Background
Adolescent idiopathic scoliosis (AIS) is a four dimensional deformity of the spine
arising in otherwise healthy children during puberty. The fourth dimension is time.
This dimension is the characteristic of our database created in 1998 with systematic
reconvening of our patients at regular intervals which increases the level of scientific
evidence [1]. The use of a brace in the conservative treatment for AIS plays an important
role and has the aim to stop the evolution of the deformity in immature adolescents
in order to prevent problems during adulthood [2, 3]. Long-term follow-ups indicate
that patients with scoliosis may have a higher prevalence of back pain and of worsening
pulmonary function if the curve becomes extremely severe [4]. A randomized control
trial BRAIST study conducted by Weinstein showed that bracing is significantly effective
in reducing the progression of AIS [5]. Previously, a Cochrane review [6] also demonstrated
the effectiveness of bracing in the treatment of AIS.
To measure the effectiveness of a brace two main factors can be involved: 1. the immediate
in-brace reduction depending how to get the three-dimensional correction and its reproducibility;
2. the patient’s adherence which depends on aesthetics and tolerance [7, 8]. Different
types of braces are used in the treatment of AIS but almost all are created on the
multiple three points system principle of applying external corrective forces across
the curve in order to stop deformity progression, produce an acceptable sagittal and
coronal contour, and delay or avoid surgical treatment [9–12]. The main biomechanical
concepts are based on: elongation along the vertical axis, lateral inflexion in the
frontal plane and derotation of the spine in order to obtain a correction of the scoliotic
curve. Derotation is the main movement along the vertical axis. The correction in
the sagittal plane is problematic because many scoliosis are accompanied by a change
in the sagittal plane with a flat back in half of the cases. All of the above mechanisms
are going in the direction of accentuation of the flat back and require significant
and uncertain changes during the manufacture of the brace. This problem has now been
finally solved thanks to segmental moulding.
Elongation
Historically, in the early twentieth century, in the United States, Sayre [13] was
the first to make a plaster cast in a standing posture using this biomechanical concept,
even if the first modern brace can be considered the Milwaukee brace, created in 1940
by Blount, which was a brace based on axial elongation between the pelvis and the
cervical collar.
In France, the Lyon brace, created in 1947 by Pierre Stagnara, was the first 3D adjustable
contention brace used after a plaster cast. With the Lyon brace, elongation occurs
between the pelvic and shoulder girdle with equal distribution of forces on the right
and on the left. The elongation requires precise adjustment of the brace during the
growth of the child [14]. Other TLSO braces introduce a new concept described by Chêneau
as the “cherry stone effect” with stretching upwards between pelvis and rib cage.
The existence of windows in the brace do not affect elongation. In contrast, with
the new Lyon brace, axial elongation type “mayonnaise tube” is achieved by the simultaneous
clamp of the two hemi polycarbonate pieces and requires the integrity of the outer
tube wall (Fig. 1).
Fig. 1
Evolution of elongation concepts along the vertical axis. At the time of the first
Sayre’s plaster cast, cervical suspension and body weight realize a bipolar overall
elongation. With the Milwaukee brace, the elongation is internal between pelvic girdle
and cervical collar. With the Lyon brace, elongation occurs between the pelvic and
shoulder girdle with equal distribution of forces on the right and on the left. The
elongation requires precise adjustment of the brace during the growth of the child.
Other TLSO braces introduce a new concept described by Chêneau as “cherry stone effect”
with stretching upwards between pelvis and rib cage. The existence of windows in the
brace do not affect elongation. In contrast, with the new Lyon brace, axial elongation
type “mayonnaise tube” is achieved by the simultaneous clamp of the two hemi polycarbonate
pieces and requires the integrity of the external tube wall
Derotation and Detorsion or Untwisting
The segmental derotation is difficult to achieve because it is done through the ribs
and could lead to an increase in a flat back. It is impossible to achieve derotation
when the rib hump is angular. The mathematical basis of the twisted column is the
circled helicoid with horizontal generating circle. The overall untwisting occurs
between the axillary and pelvic clamps and the thoracolumbar horizontal plane (Fig. 2).
Fig. 2
From segmental derotation to global detorsion or untwisting. The segmental derotation
is difficult to achieve because it is done via the ribs and could increase flat back.
