INTRODUCTION

The medical team’s desire to provide a more active and three-dimensional (3D) scoliosis brace, as well as the introduction of thermoplastics, have allowed for a proliferation of advances in scoliosis braces.  Currently the thoracolumbosacral orthosis (TLSO) Boston brace is considered the “gold standard”, and has superseded the conventional Milwaukee brace, which has been for many years the industry standard.

Thermoplastic braces such as the TLSO Boston brace places pressure pads over the convexities of the thoracic and lumbar curves in attempt to correct lateral deviation and rotation.  However, Dr. Chêneau in 1979 found these general correction principles insufficient, and as a result, he designed a Chêneau brace that endeavours to treat every aspect of the complex 3D deformity (Rigo and Chêneau, 1997).  The Chêneau brace is defined as a thermoplastic brace modelled on a hyper-corrected positive plaster-cast of the patient.  The general correction principle is that of detorsion and sagittal plane normalisation, which would correct the coronal and transverse planes, resulting in some elongation of the spine, without any significant distraction force. However, due to the complexity of rectification and fitting processes, it has been limited to specialised clinics in Europe.

2.1  ANATOMY OF THE VERTEBRAL COLUMN

The spine can be divided into anterior and posterior columns.  The anterior column consists of the posterior longitudinal ligament, intervertebral disc, vertebral body, and anterior longitudinal ligament.  The elements of the posterior column are the pedicles, laminae, transverse processes, spinous process, facet joints, and ligamentous structures, including the facet joint capsule, ligamentum flavum, intertransverse ligaments, interspinous ligaments, and supraspinous ligaments. Each anatomic component of the vertebral column has a function that contributes to the mobility and stability of a motion segment (Verbout, 1985; Lonstein et al., 1995).

2.2  PHYSIOLOGY OF THE VERTEBRAL COLUMN

The normal adult vertebral column has four curvatures in the sagittal plane, a convexity (lordosis) in the cervical and lumbar regions and a concavity (kyphosis) in the thoracic and sacrococcygeal regions.  In the coronal plane, the vertebral column is normally straight.  In the sagittal plane, both the cervical and lumbar curvatures are acquired in late foetal development when the infant begins to hold up its head to enlarge its visual environment.  The secondary lumbar curvature appears when the child begins to sit up at around 6 months, becoming more marked with standing and the onset of walking.  It is the extension of the hip, which accompanies standing and walking which tilts the pelvis forwards so that the axis of the pelvic cavity is no longer in line with that of the abdominal cavity.  The lumbar curvature develops in order to keep the trunk erect when standing.  The lumbar curvature is not fully developed until after the age of two, when a more or less adult pattern of walking is established.  In old age, the vertebral column tends to assume a gentle C-shaped curve, which is reminiscent of the fetal curve.  The reason for this is that the shape of the vertebral column is largely determined by the intervertebral discs and to a much lesser extent by the vertebras themselves.  Consequently as the discs degenerate and become thinner with increasing age, the secondary curvatures gradually disappear (Netter, 1985; Palastanga et al., 1990).

2.3  BIOMECHANICS OF THE VERTEBRAL COLUMN

Humans have an axial skeleton uniquely adapted to bipedal ambulation (White and Panjabi, 1978).  Sagittal plane contours permit the centre of mass for the head and upper torso to remain in line with the vertical axis through the centre of mass for the pelvis; therefore, a minimal expenditure of energy is required to keep the trunk upright.  The upper limbs, thus freed from the task of trunk support, are able to perform other functions associated with a complex society.  Various pathologic conditions causing abnormal sagittal plane contour, such as loss of lumbar lordosis, excessive thoracic kyphosis, or coronal plane deviation of the spine, may alter balance and coordination, interfere with visceral function, allow premature degeneration of the intervertebral disc and facet joints, and cause deterioration of neurologic function, (Netter, 1985; White and Panjabi, 1990).

