The accessory tarsal navicular is the most common accessory bone in the foot, occurring in between 4% and 14% of the population (8, 11, 28). It is frequently bilateral and occurs more commonly in females. Geist (28) recognized a higher incidence of accessory naviculars in young patients evaluated radiologically than in cadaver studies.

McKusick (29) believed that the accessory navicular was inherited as an autosomal dominant trait. Geist (28) reported that there are three types: (a) a sesamoid bone within the substance of the posterior tibial tendon, (b) a separate bone with a true articulation (synovial joint) with the navicular, and (c) an ossicle with a synchondrosis to the main navicular.

Pain, tenderness, and callus formation may develop over the firm prominence distal to the talar head on the plantar-medial aspect of the midfoot starting in adolescence. There may be a coexistent flexible flatfoot (30, 31, 32 and 33), but there is not conclusive evidence that there is a cause-and-effect relationship between the two conditions as was historically believed (12). The prominence of an accessory navicular is in close proximity to the head of the talus, which is prominent in a flexible flatfoot. Inverting and everting the subtalar joint with one’s thumb on the prominence is helpful for differentiating the two. If the prominence moves, it is an accessory navicular. If it remains stationary, it is the head of the talus.

Individuals with an accessory navicular may present for evaluation because of the prominence, but more commonly they present because of pain at the site. The typical patient is an active adolescent girl with a history of minor trauma who presents with pain, callus formation, tenderness, redness, and, occasionally, swelling over the bony prominence. Maximum tenderness is elicited by upward pressure under the prominence. Because of the frequency of this anatomic variation in the general population, one must be careful not to assume that a radiographic finding of an accessory navicular is the cause of the foot pain without thorough evaluation (34).

An accessory navicular can usually be seen on standing AP and lateral radiographs, but a lateral oblique view (opposite to the standard medial oblique view that is generally obtained) may be necessary for identification. There are three types of accessory naviculars (35) (Fig. 29-1). Type I is a rarely symptomatic, small pea-sized sesamoid bone located in the center of the most distal portion of the tibialis posterior tendon. Type II, the most frequently symptomatic type, is a bullet-shaped ossicle joined to the tuberosity of the navicular by a syndesmosis or synchondrosis. Type III is a large, horn-shaped navicular that probably results from fusion of a type II with the body of the navicular over time.

There is proliferating vascular mesenchymal tissue, cartilage proliferation, and osteoblastic and osteoclastic activity in the tissue between the ossicle and the main body of the navicular in painful type II accessory naviculars (35). These histologic findings are consistent with healing microfractures, substantiating the opinion that pain at this site is related to chronic, repetitive stress. There are at least two other possible sources of pain. One is pain from pressure on the skin overlying the bony prominence. The other is tendinitis in the tibialis posterior, the tendon in which the ossicle resides. Any or all of these sources may exist in the same painful foot.

Accessory naviculars are for the most part asymptomatic. If symptoms do occur, a period of protection from stress and injury generally returns the patient to the asymptomatic state (32).

Cavus is a manifestation of a neuromuscular disorder with muscle imbalance, until proven otherwise. At least two-thirds of patients who seek treatment for a painful high arch will have an underlying neurologic problem, and over half of these will have Charcot-Marie-Tooth (CMT) disease (43, 44). There are many other causes of cavus foot deformity. It is helpful to consider those that cause unilateral versus bilateral deformity when developing a differential diagnosis for your patient (Table 29-1). The number of cases termed “idiopathic” cavus foot continues to decrease as diagnostic methods improve.

Calcaneocavus deformity is seen almost exclusively in children with myelomeningocele and poliomyelitis due to a specific pattern of muscle imbalance seen in many of these children.

The incidence of cavus is variable and is related to the prevalence of neuromuscular disorders at any point in time.

The clinical features depend on the underlying etiology. Some of the underlying neurologic abnormalities are known at the time of presentation and some are not, some are treatable and some are not, and some are static and some are progressive. The resultant muscle imbalance, however, always leads to progressive foot deformity. The clinical manifestations are instability of gait with frequent falling, a feeling that the ankle is “giving out,” and a history of repeated ankle sprains. Instability may be secondary to muscle weakness, sensory loss, or deformity (45). Painful callosities develop under the metatarsal heads, lateral to the base of the fifth metatarsal, and over the dorsum of the proximal interphalangeal (PIP) joints of the progressively clawing toes. In the calcaneocavus foot, callus formation occurs primarily under the calcaneus; however, exaggerated callus formation under all metatarsal heads can occasionally be seen.

