Introduction
Research into the causation of adolescent idiopathic scoliosis (AIS) draws heavily from mechanical and biological disciples, but still lacks an agreed theory of etiopathogenesis [
3‐
10]. Genetic factors are believed to play an important role in the etiology of AIS with considerable heterogeneity [
5,
6]. The research problem is complicated by the suspicion that AIS may result not from one cause, but several that interact [
3,
9]. Genetic, and now genomic, research on AIS have not yet provided the therapeutically-required etiologic understanding. In other diseases and particularly diseases of developmental origin [
11] and late-onset chronic non-communicable diseases (NCDs) [
12], research on the role of environmental factors and
epigenetics after a slow start has exploded in the last decade [
12‐
17]. Not so for AIS research and the environment where, except for monozygotic twin studies, there are only sporadic reports suggesting that environmental factors are at work in etiology [
18]. Epigenetics, a relatively recent field, evaluates factors concerned with gene expression in relation to environment, disease, normal development and aging, with a complex regulation across the genome during the first decade of life. Elsewhere we commented on etiopathogenetic concepts as they may relate to normal spine development and AIS pathogenesis [
18]. Here we consider:
(1)
some theories of pathogenesis including recent multiple concepts;
(2)
blood tests for AIS based on predictive biomarkers and genetic variants that signify disease risk;
(3)
epigenetic concepts as they may relate to human development and life history phases;
(4)
AIS linked to the aging process as a non-communicable disease (NCD);
(5)
findings from epigenetic methods applied to the Silver-Russell syndrome and adiposity; and
(6)
suggestions for applying epigenetic methods to AIS etiopathogenesis in cross-sectional and longitudinal studies.
Pathogenesis - theories, hypotheses and concepts 2012
Several theories, hypotheses and concepts for AIS pathogenesis more recently implicating AIS as a systemic and/or multifactorial disorder, are providing hypotheses to test. [
3‐
5,
7]. These include:
-
Relative anterior spinal overgrowth (RASO) [
19].
Sagittal spinal profile in fathers and mothers of AIS girls. In Utrecht, The Netherlands, Janssen
et al.[
20,
21] tested the hypothesis that the familial trend of AIS is explained by the inheritance of the sagittal spinal profile, making the spine less resistant to rotatory forces. Using free-standing lateral spinal radiographs of 51 parent couples of girls with severe progressive AIS and 102 age-matched controls, they found that the fathers, but not the mothers had significantly flatter sagittal spinal profiles.
Forceful restoration of the thoracolumbar lordosis to correct double major AIS. In Enschede, The Netherlands, Van Loon [
22] reviewed factors in healthy children that can deform their spines.
The use of growth in thoracolumbar lordotic intervention for the brace treatment of AIS. In Enschede, The Netherlands van Loon
et al.[
23] evaluated old and new concepts on the pathogenesis of AIS with respect to thoracolumbar lordotic intervention for brace treatment.
Pelvic incidence and lordosis (anterior pelvic tilt). In Utrecht, The Netherlands, Janssen
et al.[
24] defined pelvic lordosis as the angle between the axis of the ischium and a line connecting the midpoint of the sacral endplate to the hip axis. Pelvic lordosis correlated strongly with pelvic incidence. Its role in spinal pathology deserves further investigation.
-
Asynchronous spinal neuro-osseous growth [
25,
26].
Typical and atypical AIS. In St Petersburg, Russia, Dudin and Pinchuk [
27] presented a theory of AIS pathogenesis. Two forms of AIS are identified: typical AIS with convex-side axial vertebral rotation; and atypical scoliosis with concave-side axial vertebral rotation. The components of the theory are:
1)
Growth under nervous and endocrine system control in pre-pubertal and pubertal periods differing in girls and boys, with “personal” programs in spinal cord and its bone-ligament-muscular sheath.
2)
Non-conjugacy of growth (growth conflict) between spinal cord and its bone sheath – termed by others asynchronous, or uncoupled, neuro-osseous growth. This non-conjugacy of growth with –
a)
excess length of vertebral column leads to convex-side vertebral rotation (typical AIS), and
b)
decreased length of vertebral column leads to concave-side vertebral rotation (atypical AIS).
3)
Formation of a scoliosis curve(s) involving Hueter-Volkmann effect and a vicious circle (Stokes),
-
Thoracospinal concept [
28].
-
Dorsal shear forces and axial rotation instability [
29].
Pre-existing axial rotation of the normal spine. In an Invited Lecture, Castelein [
30] in Utrecht, The Netherlands, reviewed their analysis of the rotational patterns of the normal growing and adult spine in relation to closure of the neurocentral junctions, organ anatomy, and the convexity of the curve in idiopathic scoliosis.
Neurocentral junction (NCJ) in the normal growing spine. In Utrecht, The Netherlands, Schlosser
et al.[
31] evaluated the closure pattern and symmetry of left and right NCJs in the normal human growing spine in relation to pre-existing axial spinal rotation and the convexity of the curve in AIS.
-
Flexural-torsional buckling from flexibility anisotropy [
32].
-
Biomechanical spinal growth modulation [
33].