To achieve the balance and mobility required for efficient energy use, humans have a multisegmented, bony spinal column.  The normal spinal column consists of 7 cervical, 12 thoracic, and 5 lumbar vertebrae connected to fused sacral vertebrae, in turn, articulate with vestigial coccygeal vertebrae.  When viewed in the coronal plane, the normal spinal contour has less than 10 degrees of lateral curvature and when viewed in the lateral plane it has physiologic cervical lordosis, thoracic kyphosis, and lumbar lordosis (White and Panjabi, 1978, 1990).

In mechanical terms, the vertebral column can be modelled as a series of semi-rigid bodies, the vertebras, separated by viscoelastic linkages and the intervertebral discs and ligaments (Pearcy, 1989).  Attached to the vertebral column are various viscoelastic and solid materials with varying mechanical properties.  They vary from the stiff ribs associated with the thoracic region, to the subcutaneous fat.  These elements form part of a body cylinder, to which the spinal brace or orthosis is applied.  The effectiveness of a spinal brace can be assessed in biomechanical terms, whether the main function is one of support, immobilisation, correction and/or pain relief.  The nature of the close-fitting orthosis establishes externally applied forces that are transmitted to the vertebral column to obtain the desired therapeutic goal.  The effectiveness of the force transmission from the orthosis to the vertebral column is determined by the mechanical properties of the human body, in particular the stiffness characteristics of the intervening biological materials.  Therefore it is more effective when the applied forces are directed through a rigid material that deforms minimally under pressure compared with less rigid material that can deform under pressure.  It can be seen why a spinal brace is more effective in holding or correcting thoracic curves, where the forces are

transmitted through the ribs, compared with the lumbar curves where the intervening soft materials are composed of muscles and viscera (White and Punjabi, 1990; Chase et al., 1993).

2.4  LITERATURE REVIEW

2.4.1  AETIOLOGY OF IDIOPATHIC SCOLIOSIS

By definition, the cause of idiopathic scoliosis is unknown (Lonstein et al., 1995).  Although research has possibly eliminated some hypothetical causes, abnormalities of disc, bone, muscle, and collagen do not appear to be aetiological factors (Abbott-Byrd, 1988; Child et al., 1999; Miller et al., 1999).  However they reflect the effects of scoliosis on normal tissues (Abbott-Byrd, 1988).  Idiopathic scoliosis is the most common type of lateral deviation of the spine (Abbott 1912), and as a result, this has prompted many lines of research, which focus on the genetic aspects (Fillio and Thompson 1971; Miller et al., 1999), growth aspects (Skogland and Miller 1980; Duval-Beaupere, 1970, 1992), structural and biochemical changes in the discs and muscle (Riddle and Roaf, 1975; Taylor et al., 1981; Drummond et al., 1984; Child et al., 1999), and on central nervous system changes (Willner, 1972; Yamada et al., 1974).

Family and population studies point to a hereditary factor to explain the well-known familial pattern (Cowell et al., 1969), however the mode of inheritance is uncertain. 

I) CURVE PROGRESSION

Growth has a definite role in idiopathic scoliosis.  Curves progress rapidly during the adolescent growth spurt, which occurs at the age of 12 years in girls and at 13 or 14 in boys (Duval-Beaupere, 1970; Stokes, 1999).  Willner (1975) reported that idiopathic scoliosis is intimately associated with growth and development, particularly during the beginning of puberty, when the rate of growth is at a maximum.  Moe (1969) claimed that the crucial year in girls is the year before menarche, (the onset of menstruation).  It is at this time that the risk of progression of the spinal curvature is at its greatest.  The Swedish study of Nordwall and Willner (1975) showed that teenage girls with idiopathic scoliosis had a skeletal age, which in early adolescence was more advanced than normal, and a menarche which did not differ from that of normal girls.

II) EFFECT ON GROWTH

Drummond et al., (1984) reported in 409 adolescents with idiopathic scoliosis that growth of children with scoliosis did not appear to differ from that of their peers.  However, when growth was compared with skeletal age, the children with scoliosis were found to be taller and heavier.  Both boys and girls with scoliosis showed a significant tendency for a delay in skeletal age and the girls showed a significant tendency for a delay of puberty.  The late skeletal and sexual development observed for the entire series was even more apparent for the girls with a Cobb angle greater than 20 degrees.

