The aim of orthotic management for the paraplegic child should be to facilitate and encourage mobility while simultaneously attempting to prevent the progression of deformity. The extent of the paralysis and corresponding deformity is known to depend on the location of the spinal lesion, its severity, and its extent. When considering the orthotic management of the paraplegic child, one should remember that any conceptual distinction that is made between the improvement of function and the prevention of deformity is artificial because both functional and preventative purposes are required simultaneously. Four clinical examples are presented that outline the progression of significant skeletal deformity and joint contracture despite the provision of orthotic management and the adoption of a reciprocal gait pattern. The inability of a particular orthosis to control the rapid progression of deformity is supported by radiographic findings and appropriate clinical review.
Key Words: Paraplegic child, progression of deformity, orthotic management.
A literature review 1-10 of approximately 1,500 reported cases of children with paraplegia illustrates that the associated problem of spinal deformity is significant. Typically, the deformity appears within the first decade of life and progresses to skeletal maturity, the most dramatic progression noted in the less-ambulatory patient. It can be assumed that the frequency of deformity relates directly to neurological deficit. Increased lumbar lordosis and hip flexion contractures are common.
Raycroft and Curtis8 reviewed the natural history of myelomeningocele spine deformity. In their review, 53 of the 103 patients without congenital vertebral anomalies other than spina bifida developed paralytic scoliosis, an incidence of 52%. Of the patients with developmental or paralytic scoliosis, 70% had uni- or bilateral hip dislocation, and 83 % had pelvic obliquity. The authors observed a correlation between the higher levels of paralysis and the incidence of scoliosis. A review1 of 268 patients with a four-year follow up indicated those patients with defects in the thoracic and lumbar spine attained a 100% incidence of scoliosis by the age of 10-14 years. Those with defects in the lumbar and. in the lumbosacral spine attained a 57% and 50% incidence, respectively. The authors reported that the incidence of scoliosis increased in each succeedingly older age group with the most dramatic increase noted in paralyzed patients who were less ambulatory or had higher level injuries. When researchers conducted a review2 of 130 patients with myelomeningocele at skeletal maturity, they observed that 101 patients, or 78%, had clinical spine deformity. A further review 3 of 82 patients with myelomeningocele reported an overall incidence of spinal deformity of 66%. The highest reported incidence of scoliosis, 100%, occurred among those patients with thoracic-level paraplegia, and the incidence gradually reduced to 70% for patients with low lumbar lesions. All curves appeared within the first decade and progressed relentlessly.
Moe et al.6 reported similar findings, with a 100% incidence of spinal deformity at or above T-11, and a rapid increase in spinal deformity during the adolescent growth spurt. Progression continued until skeletal maturity, resulting in the loss of sitting balance in the most severe cases. Piggott11 noted that in his study of 191 children, 83% had scoliosis and the frequency of deformity was related to neurological deficit.
A survey7 of 132 children with myelomeningocele examined the prevalence and progression of scoliosis as well as the results of scoliosis treatment and the effects of that treatment. In a cross-sectional clinical and radiological investigation, it was established that scoliosis occurred in 69% of these cases. The results agreed with the published reports of the last three decades: the author confirmed that the incidence varied depending on the level of dysraphism, with a 94% incidence at the thoracic level. In addition, the ambulatory status of the patients correlated with the occurrence of scoliosis: 92% of nonambulatory patients and 38% of ambulatory patients had scoliosis. Orthotic management of the scoliosis was recommended. Treatment with orthoses was associated with few problems and was successful in preventing progression, provided that the scoliosis was less than 45° at the start of treatment.
