Introduction

Different subclinical neurological dysfunction has been reported in adolescent idiopathic scoliosis (AIS), including poor postural control and asymmetric otolith vestibulo-ocular responses when compared with normal controls. The objective of this pilot study is to establish whether abnormal MRI morphoanatomical changes arise in the CNS (brain and vestibular system), among left-thoracic versus right-thoracic AIS when compared with normal adolescent controls, with use of advanced computerised statistical morphometry techniques.

Introduction

In patients with adolescent idiopathic scoliosis (AIS), anomalous extra-spinal left-right skeletal length asymmetries in upper limbs, periapical ribs, and ilia beg the question as to whether these bilateral asymmetries are connected in some way with pathogenesis. The upper arm and iliac length asymmetries correlate significantly with adjacent spinal curve severity respectively in thoracic and lower (thoracolumbar and lumbar) spine. In lower limbs, skeletal length asymmetries and proximo-distal disproportion are unrelated to spinal curve severity. Overall, these observations raise questions about mechanisms that determine skeletal bilateral symmetry of vertebrates in health and disorder, and whether such mechanisms are involved in the cause of this disease. We investigated upper arm length (UAL) asymmetries in two groups of right-handed girls aged 11–18 years, with right thoracic adolescent idiopathic scoliosis (RT-AIS, n=98) from preoperative and screening referrals (mean Cobb angle 45°) and healthy controls (n=240).

Melatonin’s concentration is high in early childhood and declines gradually thereafter. In the elderly serum melatonin levels are very low. Melatonin, the “light of night”, among other functions is involved in human sexual maturation and in osteogenesis.

Hormesis is the response of cells or organisms to an exogenous (eg drug or toxin) or intrinsic factors (eg hormone), where the factor induces stimulatory or beneficial effects at low doses and inhibitory or adverse effects at high doses [bimodal dose-response] or vice versa.

At the age around 10 years, when idiopathic scoliosis may appear, the circulating melatonin level is about 120 pg/ml – positive hormesis for menses – and menarche appears. Melatonin deficiency may result in a delay of the age at menarche and consequently the girl is susceptible to scoliosis. In these terms melatonin could be certainly involved in the scoliosis pathogenesis. Around the age of 45 years when the circulating melatonin levels are about 20 pg/ml – negative hormesis for menses, menopause starts and the woman has an increased risk for osteoporosis and fractures.

It is documented the bone-protecting effect of melatonin in ovariectomized rats which can depend in part on the free radical scavenging properties of melatonin. Additionally, melatonin may impair development of osteopenia associated with senescence by improving non-rapid eye movement sleep and restoring GH secretion. Whether modulation of melatonin blood levels can be used as a novel mode of therapy for scoliosis and augmenting bone mass in diseases deserves to be studied

This paper evaluates severe normal trunk asymmetry (TA) by higher and lower body mass index (BMI) values in 5953 adolescents age 11–17 years (boys 2939, girls 3014) whilst standing forward bending (FB) and sitting FB during screening for scoliosis. TA was measured as angle of trunk inclinations (ATIs) across the back (thoracic, thoracolumbar and lumbar) with abnormality defined as 2 standard deviations or more. The findings for sitting FB position are reported because the readings express TA free from any leg-length inequality. Relatively lower BMIs are associated statistically with

excess of abnormal TAs, and

BMI is known to be linked to puberty timing and energy balance but not to TAs in healthy students. Similar to girls with adolescent idiopathic scoliosis, we suggest that severe TA is caused by a genetically-determined selectively increased hypothalamic sensitivity to leptin with asymmetry as an adverse hormetic response, exacerbated by presumed lower circulating leptin levels associated with relatively lower BMIs. The asymmetry is expressed bilaterally via the sympathetic nervous system to produce left-right asymmetry in ribs and/or vertebrae leading to severe TA when beyond the capacity of postural mechanisms of the somatic nervous system to control the shape distortion of the trunk

The possibility that AIS aetiology involves undetected neuromuscular dysfunction is considered likely by several workers [1,2]. Yet in the extensive neuroscience research of idiopathic scoliosis certain neurodevelopmental concepts have been neglected. These include [3]:

a CNS body schema (“body in the brain”) for posture and movement control generated during development and growth by establishing a long-lasting memory, and

