Damian Mc Cormack, Mch, FRCS orth,
International Fellow in Paediatric Orthopaedic Surgery,
Atlanta Scottish Rite Hospital,
Atlanta, Georgia.

Foot and ankle deformities in spina bifida patients are both complex and variable. They differ in several respects from similar clinical deformities in otherwise normal children and in other disease processes such as cerebral palsy and polio. They challenge the clinical judgment and surgical skill of the surgeon. In doing so, however , these deformities and their management can greatly increase our understanding of foot and ankle mechanics and function.
In this review it is hoped to clarify some of the principles involved in the assessment and management of foot and ankle deformities in this population. I am in general avoiding the use of historical acronyms for procedures mentioned to avoid confusion.
The paper is written in a question and answer format which, I think, makes for a more interesting read and helps to maintain a logical approach. Too often do we know an answer without really understanding the question.

1.

What percentage of spina bifida patients will have significant foot deformities and what is the goal in their treatment?

About three quarters will have foot / ankle deformity. The goal of treatment is a mobile and braceable foot.

2.

Why are the ( foot and ankle ) deformities in spina bifida so variable and so severe when compared to those in, for instance, cerebral palsy or polio?

The severity is for the most part a temporal effect. For instance, a one year old with cerebral palsy has developed his equinus deformity since birth ( given that most cerebral insults occur late in pregnancy). A one year old with spina bifida has been developing his equinus deformity for almost two years, since early in fetal development. Closure of the neural tube should occur by day 24 to 26 of gestation.
The variability and asymmetry in bilateral foot deformity is a clinical give-away that the underlying process is probably spina bifida. This results from the irregular scarring at the myelomeningocele site in the thoracic or lumbar spine. Motor and sensory levels may therefore be different in the feet. The deformity may result from a mix of upper and lower motor neuron deficits. Thus spasticity and flaccid paresis may co-exist in the limb. One may see a valgus deformity on the right and a varus deformity on the left. Sensation may be present in part of one foot and absent in the other.

Polio creates a motor deficit and imbalance without sensory loss. Many of the procedures described for polio were tried in myelomeningocele patients and failed. A brace suitable for an equino-valgus foot in a polio victim might leave the insensate myelomeningocele foot ulcerated. The foot in c.p. patients demonstrates spasticity but will also have protective sensation.

3.

What are the practical management implications of the severity of deformity in myelomeningocele feet ?

These feet do not tolerate ill fitting braces and demand surgical correction of deformity to allow safer bracing. Clubfoot correction in an otherwise normal child requires tendon lengthening or simple tenotomies. With spina bifida sectional tendon excision is required. Whereas a simple percutaneous tenotomy may correct an equinus deformity in a cerebral palsy foot, the same procedure is less likely to have lasting success in spina bifida. Excision of the posterior ankle and subtalar capsules as well as excision of a section of the Achilles tendon may be required in the myelomeningocele foot.

4.

What neurological processes can contribute to the development of extremity deformity in spina bifida patients?

It is a mistake to think only of the failure of closure of the distal neural tube. Spina bifida is a disease of the central nervous system. The neurological problems therefore include hydrocephalus, necessitating shunt placement. Shunt blockage causes neurological deterioration. Shunt infection causes an encephalitis with subsequent neurological deterioration. The Arnold Chiari malformation ( Types I , II and III) places potentially fatal pressure on the brainstem at the foramen magnum ( FIG 1). A syrinx may be present in the spinal cord and can potentially cause neurological deterioration. About a quarter of myelomeningocele patients demonstrate upper limb deficits which may have implications for assisted ambulation, or loss thereof. The cord inevitably will scar to the placode at the myelomeningocele site and become a tether with spinal growth ( or surgical correction). Retethering is common following neurosurgical release. A lipoma may develop at the site of a meningocele and impair neurological function later in life. Another common defect is a diastomatomelia. The bony spur may be visible on plain radiographs and the split spinal cord may deteriorate with growth.

5.