In another Swedish study, Willner (1975) found that girls with adolescent idiopathic scoliosis were significantly taller than their normal peers.  These girls started their growth spurts earlier, grew for a longer period and had a skeletal age more advanced than their normal peers. At the end of growth, the heights of the girls with scoliosis and their normal peers were equal.  The levels of growth hormone in girls with scoliosis were compared to normal, and some studies showed an increase in these levels, however Misol et al., (1971) could not confirm this finding.

Investigations into collagen in the ligaments and tendons in patients with idiopathic scoliosis were compared to normal, but no differences were found (Waters and Morris, 1973; Lonstein et al., 1982,).  Muscles have been implicated as the cause of idiopathic scoliosis however electromyographic studies have been inconclusive (Zuk, 1962; Riddle and Roaf, 1975).  Increased activity on the convexity has been found by some investigators (Butterworth and James, 1969; Sahlstrand and Petruson, 1979), whereas others found no difference (Henssge, 1967; Lihvar et al., 1975).   Postural equilibrium dysfunction has been found by many authors (Herman et al., 1979; Willner, 1982), and these findings were not specific for idiopathic scoliosis.  It appears that there may be a postural equilibrium problem in idiopathic scoliosis and some authors have suggested that this may be due to the brainstem.

The cause of idiopathic scoliosis is still unknown, however despite numerous studies that have been done on the subject, it appears that the cause is multifactorial, as no single causative factor can be found (Czeizel et al., 1978; Aksenovich et al., 1988).  Genetic, growth, chemical, biomechanical and neuromuscular factors all seem to be

involved (Willner, 1982; Child et al., 1999).  It has also been postulated that a mild central nervous system abnormality is genetically determined.  With increased growth and the altered viscoelasticity of the discs, the spine is biomechanically less stable, making it susceptible to changes in postural equilibrium.  The interrelation of all these factors determines whether the curve is progressive or nonprogressive, and how much progression will occur (Lonstein et al., 1995).

III) PREVALENCE

Prevalence refers to the number of the population with the disease or disorder, therefore when discussing scoliosis, the studies give prevalence rates.  Prevalence rates vary as to the degree of Cobb angle, being 20 to 30 cases per 1000 individuals for curves over 10 degrees Cobb angle.  The number of cases reduces to three to five cases per 1000 for curves over 20 degrees Cobb angle and two to three per 1000 for curves over 30 degrees Cobb angle (Shands and  Eisberg, 1955;  Dickson et al., 1980;

Willner, 1982).  Therefore, the prevalence of idiopathic scoliosis decreases when a larger curve magnitude is considered. A study in Edinburgh of 153 patients with idiopathic scoliosis showed that 4% had infantile, 7% had juvenile, and 89% had adolescent idiopathic scoliosis (McMaster, 1983).  Mau (1981), in Germany as well as Riseborough and Wynne-Davies (1973), in North America found similar frequencies in idiopathic scoliosis.

2.4.2  NATURAL HISTORY OF ADOLESCENT IDIOPATHIC SCOLIOSIS

Much work has been done on the negative effects of untreated scoliosis, which involves back pain, cardiopulmonary problems and socio-economic effects (Nachemson, 1968, 1996; Weinstein et al., 1981).  The progression rates (in which the Cobb angle increases five degrees or more during a six-month period), are fairly well documented for adolescent idiopathic scoliosis and these vary from 5.2% to 56% (Clarisse, 1974;  Brooks et al., 1975; Rogala et al., 1978; Fustier, 1980; Lonstein and Carlson, 1984; Bunnell, 1986), with the lower rates being found in school screening studies (table 2.1).