It is impossible to obtain when the rib hump is angular. The mathematical basis of
the twisted column is the circled helicoid with horizontal generating circle. The
overall untwisting occurs between the axillary and pelvic clamps and the thoracolumbar
horizontal plane
Cotrel added a fundamental component: the flexion in the frontal plane [14, 15]. The
acronym ‘EDF’ stands for: Elongation, Derotation, Flexion. He created a framework
for three-dimensional scoliosis correction in the supine position with spine untwisting.
At the end of plaster cast weaning, the plaster mould to build the Lyon brace reproduces
the overcorrection obtained [14, 16].
Many previous studies support the positive results associated with the casting and
Lyon braces [14, 16, 17] but the difficulty and cost of making the plaster cast, administrative
economical questions and low compliance, can also explain the reasons which ultimately
have prompted the development of new design concepts with immediate in-brace correction.
It was only in 2013 that advances in computer technology with the latest generation
software (OrtenShape) allowed the superposition of different CAD/CAM moulds and a
segmental 3D reconstruction [18, 19]. The aim was to use this new software to replace
the plaster cast with a new Lyon brace: the ARTbrace. Segmental moulding is one of
the fundamental innovations of the ART brace. The overcorrection is performed in the
frontal plane and the sagittal plane precisely and individually for each child at
three levels: pelvis, lumbar spine and thoracic spine. The detorsion is obtained by
untwisting coupled movements. The Chêneau brace is also a night and day overcorrecting
brace, but the overcorrection is only made by the CPO. The ARTbrace is a custom night
and day overcorrecting brace. It is the patient himself who will determine the overcorrection
(Fig. 3).
Fig. 3
Segmental moulding. Segmental molding is one of the fundamental innovations of ART
brace. The overcorrection is performed in the frontal plane and the sagittal plane
precisely and individually for each child. The detorsion is obtained by untwisting
coupled movements. Chêneau brace is also a night & day overcorrecting brace, but the
overcorrection is only made by the CPO. ARTbrace is a custom overcorrecting brace.
It is the patient himself who will determine the overcorrection
High Rigidity
Even if the old Lyon brace in polymethacrylate was very rigid, the credit for HIGH
RIGIDITY goes to the Italian team of ISICO with the Sforzesco brace, which has proven
to be effective by avoiding plaster casts for scoliosis over 45° [20]. The acronym
ARTbrace (Asymmetrical Rigid Torsion brace) was created by Stefano Negrini. The merit
of the ARTbrace is the addition of overcorrection to the high rigidity with a global
detorsion. It is this overcorrection for small curvatures which explains the average
improvement of the in-brace correction.
Since May 2013, all patients of JCdM were treated with the ARTbrace instead of a plaster
cast which showed good initial results. Indeed, the first immediate results of the
ARTbrace have demonstrated that the in-brace correction of the Cobb angle in the first
225 cohort of patients was 70 %, a correction which is 40 % higher than with the former
Lyon brace or historical Lyon brace. The value of this correction was even higher
than for other braces published in the literature, including retrospective studies
[21, 22, 8, 23, 24].
Like the historical Lyon brace, the ARTbrace is ADJUSTABLE. Both axillary and pelvic
clamps are adjustable with a precise wrench and a bolt system and an anterior ratcheting
buckle.
Like the historical Lyon brace, the SAGITTAL PLANE is fixed by the posterior bar.
But the sagittal plane is determined by the segmental mould and the superposition
of the mouldings. In additional it is the lack of support at the sterno-clavicular
level and at the abdominal level that avoids lumbar delordosis and thoracic flat back.
In this study, early results of 148 first consecutive scoliosis treated with the ART-brace
after 1-year are reported in correlation with a matched pair control of the last 100
patients treated with the old Lyon Brace.
Material and Methods
Study design
We performed a prospective case series of 148 scoliosis with short time results after
1 year compared with a historical retrospective case series of 100 scoliosis. Consecutive
cases are recruited in both groups. Randomization was not possible due to the administrative
impossibility to perform plaster casts after May 2013. All lumbar scoliosis Lenke
5 were eliminated in the two groups as they continue to be treated by the GTB short
brace [25].