A history of progressive change in foot shape and function must be ascertained, even in situations of a known underlying neurologic abnormality. A sudden increase in cavus deformity in a child with myelomeningocele or lipomeningocele could represent evidence of a tethered spinal cord.

Medical, birth, and developmental histories, as well as a review of systems, are mandatory in searching for underlying causes of cavus foot. A family history of cavus foot deformity should be investigated as an aid to diagnosis. Charcot-Marie-Tooth disease, an autosomal dominant disorder with variable phenotypic expression, is one of the most common causes of cavus foot deformity (43). Examination of other family members can often be revealing, because a subtle cavus foot may be present.

A detailed neurologic examination, including motor, sensory, and reflex testing, of the upper and lower extremities is mandatory. The spine must be examined for deformity, midline defects, hairy patches, dimples, or other evidence of spinal dysraphism. The Trendelenburg test should be performed.

The foot is examined with the child seated, standing, and walking. Cavovarus is the most common specific form of cavus foot deformity and can be caused by many different underlying neurologic disorders. The longitudinal arch is excessively elevated along the medial border of a weight-bearing foot (Fig. 29-2). The lateral border of the foot is convex and plantigrade, and the base of the fifth metatarsal is prominent and often callused. The hindfoot is in varus alignment. Atrophy of the intrinsic muscles is apparent. There may or may not be contracture of the gastrocnemius or the triceps surae. In many cases, as in most cavovarus deformities secondary to CMT, the forefoot equinus gives the false impression of a hindfoot equinus (46). The thigh-foot angle is usually neutral, despite the internal rotation deformity in the subtalar joint. The reason is that external tibial torsion is always associated with a developmental cavovarus deformity. The two rotational deformities are in opposite directions and cancel each other out. It is important to discuss this with the patient and family, because the external tibial torsion will become obvious after correction of the foot deformity.

Despite differences in the patterns of muscle imbalance that create cavovarus, the pattern of deformity development is fairly constant (47, 48 and 49). The first metatarsal becomes plantar-flexed, giving the appearance of pronation of the forefoot on the hindfoot. The deformity is flexible at first but becomes rigid with time. The plantar fascia and the other
plantar soft tissues become contracted. The incompletely ossified bones change shape due to excessive compression on their plantar aspects (Hueter-Volkmann law). The normal tripod structure of the foot becomes unbalanced. Bearing weight on the plantar-flexed first metatarsal causes the forefoot to supinate in relation to the tibia, thereby allowing the fifth metatarsal head to touch the ground. Because the forefoot is rigidly pronated in relation to the hindfoot, the subtalar joint is thereby driven into inversion, or varus (48) (Fig. 29-3). This flexible hindfoot varus deformity eventually becomes rigid as the plantar-medial soft tissues of the subtalar joint contract. The cavovarus foot, therefore, has two major rotational deformities in opposite directions from each other: pronation of the forefoot and supination (varus and inversion are other descriptive terms) of the hindfoot. It appears as if the foot is wrung out (1). Determination of the flexibility or rigidity of each deformity is important when planning an operation. Flexible deformities are treated with tendon transfers, and inflexible deformities are treated with soft-tissue releases, osteotomies, and, occasionally, arthrodeses. Coleman and Chestnut (50) devised the block test to help evaluate the flexibility of the hindfoot (Fig. 29-4). The patient stands with a block of wood under the lateral border of the foot to recreate the tripod while allowing the first metatarsal to plantar-flex. A flexible varus deformity of the hindfoot will correct to valgus alignment. One that is already contracted and rigid will not. In the first situation, surgery for deformity correction is confined to the forefoot. In the latter case, forefoot and hindfoot procedures are needed.

The arch is elevated across the entire midfoot in the calcaneocavus deformity (Fig. 29-5). The calcaneus is dorsiflexed and vertically aligned, giving it the appearance of posterior truncation. The plantar heel pad is thickly callused from excessive pressure over a small surface area.