The factors that are related to the risk of curve progression are divided into two groups, those of curve magnitude and growth potential.  Firstly, the factors related to the curve magnitude, such as whether the curve is 25 to 29 degrees or over 30 degrees, are evaluated.  Secondly, those related to the child’s growth potential, such as age, skeletal maturity, menarchal stage (Peterson and Nachemson, 1995), and the stage of development of the apophysis of the iliac crest (the Risser sign (Risser, 1958, 1964)).  Generally the larger the Cobb angle, the greater the incidence of progression.  This also varied with the curve pattern (Clarisse, 1974).  Lonstein and Carlson (1984) found that, in curves between 5 and 29 degrees, the incidence of progression in the different curve patterns was fairly equal, except for the single lumbar and single thoracolumbar pattern.  Therefore, a double curve is more likely to progress than a single curve (Lonstein et al., 1995; Peterson and Nachemson, 1995).

                    Table 2.1  Progression of idiopathic scoliosis in relation to the Cobb angle (Lonstein et al., 1995).

When analysing the growth potential and curve progression, it is generally true that the younger the child (i.e. the greater the growth potential), the greater the incidence of progression (Goldberg et al., 1993; Lonstein et al., 1995;  Peterson and Nachemson, 1995).  This can be measured by age (chronological or skeletal), menarchal status, or Risser sign.  The Risser sign (Risser, 1958, 1964), is measured by the ossification of the iliac epiphysis (figure 2.1).  Ossification normally starts at the anterior superior iliac spine (ASIS) and progresses posteriorly to the posterior superior iliac spine (PSIS).  Risser divided the excursion into four quarters, Risser sign 1 through 4, with Risser sign 5 once complete ossification has occurred in which fusion to the iliac crest takes place. The incidence of progression is higher in adolescents with a Risser sign of 0 or 1, compared to those with a Risser sign of 2 or more (Lonstein et al., 1995).

Figure 2.1  Coronal plane view of the pelvis and the 4th and 5th lumbar vertebrae which presents iliac epiphysis.  Ossification of the epiphysis usually starts at the anterior superior iliac spine and progresses posteriorly. The iliac crest is divided into four quarters, and the excursion or stage of maturity is designated as the amount of progression. In the example shown, the excursion is 50 per cent complete, and the Risser sign is thus 2+. On the right, the excursion is complete and the epiphysis has fused with the iliac crest, this is a Risser 5+.

A useful cross-correlation of incidence rates of progression for curves under 29 degrees was reported by Lonstein et al., (1995).  The two factors taken were the curve magnitude and maturity as assessed by the Risser sign (table 2.2).  These figures are used for the natural history of incidence of progression when evaluating the effectiveness of treatment.

Table 2.2  The prediction of curve progression in untreated idiopathic scoliosis during growth (Lonstein et al., 1995).

For curves of 20 to 29 degrees in an immature child with a Risser sign of 0 or 1, the incidence of progression was 68 %.  On the other extreme, for curves less than 19 degrees in a mature adolescent with a Risser sign of 2 or more the incidence of progression was 1.6%.  In the other two groups, that is, a smaller curve (less than 19 degrees) in an immature child (Risser 0 or 1), and a larger curve (20 to 29 degrees) in a mature child (Risser 2 or more), the incidence of progression is approximately the same at 22% to 23% (Lonstein et al., 1995).

Bunnell (1988), reported that a large number of patients have minor degrees of curvature (approximately 100 per 1000 patients), although only about two per 1000 patients warrant treatment because of curve progression.  As a result, a large number of these curves do not increase even without treatment.

Kehl and Morrissy (1988) reported that, in the past, the majority of idiopathic curves in patients who presented for treatment were of a greater magnitude (40 to 60 degrees) than what is seen today.  This is because school screening programs becomes more prevalent and a large number of patients with smaller curves (10 to 30 degrees) have been identified.  Past experience also indicated that idiopathic scoliosis was a disease of relentless curve progression. Kehl and Morrissy, (1988), assumed that without early treatment, curves that had progressed to 20 degrees would continue to progress and subsequently become the larger curves that normally constituted the bulk of scoliosis practice.  As bracing was aimed at preventing the progression of scoliosis as all curves were assumed to be progressive, it was logical to begin brace treatment early for all small curves.  This concept of treating all 20-degree curves with a brace was supported

by initial reports that showed a high rate of success in controlling progression (Kehl and Morrissy, 1988).  This was an erroneous assumption demonstrating the fallacy of evaluating the treatment of a disease in which the natural history was not known.  Statistics from school screening studies would soon change this concept (Edgar et al., 1982; Goldberg et al., 1993; Peterson and Nachemson, 1995; Lonstein et al., 1995).