Population
Since May 2013, we treated more than 400 patients at the “Clinique du Parc – Lyon”
with the new Lyon brace (ARTbrace) instead of the classical EDF plaster cast followed
by the historical Lyon brace. The initial aim was to avoid a plaster cast, but very
quickly, the ARTbrace appeared to be a much more effective solution compared to the
former plaster casts and it was even better tolerated. Following the early successes
the whole treatment was continued with the same brace. In this prospective study,
only the first 148 of all patients, 17 % of males and 83 % of females with an average
age of 13.37, with a follow-up at 1 year, have been included. The patients of this
main group presented 35 thoracic primary curves and 28 lumbar or thoraco-lumbar primary
curves and 37 double major curves with a Cobb angle ranging from 20° to 53° (average
29.23° and Standard Deviation: 8.14°). These 148 patients are group A. The second
matched pair control group consisted of a consecutive series of 100 patients (22 %
males 78 % Females) and an average age of 13.6, treated with a plaster cast and the
historical Lyon brace, and controlled 1-year after brace fitting, with 41 thoracic
primary curves, 23 primary lumbar or thoraco-lumbar curves and 36 double major curves
with a Cobb angle ranging from 20° to 52° (average angle Cobb 30.4° and Standard Deviation:
9.61°). These 100 patients are group B.
All treatment parameters like indications, physiotherapy, full or part-time bracing
were identical for both groups, according to the experience of Lyon management [14,
16, 17]. The plaster cast time was replaced by an equivalent time of “full time” ARTbrace.
The study of dropouts is fundamental and we expected a high rate because the realization
of the plaster cast was a barrier that only 2/3 of children were crossing. Lyon bracing
management has always been considered as an elitist treatment. After 1 year the number
of dropouts is 14 (162-148) about 10 %. Some patients referred by colleagues from
other countries and controlled by them are not considered as dropouts.
Method
All patients were evaluated radiologically before treatment (T0), in-brace (T1), at
6 months without brace (T2) and at 1 year without brace (T3) during treatment. Clinical
evaluation at T1 is performed at the end of full time wearing. The clinical parameters
were identical for both groups and consisted of measure rib hump in millimetres, and
Bunnel ATR by Adam’s posture.
The radiological examination of ARTbrace group A was performed with an EOS micro dose
radiological System, an ultra-low dose radiation imaging system that provides simultaneous
AP and Lateral views in the standing position with 25 times less radiation than traditional
X-ray, equivalent to one week of natural Earth radiation [26–28]. The standing frontal
Cobb angle was always measured by the first author. Automatic measurement EOS cannot
be used due to some inversion of curves (Fig. 4).
Fig. 4
Overcorrection of a long lateral thoraco-lumbar curve. When the deviation is accompanied
by very little deformation of the vertebral body, it is possible to completely reverse
the curvature with an overcorrecting brace. Stereos shows the complete translation
mirrored. There is also a realignment of the curves in the sagittal plane of the spine
with recentering on the gravity line. Such result can be expected with many other
asymmetrical braces but there is only one asymmetrical high rigidity brace
An automated management system control facilitates regularity in follow up meetings.
A 3D sterEOS study was carried out every time we had the EOS radiography at T0.
The radiological follow-up of control group B patients was performed without a sagittal
view due to radiation saving habits with traditional radiology and probably also because
the correction in the sagittal plane was not perfect. The problem of radiation in
scoliosis was discussed in the consensus session of the SOSORT 2011 meeting [29].
In fact to avoid excessive radiation, exposition lateral view X-ray was not systematically
executed for most patients. On the contrary, thanks to using ultra-low dose EOS system,
a systematically sagittal analysis of spine was possible for the patients in the main
group A. The sagittal parameters like Sacral Slope (SS), Lumbar Lordosis (LL) and
Thoracic Kyphosis (CT) were automatically measured by EOS system.
All the data are recorded immediately into a database and a serial number is automatically
assigned at T1 about 3 days after bracing. All other statistical tests are done with
the package SPSS v20. The first step is to confirm the normality of distribution (Kolmogorov-Smirnov
& Shapiro-Wilk) and then use an independent-samples T test to compare Cobb angles
T0 (before brace), T1 (in-brace),T2 (at 6 months) and T3 (at 1 year). A p value of
less than 0.05 was considered to be significant. A copy of the Excel database can
be downloaded to allow any comparisons (Additional file 1).
Results
Clinical findings
We present the very short results at 6 months in both groups, demonstrating the superiority
of the new Lyon brace. The main results on Rib hump (RH) and Bunnel’s ATR (Bu) are
shown in Table 1.