Standing AP and lateral radiographs of the foot are indicated on initial evaluation. According to Meary (51), there is normally a straight-line relationship between the axis of the talus and that of the first metatarsal on the lateral view. In essentially all cavovarus foot deformities, those two lines intersect within the body of the medial cuneiform, indicating that site as the center of rotation of angulation (CORA) (52), with the apex of the angulation directed dorsally (Fig. 29-6A). A calcaneal pitch (CP) >30 degrees is also indicative of a cavus deformity. The hyperdorsiflexion of the calcaneus seen on the radiographs confirms that the apparent clinical equinus deformity is, in fact, forefoot equinus, that is, cavus (Fig. 29-6). Inversion or varus deformity of the hindfoot is indicated on the AP radiograph by parallelism between the talus and calcaneus and by adduction at the talonavicular joint. It is also manifest as the intersection of the axis of the talus and that of the first metatarsal (the CORA) (52) at the talonavicular joint or within the head of the talus (Fig. 29-7A).

I obtain a standing AP Coleman-type block test radiograph to document the flexibility of the hindfoot. I find that the clinical and radiographic block tests are more reliable if a Plexiglas block is placed under the lateral metatarsal heads alone with the heel on the X-ray plate/platform (Fig. 29-7).

An AP pelvis radiograph is indicated because of the known association of Charcot-Marie-Tooth disease with progressive hip subluxation and dysplasia. Standing AP and lateral radiographs of the spine are indicated if any physical examination findings suggest spinal dysraphism, spine tumor, or diastematomyelia.

Consultation by a pediatric neurologist is in the best interest of the child. Although most cavus foot deformities are caused by CMT, a progressive genetically determined neuromuscular disorder for which there is no known treatment, it is vital to identify and correct a treatable neurologic disorder before treating the foot. An MRI of the spine along
with DNA blood tests for CMT, electromyogram with nerve conduction studies, and muscle biopsy may be indicated.

Several patterns of muscle imbalance can create the cavovarus deformity. A common pattern is that seen in CMT (47, 49). Denervation begins in the intrinsic muscles of the foot (49). The weakened lumbricals allow the long toe extensors to extend the metatarsophalangeal joints and the long toe flexors to flex the interphalangeal joints, thereby creating claw toe deformities. These same forces create elevation of the longitudinal arch during gait by the windlass effect of the plantar fascia (53) (Fig. 29-8). The intrinsic muscles undergo atrophy, fibrosis, and shortening that lead to secondary contracture of the plantar fascia. This creates a bowstring between the anterior and posterior pillars of the arch that draws them closer and produces equinus of the forefoot on the hindfoot. The tibialis anterior, a dorsiflexor of the first metatarsal, becomes weak, while the peroneus longus, a plantar flexor of the first metatarsal, remains relatively strong (42). The extensor hallucis longus is involuntarily recruited in an attempt to provide additional dorsiflexion strength along the medial column of the foot, but it creates a paradoxical effect of plantar flexion due to the windlass effect of the plantar fascia. The first metatarsal starts to plantar-flex, and, with time, this creates more contracture and shortening along the plantar-medial than the plantar-lateral border
of the foot. The forefoot becomes rigidly pronated in relation to the hindfoot. The tripod effect (48) accounts for the varus position that the hindfoot must assume during weight bearing due to the fixed pronation of the forefoot. Also contributing to the varus deformity of the hindfoot is the muscle imbalance between the tibialis posterior, an invertor of the subtalar joint, that remains strong and the peroneus brevis, an evertor of the subtalar joint, that becomes weak (47). The subtalar joint eventually becomes rigidly deformed in varus because of contracture of the plantar-medial soft tissues, including those of the subtalar joint complex. Although the triceps surae does not become contracted in CMT, it does in some of the other diseases that cause cavus.

The calcaneocavus deformity develops when there is little or no strength in the triceps surae, but strength exists in the muscles that plantar-flex the forefoot and toes. The tibialis posterior, peroneus brevis and longus, flexor hallucis longus, and flexor digitorum longus (FDL) bypass the calcaneus and plantar-flex the entire forefoot on the hindfoot without creating varus. Contracture of the plantar fascia, elongation of the paralyzed triceps surae, and preservation of functional strength in the tibialis anterior contribute to the dorsiflexion posture of the calcaneus.

Muscle imbalance from both static and progressive neuromuscular disorders leads to progressive increase in the severity and stiffness of cavus foot deformities, though the rate of progression varies considerably.

The cleft foot is a rare condition with an overall occurrence rate of 1 per 90,000 live births (84). In the most common form of this condition, the malformation is bilateral, usually associated with cleft hands, and has an autosomal dominant pattern of inheritance with incomplete penetrance (84, 85 and 86) (Fig. 29-24). In the less common form, the cleft foot is unilateral without associated hand malformation, and there is no evidence of familial inheritance. The incidence of this form is 1 in 150,000 (84). Approximately 10% of cleft feet have no family history (87). Boys are affected more frequently than girls.