During normal development the CNS has to adapt to the rapidly growing skeleton of adolescence, and in AIS to developing spinal asymmetry from whatever cause. Examination of publications relating to the CNS body schema, parietal lobe and temporo-parietal junction [4,5] led us to a new concept: namely, that a delay in maturation of the CNS body schema during adolescence with an early AIS deformity at a time of rapid spinal growth results in the CNS attempting to balance the deformity in a trunk that is larger than the information on personal space (self) already established in the brain by that time of development. It is postulated that this CNS maturational delay allows scoliosis curve progression to occur – unless the delay is temporary when curve progression would cease. The maturational delay may be primary in the brain or secondary to impaired sensory input from end-organs [6], nerve fibre tracts [2,7,8] or central processing [9,10]. The motor component of the concept could be evaluated using transcranial magnetic stimulation [11].

Conclusion: Any maturational delay of the CNS body schema could impair postural mechanisms in girls and boys with or without early AIS deformity. The “body in the brain” concept adds a particular CNS mechanism (maturational delay) to the neuro-osseous timing of maturation (NOTOM) hypothesis for the pathogenesis of AIS [12,13]. The NOTOM hypothesis states that there are more girls than boys with progressive AIS because of different developmental timing of skeletal maturation and postural maturation between the sexes in adolescence [12,13].

In subjects with lumbar, thoracolumbar or pelvic tilt scoliosis no pattern of structural leg length inequality has been reported [1]. Forty-seven girls of 108 consecutive adolescent patients referred from routine scoliosis school screening during 1996–1999 had lower spinal scoliosis – lumbar (LS) 17, or thoracolumbar (TLS) 30 (mean Cobb angle 16 degrees, range 4–38 degrees, mean age 14.8 years, left curves 25). The controls were 280 normal girls (11–18 years, mean age 13.4 years). Anthropometric measurements were made of total leg lengths (LL), tibiae (TL) and feet (FL) by one observer (RGB) and asymmetries calculated for LL, TL and FL, as absolutes and percentage asymmetries of right/left lengths. There are no detectable changes of absolute asymmetries with age for LL, TL or FL in scoliotic or normal girls. Asymmetries are found in scoliotic girls compared with normals with relative lengthening on the right for each of LL (0.95%) and TL (0.99%) (each p< 0.001), but not FL (0.38%).

Conclusion: The relative lengthenings in the right leg are unrelated statistically to the severity or side of the lower spinal scoliosis; the cause is unknown and may be related to posture – free standing on the right leg [2] – to neuromuscular mechanisms, or to primary skeletal changes in growth plates of femur(s) and tibia(e).

Nachemson [2] suggested that there are more girls than boys with progressive AIS for the following reason. The maturation of postural mechanisms in the nervous system is complete about the same time in boys and girls. Girls enter their skeletal adolescent growth spurt with immature postural mechanisms. So, if they have a predisposition to develop a scoliosis curve, the spine deforms. In contrast, boys enter their adolescent growth spurt with mature postural mechanisms so they are protected from developing a scoliosis curve. We termed Nachemson’s concept the neuro-osseous timing of maturation (NOTOM) hypothesis and used it to propose a possible medical treatment for idiopathic scoliosis by delaying puberty through the pituitary using gonadorelin analogues as in idiopathic precocious puberty [3,4].

The prevalence of scoliosis is reported to be increased in rhythmic gymnasts (RGs) in Bulgaria [5] and in ballet dancers (BDs) in the USA [6]. Both groups exhibit delayed puberty, which, at first sight, nullifies the NOTOM hypothesis for idiopathic scoliosis. There are similarities between scoliotic RGs and BDs that include intensive exercise from a young age, dieting, delayed menarche, increased scoliosis prevalence (RGs 12%, BDs 24%), mild scoliosis curves (10–30 degrees), and presumably generalised joint laxity. Other differences in addition to country of origin and exercises, include certain anthropometric features and importantly in RGs, thoracolumbar and lumbar curves and, in BDs, right thoracic curves. While constitutional and environmental factors may determine the scoliosis, the different curve types in RGs and BDs suggest that the exercise pattern over many years determines which type of scoliosis develops, although not the curve severity.