Typically an attempt is made to help paraplegic children stand upright between 12 and 18 months of age, but currently there is controversy about the advantages of fitting such children with conventional HKAFOs. Conventional or traditional forms of orthotic management improve the function of paraplegic children; however, prevention of deformity, specifically control of the pelvis and hl s, has been problematic in this group12,13 because of the imbalance of muscle power around the hip joint. This common problem in children who are candidates for orthotic management can be readily appreciated when the differential innervation, particularly of the hip flexors versus the hip extensors, is considered. The hip flexors are at least partially innervated at the L-2 and L-3 level, whereas the hip extensors are innervated below the L-3 level. Such imbalance at the L-2 and L-3 level of involvement is the cause of the lordosis that is so often seen in these patients, which is often aggravated by significant hip flexion contractures.13
It can be argued that conventional HKAFOs provide stability and reduce the potential for hip flexion contracture. It is the opinion of this author that the pelvic band provides some mediolateral hip stability, but is typically not successful in controlling anterior pelvic tilt.14
Before the availability of reciprocating gait orthoses, conventional HKAFOs were routinely prescribed and children were encouraged to ambulate with a pivotal or swing-through gait. 15-17 Conventional orthoses with locked hip joints were traditionally fitted, but because of the tendency of patients with weak hips to fall into a flexed position while standing (Figure 1 ), these orthoses are no longer routinely prescribed. The clinical example shows the inability of this design to control the lumbar spine and pelvis, although it does provide this child with an upright posture. From a biomechanical perspective, however, the orthosis is clearly insufficient. Some means of controlling hip motion through a functional range while allowing a step to be taken with trunk control was deemed necessary.18
The HGO was developed originally for children with spina bifida.19,20 Independent ambulation as defined by Rose21 has three mandatory features: 1) low energy cost at a reasonable speed of ambulation (30%-60% of normal speed for that age); 2) independent transfer from chair to walking and vice versa; and 3) independent doffing and donning of the orthosis. Within these parameters, the team at the Orthotic Research and Locomotor Assessment Unit (ORLAU) reviewed the records of 27 paraplegic children who had used the orthosis for at least six months in an attempt to establish the degree of benefit achieved. 21 All patients were examined to determine the sensory level of dysfunction. Those with no voluntary activity at the hip were classified as flail. The spine was assessed to determine "truncal stability," defined as the ability to sit with the arms raised above the head for a prolonged period without support. The patients were also examined for hip dislocation and flexion contractures. Thirteen patients improved their ambulatory status. This review indicates that the HGO can provide, at low energy cost, reciprocal ambulation for low-thoracic and high-level lumbar lesions. All of the patients had truncal stability, and it was reported that this was a necessary prescription criterion. Two important observations can be made from this review: 1) 25 of 27 patients had been upright in some form of orthosis prior to the HGO fitting; and 2) only 2 of 27 (7.6%) of the patients had scoliosis. Further work is required to investigate whether this extraordinarily low incidence of scoliosis relates to the first observation.
The LSU RGO has been described22 as a lightweight bracing system that gives structural support to the lower trunk and lower limbs of the paralytic patient while allowing, through a cable coupling system, proper hip joint motion for walking. A report22 on a series of 138 patients, 95 with spina bifida, claimed that long-term bracing and early ambulation seemed to decrease the potential for the recurrence of deformity, possibly by preventing distortions caused by gravitational positioning of flexed joints. Although the LSU RGO has been shown to provide a reciprocal gait pattern for both adult and pediatric paraplegics, it has been commented23 that extravagant claims have been made, resulting in uncertainty about prescription criteria.
Although the LSU RGO is intended for paraplegic children and routinely used by them, there is no information that shows the effectiveness of this form of management in the prevention of deformity, a fundamental orthotic principle. It is important to note that prevention of deformity in the traumatic paraplegic adult is perhaps less important than in children, because the skeletal frame in the adult is typically fully developed at the time of injury. Bone tissue is not typically malleable and epiphyseal plates are ossified, thus reducing the potential for joint deformation.
Few reports have been published on RGOs. Yngve et al.24 measured the effectiveness of RGOs for the spina bifida patient by reporting on 17 patients. Of the patients studied, 75 percent had at least L-3 motor levels, and only 18 percent had complete paralysis of hip musculature. No details were given relative to the eight patients selected for gait analysis.
A retrospective review25 was undertaken to evaluate the long-term usage pattern of patients fitted with the LSU RGO. Twenty-one patients who had used the orthosis were recalled for evaluation. The group consisted of 13 boys and 8 girls with an average age of eight years and nine months. The average usage time of the orthosis was 25.8 months. Nine of the children had thoracic-level lesions and 12 had lumbar lesions. Significantly, it is reported that all 21 patients required surgical correction of the lower limbs or spine prior to or during the bracing period, with 17 patients exhibiting residual contracture.
A further study26 analyzed the effectiveness of the orthosis to provide a reciprocal gait for patients with a thoracic or high-lumbar level of paralysis. This report concluded that, for the child with paraplegia, age and level of paralysis are the most important determinants of successful use of the LSU RGO and that most children in the study discontinue orthosis use between 7 and 11.5 years of age.
Four cases are presented after consideration of variables such as pathologic lesions, lesion level, age, intellectual status, ambulatory status classification and duration of orthosis usage.