The apical ectodermal ridge (AER) induces normal development of limb buds by interaction with the underlying mesenchyme (88). A defect of the AER, by genetic or toxic influences, could induce osseous syndactyly by deficient differentiation, polydactyly by excessive differentiation, or a central defect by another mechanism. Watson and Bonde (89) proposed that cleft formation was the result of selective damage to a specific region in the AER localized to the second or third ray, which is the typical area for deficiency in a cleft foot. They further proposed that the extent and duration of damage to the AER would relate to both the width and depth of the defect.

Variations of the malformation known as cleft foot range from a mere deepening of the interdigital commissure to the typical central ray deficiency to the monodactylous foot (90). The width of the foot at the metatarsal heads is excessive in comparison with the hindfoot, particularly in those feet with a greater deficiency of rays. And in these feet with greater ray deficiencies, there are often exaggerated hallux valgus and fifth toe varus orientations (Fig. 29-24). Anomalies associated with cleft foot include cleft hand, cleft lip and palate, deafness, urinary tract abnormalities, triphalangeal thumb, and tibial hemimelia.

Blauth and Borisch (90) have classified cleft foot into six groups based on the number of metatarsal bones (Fig. 29-25). They identified two additional forms: a polydactylous type and a diastatic type. The latter is characterized by monodactyly with lower leg diastasis or tibial aplasia or both. Other variations of morphology in this condition, such as crossbones and synostoses, are also described by the authors (Fig. 29-26). In addition to radiographs of the foot, renal ultrasonography is indicated to rule out renal abnormalities, as they are frequently associated with cleft foot.

The pathoanatomy is one of variable degrees of failure of formation with occasional duplication and malorientation of bones. Synostoses may be seen at the margins of the cleft and in the tarsals (90). Often a hallux valgus deformity results from the pull of conjoined flexor and extensor tendons, resulting in an abduction force.

Many patients with cleft foot function well and can wear regular shoes without pain or compromise of function, while others do well with accommodative shoe wear. For such individuals, no surgical treatment is required.
However, in many cases of cleft foot, there is a marked increase in the width of the forefoot (Fig. 29-24), making it difficult or impossible to find shoes that fit comfortably. Painful callosities develop over the medial and lateral metatarsal heads.

The incidence of clubfoot is 0.93 to 1.5 per 1000 live births in Caucasians (93, 94 and 95), 0.6 per 1000 in Asians, 0.9 per 1000 births in Western Australia (96), and 6.8 per 1000 in Hawaiians, Polynesians, and Maorians (97). Boys are affected twice as often as girls are, and bilateral involvement is seen in approximately 50% of cases (93, 94, 98).

Wynne-Davies (94, 95) determined that the occurrence rate is 17 times higher than the normal population for first-degree relatives, is 6 times higher for second-degree relatives, and approximates the normal population risk for third-degree relatives. Unaffected parents with an affected son have a 1 in 40 (2.5%) chance of having another son with the disorder, while the risk to a subsequent daughter is very low. Unaffected parents with an affected daughter have a 1 in 16 (6.5%) chance of having a son with clubfoot and a 1 in 40 (2.5%) chance of seeing the deformity in another daughter. The chances are about one in four that a subsequent child will have a clubfoot if a parent and child have the disorder. There is a 32.5% rate of concordance in monozygotic twins and a 2.9% rate in heterozygotic twins.

Clubfoot is probably etiologically heterogeneous. Genetics clearly plays a part. The observations of Wynne-Davies (94, 95) on occurrence rates led her to propose that clubfoot is inherited as a dominant gene with reduced penetrance or multifactorial inheritance. Cowell and Wein (98) concluded that the data could be accounted for using a multifactorial inheritance model. Complex segregation analysis using a regressive
logistic model was used by Rebbeck et al. (99) to conclude that the probability of having clubfoot was explained by the mendelian segregation of a single gene with two alleles plus the effects of some other factors that are yet to be elucidated.

The proximate cause of the genetic message is debated with fervor, based on a small number of personal observations by a large number of investigators. There may be validity in many of the discordant studies if, in fact, clubfoot is a clinical manifestation that can result from multiple causes and mechanisms. Among the proposed theories are in utero molding (100), primary muscle lesion (101), primary bone deformity (germ plasm defect) (102), primary vascular lesion (103), intrauterine enteroviral infection (104), developmental arrest (105), primary nerve lesion (106), abnormal tendon insertion, retracting fibrosis (107), and abnormal histology (108).