Conclusion: The view that scoliotic RGs should be included in a group of sports-associated scoliosis separate from idiopathic scoliosis [5] is supported. We suggest that most BDs who develop mild-moderate scoliosis do not have idiopathic scoliosis but a scoliosis related to intensive exercises over many years acting on a particular phenotype and genotype, similar to the sports-associated scoliosis. In this context the delayed puberty of RGs and BDs with scoliosis does not nullify the NOTOM hypothesis. There is a need to focus research on such subjects who have defined constitutional and environmental factors related to their scoliosis.

Objective: Infants introduced to indoor heated swimming pools in the first year of life show an association with progressive adolescent idiopathic scoliosis (AIS). Similarly control children exposed in this way show an association with vertical spinous process asymmetry. A new method of assessment was used on these controls who were standing in an upright position. Overall, our evidence suggests that indoor heated swimming pools contain a risk factor that predisposes some infants to develop spinal asymmetries years later – progressive AIS in a few and off-vertical spinous process asymmetry in the many. What the risk factor may be and its possible portal of entry into the infant’s body are unknown and possibilities are examined. A subsequent new group of control children confirms the association of indoor heated swimming pools and vertical spinal asymmetry.

Risk factors: An irritant gas trichloramine (nitrogen trichloride) has been found to contaminate the air of indoor-chlorinated pools which Bernard et al link to asthma and chronic airway inflammation. Besides the lungs the skin in infants may provide another portal of entry of any chemical risk factors for spinal asymmetries. In connection with a chemical risk factor Nachemson anecdotally noted the development of scoliosis in salmon fry at a fish farm who were exposed to water contaminated after the re-painting of a water regulating dam.

Environmental epigenomics and disease susceptibility: Barker and his colleagues and others have shown that the origins of important chronic diseases of adult life may lie in foetal responses to the intrauterine environment and in infants to early postnatal life. Currently, there are British and US medical research projects to gather information on how human genes and environment interact over the years to cause disease; the British project is called Biobank. Another aspect concerns disease susceptibility by spotting gene variants in people who already have specific diseases. Do the suspected risk factors of indoor-chlorinated pools for spinal asymmetries need to be included in such studies? Is there potential for prevention?

In our earlier study we found 61% of the controls taken swimming in the first year of life had vertical spinous process asymmetry. In the subsequent smaller study the incidence even higher (83%).

Conclusions:

The evidence reported in our earlier paper suggests that infants introduced to indoor heated swimming pools in the first year of life show an association with spinal asymmetries including progressive AIS and in controls vertical spinous process asymmetry.

Patterns of extra-spinal skeletal length asymmetry have been reported for upper limbs [1] and ribcage [2] of patients with upper spine adolescent idiopathic scoliosis. This paper reports a third pattern in the ilia. Seventy of 108 consecutive adolescent patients referred from routine scoliosis school screening during 1996–1999 had lower spine scoliosis – lumbar (LS), thoracolumbar (TLS), or pelvic tilt scoliosis (PTS). Radiologic bi-iliac and hip tilt angles were both measurable in 60 subjects: LS 18, TLS 31, and PTS 11 (girls 44, boys 16, mean age 14.6 years). Cobb angle (CA), apical vertebral rotation (AVR) and apical vertebral translation from the T1-S1 line (AVT) were measured on standing full spine radiographs (mean Cobb angle 14 degrees, range 4–38 degrees, 33 left, 27 right curves). Bi-iliac tilt angle (BITA) and hip tilt angle (HTA) were measured trigonometrically and iliac height asymmetry calculated as BITA minus HTA (corrected BITA=CBITA) and directly as iliac height asymmetry. Iliac height is relatively taller on the concavity of these curves (p< 0.001). CBITA is associated with Cobb angle, AVR and AVT (each p< 0.001).

Conclusion: The relatively taller concave ilium may be 1) real from primary skeletal changes or asymmetric muscle traction on iliac apophyses [3], or 2) apparent from rotation/torsion at the sacro-iliac joint(s).

