skwara et al assessed the foot loading characteristics and foot function in 10 patients (15 feet) who underwent surgical correction for tarsal coalition. In 12 feet, calcaneonavicular coalition was present; in three, talocalcaneal. Overall, the clinical results were fair regarding improvement of pain, range of motion (ROM), and walking distance. The American Orthopaedic Foot and Ankle Society (AOFAS) scores were fair as well (mean, 78.1) at follow-up. Gait analysis revealed alterations in kinematic and kinetic parameters for the operated foot, and pedobarographic analysis showed altered loadings for heel and forefoot.[19]

Khoshbin et al used an administrative database to follow up a cohort of patients who had undergone surgical resection of tarsal coalition with the aim of determining the rates and risk factors for subsequent resection or arthrodesis.[20]  Their study cohort consisted of 304 patients aged 8 years or older (average, 24.2 y) who were treated between July 1994 and August 2009. Patients with nonidiopathic coalitions were excluded. Of the 304 patients, 26 (8.6%) underwent subsequent resection, and 16 (5.3%) underwent mid- or hindfoot arthrodesis.

According to this study, the need for future fusion was more likely only if the primary resection had been performed at an academic hospital or if the patient had undergone concomitant arthrodesis at primary resection of the coalition.[20]  The incidence of reoperation after primary tarsal coalition resection was low. More than 85% of the patients had not required an additional operative intervention at an average of 9 years after the initial resection. The investigators suggested that primary treatment of tarsal coalition with resection and concomitant arthrodesis increases the risk of requiring a second fusion in the future.

Mahan et al evaluated patient-reported data on functional outcome after tarsal coalition resection in children and adolescents.[21]  In addition to assessing medium-term (>2 y) outcomes, they compared outcomes of calcaneonavicular coalitions with those of talocalcaneal coalitions. In 73% of their patients, activities were not limited by pain. No significant differences in outcome were noted between patients treated for calcaneonavicular coalitions and those treated for talocalcaneal coalitions.

Yildiz et al evaluated long-term (minimum follow-up, 5 y) functional and radiologic outcomes after talocalcaneal and calcaneonavicular coalition resections and found the outcomes to be similar for the two types of resection.[22] Subtalar joint osteoarthritis developed in all of the patients who underwent talocalcaneal resection and most of those who underwent calcaneonavicular resection, but no functional impairment was noted.

Patients with tarsal coalition usually present during the second decade of life,[23] but presentations in adulthood have been documented.[24] Complaints include mild pain deep in the subtalar joint and limitation of range of motion (ROM). The patient usually presents after some traumatic event, such as an ankle sprain. Indeed, what seem to be repetitive sprains should raise suspicions of tarsal coalition. Often, the symptoms are relieved by rest and aggravated by prolonged or heavy activity.

In cases of calcaneonavicular coalition, pain may be more superficial and may originate from the area of the coalition in the sinus tarsi. Palpation may elicit pain at the calcaneonavicular junction laterally. In cases of talonavicular coalitions, the pain is usually more vague, but tenderness may be elicited with palpation of the middle facet region, just anterior to the medial malleolus.

Different coalitions ossify at different ages, possibly explaining the difference in ages of presentation of different coalitions. Slight limitations of ROM and mild valgus are thought to be possibly all that is present prior to ossification. Talonavicular coalitions begin to ossify in children aged 3-5 years; calcaneonavicular coalitions begin to ossify in children aged 8-12 years; and talocalcaneal coalitions begin to ossify in adolescents aged 12-16 years.

Loss of subtalar motion and valgus position of the hindfoot become more apparent as the coalition ossifies, leading to the appearance of pes planus. Middle-facet talocalcaneal coalitions are associated with the greatest loss of subtalar motion and are the most likely to generate valgus.

This loss of subtalar motion may be evaluated with the heel-tip test, wherein the examiner supinates the foot of a standing patient by raising the medial border of the forefoot and keeping the heel and lateral border in contact with the floor. In patients with limited or decreased subtalar motion, compensatory external tibial rotation is decreased, and the patella is not observed to rotate outward as in persons without tarsal coalition.

Other tests include toe standing and the Jack toe-raise test; both demonstrate the fixed nature of the pes planus and the loss of hindfoot inversion.[25]

Studies have shown that patients with neutral hindfeet have fewer symptoms; this is likely secondary to decreased peroneal shortening. Cases of hindfoot varus have been reported in the literature in conjunction with calcaneonavicular coalitions. Repeated sprains may actually mask loss of subtalar motion by allowing motion because of ligamentous laxity.