Environmental factors may modulate the genetic expression of the disorder. Skelly et al. (109) demonstrated an increased risk of clubfoot for mothers who smoked during pregnancy. The data were particularly convincing because the risk increased with increased numbers of cigarettes smoked per day. Honein et al. (110) also identified maternal smoking as a significant etiologic factor. The joint effect of family history of clubfoot and maternal smoking during pregnancy was more than additive, suggesting a genetic-environmental interaction.

Other associations with clubfoot have been identified, such as increased ligament laxity in families of children with clubfoot (111) and increased internal hip rotation in a limb with a clubfoot (112). The implications of these findings are yet unknown.

The deformities of the clubfoot fit the acronym CAVE. They are cavus (plantar flexion of the forefoot on the hindfoot), adductus of the forefoot on the midfoot, varus (or inversion) of the subtalar joint complex, and equinus of the ankle. These deformities are not passively correctable. The severity of the deformities and associated findings vary from foot to foot, even in bilateral cases. There is a single (occasionally double) posterior ankle skin crease (Fig. 29-29). The calcaneus is difficult to palpate within the fatty heel pad, resulting on the so-called empty heel pad sign. A deep transverse skin crease crosses the midfoot and extends under the longitudinal arch. Assessment of tibial torsion in a newborn with a clubfoot is unreliable. The head of the talus can be seen and palpated on the dorsolateral aspect of the midfoot/hindfoot just anterior to the ankle joint. This is due to excessive inversion of the subtalar joint complex around the talus. An examiner’s thumb can be placed on the dorsolateral aspect of the head of the talus as a solid fulcrum around which one can attempt to evert the subtalar joint. In an idiopathic clubfoot, the navicular will not fully align with the head of the talus and displace the examiner’s thumb.

The foot and calf are smaller than the contralateral side. Little and Aiona (113) noted a leg length discrepancy of >0.5 cm in 18% of children with unilateral clubfoot and 4% with bilateral deformities. In those with unilateral deformity and discrepancy, the tibia is short in 89% and the femur is short in 43%. Spiegel and Loder (114) found a leg length discrepancy of >0.5 cm in 32 of 47 patients with a unilateral clubfoot, a frequency that is perhaps more accurate.

In 2009, Howlett and Mosca (112) reported increased internal hip rotation in limbs with clubfoot deformity. Because this was a clinical study, it was not possible to determine if this finding was due to femoral anteversion, acetabular anteversion, or both. There was at least 10 degrees greater internal hip rotation, and 10 degrees less external rotation, ipsilateral to a unilateral clubfoot compared to the nonaffected limb in >80% of cases. Children with bilateral clubfoot deformities had at least 10 degrees greater internal hip rotation than that reported for age-matched normal controls (115). There might be genetic implications for this finding that are yet to be elucidated, but the clinical implications are significant. An in-toeing gait in a child with a clubfoot could represent persistence or recurrence of the clubfoot deformity, but, if the in-toeing is due to increased internal hip rotation, it must be identified and differentiated before considering inappropriate treatment of a well-corrected clubfoot.

A complete physical examination of the child is indicated to rule out a neurogenic or syndromic etiology for the deformity (Table 29-2). Muscle testing and sensory examination should be part of the initial examination in all patients, as clubfoot has been found to be associated with absent anterior compartment muscles and lesions involving the innervation to the anterior and lateral compartment muscles. A finding as subtle as adducted and contracted thumbs across the palms will identify a child with clubfeet as having arthrogryposis. Examination of the hips in a newborn is an important part of the musculoskeletal screening exam, but there is no reported increased risk for developmental dysplasia of the hip in children with clubfoot (116, 117).

Classification of the severity and rigidity of the clubfoot is important for the comparison of treatment modalities. Several classification systems have been proposed (120, 121 and 122, 125, 163, 173, 528), but none has been universally accepted. They all suffer from excessive subjectivity and only fair reproducibility.