In schoolchildren screened for scoliosis about 40% have minor, non-progressive, lumbar scolioses secondary to pelvic tilt with leg-length and/or sacral inequality [1] not reported with preoperative thoracic curves [2]. Forty-nine of 108 consecutive adolescent patients referred from routine scoliosis school screening during 1996–1999 had lower spinal scoliosis with measurable radiological sacral alar and hip tilt angles – lumbar scoliosis 18, thoracolumbar scoliosis 31 (girls 41, boys 8, mean Cobb angle 16 degrees, range 4–38 degrees). In standing full spine antero-posterior radiographs measurements were made of Cobb angle and pelvic asymmetries as sacral alar and iliac heights (left minus right). From anthropometric measurements derivatives were calculated as ilio-femoral length (total leg length minus tibial length) and several length asymmetries, namely: ilio-femoral length asymmetry, total leg length inequality and tibial length asymmetry (all left minus right). Ilio-femoral length asymmetry correlates significantly with sacral alar height asymmetry (girls negatively r= − 0.456, p=0.002, boys positively r=0.726 p=0.041) but not iliac height asymmetry (girls p=0.201) from which three types are identified. Total leg length inequality but not tibial length asymmetry in the girls is associated with sacral alar height asymmetry (r= − 0.367 p=0.017 & r=0.039 p=0.807 respectively). Interpretation is complicated by total leg lengths each including some ilium in which there is asymmetry [3]. But lack of association between ilio-femoral length asymmetry and iliac height asymmetry suggests that the femoral component is more important than iliac component in determining the associations between sacral alar height asymmetry and each of ilio-femoral length asymmetry and total leg length inequality.

Conclusions:

Sacral alar height asymmetry and leg length asymmetries. The evidence suggests that sacral alar height asymmetry is not secondary to the leg length inequalities at least in most girls (negative correlations) and is more likely to result from primary skeletal changes in femur(s) and sacrum.

The side distribution of single spinal curves in our school screening referrals for 1988–99 (n=218) suggests that the mechanism(s) determining curve laterality for the upper spine differs from those for the lower spine. We address here the laterality of right thoracic AIS. In the search to understand the aetiology of AIS some workers focus on mechanisms initiated in embryonic life including a disturbance of bilateral symmetry. The normal external bilateral symmetry of the body, highly conserved in vertebrates, results from a default process involving mesodermal somites. The normal internal asymmetry of the heart, major blood vessels, lungs and gut with its glands is also highly conserved among vertebrates. There is recent evidence that vertebrates retain an archaic asymmetric visceral organization in thoracic and abdominal organs (Cooke). In early embryonic life the visceral asymmetry develops from the breaking of the initial bilateral symmetry by a binary asymmetry switch producing asymmetric gene expression around the embryonic node and/or in the lateral plate mesoderm. In the mouse this switch occurs during gastrulation by cilia driving a leftward flow of fluid and morphogen(s) at the embryonic node (nodal flow) favouring precursors of heart, great vessels and viscera on the left. Based on the non-random laterality of thoracic AIS curves, we suggest that the binary asymmetry switch – through genetic/environmental factors extending to involve anomalously left-sided mesodermal precursors of vertebrae, ribs and/or muscles (positively or negatively), explains the distribution of right/left thoracic AIS. Some support for this hypothesis is the prevalence of scoliosis curve laterality associated with situs inversus.