Given the complexity of the neurological pathology how is the neurological deficit in spina bifida best classified?

Sharrard tried to formulate a classification based on motor levels, similar to that used in spinal trauma patients. However the useful functional level may be much higher than the lowest observed muscle movement. This may be due to problems at the level of the cord, brainstem or cortex. For instance a motored but insensate ankle is not a functional ankle! Motor function may be present to command but inadequate for complex movements. Abnormal co-contraction of flexors and extensors around a joint negate the benefits of motor function at that level. The patients intelligence may make ambulation impossible despite the presence of adequate limbs. It is far better to use sensory levels to group these patients.

6.

What foot and ankle deformities are seen in myelomeningocele patients?

The deformities include;
Equinus deformity
Clubfoot or talipes equino varus deformity.
Calcaneal deformity.
Cavus or cavo-varus deformity.
Plano-valgus deformity.

7.

Why not simply do arthrodeses or tendon transfers in these feet?

An arthrodesis may work but will frequently cause problems primarily because the feet may lack sensation. A mobile foot is easier and safer to brace and the brace is necessary because some or all of the foot is paralysed.
A tendon transfer will drop the power in the transferred muscle at least one level. Central control of the transfer may be poor. Spasticity and co-contracture interfere with the transfer. Therefore it is far better to avoid transfers.

8.

So what’s left in the armamentarium, if tendon transfers and arthrodeses are out?

Deformity correction or improvement in myelomeningocele patients can be achieved using osteotomies, excisional tenotomies and tenodeses as described below.

9. How does management of clubfeet in the 30% to 50% of myelomeningocele patients who have this deformity differ from that of idiopathic clubfoot?

These feet are severe, like those in arthrogryposis. Their management differs from that of idiopathic club feet (FIG 2) in several respects. There is relative tibialis anterior overactivity. There is a high recurrence rate after surgery. Prolonged casting is often unsuccessful and may make subsequent surgery even more difficult.
During surgery tendons must be excised rather than incised. The tibialis anterior tendon must be released ( not usually done in club foot surgery). The post operative casts are exchanged for braces ( ankle foot orthoses; AFO’s) which are worn indefinitely.
The hindfoot may not correct from equinus to neutral despite repeat extensive soft tissue release. As in arthrogrypotic feet a talectomy shortens the hindfoot, relieves tension of the skin and allows excellent hindfoot correction. The results of talectomy are surprisingly good. One problem is subsequent residual talar fragment growth. Another is that the midfoot and forefoot deformity is not addressed and must be corrected, usually as a staged procedure.

10.

How does one address the residual severe varus / adduction deformity in the mid foot and forefoot in this situation?

A residual forefoot adductus deformity in an idiopathic club foot may be addressed, for instance, by transfer of part of the tibialis anterior tendon from its dorso-medial insertion to a dorso-lateral site. This tendon transfer would probably fail in the myelomeningocele foot for reasons described above.
Another option in the idiopathic foot is to adjust the lengths of the medial and lateral columns of the foot by opening and closing wedge osteotomies. A closing wedge osteotomy on the lateral side ( which is relatively too long) and an opening wedge on the medial( short) side will correct the mid/forefoot adductus. The wedge may be taken from the (lateral) cuboid and inserted in the (medial) medial cuneiform.
This principle is employed in the spina bifida foot but shortening of the lateral column alone may suffice if combined with a medial soft tissue release. The lateral column may be shortened by a closing wedge in the cuboid, as described, or by excision and fusion of the calcaneo-cuboid joint ( but stiffness results) or excision of the antero-lateral aspect of the calcaneus, but preserving the articular surface of the calcaneo-cuboid joint.( FIG 3)

11.

How does one address a calcaneal deformity in a spina bifida patient?