With respect to classification of clubfeet, important clinical features must be documented. Included in these features are

Although there are a number of classification systems in use, two of them seem to be of particular value in attempting to classify clubfeet at the initiation of treatment. One of these classification systems was defined by Dimeglio et al. (121) and the second by Pirani (122, 123). The classification systems apply a point score to a number of physical findings, which, when totaled, leads to a “grade of involvement.” Flynn et al. (124) showed a good correspondence between the classification systems of Dimeglio and Pirani. They found a correlation coefficient of 0.83 with the Dimeglio system and 0.9 with the Pirani system when applied by three individuals. The correlation improved in both systems after the initial 15 feet were scored. Wainwright et al. (125) compared four classification systems: Dimeglio, Catteral, Harold and Walker, and Ponseti and Smoley, and determined that the Dimeglio system was the most reproducible. Both the Dimeglio and the Pirani point systems attempt to differentiate between mildly affected feet requiring little treatment and those that are extremely severe. If the outcomes of treatment are to be compared, a valid classification system must be employed before the initiation of treatment.

One should create a checklist with either or both of the systems and attempt to score clubfeet at the initiation of treatment. Each of the scoring systems can be applied in <2 minutes. In the Pirani system, isolated physical findings, including the severity of deformity, the depth of skin creases, and the degree of certain pathoanatomic variations of the midfoot and the hindfoot, are each given severity scores of 0, 0.5, or 1 with a maximum possible score of 6. In the Dimeglio system (Fig. 29-30), the deviations of the foot—equinus, varus, adductus, and foot rotation—are scored based on the degree of deformity. Scores for the depth of skin creases, presence of cavus, and the condition of the muscles are combined to give a maximum possible score of 20. Choose one and attempt to grade clubfeet on the basis of these clinical criteria.

There is no consensus on the role of radiographs in the diagnosis and management of the clubfoot (121, 122). However, it should be stated that the diagnosis of clubfoot in the newborn can and should be based solely on clinical findings. The intended role of radiographs in the assessment of foot deformities is to demonstrate the relationships between bones. This is accomplished by first drawing the axis of each bone, and herein lies the limitation of this imaging modality. There is little ossification of the bones in the normal newborn foot, and there is a delay in ossification in the clubfoot (126). The ossific nucleus of the talus is not centrally located in the cartilaginous anlage (127, 128). The ossific nucleus of the talus is between the head and neck and may be spherical in shape for the first several weeks of life. Ossification of the navicular does not begin until age 3 to 4 years in children with clubfoot and even then is eccentric. Brennan et al. (129) identified an even more fundamental problem when they showed very poor reproducibility in positioning the clubfoot for the radiograph. These factors make it unrealistic to consider radiographs of the newborn and infant clubfoot as objective data. The age or point at which radiographs become reliable is unclear.

Despite these limitations, there might be a role for radiographs to confirm correction of the clubfoot deformities or to help identify the site(s) of residual deformity in the child who is several months old and has been undergoing serial manipulation and casting. In the latter situation, the information can be helpful for surgical planning, particularly if one ascribes to à la carte surgery (118, 119, 130). The AP view is obtained with the foot pressed against the radiographic plate with a dorsiflexion and external rotation force (120, 131) (Fig. 29-31A). This will place the subtalar joint in its most everted and corrected position. The lateral radiograph is obtained with the foot dorsiflexed and maximally everted, but also with the leg internally rotated to assure a true lateral view of the ankle (manifest by the projection of the fibula within the posterior half of the tibia) (Fig. 29-31B). The talocalcaneal and talus-first metatarsal angles are measured on both views (10). The axis of the talus and calcaneus normally diverge from each other and the axis of the talus and the first metatarsal normally form a nearly straight line on both views. The tibiotalar and tibiocalcaneal angles are measured on the lateral view (10). The axis of the talus normally aligns almost perpendicular to the tibia and the calcaneus dorsiflexes above a right angle with the tibia. The alignment of the calcaneus and cuboid are assessed on the AP view (132).

A second point at which radiographs may be helpful is intraoperatively to confirm the adequacy of correction of the deformities. The low dose radiation and convenience of minifluoroscopy make that technology desirable. The third indication for radiographs could be at some substantial time after surgery to confirm maintenance of deformity correction. Alternatively, the third point at which radiographs are obtained is when recurrence or other secondary deformities are identified.

In response to the limitations of radiographs, ultrasound techniques are evolving for the assessment of the infant clubfoot during nonoperative and operative treatment (133, 134, 135, 136 and 137). Early results seem promising; however, availability of the technology in orthopaedic outpatient clinics
is currently limited. Arthrography, computerized tomography (138), and magnetic resonance imaging (127, 139, 140) may have a role in research or in the evaluation of postsurgical deformities, but do not have a role in the routine assessment of the idiopathic clubfoot.