Most workers consider that ribcage changes in AIS are secondary to spinal deformity. Others claim that ribs are pathogenic in curve initiation or aggravation. In 117 consecutive patients referred from school screening in 1996–99 and routinely scanned by ultrasound, 24 had thoracic and 33 thoracolumbar scolioses (right 37, left 20; mean age 14.9 years, range 12–18 years, girls 44 postmenarcheal 37, boys 13). On anteroposterior standing radiographs, Cobb angle (CA), apical vertebral rotation (AVR, Perdriolle) and apical vertebral translation (AVT from the T1-S1 line) were measured (mean & range: CA 19°, 6–42°; AVR 15°, 0–39°; AVT 17 mm, 0–38 mm). Real-time ultrasound in the prone position recorded laminal rotation (LR) and rib rotation (RR) segmentally and the spine-rib rotation difference (SRRD) as LR minus RR to estimate the combined rib deformity in the transverse plane using for thoracic curves apical LR and RR and for thoracolumbar curves T12 LR and T12 RR (mean LR 8.3°, RR 3.8°, SRRD 5.2° absolute). All deformity parameters, radiological and ultrasound, are unrelated to age. SRRD correlates significantly with each of AVR (r=0.753 p< 0.0001), Cobb angle (r=0.738 p< 0.0001), and AVT (r=0.725 p< 0.0001). Partial correlation analysis shows AVR rather than AVT is associated with the transverse plane rib deformity (SRRD/AVR controlling for AVT r=0.386 p=0.004; SRRD/AVT controlling for AVR r=0.257 p=0.058; SRRD/CA controlling for AVR r=0.260 p=0.055 and for AVT r=0.223 p=0.101). These and other findings suggest that rib rotation in thoracic curves is associated with AVR and AVT and in thoracolumbar curves more with AVR than AVT each within the 4^th column of the spine.

Several workers consider that the aetiology of adolescent idiopathic scoliosis (AIS) involves undetected neu-romuscular dysfunction. During normal development the central nervous system (CNS) has to adapt to the rapidly growing skeleton of adolescence, and in AIS also to developing spinal asymmetry from whatever cause. A new etiologic concept is proposed after examining the following evidence:

anomalous extra-spinal left-right skeletal length asymmetries of upper arms, ribs, ilia and lower limbs suggesting that asymmetries may also involve vertebral body and costal growth plates;

The central of four requirements is maturational delay of the CNS body schema relative to skeletal maturation during the adolescent growth spurt that disturbs the normal neuro-osseous timing of maturation. With the development of an early AIS deformity at a time of rapid spinal growth the association of CNS maturational delay results in postural mechanisms failing to balance a lateral spinal deformity in an upright moving trunk that is larger than the information on personal space (self) established in the brain by that time of development. It is postulated that CNS maturational delay allows scoliosis curve progression to occur – unless the delay is temporary when curve progression would cease. The concept brings together many findings relating AIS to the nervous and musculoskeletal systems and suggests brain morphometric studies in subjects with progressive AIS.

Nachemson [2] drawing upon the theses of Sahlstrand [3] and Lidström [4] articulated the view there are more girls than boys with progressive AIS for the following reason. The maturation of postural mechanisms in the nervous system is complete about the same time in boys and girls. Girls enter their skeletal adolescent growth spurt with immature postural mechanisms – so if they have a predisposition to develop a scoliosis curve, the spine deforms. In contrast, boys enter their adolescent growth spurt with mature postural mechanisms so they are protected from developing a scoliosis curve. We term Nachemson’s concept the neuro-osseous timing of maturation (NOTOM) hypothesis [1,5] The earlier sexual and skeletal maturation of girls may have an evolutionary basis through natural selection. Curve progression in AIS is associated with acceleration of the adolescent growth spurt [6]. Postural sway involves proprioceptive, vestibular and visual input to the central nervous system. In normal children there is a significant reduction in postural sway amplitude between six to nine years and 10–14 years [7,8]. In 1071 normal children aged 6–14 years postural sway is more stable in girls from 6–9 years and over 10 years there is no sex effect [9]; all these findings fit the Nachemson concept. But in view of a subsequent report on 64 normal children aged 3–17 years showing the change with age is limited to boys [10] the age and sex effect of postural sway in healthy children needs further evaluation. In AIS children stabilometry findings are conflicting and observed greater postural sway may be secondary to the curve. In the siblings of scoliotics Lidström et al [11] concluded that postural aberration is a factor in the aetiology of AIS.

Conclusion: The NOTOM hypothesis suggests a treatment to prevent progression of late-juvenile idiopathic scoliosis, early-AIS, and some secondary scolioses. It is based on delaying the onset of the adolescent growth spurt and puberty as used therapeutically in children with idiopathic precocious puberty (IPP)[12]. The proposal is to administer a gonadorelin analogue which in the pituitary down-regulates receptors to hypothalamic gonadotropin-releasing hormone (GnRH) causing a fall in both luteinizing hormone (LH) and follicle-stimulating stimulating hormone (FSH); in turn this causes a fall in oestrogens and androgens and thereby delays or stops menarche and slows bone growth – as in girls and boys with IPP [13]. Expert paediatric opinion is supportive. King [14] has suggested the use of a gonadorelin analogue (Lupron) to delay the onset of the adolescent growth spurt in progressive AIS.