Table 1
Average and Standard Deviations of Rib hump and Bunnel ATR before bracing and at 6
months for group A (ARTbrace) and Group B (old Lyon brace)
RH T0
Bu T0
RH T2
Bu T2
T – ARTbrace (A)
23.44 ± 9.4
9.75 ± 4.1
10.33 ± 6.6
5.14 ± 3.3
T - Old Lyon (B)
23.56 ± 8.6
10.55 ± 3.9
16.7 ± 8.5
7.95 ± 4.0
L – ARTbrace (A)
17.21 ± 7.8
7.51 ± 3.6
4.65 ± 4.5
2.06 ± 2.4
L – Old Lyon (B)
16.41 ± 7.4
7.47 ± 3.4
9.41 ± 6.3
4.51 ± 3.2
There was not a significant difference in the score of Thoracic rib hump before brace
for control group with the old Lyon brace (M = 23.56, SD = 8.61) and the ARTbrace
group (M = 23.44, SD = 9.43, t(176) = 0.089, p = 0.929.
There was not a significant difference in the score of Thoracic Bunnel ATR before
brace for control group with the old Lyon brace (M = 10.55, SD = 3.85) and the ARTbrace
group (M = 9.75, SD = 4.10, t(176) = 1.307, p = 0.193.
There was not also a significant difference in the score of Lumbar rib hump before
brace for control group with the old Lyon brace (M = 16.41, SD = 7.36) and the ARTbrace
group (M = 17.21, SD = 7.76, t(154) = -0.612, p = 0.541.
There was not also a significant difference in the score of Lumbar Bunnel ATR before
brace for control group with the old Lyon brace (M = 7.47, SD = 3.378) and the ARTbrace
group (M = 7.51, SD = 3.63, t(154) = -0.072, p = 0.943.
There was a significant difference in the scores for thoracic rib hump and Bunnel
ATR and for lumbar rib hump and Bunnel ATR, at 6 months between the two groups.
Thoracic rib hump: t(176) = 5.651, p = 0.00
Thoracic Bunnel ATR: t(176) = 5.104, p = 0.00
Lumbar rib hump: t(155) = 5.459, p = 0.00
Lumbar Bunnel ATR: t(155) = 5.304, p = 0.00
Group A (ARTbrace)
At the thoracic level the percentage improvement is: 57 % for rib hump and 51 % for
ATR
At the lumbar level the percentage improvement is: 79 % for rib hump and 86 % for
ATR
Group B (Historical Lyon brace)
At the thoracic level the percentage improvement is: 27 % for rib hump and 25 % for
ATR
At the lumbar level the percentage improvement is: 53 % for rib hump and 49 % for
ATR
The percentage improvement between the old and the new Lyon brace is near 30 % for
both rib hump and ATR. It is better for the lumbar area compared with the thoracic
one.
Radiological findings
Frontal correction
The main group A of 148 patients (ARTbrace) had 195 primary curves from 20° to 55°:
63 curve Thoracic, 42 curve lumbar with 45 double major curves. Only primary curves
were selected.
The control group B of 100 patients (Historical Lyon brace) had 136 curves from 20°
to 50°: 41 Thoracic curves 23 lumbar curves with 36 double major curves. Only primary
curves were selected (Table 2).
Table 2
Average and Standard Deviation of Cobb angle at T0 (before bracing), T1 (in-brace),
T2 (6 months), T3 (1 year), for group A (ARTbrace) and Group B (old Lyon brace)
Cobb angle
T0 Initial
T1 In-brace
T2 6 months
T3 1 year
A-Tho ART (n = 108)
30.03 ± 9.6
11.26 ± 8.65
20.25 ± 11
21.47 ± 11
A-Lumb ART (n = 87)
27.83 ± 7.5
6.64 ± 8.8
15.50 ± 9.1
16.40 ± 9.38
B-Tho Hist (n = 76)
31.14 ± 9.6
16.96 ± 9.3
23.96 ± 11
26.95 ± 11.9
B-Lumb Hist (n = 59)
29.69 ± 7.7
12.32 ± 7.9
18.81 ± 9.4
20.41 ± 11
The percentage of improvement was calculated using the following formula: (average
T0 – average T1)/average T0 and so on for T2 and T3 (Table 3).
Table 3
Percentage improvement relative to the initial angle at T1 (in-brace), T2 (6 months),
T3 (1 year), for group A (ARTbrace) and Group B (old Lyon brace)
Percentage improvement
(T0-T1)/T0
(T0-T2)/T0
(T0-T3)/T0
A-Tho ART (n = 108)
62.5 %
32.6 %
28.5 %
A-Lumb ART (n = 87)
76.1 %
44.3 %
41.1 %
B-Tho Hist (n = 76)
45.5 %
23.0 %
13.5 %
B-Lumb Hist (n = 59)
58.5 %
26.5 %
31.2 %
To compare the progression between the two groups the differential was calculated
using the following formula: (percentage A – percentage B)/percentage B for T1, T2
and T3 (Table 4).