Effects of carrying weight on posture. In Middlesborough, UK, Bettany-Saltikov and Cole [
34] evaluated the effects of carrying weights (front packs, shoulder bags and hand held bags, each about 15% of body weight) on back shape and posture using an ISIS2 scanner on 25 university students. The shoulder and hand held bags produced postural deviations in all planes; these may cause stress and strain on the spine and lead to scoliosis curve progression.
Lumbo-sacral-joint efforts during gait and Hueter-Vollkmann effect. In Montreal, Canada, Brussels, and Louvain, Belgium, Raison
et al.[
35] studied adolescents with left lumbar and thoracolumbar AIS and controls using an acquisition system involving body joint motion via optokinetic sensors and ground reaction forces via a treadmill fitted with force sensors. At L5-S1 subjects with severe AIS had higher medio-lateral forces than the controls which could lead to asymmetrical vertebral growth modulation and curve progression (Hueter-Volkmann effect).
Growth-plate mechanobiology. In Montreal, Canada, Menard
et al.[
36] studying rat caudal vertebrae found dynamic loadings modulated growth with less damage to growth plates than static loading.
Effect of torque on growth in caudal vertebrae. In Milwaukee, USA, Rizza
et al.[
37] using the rat tail, applied torsional loads to vertebrae that led to curvature in growth-plate morphology and remarkably increased growth-plate thickness.
Osteoblasts, biomechanical stress and oestrogens. In Montreal, Canada, Moldovan
et al.[
38] in AIS subjects demonstrated that biomechanical stress and estradiol are involved in the expression of certain genes. Cultured osteoblasts subjected to biomechanical stress showed increased levels of NO, COX-2, OPN, and ATP levels in both control cells and AIS cells with significantly higher levels of NO and COX-2 in AIS cells.
-
Biomechanical theory [
39].
Causative role of gait and standing at ease on right leg. In Lublin, Poland, Karski [
40] outlined his theory that idiopathic scoliosis is connected with gait and persistent standing at ease on the right leg. Scoliosis, results from asymmetry of function proving the possibility of causative prophylaxis.
-
Intervertebral disc disorder [
41‐
47].
-
Deforming three joint complex hypothesis [
48].
-
Motor control disorder [
49‐
51].
Hand grip strength. In Hong Kong, China, Yu
et al.[
52] fond that hand grip strength was lower than controls suggesting muscle dysfunction in AIS.
-
Sensorimotor integration disorder & dystonia [
51,
53].
-
Sensory integration disorder [
54].
-
Vestibular disorder [
55].
Semicircular canals. In Hong Kong, China, Chu
et al.[
56] used MRI to evaluate the morphology of semicircular canals (SSCs) in subjects with right thoracic AIS and controls. Significant differences in the shape of left, but not right, SSCs were found between AIS and controls. It was suggested that these morphological changes are likely to be related to subclinical postural, vestibular and proprioceptive dysfunctions in AIS subjects.
-
Body spatial orientation disorder [
57].
-
Neurodevelopmental disorder [
58].
-
Systemic and metabolic disorders involving -
-
○ Platelet calmodulin [
59,
60].
-
Melatonin receptors. In Hong Kong, China Yim
et al.[
65] extended their previous research which showed that melatonin receptor 1B (MTNR1B) was not detected in osteoblasts of some AIS girls. In 41 AIS girls, while MTNT1A and MTNR1B were found in all, that of MTN1RB was lower, suggesting a quantitative rather than a qualitative difference linked to the pathogenesis of AIS.
-
○ Melatonin-signalling defect (MSD) [
66,
67].
-
○ Osteopontin (OPN) and soluble CD44 (sCD44). Azeddine
et al.[
68] and Moreau
et al.[
69] reported mean plasma OPN levels to be increased in:
-
patients with idiopathic scoliosis, correlating significantly with curve severity, and
-
“an asymptomatic at-risk group” (offspring born from at least one scoliotic parent); this finding, if confirmed, suggests predictive biomarkers and possibly a prodromal stage with the prospect of intervention early in deformity evolution.
In contrast, mean plasma levels of sCD44 were significantly lower in patients with Cobb angles of 45 degrees or more. Drawing on evidence from mouse models, it was concluded that OPN is essential to induce scoliosis formation and curve progression through interactions with CD44 receptors,
“….thus offering a first molecular concept to explain the pathomechanism leading to the asymmetrical growth of the spine in idiopathic scoliosis.”[
69].
Moreau and colleagues report that blood tests could be useful markers for the diagnosis of idiopathic scoliosis and the prognosis of curve progression: a functional scoliosis test [
66‐
70], further refined recently using a more accurate technology called cellular dielectric spectroscopy [
71]; and a biochemical scoliosis test using raised plasma OPN and lower sCD44 values. Moreau [
72] states that OPN and sCD44 are not disease-specific but when observations of both are combined they become highly specific for idiopathic scoliosis. By binding free OPN, sCD44 can prevent OPN from triggering scoliosis or curve progression. Moreau considers that environmental factors could potentially affect the circulating levels of OPN in humans. With colleagues he is conducting tests to identify potentially useful therapeutic agents [
72].