Kehl and Morrissy (1988) claimed that although the understanding of the natural history of adolescent idiopathic scoliosis has increased, it remains incomplete.  This information implies that the high rate of success seen in bracing in idiopathic curvature was more likely to be secondary to the favourable natural history associated with these cases than the effects of wearing a brace.  It also implies that the frequency of the need for bracing is much less than has previously been proposed (Kehl and Morrissy, 1988).

Therefore, by claiming that if all idiopathic curves less than 25 degrees were routinely braced, simply to prevent a progression from occurring, then a very large percentage of patients treated would be braced unnecessarily.  This unnecessary bracing would subject many adolescent patients and their families to unneeded financial cost and potential psychological harm without adding any potential benefit to the patient’s final outcome (Kehl and Morrissy, 1988).

Goldberg et al., (1993) found that observation of progression of at least 10 degrees, which occurred in 14% of the group of 339 girls with adolescent idiopathic scoliosis, depended on the timing of diagnosis and related primarily to the child’s position on her growth rate curve as well as her pubertal status, and much less to her skeletal maturity as interpreted by iliac crest ossification or bone age.

2.4.3  PHYSIOTHERAPY

Physiotherapy programs such as the Schroth technique have been advocated as a necessary part of brace treatment in idiopathic scoliosis (Weiss and Werkmann, 1996; Rigo 1996, 1997, 1999a).  Exercise is believed to improve the brace treatment of scoliosis by maintaining flexibility of the spine as well as preventing paraspinal muscle atrophy secondary to the immobilising effects of the brace.

2.4.4  SCOLIOSIS:  3D DEFORMITY

According to Dubousset (1992), from an anatomical point of view, a scoliotic deformity can be described as a series of vertebral segments placed in extension or lordosis, which deflect and axially rotate towards the same side.  Lateral curve, anatomical lordosis and axial rotation are the three elements of a scoliotic lesion.

In all cases, the intention of bracing is to change the shape of the spinal column.  The shape of a scoliotic spinal column is best defined from a geometric rather than an anatomical point of view.  Aubin et al.,(1997) recently described scoliosis as a complex process of trunk deformation including morphological changes and a global transformation of the shape of the vertebral column, which moves from its original position in the sagittal plane, to a complex torsional geometry in the three dimensions of space.

The term torsion, has two meanings.  One meaning is mechanical torsion, which refers to the torsional deformity of the spinal column being considered as a plastic structure and which includes an intravertebral torsion and an intervertebral torsion.  In 50% of patients, intravertebral torsion is responsible for approximately 45% of total axial rotation (Aubin et al., 1997).  For this reason, braces cannot correct this aspect as effectively as they correct the lateral curve.  The term is also used as a geometrical or helicoidal torsion, which refers to the twisting of the spine considered as a line in space.  The vertebral column, moved by deforming forces, changes its physiological shape in the coronal, transversal and sagittal planes, adopting extremely diverse anatomoradiological patterns.

2.5  CLASSIFICATION OF CURVE PATTERNS

Classification of curve patterns is necessary to allow comparisons and prognoses of various patients.  In idiopathic scoliosis, the curve pattern generally does not change from that noted at the onset of the deformity.  Although the sagittal plane X-ray is often evaluated for the identification of sagittal plane deformities, (i.e. hypokyphosis or hypolumbar and hyperlumbar lordosis), a scoliotic curve is classified by evaluation of the dorsal aspect of the coronal plane X-ray.  Also, it is important to identify the direction of the curve or curves of the scoliosis pattern as a left or right.  A left curve has the direction of the convexity to the left and its concavity on the right of the vertebral column.  A right curve has the direction of the convexity to the right and its concavity on the left of the vertebral column.  In all types of classification of scolioses patterns, the vertebral column is always viewed by its dorsal aspect or posterior view of the pelvis and vertebral column.