Although peroneal spasticity was often attributed to tarsal coalitions in the older literature, this actually is relatively infrequent. Peroneal spasticity may be intermittent or continuous, varying with activity or stress of the ankle joint.

Tibialis spastic varus foot has also been described, albeit rarely. Kurashige et al reported a case of tibialis spastic varus foot with calcaneonavicular coalition in an 11-year-old boy with intellectual disability.[26]  Conservative treatment failed, but surgical resection of the coalition led to a good outcome.

Rocchi et al proposed a new physical finding in talocalcaneal coalition: the so-called double medial malleolus, a palpable medial prominence just below the medial malleolus.[27]  They determined this finding to be strongly associated with the presence of talocalcaneal coalition and suggested that it could be useful to help guide the ordering of confirmatory diagnostic imaging (eg, computed tomography [CT] with three-dimensional [3D] images) and surgical planning.

Standard plain anteroposterior (AP) and lateral radiographs of the foot do not visualize tarsal coalition well and thus are usually not diagnostic; however, they may be suggestive of this condition on the basis of secondary signs such as the talar beak or the anteater-nose sign.

Classic secondary signs of tarsal coalition are mainly seen in the lateral view. These include the following:

• Talar beaking seen on the anterior talar side of the talonavicular junction
• Broadening and rounding or flattening of the lateral talar process
• Narrowing of the posterior facet

The last two are signs of degenerative changes. Any rotation of the foot may result in the appearance of a tarsal pseudocoalition on the lateral view secondary to off-plane view of the subtalar joint facets.

Some have suggested that the C-sign of Lateur, a C-shaped line composed of the dome of the talus and the inferior outline of the sustentaculum tali, is pathognomonic for subtalar coalition.[31]  A study by Sakellariou et al examined lateral radiographs of 20 patients with clinical and radiographic diagnosis of talocalcaneal coalition and compared them with those of 22 asymptomatic volunteers.[32]  Computed tomography (CT) was used as the diagnostic standard. The authors concluded that the C-sign was highly sensitive and specific for the diagnosis.

However, in a retrospective review of 48 patients who had lateral ankle radiographs and CT scans for nontraumatic indications, Brown et al found that the C-sign was specific but not sensitive for a flatfoot deformity and was neither specific nor sensitive for talocalcaneal coalition.[33]

Calcaneonavicular coalition (see the image below) may be suggested by the presence of the anteater-nose sign on the lateral view corresponding to an elongated anterior calcaneal process.[34]  This radiographic sign may be found on the lateral radiographs of patients aged 9 years or older with tarsal coalition.

Oblique 45º views of the foot demonstrate a calcaneonavicular coalition 90-100% of the time. Only approximately 10% of the cases demonstrate a frank osseous coalition, with the remainder demonstrating increased proximity of the two tarsals, indistinct juxtaposed cortices, hypoplasia of the head of the talus, and flattening or narrowing of the navicular as it approaches the anterior calcaneus process.

A talocalcaneal coalition is best seen with the Harris-Beath axial or "ski-jump" view. This is taken with the patient standing on the cassette, bending 10º at the ankle. Harris and Beath recommended a 45º beam view originally, but they later expanded their views to beam angles of 30º, 35º, and 45º. Cowell recommended first taking the 45º beam view and then, if the middle and posterior facets are not well visualized, using a standing lateral view to measure the appropriate angle for the axial view.[35]

In cases where the middle and posterior facets are not parallel, two different angles would have to be measured and two corresponding axial views taken. If a middle facet coalition is present, the coalition is seen on the medial side. If the coalition is osseous, no joint is visualized. If it is fibrous or cartilaginous, the joint appears irregular and angled inferior-medial. In normal feet, the middle facet is usually parallel to the posterior facet on axial views.

In the past, tomograms were necessary to demonstrate anterior facet coalitions and to confirm the presence of more difficult middle- or posterior-facet coalitions. Since the mid-1980s, however, coronal CT has become the criterion standard in the evaluation of tarsal coalitions, particularly for assessing talocalcaneal coalitions.[36]  (See the image below.)

In 1986, Herzenberg et al correlated the use of coronal CT views to evaluate tarsal coalitions with cadaver specimens. In their study, the patient's feet were placed in a plantar position on the gantry with the CT ring in the neutral position.[37]  Other studies have used a special apparatus to dorsiflex the foot and rotate the ring to maintain a coronal axis of the subtalar joint. The advent of high-speed spiral (helical) CT scanners and advanced image reconstruction software prompted some to accept coronal reconstructions of noncoronal views, but equal diagnostic ability has yet to be conclusively established.