This child was initially seen for orthotic evaluation at the age of three years September 1984. Radiographic evaluation showed congenital absence of vertebral posterior elements from T to the level of S-1. This finding corresponded with the child's known diagnosis of myelomeningocele that was closed at birth. The patient was classified as a complete paraplegic at lower thoracic level, indicating no voluntary control of the hips or lower extremities. An LSU RGO was p scribed for this child. A review of chart entries between September 1984 and June 1985 indicated that, after some initial therapy and adjustment, she was making appropriate progress, using the orthosis regularly, and ambulating in reciprocal fashion.
Her clinical condition was described by her orthopedic surgeon: "orthopedic problems are minor, she has mild scoliosis, her hips appear stable." Radiographic evaluation in June 1985 indicated a mild thoracolumbar scoliosis convex to the left measuring 16° (Figure 2 ). In November 1985, the orthotist stated, "things are going well, her ambulation pattern continues to improve."
In May 1986, the child was evaluated clinically and radiographically, 10 months after taking delivery of the orthosis. The findings were significant: mild thoracolumbar scoliosis rapidly progressed to 48 degrees (Figure 3 ) and, in addition, her orthopedic surgeon indicated that she had developed a 30° left-sided hip flexion contracture.
In June 1986, this patient underwent bilateral surgical releases of her hips. On recovering from the surgical procedure, the patient received a new LSU RGO and commenced walking once gain in a reciprocal fashion. During the next eight months the child used the orthosis consistently; however, when she was evaluated in February 1988, it was noted she was ambulating with some difficulty and had once again developed a contracture of the left hip, which was described as "abducted, flexed, and externally rotated." In the coronal plane, the spine remained basically unchanged, measuring 47°; however, when it was viewed in the sagittal plane, a marked increase in lumbar lordosis was noted (Figure 4 ).
A further radiographic review of her spine in November 1988 indicated progression of the scoliosis to 53°. Her orthopedic surgeon commented that "the spine is progressing both clinically and radiographically; eventually she will require a combined anterior and posterior spinal fusion, but as she is only eight I would like to defer this for at least three years." After minimal use of her RGO, the child discontinued orthotic management, finding a wheelchair a more effective means of "ambulation."
In August 1991, after progression of her scoliosis to >90° (Figure 5 ), the child underwent a combined anterior/posterior spinal fusion. She was 10 years old at the time of surgery.
This patient was initially seen for orthotic evaluation in November 1982. At that time his lesion level was approximated by his physician to be "high lumbar," with the possibility of a trace of hip flexor strength. He had a confirmed diagnosis of myelomeningocele and was found to have a mild thoracolumbar scoliosis and hip flexion contracture. Because of the contracture, a reciprocating gait orthosis was contraindicated and the patient received conventional HKAFO, which he used for approximately 2.5 years. In early 1985, the child underwent a surgical release of his right hip, primarily to facilitate LSU RGO fitting.
In July 1985, he was ambulating reciprocally in an RGO of the LSU design and making appropriate progress. In November at a routine clinical review, it was noted that the patient had a thoracolumbar scoliosis with a noticeable increased lumbosacral lordosis. In May 1986, radiographic evaluation (Figure 6 ) indicated a 21° thoracolumbar scoliosis. During the next two years, chart entries indicated he was "doing very well" and was consistently using his RGO for ambulation. It is, however, obvious from chart review that cable breakage was frequent and that "considerable difficulty was experienced in preventing the patient from adopting a flexed posture at the hips." Radiographic evidence obtained in August 1988 confirmed a marked increase in lordosis that would correlate directly with the flexed posture and mechanical problems.
In November 1988, the scoliosis had progressed to 38° and by September 1989 had progressed to 45°. A decision was made during this period that the child should revert to a conventional HKAFO. Ambulation became less frequent and the magnitude of the spinal deformity increased, and by April 1990 the curve had progressed to 74°. In January 1991, the child was having respiratory problems and difficulty even in sitting in a wheelchair; he underwent a combined anterior/posterior fusion at this time. He was 11 years old at the time surgery and the scoliosis had progressed to >100° (Figure 7 ).
Initially seen for evaluation in January 1984, the patient had a known diagnosis of myelomeningocele confirmed to be at the L-1 level; radiographically, a thoracolumbar scoliosis measuring 21° was evident. An LSU RGO was prescribed for this child, and it was reported in April 1984 that "the patient is undergoing a therapy program and with minimal assistance the orthosis works very well with this child." Although ambulating consistently well during the next year, the child was seen frequently for "adjustment and retightening and repair of the cable mechanism" as serious attempts were made to prevent the child from adopting a flexed attitude at the hips.