Table 4
Differential percentage between group A (ARTbrace) and Group B (old Lyon brace) at
T1 (in-brace), T2 (6 months), T3 (1 year)
Differential Percentage
(%A-%B)/%B at T1
(%A-%B)/%B at T2
(%A-%B)/%B at T3
Differential Tho (n = 184)
0.37
0.417
0.715
Differential Lumb (n = 146)
0.301
0.672
0.317
The results were reported for the thoracic and lumbar curves. We find that the extra
in-brace correction obtained persists at 6 months with even a tendency to improve
after 1 year (Fig. 5).
Fig. 5
Evolution of Cobb’s angle average in the frontal plane during the follow up of 1 year.
Between T0 and T1, we see both for lumbar and thoracic curves a large divergence.
The new Lyon brace is much more corrective than the former plaster cast. This difference
decreases when checking at 6 months (T2), but there is still a significant difference
between the two groups. Between 6 months (T2) and one year (T3), there is again a
divergence between the two curves, especially at the thoracic level, which means that
the corrective effect continues and modifies the final result. A difference of 5°
between the two groups is highly significant
With SPSS we can confirm with two tests: Shapiro-Wilk, and Kolmogorov-Smirnov, that
the data comes from a normal distribution (Additional file 2).
We also use SPSS comparison of means tests to compare the two independent groups and
answer the following questions (Additional file 3).
There was not a significant difference in the score of Thoracic Cobb angles before
brace for control group with the old Lyon brace (M = 31.14, SD = 9.62) and the ARTbrace
group (M = 30.03, SD = 8.30, t(182) = 0.834, p = 0.405.
There was not also a significant difference in the score of Lumbar Cobb before brace
for control group with the old Lyon brace (M = 26.69, SD = 7.72) and the ARTbrace
group (M = 27.82, SD = 7.51, t(144) = -0.884, p = 0.378.
There was a significant difference in the scores for thoracic and lumbar curves: in
brace, at 6 months and after 1 year.
Thoracic in-brace: t(182) = 4.254, p = 0.00
Lumbar in-brace: t(144) = 3.993, p = 0.00
Thoracic at 6 months: t(182) = 2.284, p = 0.023
Lumbar at 6 months: t(144) = 2.131, p = 0.035
Thoracic at 1 year: t(182) = 3.205, p = 0.02
Lumbar at 1 year: t(134) = 2.463, p = 0.015
The distribution of improvements at the 1 year follow-up was first studied with standardized
SRS criteria [30] with a worsening of more than 5°, stability at ± 5° and improvement
of more than 5°. But it seems more pertinent to create a new class of improvement
of 10° and more to improve the readability of the chart (Fig. 6).
Fig. 6
Angular distribution of improvements at 1 year. The one-year results were initially
grouped according to the criteria of the SRS. But he scoliosis group whose angulation
improved by more than 5° was too large with bad readability, so we have created a
4th group of angular correction of 10° or more. The 10° and more group is highly improved
by the new Lyon brace
Sagittal correction
The radiological follow-up of control group patients was performed without lateral
X-ray and therefore it is not possible to make a statistical comparison. But thanks
to the use of the micro dose EOS system a systematic sagittal analysis was possible
for the main group of patients (ARTbrace).
In a previous study, we showed that the average thoracic kyphosis angle with the upper
limit T4 was 37° [31]. For this study, we set the cut off at 30° for hypokyphosis
or flat back.
The results are the subject of a separate presentation we can summarize.