In idiopathic scoliosis the detection of extra-spinal left-right skeletal length asymmetries in the upper limbs, ribs, ilia and lower limbs [1–7] begs the question: are these asymmetries unconnected with the pathogenesis, or are they an indicator of what may also be happening in immature vertebrae of the spine? The vertebrate body plan has mirror-image bilateral symmetries (mirror symmetrical, homologous morphologies) that are highly conserved culminating in the adult form [8]. The normal human body can be viewed as containing paired skeletal structures in the axial and appendicular skeleton as a) separate left and right paired forms (e.g. long limb bones, ribs, ilia), and b) united in paired forms (e.g. vertebrae, skull, mandible). Each of these separate and united pairs are mirror-image forms – enantiomorphs. In idiopathic scoliosis, genetic and epigenetic (environmental) mechanisms [9–11] may disturb the symmetry control of enantiomorphic immature bones [12–13] and, by creating left-right endochondral growth asymmetries, cause the extra-spinal bone length asymmetries, and within one or more vertebrae create growth conflict with distortion as deformities (= unsynchronised bone growth concept) [14].

Conclusion: This enantiomorphic disorder concept applied to the axial skeleton during infancy, juvenility and adolescence – through reductionism into the molecular mechanisms of growth plate responses to different hormones at successive phases of development – provides a new theoretical insight to explain the whole body deformity of AIS. The concept suggests preventive surgery on spine and ribs.

Left-right skeletal length asymmetries in upper limbs related to curve side have been detected with adolescent thoracic idiopathic scoliosis (AIS). In school screening referrals with thoracic scoliosis we find apical vertebral rotation (AVR, Perdriolle) is associated significantly with upper arm length asymmetry. Sixty-nine of 218 consecutive adolescent patients referred routinely during 1988–1999 had idiopathic thoracic scoliosis of whom 61 had left and right upper arm lengths measured with a Holtain anthropometer (right curves 49, left curves 12, mean age 14.9 years, girls 38 postmenarcheal 34, boys 23). The controls are 278 normal girls and 281 boys (11–18 years, mean age 13.5 years). The mean value for Cobb angle is 18 degrees (range 4–42 degrees), AVR 13 (range 0–34 degrees), Cobb angle (CA) and AVR are each positively associated with upper arm length asymmetry (p=0.001 & p< 0.0001 respectively) and after correcting for each of Cobb side, apical level, sex and handedness, AVR and upper arm length asymmetry are still significantly associated (p=0.004 ANOVA). Partial correlation analysis shows AVR is associated with upper arm length asymmetry after controlling for CA (p=0.033); but not CA and upper arm length asymmetry after controlling for AVR (p=0.595). The reason why a larger AVR to the right is associated with a relatively longer right upper arm is unknown. Possibilities include neuromuscular and skeletal mechanisms, the latter relative concave overgrowth of neurocentral synchondrosis and/or of periapical ribs. We suggest consideration be given to combining convex vertebral body stapling (Betz) with concave periapical rib resection (Sevastik and Xiong) for right thoracic AIS in girls.

Background: In lumbar scoliosis curves of school screening referrals were evaluated (1) for the possible relation of pathomechanisms to standard and non-standard vertebral rotation (NSVR) [1], and (2) the relation between apical lumbar axial vertebral rotation and the frontal plane spinal offset angle (FPTA) [2].

Methods: Consecutive patients referred to hospital during routine school screening using the Scoliometer were examined in 1996–9. None had surgery for their scoliosis. There are 40 subjects with either pelvic tilt scoliosis (11), idiopathic lumbar scoliosis (19), or double curves (10)(girls 31, postmenarcheal 25, boys 9, mean age 15.3 years). One observer (RGB) measured: 1) in AP spinal radiographs Cobb angles (CAs), apical vertebral rotations (Perdriolle AVRs), and trigonometrically sacral alar tilt angle (SATA), and FPTA as the tilt of the T1–S1 line to the vertical; and 2) total leg lengths (tape).