The intrauterine diagnosis of clubfoot has become increasingly frequent with the routine use of fetal ultrasonography during pregnancy. This has implications for the orthopaedist, who is being consulted by prospective parents regarding the diagnosis, possible relationship to syndromes, treatment options, and prognosis. It appears that the earliest that a clubfoot can be diagnosed by ultrasound with accuracy is 12 weeks of gestational age. With sequential studies, there is an increased ability to visualize the deformity, relating either to the progressive development of a clubfoot deformity or perhaps the accuracy with which it can be seen. According to Keret et al. (141), the clubfoot deformity has been detected in routine studies in approximately 60% of cases, which is indicative of some degree of false-negative assessment. In 86% of cases, the deformity is identified by 23 weeks of gestational age, but still others are recognized up to 33 weeks. The diagnosis is made on ultrasound by the fixed position of the foot in an equinovarus position, not deviating from this on sequential observations (Fig. 29-32). Three-dimensional ultrasound may provide a more accurate diagnosis than standard ultrasound studies (142).

In studies of large populations using routine in utero ultrasound (143), the recognition of clubfoot deformity varies from 0.1% to 0.4% (144). Because postnatal studies suggest an occurrence rate at birth between 0.1% and 0.6%, one can assume a rather low false-negative rate. The false-positive rate for in utero diagnosis of clubfoot using ultrasound varies from 30% to 40%, depending on the series and the criteria (144, 145, 146 and 147). A term functional false-positive rate has been used in cases in which a foot may have the appearance of remaining in a plantar-flexed, varus, and medially deviated position but can passively be corrected to neutral during exam just following birth. The foot is characterized with a score of 0, 1, or 2 using the Dimeglio classification system and has been classified by some authors as a positional clubfoot. Such a foot requires only parent-administered exercise, and no long-term deformity results. With the advancement of ultrasound and increase in experience, the accuracy of diagnosis will steadily increase.

The ability to recognize syndromes associated with skeletal malformations is also increasing with time (142, 148). The combination of technologic advances and improved expertise in obtaining and interpreting images will certainly lead to further progress in recognizing fetal structural abnormalities. This brings one to the question of the need for amniocentesis and karyotyping if an isolated clubfoot deformity is found. In 1998, Shipp and Benacerraf (149) and Rijhsinghani et al. (144) suggested that amniocentesis and karyotyping were necessary to identify associated syndromes when clubfoot was identified. Malone et al. (150), in the year 2000, showed that in 57 cases of isolated clubfoot deformity out of 27,000 prenatal exams, there were no unrecognized associated abnormalities. Therefore, the recommendation is that karyotyping not be done in cases where a diagnosis of isolated clubfoot deformity was made. This still appears controversial, and a geneticist should be consulted about the need for amniocentesis if the question arises.

There is no attempt to provide any therapeutic intervention once an intrauterine diagnosis of clubfoot is made. The orthopaedist is only involved in counseling the family about the etiology, treatment, and prognosis, which generally alleviates fears and guilt, dispels myths, allows the parents to make personal decisions concerning the pregnancy, and allows for an improved emotional state for the family during the remainder of the pregnancy and the delivery of their child.