73/148 patients (i.e. 49.4 %) had initial thoracic kyphosis below 30° (m = 19.6°,
SD = 6.77)
In-brace angulation (m = 28.45°, SD = 5.84°) improvement in ARTbrace is 8.84°, significant
(p = 0.000)
50 patients were monitored with sagittal EOS without brace at the 1 year follow-up
(m = 27.3°, SD = 5.40). For this specific group: initial kyphosis (m = 18.58°, SD = 6.63),
in-brace kyphosis (m = 28.06°, SD = 5.45) and last follow-up without brace (m = 27.15°,
SD = 5.45) were analysed (Fig. 7)
Fig. 7
Improved initial flat back in-brace and at 1 year follow-up. Nearly half of scoliosis
have a flat back with thoracic kyphosis angulation of less than 30°. In ARTbrace,
back flat improved significantly of more than 8° and this improvement was maintained
at 1 year follow-up without brace. There is no significant difference between the
in-brace angle of thoracic kyphosis and at 1 year follow-up without brace
The in-brace improvement rate is 50 % and very significant (p = 0.000), and without
brace at the last follow up, the improvement rate is 46 % (p = 0.000). There was no
statistical difference between the in-brace group and 1 year after when not wearing
a brace (p = 0.289). The in-brace improvement is therefore maintained at the 1 year
follow up when the brace is off.
38 % of patients showed an improvement of 10° or more, 36 % of patients showed an
improvement between 5° and 9°, 26 % of cases with stability, no back flat worsening.
Comparing plaster cast and ARTbrace
Treatment was started with the EDF plaster cast and former Lyon brace. The translation
along the vertical axis {Elongation) takes place to the detriment of the lumbar lordosis
and the thoracic kyphosis. In ARTbrace, the sagittal plane is determined by the posterior
metal bar and ahead, expansions at both ends allow active 4D correction (Fig. 8).
Fig. 8
Comparing EDF plaster cast and ARTbrace. In the sagittal plane, the main difference
with the EDF plaster cast and the former Lyon brace is that there are now two expansions
in the sagittal plane. One at the sternoclavicular level for high thoracic kyphosis
and the other at the abdominal level facilitating lumbar lordosis
In this case n° 389, the improvement in the sagittal plane is important and harmonious.
The spine follows the curve imposed by the posterior metallic bar (Fig. 9).
Fig. 9
Case n° 389. Before bracing, there is an overall decrease in sagittal curvatures.
In ARTbrace, it is possible to restore physiological sagittal curvatures
Horizontal plane in-brace correction
Thanks to the use of the EOS system and reconstruction in 3D, in 15 characteristic
cases, the effect of the ARTbrace in the horizontal plane could be studied. SterEOS
gives us the position of each vertebra in the horizontal plane thanks to an upper
view. The first visible effect of the brace is the translation of the spine towards
the vertical axis due to the “mayonnaise tube” effect characteristic of high rigidity
braces. The segmental rotation is automatically calculated and we define a global
torsion index which is the average of all 17 segmental rotations (Fig. 10).
Fig. 10
Improvement of segmental rotation and global detorsion. The sum of initial segmental
rotations is 92° i.e. an average overall torsion of 5.4°. In-brace, the arithmetic
sum (negative sign if the rotation is inverted) is 15° i.e. an overall torsion of
less than 1°. The overall untwisting in this case exceeds 80 %. Qualitative vectorial
representation confirms the translation of the vertebral bodies along the midline
with overcorrection
Case n° 401
This patient of 13 years has a very significant worsening of a thoracic scoliosis
of 57° despite a Chêneau brace well executed and worn appropriately during three years.
Parents wanted to wait before surgery and her surgeon asked us to try out the new
Lyon brace on her. The immediate in-brace reduction obtained was 21°, and with sterEOS
we see a good translation and reharmonisation of the spine (Fig. 11).
Fig. 11
Case n° 401. In-brace correction after failure of 3 years Chêneau brace. Despite surgical
indication, the parents refuse surgery and ask us to continue treatment with the new
ARTbrace. The angular correction in the frontal plane remains above 50 %. The geometrical
detorsion is satisfactory
The top view confirms the vertebral vector projection (Fig. 12).
Fig. 12
Da Vinci view of Case n° 401. The Da Vinci view confirms the geometric detorsion with
recentering of the vertebral bodies along the midline and improved segmental derotation
The calculated overall untwisting which is the difference of both arithmetic average
and all of the segmental rotations before and in-brace is 33 % (Fig. 13).
Fig. 13
Global Torsion Index and percentage detorsion of case n° 401. The overall detorsion
index has improved of more than one third in this case
The average for the 15 patients studied was 37 %.
Discussion
Adolescent idiopathic scoliosis (AIS) is a structural three-dimensional deformity
of the spine arising in otherwise normal children during puberty. The use of brace
in the conservative treatment for AIS [32] plays an important role and is meant to
stop the evolution of the deformity in immature adolescents in order to prevent problems
during adulthood. Different types of braces have been used in the treatment of AIS.