A radiograph taken in August 1985 showed progression of the curve to 33°. Thereafter the quality of the patient's gait, specifically her ability to clear her left leg, diminished. Clinically she had developed a right hip flexion contracture and attempts made to alleviate the walking difficulties by adding a right-sided shoe raise were perhaps misguided, considering her scoliosis, hip flexion contracture, and associated pelvic obliquity. The child, however, continued to ambulate, with numerous adjustments being made to her orthosis to facilitate this. In May 1987, the child underwent a surgical procedure in an attempt to realign her lower extremities. Specifically, she had a left proximal varus shortening osteotomy and a right distal derotational and extension femoral osteotomy; in December 1987, the child additionally underwent an anterior release of her right hip. On recovering from the surgery, the child took delivery of a new LSU RGO and regained her status as a community ambulator in her orthosis. Her spine continued to deteriorate and progressed to 85' prior to surgery in October 1988 (Figure 8 ). She was 12 years old at the time of surgery.
This patient (date of birth, 3.25.77) presented with a known diagnosis of myelomeningocele. At the time she presented for clinical and orthotic evaluation in August 1984, she was already 7 years old. Clinically she was described as having a "complete lesion at the T-12 level with tight hip flexors and an apparent leg length inequality as a result of pelvic obliquity; her spine is in sufficient alignment and requires no support." An LSU RGO was prescribed and she took delivery of it in November of the same year. There is no indication that her spine was radiographically evaluated.
The pattern that has emerged during the first three case presentations is repeated for this case. The child seemed to do very well initially, followed by numerous adjustments that coincided with clinical deterioration and resulted in hip flexion contracture, increased lumbar lordosis, and progression of her scoliotic curve to an excess of 80° This child underwent spinal fusion in January 1990. At the time of surgery, she was 12 years old and had been ambulating consistently in a reciprocal pattern for six years.
Reciprocating gait orthoses improve the function of paraplegic children. Numerous clinical examples show that children stand and walk who are unable to do so without an orthosis. It is also clear that this is possible with varying designs and configurations. The literature confirms that contemporary forms of orthotic management, specifically the HGO and the LSU RGO, improve the function of the paraplegic child. This claim has been supported by this author's own research14 and by clinical experience; however, the ability of these accepted designs to influence the development of joint contracture and the progression of deformity is questionable. It is somewhat surprising that no reports can be found that demonstrate the effectiveness of this form of management relating to the fundamental orthotic principle of prevention of deformity.
The largest user group of reciprocating gait orthoses is undoubtedly children who have spina bifida. The associated problem of progressive spinal deformity has been shown, and the correlation between deformity and neurosegmental function should be accepted. It would seem reasonable that any orthosis designed for this population would have the inherent capability of providing spinal orthotic management, either present or potential. The LSU RGO does not have this capability. The ability to conservatively influence the alignment of the spine together with a mechanical means of reducing or preventing contracture would seem to be necessary biomechanical characteristics that are largely ignored.
The importance of the proximal component cannot be understated or generically designed if factors such as the magnitude of deformity, the skeletal maturity of the patient, and the level of neurosegmental function are considered. The inherent biomechanical control of the orthosis used has been predetermined without regard to any clinical or radiographic evaluation of the spine. It is therefore no real surprise that it appears to be of little benefit in preventing the progression of deformity. Of greater concern is the current availability of "new and advanced" components for this group of children, where the generic nature of the proximal component is continued and promoted as being appropriate.
Four clinical examples have been presented that outline the progression of significant skeletal deformity and joint contracture despite the provision of orthotic management and the adoption of a reciprocal gait pattern. The inability of a particular orthosis to control the rapid progression of deformity is supported by radiographic findings and appropriate clinical review. It is evident that orthotic prescription and corresponding orthosis design should be based on accurate assessment of neurological deficit and resulting musculoskeletal impairment; only these factors should determine and correlate to the biomechanical properties of the orthosis. If this basic tenet of orthotic science is accepted, it is unreasonable to assume a single orthotic design containing the same biomechanical properties would be the best option for both pediatric and adult paraplegic patients.
We have come to accept that paraplegic children will abandon their orthoses and not ambulate past their early teenage years. Clinicians speculate about why these children have no interest in walking. It may be more than a coincidence that many of them have significant deformity and contracture despite several years of contemporary and aggressive orthotic management.