Chapter 10 Miscellaneous Conditions Associated with Spine Deformities

10.1Heritable Disorders of Connective Tissue

10 Miscellaneous Conditions Associated with Spine Deformities

In this chapter we have gathered together a number of different conditions, some much more prevalent than others, with the common denominator that they may well present to the scoliosis surgeon with a spinal problem and in some, such as Marfan syndrome, the scoliosis surgeon may be the first clinical port of call while others are also of clinical interest. In many of these it is not a question of surgery for a spinal deformity but rather the scoliosis surgeon may be part of a multidisciplinary team to provide the best holistic management of these children. This is not an unnecessarily exhaustive list, rather it comprises some of the more important conditions with spinal problems.

10.1 Heritable Disorders of Connective Tissue

These disorders of bone matrix are proven or presumed errors of metabolism affecting either the collagenous or noncollagenous components of the matrix. The connective tissue disorders are problems of collagen formation whereas the mucopolysaccharidoses are problems of break down. In osteogenesis imperfecta the features are principally skeletal whereas in the other connective tissue disorders, such as Marfan syndrome and Ehlers-Danlos syndrome, most are extraskeletal.

10.1.1 Osteogenesis Imperfecta: Brittle Bone Syndrome

This is a group of disorders that arise from primary inherited defects in collagen synthesis and have the common feature of bone fragility. Although the condition dates from antiquity—a 3,000-year-old Egyptian specimen resides in the British Museum ¹—the first clinical description was attributed to Ekman. ² Early classifications of the brittle bone syndrome recognized a severe and lethal sporadic form (osteogenesis imperfecta [OI] congenita) with intrauterine fractures and an early death and a milder dominantly inherited form (OI tarda) which was further subdivided into gravis or levis types according to whether fractures present within the first year of life or thereafter. ³^, ⁴ Others favored subdivision into the mild and severe types based upon the presence of long bone deformity (▶ Fig. 10.1 and ▶ Fig. 10.2) and this appears to correlate better with clinical features. ⁵^, ⁶

Fig. 10.1 The radiological classification of osteogenesis imperfecta. (a) Mild disease showing osteoporosis and Harris’ lines. (b) Severe disease with multiple fracturing and gross long bone deformity.

Fig. 10.2 (a,b) Clinical appearance of a young boy with severe osteogenesis imperfecta. (c) Varus femurs following repetitive fracturing. (d) The boy also had dentinogenesis imperfecta.

The incidence of the condition is 1 in 20,000 at birth with about 80% of cases caused by autosomal dominant mutations of the type I collagen genes, type I collagen being the principal collagen of bone, dentine, and sclera and so these are the tissues most commonly affected in OI. ⁷ These mutations produce two different protein defects, the first in activating one allele with a 50% reduction in type I collagen in bone while the other has a reduced amount of normal collagen in addition to type I collagen molecules containing mutant collagen genes. ⁸ As a result the first type produces the mild type I OI form of the disease while the more severe category produces the worst types.

There are now some eight types of OI in up-to-date classifications. ⁷ Type I, with bone fragility but with blue sclerae and autosomal dominant inheritance, is the most common and mildest form. Fracturing starts to occur when the child starts to walk and the fracture types are similar to those seen in normal children but of course they fracture more easily. The most common biochemical abnormality is defective synthesis of sufficient type I collagen, particularly the alpha 1 (1 chain), which obviously will affect the skeleton as bone contains type I collagen only. ⁶

As regards the spine the radiographic changes are indistinguishable from those in juvenile osteoporosis, with multiple compression fractures of the vertebrae that are biconcave ⁹^, ¹⁰ (▶ Fig. 10.3). About 70% of patients with OI can be shown to have a scoliosis ¹¹^, ¹² (▶ Fig. 10.4) and a similar majority can be shown to have anterior chest deformities in the nature of pectus carinatum, excavatum, or an increased sagittal chest diameter. ¹² In a multicenter study of OI in North America involving 544 OI subjects it would seem that type III cases had a higher prevalence rate of severe scoliosis and long bone deformities than types I and IV. ¹³ Meanwhile in a study of spinal curvature during growth in 316 OI patients scoliosis had a prevalence rate of 50% overall but type III had the highest prevalence rate of 68% with a mean progression rate of 6% per year, while with type IV cases the prevalence rate was 54% and the progression rate 4% per year. The lowest prevalence rate was seen in type I OI with a 39% prevalence rate with progression of 1% per year. For type III patients treatment with bisphosphonates significantly decreased the progression rate. ¹⁴

Fig. 10.3 (a) Lateral radiograph showing the typical osteoporosis and biconcave compression fractures of osteogenesis imperfecta. (b) Lateral radiograph of severe spinal disease indistinguishable from idiopathic juvenile osteoporosis.