The Lyon brace, created in 1947 by Pierre Stagnara, has been the first 3-points system
adjustable brace, used after a plaster cast reduction [14, 16, 17].
Many previous studies support the positive results with the casting and Lyon braces
[14, 16, 17]. The serial derotational plaster cast is commonly used for early onset
scoliosis to create asymmetrical growth and remodelling [33]. The idea is to recreate,
with a removable brace, the same derotational forces. In this study we report short
time prospective results after 1 year of 148 scoliosis cases treated with the new
Lyon ARTbrace, in correlation with a matched-pair control with the old Lyon Brace.
For ethical reasons, comparative studies of braces are exceptional. The originality
of this study is the instantaneous transition from the old to the new Lyon brace in
one step, for administrative reasons. Since May 2013 all patients (450) of JCdM at
the ‘Clinique du Parc – Lyon’ were treated with the new Lyon brace (ARTbrace) instead
of the classical EDF plaster cast. In fact from the first patient being treated with
the new brace, the ARTbrace appeared to be a much more effective solution compared
to the former plaster cast and it was even better tolerated. JCdM [7] has reported
recently the early radiological results of 225 scoliosis cases treated with new Lyon
brace (ARTbrace), matched with group control SRS & SOSORT criteria. This first immediate
results have demonstrated that the in-brace correction of patients’ Cobb angle was
70 % with a correction 40 % higher than the former Lyon brace or historical Lyon brace
and the other braces published in the literature, including retrospective studies.
These results can be explained by new biomechanical concepts based on scoliosis detorsion
such as:
Segmental moulding with individual correction of the frontal and sagittal plane,
Individual shape superposition
Fixed sagittal plane with simultaneous correction of flat back or hyperkyphosis
Night and day overcorrecting brace
Axilla baby lift concept
Coupled movement with biomechanical helicoidally detorsion of spine,
High rigidity asymmetric polycarbonate with “mayonnaise tube” elongation effect
Soft contact to increase tolerance and compliance
4D action with breathing toward the lateral expansions
One limit of this study was to only value the results of the immediate in-brace reduction
of scoliotic curves by the ARTbrace, so in a limited time frame (3 day follow-up).
In our study, instead, we report the early clinical and radiological results of group
A (of 148 patients) with a follow-up after 1 year, so over a longer period of time,
in correlation with a matched-pair control old Lyon brace group B.
In frontal radiological outcomes of this study, significant differences in the scores
for Thoracic and Lumbar curves at T0, T1, T2, T3 are reported.
T1 initial in-brace correction
For thoracic curves, the percentage improvement between the two groups was 37 %, confirming
the initial results.
T2 at 6 months
X-rays at 6 months are conducted without the brace. In both groups we observe an angular
recurrence linked to the elasticity of the scoliotic curvature with and without brace,
which is normal. Overall, this elasticity is statistically lower for group A, as if
a better in-brace correction decreased elasticity of scoliosis i.e. the difference
between in-brace angulation without brace. These results confirm the BRAIST study
that retains the importance of in-brace correction brace as a fundamental criterion
of the final outcome treatment. If we compare the two groups, the percentage of improvement
is even slightly higher.
T3 after 1 year
Many authors consider the results at 6 months as an excellent point in time to predict
the final outcome [8, 34, 21, 35]. But above all, it is interesting to assess the
evolution of the 2 curves that can be parallel, convergent or divergent, which has
never been described to date. In our study, we find a divergence between both thoracic
curves and lumbar curves in favor of group A, which would tend to prove that the initial
efficiency of the brace is continuing with time. The average angular improvement is
5° between the two groups.
Certainly angular reduction after 1 year is not the final treatment outcome (2 years
after weaning of brace), but we can identify some trends.
In the sagittal plane, this is the first brace to significantly improve the flat back
and delordosis tendency.
This improvement with the ARTbrace can be related to segmental moulding with fixed
sagittal correction but above all to the unscrewing or untwisting effect of the spine
with translation of the vertebral bodies near the midline. In the literature, studies
on improvement to the sagittal plane due to brace effect do not exist. Instead, many
authors report accentuating brace effect on the flat back, probably related to axial
stretching due to the overcorrection in the frontal plane [36, 37]. Analysing the
effect of a brace on 38 patients treated with Chêneau using MR animation, shows a
significant reduction of the mean Cobb angle of thoracic curves in-brace in MR animation
coronal 0° projection (simulating A-P view in X-ray) but in -90° projection, simulating
a lateral X-ray view, reported a reduction Kyphosis Cobb angle in 33/38 patients.