High osteopontin (OPN).l levels and bone mineral density. In China and Montreal, Sun
et al.[
73] in 45 AIS girls demonstrated that low cortical bone mineral density in the distal radius is significantly associated with high OPN levels. It is stated that high osteopontin levels in plasma may reflect underlying abnormalities of bone mineralization in AIS subjects.
PTPx and HSJ family members as disease-modifying factors in AIS. In Montreal, Canada, Elbakry
et al.[
74] stated that their previous research had found that AIS patients have a Gi protein signalling defect and high levels of circulating osteopontin (OPN). Here, on AIS and control osteoblasts, they investigated Protein Tyrosine Phosphatase x (PTPx) and HSJ-1 related family members to determine their potential contribution to OPN receptor activity. Bipedal PTPx knock-out mice were also evaluated. It was found that PTPx and HSJ-1 messenger RNA and protein levels were decreased in all 34 AIS patients compared with 17 controls. They concluded that PTPx and some of the HSJ family members have potential roles in AIS etiopathogenesis as disease-modifying factors exacerbating scoliosis development triggered by OPN.
Oestrogen receptor 2 (ESR2) expression in back muscles. In Poznan, Poland, Rusin
et al.[
77] reported asymmetric ESR2 in deep paravertebral muscles more on the convexity than the concavity. With ESR2 convex/concave ratios of 1 or more, the ratio correlated with Cobb angle.
○ Leptin [
7,
76,
78,
79] (see 9.1.3)
Leptin and bone mineral density. In Hong Kong, China, Tam
et al.[
80,
81] in AIS and control girls, evaluated correlations between each of leptin and soluble leptin receptor (sOB-R) and volumetric bone mineral densities (vBMDs) of various bone compartments scanned at the distal radius using high resolution peripheral quantitative computed tomography (HR- pQCT). They stated that serum total leptin has an anabolic effect on BMD and suggested that AIS girls have an abnormal bone metabolic response to serum leptin. It is unclear whether this represents signalling dysregulation, or abnormal leptin bioavailability between bone compartments. (compact and trabecular).
Leptin and cellular dysfunction. In Hong Kong, China, Tam
et al.[
82] reported that AIS girls have lower BMI, free leptin index (leptin/sOB-R), and higher sO-BR than controls. sOB-R expression correlated abnormally with each of fat content, BMI and leptin compared with controls. The authors speculate firstly, that AIS girls have cellular dysfunction that results in abnormal sOB-R expression which in turn may cause abnormal leptin bioavailability; and secondly, that sOB-R may predict curve progression.
High central leptin activity in a mouse scoliosis model. In Nanjing, China, Wu
et al.[
83] reported results which indicated that high central leptin activity might increase the risk of developing a scoliosis in bipedal mice and contribute to scoliosis progression.
Serum ghrelin levels. In France, de Gouzy
et al.[
84] found higher average levels of total serum ghrelin in AIS subjects compared with controls after adjusting for BMI. They suggested that ghrelin participates in the pathophysiology of AIS possibly involving ghrelin-like cell resistance to melatonin hormone.
Neurohormonal regulation. In St Petersburg, Russia, Khaymina
et al.[
85] evaluated blood serum from 120 children age 8–15 years with AIS, Biotesting was performed on male Wistar rats with thoracic spinal cord transection. After 30 minutes and injecting the serum into the lumbar spinal canal, EMG activity of both hind limbs was recorded as a coefficient of movement disorder (CMD). The CMDs revealed scoliosis type by progression that are claimed to provide new opportunities for understanding the pathogenesis and methods of treatment for AIS.
Bone mineralization. In China, Sun
et al.[
89] evaluated cancellous bone from AIS and normal age-matched healthy subjects. Bone mineral density (quantitative microCT) and undecalcified histomorphometry were studied. The findings showed that AIS girls had lower bone mineralization than normals. It is suggested that an abnormality of bone matrix may play an important role in the etiopathogenesis of AIS.
Bone quality. In Hong Kong, Yu
et al.[
90,
91] reported that bone quality in osteopenic AIS girls was uniquely different from that of osteopenic non-AIS controls. Trabecular compartment alterations with osteopenia were only present in AIS girls, a finding that needs etiopathogenic assessment.
-
Developmental instability & symmetry control dysfunction [
92,
93].
-
Intrinsic growth plate asymmetry hypothesis [
94,
95].
-
Multiple pathogenetic processes.
-
○ Double neuro-osseous theory [
7,
96].
-
-
○ Four components [
58,
97,
98].
Is AIS under 20–30 degrees a chaotic dynamical system? In Lyon and Toulon, France, de Mauroy and Ginoux [
99] outlined the relation of AIS to chaos.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
The original “Whither” article was written by GB and presented at IRSSD by PD. JC initiated this review of the selected scoliogenic presentations at the 2012 IRSSD Meeting to which the 2012 SRS Meeting was added by GB. Copies of the Final SRS Program were given to GB by Mrs MJ McMaster and Dr NS Harshaavardhana. The selection of abstracts for inclusion was made by GB and discussed with JC, PD. AM and TG. All authors contributed their professional skills to the inclusions of the text. All authors have read and approved the final manuscript.