Fig. 10.4 (a) Scoliosis in association with severe osteogenesis imperfecta O.I. with painful costovertebral impingement. (b) PA radiograph showing the severe deformity. (c) After Harrington rod fixation with cementation of the upper and lower hooks. (d) Five years later this correction was sustained and his symptoms remained relieved.

With so many other dysmorphic problems, altering spinal shape is clearly not a priority and Cobb angle would not be an important variable determining the need for surgery. In addition the history of surgical treatment for scoliosis has demonstrated just the sort of problems that one would anticipate. The earliest report of substantial numbers was by the Scoliosis Research Society. ¹⁵ Posterior stabilization was the method of choice, 39 of 55 having Harrington instrumentation with only 4 anterior fusions and 1 combined two-stage procedure. Cobb angle correction was a mere 36 degrees with significant complications in one-third of cases with loss of instrument attachment to bone and pseudarthrosis being the major culprits. ¹⁵ Waugh ¹⁶ suggested the use of methyl-methacrylate cement to strengthen hook sites and this has been used in patients with OI. ¹⁷ However, with cement augmented pedicle screw instrumentation for the surgical treatment of a scoliosis in OI quite promising results have been reported by Yilmaz et al. ¹⁸ There were 10 operated patients, 7 of whom had cement augmented screw insertion at the proximal and distal foundations. The mean Cobb angles to start with were 84 degrees and at follow-up 40 degrees with no instrumentation failures.

10.1.2 Marfan Syndrome

Marfan syndrome is an autosomal dominant disorder in which there is an abnormality in the Fibrillin-1 gene so that this is found in soft tissues. ⁷ The fibroblast-produced collagen appears unusually soluble. The typical patient is tall and thin with arachnodactyly (hands and feet) and pectus carinatum or excavatum. The tall stature is disproportionate with a reduced upper to lower segment ratio. There is a long narrow face with a high arched palate and eye problems are common including dislocated lenses and retinal detachment (▶ Fig. 10.5). Joints are hypermobile but the most significant problems affect the cardiovascular system with aortic incompetence and dissecting aneurysm as well as mitral and tricuspid valve disease. These are the usual causes of early death often in the forties.

Fig. 10.5 The typical appearance of Marfan syndrome: tall, disproportionately long slim limbs, arachnodactyly, and an increased upper to lower segment ratio.

The condition was first described in 1896 ¹⁹ although one of Marfan’s cases was certainly congenital contractural arachnodactyly. ²⁰ The differential diagnosis is therefore important and includes homocystinuria and congenital contractural arachnodactyly. ²¹ Homocystinuria has only been described recently and can be differentiated from Marfan syndrome by the presence of widened epiphyses and metaphyses, frequent osteoporosis of the spine in particular, and an increased incidence of thromboembolism. ²² Arachnodactyly and scoliosis are less common in homocystinuria. Congenital contractural arachnodactyly is characterized by multiple joint contractures present at birth with elbow, knee, and ankle particularly affected whereas if contractures develop in Marfan syndrome they do so later. There is also the characteristic abnormality of the ears and there is no lens dislocation, heart disease, or ligament laxity.