So the MRI animation analysis confirms the straightening effect of the brace leading
to the flattening of the sagittal spinal profile.
In the horizontal plane, this is the first time we no longer speak of segmental derotation
but overall untwisting of scoliosis measured automatically by SterEOS (index of global
detorsion).
As scoliosis is a structural three-dimensional deformity, the development of the EOS-system
has allowed us to study better transverse plane analysis. It is a concept initiated
by [38] his “torsiometer” is still considered the most accurate method of measurement
of axial vertebral rotation on 2D A-P radiographs [39, 40]. Then MRI and CT have improved
accuracy of vertebral rotations measurements but their clinical relevance is limited
by the supine position of the patient for MRI and radiation exposition for CT. Today
the development of the EOS-system has allowed us to study better transverse plane
analysis and improve our knowledge. Courvoisier, analysing the transverse plane pattern
of 111 patients with mild scoliosis in 3D by the EOS-system, combining apical axial
rotation, the intervertebral axial rotation at junctions and the torsion index, has
demonstrated that it is independent of the scoliotic curve location but above all
significant in the determination of the progression risk of mild scoliosis [26].
Simplicity of classification
The former Lyon brace requires the use of Lenke classification adapted to bracing.
The segmental moulding of the new Lyon brace is much simpler and requires only two
classes: C and S shaped scoliosis. The overall alignment is provided by the first
moulding. For C shaped scoliosis, the thoracic and lumbar shifts will be realized
in the same direction, while it will be carried out in opposite directions for the
S curves. The lumbar oblique tilt and high curvatures are also much easier to manage.
Overcorrection
The interest of overcorrection is not obvious especially for a specialist symmetrical
brace such as the historical Lyon brace. However, we were using the night overcorrection
for small thoracolumbar curves with good results. The Chêneau brace experience also
goes in the direction of overcorrection. The ability to manually perform that overcorrection
directly on the child is an advantage. Toru Maruyama showed us the interest of the
shift in the scoliosis correction [41, 42] and the patient’s posture during the segmental
moulding is very close to some of Schroth’s specific exercises [43].
Lyon brace Management
So far we have not changed the overall management of the Lyon bracing treatment following
the guidelines set by the SOSORT in 2011 [44].
The indications are the same with wear time in the day depending on the initial angulation.
The management is the same during treatment with adaptation of wearing time according
to the elasticity of the scoliosis (X-ray without brace compared to in-brace X-Ray).
Indeed, we believe that the spine is not made to grow properly under a rigid brace,
and a very effective brace worn for a shorter period during the day is better than
a less effective brace worn 23h/24. Although the consequences of the brace on bone
mass is not obvious, the precautionary principle is required [45].
Physiotherapy has remained the same, but the brace asymmetry and preparation for segmental
moulding are closer to asymmetric methods and probably the current protocol will change
in the future.
The continuation of sports during treatment is a characteristic of the Lyon bracing
treatment because the plaster cast causes a realignment of tension along the spine.
The results with the historical Lyon brace were better when the children were practicing
at least 5 h of sport each week. The initial full time wearing has the same creep
effect. There is also another advantage with increase of skin tolerance (watch effect
of Manuel Rigo). However, tolerance is worse in part time wearing.
Compliance
Compliance is a key element of the final results of the treatment with immediate in-brace
reduction. Compliance depends on the child and the family, but also on the brace which
should be light, aesthetics and the allowance of normal breathing by expansions. The
low dropout rate may also be due to changes in the brace.
Conclusion
This study demonstrates not only good results of the ARTbrace about immediate in-brace
reducibility of scoliosis described in previous studies published by the same author,
but this trend is maintained further at 6 months and at 1 year. So the new concepts
and first results of ARTbrace, defined as a modified or “new” Lyon brace, confirm
that it can completely replace the casting and old Lyon brace process.
Finally, angular reduction at 1 year is not certainly the final treatment outcome
(2 years after weaning of brace) even if some authors are using the reduction at this
point in time as a predictive criterion. So, future studies could confirm if this
criterion is valid and consequently the effectiveness of the ARTbrace.
Availability of supporting data
The data set supporting the results and all SPSS statistics results are included within
the article.
Consent
All patients were informed and have given their consent for this work. This monocentric
study was the object of a declaration to the CNIL under the number 1831534 in France
and the procedure accepted.