2.5.1  CLASSIFICATION OF COBB

Cobb (1948) reported the terms of major and minor curves, which are frequently used in the United States and United Kingdom.  A major curve is more structural and deforming, whereas a minor curve is less structural and less deforming and is often called the compensatory curve.  Cobb classified the following major and minor curves, which are helpful in outlining treatment and prognosis (Cobb, 1948):

1.   Single major high thoracic curve:  A high thoracic curve with the apex in the upper thoracic spine, which may appear without a lower thoracic curve or with a small, flexible curve below.

2.   Single major thoracic curve:  The apex lies within the thoracic spine, the upper end vertebra is between  the 4th to 6th thoracic vertebra (T4 to T6), and the lower end vertebra is between the 11th thoracic to 2nd lumbar vertebras (T11 to L2).  The majority of these curves are convex to the right and present hypokyphosis.  This pattern is usually associated with a rotation prominence; the magnitude varies significantly from curve to curve and is not related to the degree of Cobb angle or to the rotation seen on the coronal plane radiograph.

3.  Single major thoracolumbar curve:  This is a single curve with its upper end vertebra between T8 to T10 and the lower end vertebra at L3.  The apical vertebra is T12 or L1.  Both the upper thoracic and lower lumbar spine may show small compensatory curves, which are usually completely flexible.

4.   Single major lumbar curve:  This is a lumbar curve, often small and flexible with the apex usually at L2.  The upper end vertebra is between T11 to L1 and the lower end vertebra at L 4 or L5.  A pelvic tilt is often present with this curve pattern.

5.   Major thoracic and minor lumbar curves:  This curve pattern, which is commonly seen, consists of an upper curve with the upper end vertebra at T4 or T5 and the lower end vertebra at T12, and a lower curve has the upper end vertebra at T12 and the lower end vertebra L4 or L5.  The upper curve is larger and more structural of the two.

6.   Double major thoracic and lumbar curves:  This pattern consists of both thoracic and lumbar curves, both of which appear at the some time, usually during the juvenile years.  Both curves are of nearly the same degree of Cobb angle and rigidity.  The thoracic curve is generally convex to the right (but could be to the left) with the apical vertebra at T7 or T8, the upper end vertebra at T4, T5, or T6, and the lower end vertebra at T10, T11, or T12.  The lower curve is generally convex to the left (but could be to the right), having its apical vertebra at L1 or L2 and extending to L4 or occasionally L5.

7.   Double major thoracic and thoracolumbar curves:  The thoracic curve has its upper end vertebra at T4 and the lower end vertebra at T9 or T10, with its apex at T6 or T7.  The convexity is usually to the right with minimal associated rib prominence.  The thoracolumbar curve has its upper end vertebra at T9 or T10 and the lower end vertebra at L3, with an apex at the T12 - L1 disc space.

8.      Double major thoracic curves:  This curve pattern is commonly presented as a left upper and right lower thoracic curves (however they could be to the right and left respectively).  The upper curve has its upper end vertebra at T1 or T2 and the lower end vertebra at T5 or T6, with its apex at T3 or T4.  The lower curve has its apex within the thoracic spine, with the upper end vertebra at T5 or T6 and the lower end vertebra between T 11 to L2. 

9.   Multiple curve patterns:  Multiple curves other than those described above do occur, but they tend to be short and nondeforming.

2.5.2  CLASSIFICATION OF CHÊNEAU

The classification used by Chêneau (1990, 1996a, 1996b) simplifies the patterns into two groups, 3-curve and 4-curve scoliosis, except for the rare exception of the “C” shaped scoliosis that can be found in spastic patients.  Both the 3-curve and 4-curve patterns present two compensatory curves or half curves, one is located above and the other below the principal curve or curves of the coronal plane (Moe and Kettleson, 1970). Therefore a pattern with one primary curve with two compensatory curves is a 3-curve scoliosis and a pattern with two primary curves with two compensatory curves is a 4-curve scoliosis.