In Marfan syndrome severe planovalgus feet are very common in association with the arachnodactyly of the digits (▶ Fig. 10.6). The prevalence rate of scoliosis in Marfan syndrome varies from about one-third to three-quarters in series of reasonable numbers of patients suggesting an average figure of about 50%. ²³^– ²⁸ The deformity is very similar to that in idiopathic scoliosis (lordoscoliosis). An increased prevalence rate is however readily explicable on the basis that the less strong soft tissue support for the spine increases the intrinsic load on the spine rendering it less able to resist buckling deformation (lordoscoliosis) or simple angular collapse (kyphosis). Of 35 patients reported by Winter and Moe in 1969, ²⁵ 18 had a scoliosis with a curve size varying from 15 to 185 degrees. There were two typical curve patterns, right thoracic or thoracolumbar and double-structural—the latter giving rise to the typical flat-back appearance (▶ Fig. 10.7). Interestingly the curve was painful in five of their cases. The progression rate was similar to that in idiopathic scoliosis and treatment was along the same principles. In a later report of 35 cases of scoliosis in association with Marfan syndrome three-quarters were painful and almost half started in the infantile or juvenile periods. ²⁸ There was no evidence, not surprisingly, that the 14 who were braced derived any benefit therefrom. The adolescent with Marfan syndrome can undergo precisely the same type of surgical correction as the idiopathic counterpart and there would appear to be no significant surgical risk from a delicate cardiac status. However, prior to surgery, cardiovascular assessment including ultrasonography of the heart valves and aorta should be undertaken, regularly in childhood anyway, and essentially before undertaking any major surgery in older children and teenagers. Aortic replacement with or without aortic valve replacement is frequently required and is best undertaken early because vascular complications are the usual cause of the premature death. In this regard hypotensive anesthesia is to be recommended. ⁶

Fig. 10.6 The typical deformities of Marfan syndrome. (a) Severe planovalgus feet. (b) Arachnodactyly. (c) Congenital dislocation of the radial head. (d) Typical joint hypermobility in Marfan syndrome.

Fig. 10.7 (a) PA radiograph showing the typical double-structural curve with Marfan syndrome. (b) Lateral radiograph showing the severe flat-back deformity of a double lordosis in Marfan syndrome.

Whereas in Minneapolis, Louisville, Leeds, and Karlsbad, we have not encountered any surgical problems peculiar to Marfan syndrome rather than those of the late-onset idiopathic counterpart; some more recent reports seem to raise concern. Whereas in China, Li reported 12 consecutive cases with posterior segmental instrumentation correction and observed no significant complications, ²⁹ a multicenter study from several orthopaedic centers in the United States suggested differences in outcome compared to similar adolescent idiopathic cases ³⁰; as did that of Zenner et al. ³¹ One of the differences observed by Gjolaj et al ³⁰ was that many more patients with Marfan syndrome had a kyphosis of more than 50 degrees. One does find it slightly irritating to put it mildly that one still hears of kyphosis when we have published so many papers establishing that all structural scolioses are lordotic and what is being perceived as kyphosis is just a severely rotated lordoscoliosis (see ▶ 3). Yes, the spine is pointing backward but it is the front of the spine pointing backward and not the back of the spine (▶ 2 and ▶ 3 as well should remove all reasonable doubt). These sizeable pseudokyphoses do indicate the need to go anteriorly with instrumentation as well as posteriorly if necessary, as with neurofibromatosis 1 (NF1) (Fig. 8.7).

As multiple curve patterns are particularly prevalent in Marfan syndrome then lengthier instrumentations are bound to be required, but this is not a complication—rather, it is an idiopathic-type curve pattern seen more frequently in Marfan syndrome. There were three cerebrospinal fluid (CSF) leaks in the 34 Marfan patients with none in the idiopathic group. Dural ectasia has been reported in the Marfan syndrome ³² and there is no reason of course why the soft tissue of the dura should not be subject to the same Marfan soft tissue weakness (as with the cardiovascular system). The problem of “adding on” should be obviated with rigorous attention to LIV selection. In seven cases instrumentation was taken down into the pelvis, four primarily and three for revision, the need for which we simply cannot envisage. Unlike paralytic curves where maybe significant pelvic obliquity in both Marfan and idiopathic scoliosis L5 is always the robust LIV with immensely strong attachments to the pelvis via the iliolumbar and lumbosacral ligaments, as well as stronger bone than the pelvis.

Meanwhile Zenner et al ³¹ reported on 23 Marfan patients while using unnecessarily expressive language such as “the individual challenges of the underlying desmogenic disorder” went on to demonstrate that excellent/good outcomes were noted overall in 80% of patients.