Results: Excluding the double curves there are 16 left and 14 right lumbar curves mean CA 11 degrees (range 4–24 degrees), mean AVR 9 degrees (concordant to CA in 18/30, discordant in 7/30), SATA 2.8 degrees (range 0.2–7.7 degrees associated with CA side and severity, p=0.0003), and leg-length inequality 0.7 cm (significantly shorter on left, p< 0.0001 and associated with SATA (p=0.02) but not CA). Neither CA nor AVR in each of the laterality concordant and discordant lumbar or thoracic curves is significantly different. Twenty-six subjects have thoracic curves (16 right) 22 with AVR (mean CA 11 degrees, range 4–17 degrees, AVR 9 degrees, n=22) the CA being associated with each of lumbar CA and SATA (respectively p< 0.0001, p=0.003, n=26). Thoracic curve laterality of CA and AVR is concordant in 12/26 curves and discordant in 10/26 and for concordance/discordance neither is significantly different; thoracic AVR sides with laterality of lumbar curve AVR shown by thoracic AVR (but not CA) being greater in lumbar discordant than in lumbar concordant curves (14 & 7 degrees respectively, p=0.03, n=18 & 7). Both for lumbar curves alone and for lumbar with double curves, AVR by side is significantly associated with FPTA by side (r= −0.568, p=0.001, n=30; r=−0.560, p=0.0002, n=40).

Conclusion: (1) It is hypothesized that different pathomechanisms may separately affect the frontal (CA) and transverse (AVR) planes: in discordant curves these mechanisms may neutralize each other and limit curve progression; concordant curves require these biplanar mechanisms to summate and facilitate curve progression. (2) The association of frontal plane spinal tilt angle and lumbar AVR may result from balance mechanisms affecting trunk muscles – mechanisms that may underlie the complication of post-operative frontal plane spinal imbalance or decompensation [2].

Background: The study evaluates the inter-relationships between Cobb angle (CA), apical vertebral rotation (AVR), apical vertebral translation (AVT) and maximal angle of trunk inclination (max ATI). The effects of sex, curve laterality, curve type and apical levels will be studied

Methods: This is a study of consecutive pre-operative AIS patients. There are 122 pre-operative AIS patients (106 thoracic, 16 thoracolumbar), with a mean age of 15.6 years. From the pre-operative AP radiograph, CA, AVR (Perdriolle) and AVT are measured. The max ATI is measured using the Scoliometer with the patient in a standing forward bending position. Ratios between the measurements are calculated to allow comparison between different curve types and curves at different apical levels.

Results: For a given Cobb angle, each of AVR, AVT and max ATI are largest in King type IV curves, less in King type III curves and smallest in King type II curves (p=0.001 to 0.015). For curves without a significant compensatory curve, for a given AVR, the max ATI reduces significantly as the curve apex passes caudally (p=0.002 to 0.019). Sex and curve laterality are not significant factors.

Conclusion: It is suggested that as a curve develops, the interaction between the measurements in different planes may be responsible for determining the curve type (presence or absence of a compensatory curve). The smaller surface hump as the curve apex passes caudally is probably due to the transition from fixed ribs to floating ribs to no ribs. These finding also have implications for surgery. In King type IV and III curves, the emphasis should be on correcting translation and derotation perhaps with a primary costoplasty whilst in King type II curves, the emphasis should be on the correct selection of fusion levels and achieving a balanced spine.

Background: In preoperative thoracic (TC) and thoracolumbar (TLC) AIS curves to evaluate periapical rib-vertebra angle asymmetry [1] and rib-spinal angle asymmetry in relation to the spinal deformity and the 4th column support of the spine [2].