I) 3-CURVE PATTERN

This presents as a long thoracolumbar curve with the apex around T9 or T10, which is generally convex to the right (figure 2.2a).  The thoracolumbar curve could be convex to the left, however it is more often to the right.  To the opposite side of the thoracolumbar curve, cephalically there is a high thoracic (compensatory) hemi-curve with the apex around T1 to T4.  To the opposite side of the thoracolumbar curve, caudally there is a low curve, which deviates and rotates the pelvis to the left.  In the case that the thoracolumbar curve is to the left, the pelvis deviates and rotates to the right, (Chêneau, 1990, 1996a).

II) 4-CURVE PATTERN

This presents as two primary curves, one in the thoracic region and the other in the lumbar region (figure 2.2b).  These double curves have two small compensatory curves or hemi-curves, one cephalic and the other caudal.  The two compensatory curves or hemi-curves are not easily seen by viewing the pattern.  The upper primary curve has a thoracic curve with the apex at T8 or T9 and the lower primary curve has a lumbar curve with the apex around L2.  These double curves generally have the convexity of the thoracic curve to the right and the convexity of the lumbar curve to the left.  The opposite curve direction could occur (i.e.: convex to the left in the thoracic region and convex to the right in the lumbar region); however it is less common.

To the opposite side of the thoracic curve, cephalically there is a high thoracic (compensatory) hemi-curve with the apex around T1 to T4.   To the opposite side of the lumbar curve, caudally there is a low compensatory curve with the apex at L4, L5, or the 1st sacral vertebra (S1).  The pelvis could be deviated to the right, not rotated or slightly rotated. (Chêneau, 1990, 1996a).

Figure 2.2 (a) Posterior view of a 3-curve scoliosis pattern, in this case , the thoracolumbar curve is convex to  the right with the apex at T9 or T10, (represented as 2).  The high thoracic (represented as 1) and low lumbosacral (represented as 3) curves are convex to the left with the apexes around T1 to T4 and L3 to S1 respectively.  (b) Posterior view of a 4-curve scoliosis pattern, in this case , the thoracic (represented as 2) and lumbosacral (represented as 4) curves are convex to the right with the apexes around T7 to T9 and L2 or L3 respectively.  The high thoracic (represented as 1) and low lumbar (represented as 3) curves are convex to the left with the apexes around T1 to T4 and L3 to S1 respectively. 

2.5.3  CLASSIFICATION OF KING

The classification of King, reported by King et al., (1983) distinguishes five different curve patterns or types (King type I through King type V) in a series of 405 patients, miscellaneous types were presented in four of 405 patients (Table 2.3).  The King type is based on the clinical appearance and radiographic evaluation (Winter and Denis, 1994).  The King type pattern is mainly classified by the characteristics of the coronal plane curve in reference to the midline.  When using the classification of King, the midline refers a vertical line taken from the midline of the vertebral column, in which S1 or L5 is utilised at the centre of the pelvis.

 Table 2.3  King type I through King type V (King et al., 1983).

2.5.4  CLASSIFICATION OF CURVE RELATIONSHIP

The classification of Cobb uses the term major and minor to classify the main and secondary scoliotic curves respectively by their location and flexibility (structural) in the vertebral column.  However, the classification of Chêneau simplifies the scoliosis patterns into two groups by identifying the scoliosis as having one or two primary curves. The primary curve or curves always have two compensatory curves or half

curves, one is located above and the other below the primary curve or curves.  Therefore a pattern with one primary curve is a 3-curve scoliosis and a pattern with two primary curves is a 4-curve scoliosis.  Whereas the classification of King has classified scoliosis into five patterns or King types by the characteristics of the curve in reference to the midline of S1 or L5.  In addition to the reference of the midline of S1 or L5, King type I and King type II use the size of the lumbar and thoracic Cobb angles for classification.