The deformities of the clubfoot are created, in part, by malalignment of the bones at the joints and, in part, by deformation in the shapes of the bones (102, 127, 138, 139 and 140, 151, 152, 153, 154, 155, 156, 157, 158, 159 and 160). The neck of the talus is short and deviated plantar-medially on the body of the talus (102, 153, 157, 158, 159 and 160). This directs the articular cartilage of the head of the talus in the same plantar-medial direction. Anatomic dissections (158) and MRI scan images (139, 140, 153) confirm that there is a varus deformity of the distal end of the calcaneus creating a medial tilt of its articular surface at the calcaneocuboid joint. Howard and Benson (157) and Epeldegui and Delgado (156) found that the anterior facet of the calcaneus is tilted medially in relation to the middle facet in the clubfoot, indicating that the location of the varus deformity is between the two facets (Fig. 29-33). Epeldegui and Delgado (156) performed elegant microdissections of the feet of 75 stillborns, some of which had clubfoot deformities. He specifically studied the bony and softtissue anatomy of the talocalcaneonavicular joint, which he, like Sarrafian, called the AP and Scarpa (14) termed the pes acetabulum. The AP is an ellipsoid articular cavity that holds and rotates around the head of the talus, comparable to the relationship between the pelvic acetabulum and the femoral head at the
hip joint. Its bony elements are the posterior articular surface of the navicular and the anterior and posterior articular facets of the calcaneus. Epeldegui found that the soft tissues of the AP were likewise markedly different in shape and orientation in the clubfoot from the normal foot. The shape of the medial cuneiform has not been studied in the newborn, but it is trapezoid shaped in the older child with residual forefoot adductus deformity. The subtalar joint complex is severely inverted, a combination of internal rotation and plantar flexion. The axis of rotation is in the interosseus talocalcaneal ligament. The AP is inverted around the plantar-medially-deviated head and neck of the talus, thereby aligning the navicular at or near the medial malleolus. The calcaneus is rotated downward and inward resulting in parallel alignment with the talus in the frontal and sagittal planes. The posterior part of the calcaneus is tethered to the fibula by the calcaneofibular ligament. There is a varus deformity of the distal end of the calcaneus with medial deviation of a congruous calcaneocuboid joint in many clubfeet (78, 139, 140, 153, 155, 156, 157, 158 and 159, 161). There may be medial subluxation of the cuboid on the distal calcaneus in some feet (152, 162). The plantar fascia, short plantar muscles, and spring ligament are contracted. The Achilles, tibialis posterior, flexor hallucis longus, and flexor digitorum communis tendons are contracted. The posterior capsules of the ankle and talocalcaneal joints are contracted.

Tibial torsion and the position of the talus in the ankle are debated. McKay (154) believes the talus is in neutral alignment, Goldner (163) believes that it is internally rotated, and Carroll (151, 152, 155) believes that it is externally rotated. A recent three-dimensional MRI study of clubfeet revealed an externally rotated position of the talus in clubfoot compared with normal feet (127).

The muscles are abnormal in both anatomical insertion and intrinsic structure (101, 106). Muscles in clubfoot are smaller than normal and there is an increase in intracellular connective tissue within the gastrocsoleus and posterior tibial muscles. A predominance of type I muscle fiber has been seen in posterior and medial muscle groups. Electron microscopic studies have shown loss of myofibrils and atrophic fibers, suggesting a regional neuronal abnormality as well (108).

The ligaments are thick, with increased collagen fibers and increased cellularity (107). This is particularly true of the calcaneonavicular ligament or spring ligament and the posterior tibial tendon sheath (164). An electron microscopic study of medial ligaments in clubfoot identified myofibroblasts, which could be responsible for fibroblastic contracture in the postoperative clubfoot. Fukuhara et al. (165) showed myofibroblast-like cells in the deltoid and spring ligaments. Together, the thickened and shortened ligaments with contractile fibroblasts may
produce a significant component of the clubfoot pathology. Sano et al. (166) confirmed these findings, showing that cells of the medial ligamentous structures contained vimentin uniformly and myofibroblasts in some cases. More recently, Khan et al. (167) were unable to show myofibroblast-like cells in clubfeet, and van der Sluijs and Pruys (168) demonstrated normal collagen cross-linking in clubfeet.

Finally, abnormal vasculature in the foot is frequently present (169). The dorsalis pedis artery in many cases is absent or altered. Katz et al. (170) showed deficient dorsalis pedis flow in 45% of clubfeet compared to 8% of normal controls. In the more severely affected feet requiring surgery, the incidence of dorsalis pedis abnormality was 54%, whereas those successfully treated with cast therapy had an abnormality in dorsalis pedis flow in only 20% of cases. These data suggest that the severity of clubfoot may in some way relate to the vascular abnormality frequently seen in this condition.

The untreated clubfoot persists as a rigid, unsightly deformity. A large, callused bursa develops over the dorsolateral aspect of the hyperflexed midfoot which functions as the weight-bearing surface of the foot (Fig. 29-34). In the most extreme cases, the toes point backward during ambulation. Specially made footwear is required to accommodate the deformity. Surprisingly, untreated adults in certain cultures and environments will have little pain for many years and can function adequately. Their function is similar to that of individuals with Syme amputations when not wearing their prostheses. City-dwelling adolescents and adults with untreated clubfoot experience pain and disability with ambulation on paved sidewalks and hard floors.

The laundry list of operative complications was mentioned earlier. The approach to management of some of the postoperative deformities will now be discussed.

Surgical Correction of Clubfoot (Figs. 29-44, 29-45, 29-46, 29-47, 29-48, 29-49, 29-50, 29-51, 29-52, 29-53 and 29-54)