Upasani et al evaluated three-dimensional (3D) multiplanar CT images of calcaneonavicular coalitions and adjacent tarsal relationships in 74 feet (37 patients).[36]  In 32 of the 37 patients (86%), bilateral involvement (69 coalitions) was present. Coalitions were categorized into four types as follows:

• Type I (forme fruste), 28%
• Type II (fibrous), 23%
• Type III (cartilaginous), 45%
• Type IV (osseous), 4%

The average shape of the coalition was found to be a curved wedge, which was, on average, 16 mm wide dorsally, 7 mm wide on the plantar surface, 10 mm in length, and 25 mm in depth.

According to the authors, the shape of the cuboid correlated with the extent of ossification.[36]  In type I or type II coalitions, the cuboid extended medially plantar to the fibrous connection; in more complete type III and type IV coalitions, the cuboid was squared off and remained lateral to the osseous bridge. The authors noted that it is important to understand 3D anatomy when diagnosing milder forms of coalitions and during resection so as to avoid iatrogenic injury to the calcaneus, the head of the talus, or the cuboid.

Magnetic resonance imaging (MRI) has been advanced as another tool for evaluating cases of fibrous or cartilaginous coalitions that may not be well seen on CT; however, the evidence to date has not proved it to have a significant diagnostic advantage over CT. Although MRI has been found to be very good at detecting tarsal coalition and determining the exact type (as well as identifying associated soft-tissue abnormalities),[38]  CT is still considered to be the criterion standard and to be more cost-effective than MRI.[39]

Bone scintigraphy has been advanced as a way of confirming suspected coalitions that are not well seen on plain radiography or as a screening tool. However, the decreasing expense of CT and the ability of CT to depict detail has tended to reduce the utilization of scintigraphy.

It has been suggested that bone scintigraphy with 99mTc-diphosphonate single-photon emission CT/CT (bone-SPECT/CT) can be a useful complement to other imaging techniques for tarsal coalition, in particular by providing accurate bone pathology characterization, excluding synchronous lesions elsewhere, and offering insight into osseous structures and 3D localization of bone metabolism for surgery planning.[40]

The tissue of a tarsal coalition may be osseous, cartilaginous, or fibrous. Histopathologic analysis of resected nonosseous coalitions has revealed no evidence of neural elements. This absence of nerve tissue argues against the abnormal coalition tissue acting as a primary pain generator. Microfractures and histologic signs of bone remodeling near a coalition's boundary with normal bone have been identified and are likely pain generators via periosteal nerve fibers.

Conservative treatment of tarsal coalition includes soft shoe inserts and walking-cast immobilization. Surgical treatment is indicated in cases where conservative treatment has failed and symptoms persist.[41] Surgical treatment includes resection of the coalition before onset of degenerative changes and subtalar fusion in the case of talocalcaneal coalitions. Arthroscopic approaches to tarsal coalition resection have been developed and appear to be feasible and effective.[42]

Coalition resection is contraindicated when significant degenerative change is present in the joints adjacent to the coalition. Arthrodesis is preferable in such cases. Subtalar arthrodesis alone is contraindicated in cases of talocalcaneal coalitions in which midfoot joints show signs of degenerative change. Triple arthrodesis is indicated instead. Finally, in patients with multiple anomalies, individual resections are unlikely to establish a normal-appearing and normal-functioning foot and should therefore not be pursued.

Currently, it is not clear what should control the use of talocalcaneal coalition resection; the limiting factors of heel position and percentage of involvement do not consistently segregate good and bad outcomes. However, once global degenerative changes have begun, triple arthrodesis almost always is indicated.

Calcaneonavicular coalition

Initial conservative treatment of calcaneonavicular coalitions may include soft shoe inserts or a trial of walking-cast immobilization. These treatments have been described in the literature to extend from 3 to 6 weeks each. Cast immobilization trials have been repeated once prior to surgery if the first attempt did not relieve symptoms. Immobilization must be with the hindfoot in neutral and away from valgus.

In a number of studies involving symptomatic calcaneonavicular coalitions, conservative treatment yielded poor results. However, some authors have found that nonoperative treatment is often effective in this setting.[43, 44]

Talocalcaneal coalition

For talocalcaneal coalitions, conservative treatment decreasing subtalar motion or stress has been found to have good results in cases of first presentation with no evidence of arthritic changes.