Calcaneal posturing is so common in a normal neonate that it absence may alert one to the possibility of an equinus or clubfoot deformity. In the normal child this posture resolves spontaneously in the majority. In the myelomeningocele patient the calcaneal deformity is due to unopposed contracture or spasticity of the anterior muscles, with a varus component if tibialis anterior dominates and a valgus component in the peroneals are relatively strong or spastic ( FIG 4).
The deformity progresses. The patient crouches to maintain forefoot contact with the floor ( by bending hips and knees). The point contact at the heel creates ulceration. In the infant casting or bracing may help. Transfer of the tibialis anterior back through the interosseus membrane to the calcaneus may succeed in polio but has less success in this situation. Segmental excision of the anterior ankle tendons ( tib’ ant’ and the long toe extensors and peroneals if necessary) is more certain to achieve the desired effect; a mobile braceable foot.
In the older child a calcaneal osteotomy with posterior displacement of the prominent posterior fragment helps relieve skin pressure.

12.

How does one assess and manage valgus foot deformity in these patients?

The patient with a valgus deformity will, if ambulatory, load the medial side of his foot, with potential for ulceration beneath his talar head. The first realization is that there are two places where a clinical valgus deformity can arise. The first is at the ankle . The second is at the subtalar joint. It is important to make the correct diagnosis.

13.

How can one manage ankle valgus deformity?

A valgus deformity may result through lateral wedging of the distal tibial epiphysis. This ankle valgus is clinically apparent because the medial malleolus will be at a relatively low level compared to the lateral malleolus. An ankle anterior mortice view will demonstrate the deformity. There is also relative shortening of the lateral malleolus and lateral shift of the calcaneus in the mortice.

The surgical correction of ankle valgus is interesting. One procedure gained accidental fame. The insertion of the proximal aspect of the Achilles tendon into the distal fibula , creating a tenodesis effect, was mistakenly performed by a junior who was supposed to insert it into the tibia to control a calcaneus deformity! However the result was elongation of the fibula and lateral malleolus, possibly because of increased local blood flow. This lateral malleolar elongation can correct ankle valgus however and the procedure was adopted for a time. The results are variable and this Achilles tendon to fibula tenodesis is no longer prefered.
A supra-malleolar medial closing wedge osteotomy is a more certain way to correct the distal tibial valgus. It typically leaves the medial malleolus prominent however (FIG 5). Therefore Wiltse described a modification by which the distal tibial osteotomy was fashioned to allow upward and lateral shift of the distal fragment. This reduces malleolar prominence (FIG 6).
A simpler solution is to perform a more distal osteotomy. In the older child a transphyseal closing wedge will effect the correction without medial malleolar prominence. The resulting growth arrest is of little consequence in this situation.
An even simpler solution is to create a medial hemiepiphysiodesis to stop medial growth and ( hopefully) allow the lateral side to grow and correct the valgus deformity. This medial hemiepiphysiodesis can be performed by percutaneous placement of a transphyseal medial malleolar screw.

14.

How does one assess and manage hindfoot valgus in myelomeningocele patients?

In an otherwise normal individual hindfoot valgus places the center of the calcaneus lateral to the center of the ankle. The result is passage of the mechanical axis of the limb medial to the calcaneus. There is no bony block to progressive valgus deformity because the joints of the foot and ankle relax in valgus. On the contrary varus deformity locks up these joints which therefore limit further varus deformity. The valgus foot deformity is more obvious, may be associated with a very flexible flat foot, and is often asymptomatic.
In the ambulatory myelomeingocele patient with diminished foot sensation the valgus deformity can cause pressure necrosis under the prominent talar head. Furthermore lateral muscle spasticity and posterior tibial paralysis may create a valgus deformity which is stiff and difficult to brace. However bracing may be attempted in the infant.
Surgical correction of hindfoot valgus requires that the calcaneus be replaced beneath the talus. This requires an extensive soft tissue release ensuring that deforming forces are neutralized. A subtalar or triple arthrodesis may succeed but problems include forshortening of the immature foot and the risks of a stiff hindfoot. An extraarticular subtalar arthrodesis ( Grice) is popular because it avoids premature growth plate closure.
Another alternative is a medial displacement sliding osteotomy of the calcaneus which maintains hindfoot flexibility.