Methods: Consecutive preoperative AIS patients having spinal instrumentation and fusion were assessed using radiographs and ultrasonographs. Twenty-eight preoperative patients with AIS were studied (TC 19, apex T8-9 in 15, TLC 9, apex T12 in 2, L1 in 7, mean Cobb angle 51 degrees). In AP radiographs the following were measured by one observer (RGB): Cobb angle (CA), apical vertebral rotation (AVR) and apical vertebral translation (AVT) from the T1-S1 line; in TC at 6 levels about the apical vertebra (3 above, at and 2 below) for each of 1) rib-vertebral angles (RVAs) and difference (RVAD=concave minus convex RVA), 2) rib-spinal angles (RSAs) to the T1-S1 line and difference (RSAD), and 3) vertebral tilt; and in TLC the RVAs, RVADs, RSAs and RSADs of ribs 11 & 12. The ultrasound apical spine-rib rotation difference (SRRD) was obtained as a measure of transverse plane rib deformity. With the subject in a prone position and head supported, readings of laminal rotation (LR) and rib rotation (RR) were made on the back at 12 levels by one of two observers (RKA, ASK) using an Aloka SSD 500 portable ultrasound machine with a veterinary long (172mm) 3.5 MHz linear array transducer. The maximal difference between LR and RR about the curve apex was calculated as the apical spine-minus-rib rotation difference (SRRD).

Results: Thoracic curves. The RVADs (but not the RVAs, RSAs or RSADs) only at 2 & 3 levels above the apex correlate significantly with each of CA (p=0.054), AVR (p=0.047), AVT (p=0.014, after controlling for CA p=0.131) and vertebral tilt (p=0.032) but not SRRD (all two levels above apex). Thoracolumbar curves. The 11th RSAD (but not RVAD or RSAs) correlates significantly with each of AVR (r= −0.776, p=0.014, after controlling for CA p=0.022) and SRRD (r= −0.890, p=0.001, after controlling for CA p=0.003) that together correlate significantly (r=0.672, p=0.048).

Conclusion: In TC supra-apical rib asymmetry (RVAD) in sternally-stabilized [2] and longest levers of the sternal-rib complex is associated with spinal deformity; in TLC supra-apical rib asymmetry (11th RSAD) is associated with transverse plane deformity of each of the apical vertebra (mainly L1) and 12th ribs. These rib associations, probably secondary to the spinal deformity, may involve a primary rib component in the 4th spinal column. The prognostic value of supra-apical RVAD and RSAD for progressive AIS needs to be evaluated.

Objective: To assess the radiological and back surface correction achieved following anterior USS in the treatment of thoracic adolescent idiopathic scoliosis (AIS).

Design: Prospective study of back surface correction, retrospective radiological review.

Subjects: 14 patients with thoracic AIS (age 11–18 yrs) were treated with anterior USS between 1995 and 2000. There are 12 females and 2 males, all with 2 year follow-up. 8 patients have complete surface data. Data from a further 6 patients will shortly be available as they reach 2 year follow-up.

Outcome measures: Cobb angle, apical vertebral rotation (AVR), apical vertebral translation (AVT), frontal plane imbalance, kyphosis and lordosis were measured from the radiographs. A Scoliometer was used to assess the maximal angle of trunk inclination (max ATI) in the thoracic region. All measurements were obtained before surgery and at 8 weeks, 1 year and 2 years after surgery. Complications were recorded.

Results: Significant initial corrections are observed for each of: Cobb angle (51%, p< 0.001), AVR (40%, p=0.003),AVT (64%,p< 0.001),maxATI (47%,p=0.001). There is no significant correction loss during the 2 year follow-up. Three patients had spinal imbalance (> 2cm) before surgery with one patient after surgery. The kyphosis significantly increased from 24° to 29° immediately after surgery with no significant change during follow-up. There was no change in lordosis. There were no neurological complications and no instrumentation failures were observed. In two cases the upper screw partially pulled out of T5 with some loss of correction.

Conclusions: Anterior scoliosis correction for thoracic AIS achieves good and stable radiological and particularly back surface corrections (max ATI – 47% compared with 22% correction after posterior surgery). Rigid anterior instrumentation has eliminated the 20% rod failure seen with Zielke. New techniques for preventing upper screw pull out will be discussed and new retractor systems allow smaller thoracotomies. There remains a small but significant increase in kyphosis which is less of a problem in the thoracic spine than at the thoracolumbar junction where anterior scoliosis correction is most commonly advocated.

Anterior instrumentation for thoracic AIS has advanced to a point where it can be widely adopted, particularly if the patient expresses concerns regarding the rib hump or is hypokyphotic.