There is absolutely no reason why the Marfan patient should not be treated exactly as the idiopathic counterpart with the same excellent expectations.

Whereas scoliosis is less frequent than in homocystinuria it is important to make sure that the suspected Marfan case does not have homocystinuria (detectable in the urine by the nitroprusside test) because vascular damage leading to thrombosis would be a surgical contraindication.

10.1.3 Mucopolysaccharidoses

These disorders affect the skeleton by way of failure of the normal breakdown of complex carbohydrates which therefore accumulate in the tissues and appear in excess in the urine. The last several decades have seen knowledge of these conditions change from purely descriptive aspects to a more clear understanding of the biochemical defects involved. ⁷ There are two groups: mucopolysaccharidoses (MPS) and mucolipidoses. The dreadful prognosis of the latter makes them very unlikely to be encountered by orthopaedic surgeons. The incompletely broken down mucopolysaccharides accumulate in the lysosomes of cells and many tissues can therefore be affected (e.g., cartilage, bone, liver, central nervous system).

The typical deformity of the MPSs is a thoracolumbar kyphosis due to apical developmental wedge-shaped vertebrae. Although spinal surgeons are unlikely to encounter MPSs other than Morquio’s syndrome, in MSII Hunter syndrome—which is milder physically and mentally than Hurler syndrome (MPS I)—there is again a thoracolumbar kyphosis with characteristic upper lumbar vertebral beaking (▶ Fig. 10.8).

Fig. 10.8 Hurler’s syndrome. Lateral radiograph showing the characteristic thoracolumbar kyphosis with the apical bullet-shaped vertebra.

10.1.4 MPS IV: Morquio’s Syndrome

This condition was first described in 1929 by Morquio of Uruguay ³³ who described four children, two girls and two boys, out of a family of five. At the same time in Birmingham, Brailsford published a case of Naughton Dunn’s. ³⁴ The condition is caused by a deficiency of the enzymes responsible for the degradation of keratan sulphate. The condition is autosomal recessive with a normal intelligence and a variable severity of associated physical disability (▶ Fig. 10.9). The major problems are skeletal with deformity and short-trunk dwarfism—patients rarely achieving a height greater than 1.2 m. There is a flexed stance with genu valgum and there is often hyperextensibility of joints and skin as well as corneal clouding. The prognosis is variable and death often occurs early from cardiorespiratory failure or spinal cord compression.

Fig. 10.9 Two characteristic skeletal features of Morquio’s syndrome. (a) PA of the pelvis showing a hypoplastic pelvis with iliac flaring and a wine glass–shaped pelvis with marked dysplasia of the femoral capital epiphysis. (b) The characteristic radiographic appearances of the knees in Morquio’s syndrome. There is deficient lateral ossification of the femoral and tibial epiphyses and metaphyses which leads to genu valgum.

As regards the spine there is an early thoracolumbar kyphosis with later platyspondyly and anterior vertebral beaking (▶ Fig. 10.10). There are two major problems: the thoracolumbar kyphosis and atlantoaxial instability as a result of a deficient odontoid (▶ Fig. 10.11). The thoracolumbar kyphosis is important because it is both common and dangerous from the point of view of spinal cord compression. Melzak reported two sisters with Morquio’s syndrome who both had spinal cord compression from thoracolumbar kyphoses. ³⁵ More recently in a study of 80 dwarfs of various types, 18 were MPSs and 14 of these had a thoracolumbar kyphosis measuring between 14 and 53 degrees. ³⁶ Thoracolumbar kyphosis is associated with a barrel-shaped chest, a prominent sternum above, and premature fusion of the sternal segments with an immobile chest which can lead to severe chest dysfunction resulting in early death.

Fig. 10.10 The spine in Morquio’s syndrome. (a) PA radiograph showing the severe degree of platyspondyly. (b) Lateral myelogram showing indentation of the dye column over the region of the kyphosis thought to be due to meningeal thickening. Note the apical vertebra hooked anteriorly.

Fig. 10.11 Lateral radiograph of the upper cervical spine showing a deficient odontoid, the cause of atlantoaxial instability (AAI). The neck is extended and this reduces the AAS.

Only gold members can continue reading. Log In or Register to continue