15.

How is a varus foot deformity managed in the myelomeningocele patient?

A pure hindfoot cavus deformity is uncommon and can be addressed by a lateral closing wedge osteotomy of the calcaneus or a lateral sliding calcaneal osteotomy.

16.

How is the more common cavo-varus deformity addressed?

The cavus comes first and should be addressed first. If corrected it may cause automatic resolution of the varus deformity in the hindfoot. This will happen if the varus deformity is flexible and this can be assessed by examination of subtalar motion with the patient prone and knee flexed. By manual correction of the forefoot deformity to neutral the heel is observed to see if the hindfoot varus corrects or remains.
Coleman tried to explain this simple examination to some students and placed a wooden block on the floor to demonstrate the same maneuver with the patient erect. The block beneath the lateral border of the foot corrects the forefoot to neutral ( by allowing the abnormally plantarflexed first ray to sit on the floor without forcing the forefoot into supination). Unfortunately the Coleman block test caught on and has confused generations of trainees. He allegedly denies culpability and says he doesn’t even use the test! Prone examination is easier.
If passive correction of the forefoot cavus (during the prone examination) causes hindfoot correction from varus to neutral then only the forefoot cavus is addressed surgically. If the hindfoot does not correct then both forefoot and hindfoot require surgical correction.

In the myelomeningocele foot with a flexible cavus deformity a radical plantar release may suffice. Otherwise an dorsal closing wedge osteotomy of the first metatarsal or mid tarsal joint complex is necessary to correct the cavus deformity. If the hindfoot is stiff as described above a calcaneal osteotomy to correct the varus deformity is added.
Any muscle imbalance is corrected by tenotomies at the same time of in a prior procedure. Again a triple osteotomy, with wedge resections to correct cavus at the midfoot level and varus at the subtalar level may succeed but may require revision for reasons alluded to above.

17.

What specific global precaution should be taken during surgery for these spina bifida patients?

Latex allergy precaution.

18. What else?

Needless to say there is much more to spina bifida and there are numerous other foot and ankle deformities. A full discussion of foot deformities in the spina bifida patient can be found the following references.

References.

1. Lindseth R. Myelomeningocele, Lovell and Winter’s Pediatric Orthopaedics, Vol 1, 4th edition. 523-533.

2. Dias L. Myelomeningocele, in Pediatric Orthopaedics; Ed. Canale: 739-773.

Legends to figures:

Figure 1.
Arnold Chiari malformation type II.
This is the commonest form of A-C malformation in spina bifida. Type I is less severe and type III more severe with herniation of the cerebellum into the foramen magnum.

Figure 2.
Idiopathic clubfoot deformity. In myelomeningocele the deformities may be less symmetrical and more severe. The feet may be insensate. Tendon resections rather than lengthening or simple tenotomies are necessary in correction of neuropathic deformities.

Figure 3.
Four ways to shorten the lateral column of the foot. Simons described resection of a wedge from the antero-lateral part of the calcaneus with preservation of the calcaneo-cuboid joint.
Lichtblau described resection of the calcaneal aspect of the calcaneo-cuboid joint.
Evans reported resection of the calcaneo-cuboid joint and fusion.
A fourth alternative is to resect a wedge from the cuboid avoiding the joints. This wedge may be placed as graft in an opening wedge of the medial cuniform to elongate the medial column.

Figure 4.
A calcaneal deformity in a child with spina bifida. This unbraceable foot results from imbalance of the anterior and posterior musculature at the ankle from spasticity or paralysis or both. The heel is vulnerable to skin breakdown in the insensate foot.

Figure 5.
A closing wedge supra malleolar osteotomy for ankle valgus will translate the medial malleolus causing prominence and potential skin problems in brace.

Figure 6.
Wiltse’s variation reduces medial prominence. Medial hemiepiphysiodesis or a more distal transphyseal osteotomy are easier and more successful options in the older child.