Calcaneonavicular coalition

Given the ease and relatively good results of surgical treatment of calcaneonavicular coalitions (see the image below), coalition resection prior to onset of arthritic changes has become more commonly indicated.[45]

The procedure, as described by Badgley[7] and Cowell[35] and others, includes the following:

• Anterolateral approach over the coalition
• Resection of at least 1 cm of the coalition
• Resection of a block rather than wedge section
• Interposition of the head of the extensor digitorum brevis
• Avoidance of the talonavicular joint to prevent theoretical subluxation of the navicular over the talar head

Some modifications of the original technique include use of bone wax[46] or electrocautery to treat the remaining surfaces after bar resection and tying the interposition sutures over the plantar fascia and under the skin rather than securing them with a button over the skin.

Various long-term studies of this technique have shown excellent or good results in 77-100% of patients.

Talonavicular and talocalcaneal coalition

Surgical treatment for symptomatic talonavicular coalition traditionally has consisted of triple arthrodesis. Before the advent of computed tomography (CT), resection of the middle-facet coalition was not very successful, because of poor visualization of the coalition. Since then, such resections have become more popular and are indicated in cases in which conservative treatment has failed, visualization of the middle-facet coalition is good, and no evidence of arthritic changes is present.[47, 48]  Many authors no longer consider talar beaking to be evidence of degenerative change.

Comfort and Johnson found that there was an 80% success rate when the coalition involved one third or less of the total surface area of the subtalar joint on CT.[49] Wilde et al found that a valgus greater than 16º and a coalition surface area greater than 50% of the posterior facet on CT were predictors of poor results after resection.[50]

Luhmann and Shoenecker found that although an association existed between poor results and a heel valgus of 21º or a coalition greater than 50% of the posterior facet, some patients still had good postoperative results.[51] They therefore recommended that resection be tried as a first procedure and that patients be informed that they could still have a good result in cases where the poor predictive factors are present. Various long-term studies have shown excellent, good, or improved rates of 80-100%.

The resection is approached medially, distal to the medial malleolus. The middle facet is exposed by retraction of the flexor hallucis longus tendon. The prominent joint is resected with a rongeur, and fat is interposed.

In cases where a resection is not possible or desired, Mann and Baumgarten proposed fusion of the subtalar joint only, instead of the traditional triple arthrodesis, reasoning that any motion saved in the midtarsal joints would maintain force transfer during motion, decreasing or slowing degenerative changes in adjacent joints.[52] However, in cases where degenerative changes are apparent, triple arthrodesis is indicated, in that isolated subtalar fusion would only accelerate the changes in the midtarsal joints. In cases where skeletal immaturity is present, a Grice-Green extra-articular arthrodesis may be indicated as an intermediate procedure.

De Wouters et al reported on the use of a patient-specific, made-to-measure surgical guide for tarsal coalition resections in nine consecutive patients.[53] The guide was created by means of three-dimensional (3D) modeling from a CT scan of the foot. After being placed on the bone surface, the guide was used to orient the saw blade to resect the bone bridge at the appropriate depth. A fascia lata allograft was interposed.

Complete resection and absence of recurrence were checked on postoperative CT for patients with talocalcaneal coalitions and on radiography for those with calcaneonavicular coalitions.[53] In all cases, resection was complete, and there were no recurrences at the final follow-up. The results suggested that use of such a guide can make tarsal coalition resection easier and more reliable and improve precision.

Aibinder et al described an approach to resection of talocalcaneal coalitions that combined intraoperative CT with the use of a navigated instrument system.[54] The use of the navigated probe and burr were employed to help define the most anterior, posterior, and medial extents of the coalition, with the aim of minimizing residual coalition. The authors found that this technique reduced morbidity, with less bone removed and intact subtalar articulations preserved, and allowed an efficient, thorough, and controlled resection.

Postoperative treatment includes immobilization for 3 weeks in a nonweightbearing cast followed by 3 weeks of partial immobilization with a nonweightbearing ankle cast and range-of-motion (ROM) exercises out of the cast. This is then followed by gradual advance to full weightbearing and ROM exercises with physiotherapy. Bilateral procedures are staged so as to allow full recovery of the first foot before surgery of the second foot.

Infections and wound complications are possible with surgical resection. As previously stated, in the event of a failure to resolve symptoms in cases treated with resection of the coalition, various arthrodeses are appropriate as salvage or next-step treatment.

In the case of calcaneal navicular bar resection, violation of the talonavicular capsule may result in subluxation of the navicular on the talus, which may lead to abnormal motion of the midfoot and risk of pain and degenerative changes.

In the case of talocalcaneal coalition resections, it was found that coalitions 50% or greater in surface area than the posterior facet had a tendency to yield a poor result. This would imply that accidental resection of the middle facet 50% or greater than the posterior facet may yield a poor result.