Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal

Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal

SPINE Volume 31, Number 17, pp 1933–1942
©2006, Lippincott Williams & Wilkins, Inc.
Progression Risk of Idiopathic Juvenile Scoliosis During
Pubertal Growth
Yann Philippe Charles, MD,* Jean-Pierre Daures, PhD,† Vincenzo de Rosa, MD,*
and Alain Diméglio, MD*
Study Design. A retrospective study investigated the
progression risk of juvenile scoliosis until skeletal maturity or spinal fusion.
Objectives. To define risk factors of curve progression
during pubertal growth and analyze the timing of arthrodesis.
Summary of Background Data. Juvenile scoliosis is
characterized by a major, extremely variable progression
risk. Peak growth velocity is the most critical period. Curve
progression related to growth needs to be analyzed critically
for an adequate treatment.
Methods. A total of 205 patients, including 163 girls and
42 boys, with juvenile scoliosis were reviewed at skeletal
maturity. The scoliosis was divided into juvenile I with an
onset of 4 –7 years (52 patients) and juvenile II with an onset
of 8 –10 years (153). Standing and sitting height, weight,
Tanner signs, skeletal age, and menarche were regularly
assessed. Topographies and Cobb angles of primary and
secondary curves were referred to the pubertal growth diagram.
Results. Of 205 patients, 99 (48.3%) were operated on.
Of 109 curves ⱕ20° at onset of puberty, 15.6% progressed
⬎45° and were fused. Of 56 curves of 21° to 30°, the
surgical rate increased to 75.0%. It was 100% for curves
⬎30°. Curves ⬎20°, which increased and were operated
on, progressed significantly during peak growth velocity
(P ⫽ 0.0014). Curves that progressed by 6° to 10°/y were
fused in 70.9%, curves which increased ⬎10°/y in 100% of
cases (P ⫽ 0.0001). This risk was highest for primary
thoracic curves: King V, III, and II (P ⫽ 0.0001). There was
no difference between males and females or juvenile I and II.
Conclusions. Curve pattern, Cobb angle at onset of
puberty, and curve progression velocity are strong predictive factors of curve progression. Juvenile scoliosis
⬎30° increases rapidly and presents a 100% prognosis for
surgery (curve ⬎40° to 45°). Anticipation is necessary if
the scoliosis progresses during the first year of puberty.
The prediction is difficult for curves of 21° to 30° during
the first 2 years of puberty. Curve pattern and curve progression velocity are useful to detect which curves are
likely to progress. From this retrospective analysis, spinal
fusion could have been indicated earlier sometimes. An
earlier intervention is probably preferable to obtain better
From the *Service d’Orthopédie Pédiatrique, Centre Hospitalier Universitaire, and †Institut Universitaire de Recherche Clinique, Faculté de
Médecine, Montpellier, France.
Acknowledgment date: May 5, 2005. First revision date: September 4,
2005. Second revision date: October 23, 2005. Acceptance date: October 24, 2005.
The manuscript submitted does not contain information about medical
device(s)/drug(s).
No funds were received in support of this work. No benefits in any
form have been or will be received from a commercial party related
directly or indirectly to the subject of this manuscript.
Address correspondence and reprint requests to Yann Philippe Charles,
MD, Service d’Orthopédie Pédiatrique, Hôpital Lapeyronie, 371, Av
du Doyen G. Giraud, 34295 Montpellier Cedex 5, France; E-mail:
[email protected]
curve reduction on a supple spine, even if a perivertebral
fusion is necessary. We use the 3 parameters for operative indications. If an early spinal fusion leads to better
curve correction needs to be verified on prospective data.
Key words: juvenile scoliosis, pubertal growth,
curve progression, surgical indication. Spine 2006;31:
1933–1942
Juvenile scoliosis represents a particular entity within
idiopathic scoliosis. It is characterized by an early deformity that leads to a major but extremely variable progression risk throughout the pubertal growth spurt. Remaining growth is an essential parameter that must be
considered in the evaluation of curve progression risk.
Lonstein and Carlson1 pointed out that 3 strong progression factors for idiopathic scoliosis were the curve magnitude along with the patient’s chronologic age and the
Risser sign. Duval-Beaupère et al2 showed that the main
progression of idiopathic scoliosis occurs at the most
rapid adolescent skeletal growth, which is a critical period in the evolution of spinal deformity and for its final
outcome.
Little et al3 confirmed that peak height velocity is a
reliable clinical marker for the prediction of remaining
growth and progression of scoliosis. This time of peak
growth velocity occurs around 11–13 years of skeletal
age in girls and 13–15 years in boys, and it is characterized by a gradual increase in the spinal growth rate.4
Patients presenting with juvenile idiopathic scoliosis will
go through this entire period of pubertal growth and,
therefore, need systematic follow-up on their growth
curve. The determination of secondary sexual characteristics as well as skeletal age are helpful in the evaluation
of skeletal maturity in addition to 6 monthly height measurements.5 The radiographic evolution of the scoliosis
can then be plotted against these maturity indicators to
obtain precise information about remaining growth and
the potential curve progression risk.
The purpose of this study is to analyze the evolution of
idiopathic juvenile scoliosis until skeletal maturity in
conservatively treated patients and until the time of spinal fusion in those who underwent surgery. Risk factors
of curve progression related to pubertal growth parameters, particularly during the phase of peak growth velocity, as well as curve pattern, onset of scoliosis, and
gender are determined. Furthermore, the timing of spinal
arthrodesis is then analyzed retrospectively for operated
scoliosis, and discussed within the context of specific risk
factors and the degree of skeletal maturity.
1933
1934 Spine • Volume 31 • Number 17 • 2006
Materials and Methods
The medical records and radiographs of 444 consecutive patients with juvenile idiopathic scoliosis followed at our Pediatric Orthopedic Department between 1988 and 2004 were reviewed. Derived from this cohort, 205 patients were observed
regularly every 6 months from onset of scoliosis to skeletal
maturity, which was defined as clinical cessation of trunk
growth and Risser 5. Usually the patients were seen during the
month of their birthday, which made regular follow-up easier.
The complete evolution during the growing period of these
patients with scoliosis was well documented and their data
analyzable in the present study. The scoliosis was divided into
2 groups: “Juvenile I” with early onset from 4 to 7 years and
“Juvenile II” with later onset from 8 to 10 years. Scoliosis with
an underlying neurologic disorder or syndrome was excluded
from the study protocol. At our institution, magnetic resonance
imaging is systematically performed in addition to clinical and
neurologic examinations to detect any neural axis abnormality
in juvenile scoliosis.
The patients received treatment according to their curve and
skeletal maturity. In general, curves less than 20° were observed. Curves ⱖ20° were braced. The Milwaukee brace was
mostly used for smaller children during the prepubertal period
and later for high thoracic curves. For older children and adolescents, thoracolumbosacral orthosis and Charleston bending
braces were applied. Surgery was indicated in progressive
curves exceeding 40° to 45° and consisted of a posterior spinal
fusion combined with an anterior growth arrest if the Risser
was zero. A single posterior instrumentation was performed in
adolescents from Risser 1 if the curves were still relatively reducible. A previous anterior release was indicated for structural
curves usually more than 70°. Anterior instrumentation was
not performed in this cohort.
We regularly use a checklist on a computerized database at
our clinics to assess the evolution of growth parameters and
developmental stages of puberty. Standing and sitting height as
well as weight were always measured the same way and using a
fixed graduation and scales in our clinics. Secondary sexual
characteristics according to the Tanner stages6 and menarche
were also documented. These clinical parameters were combined with skeletal age assessment. The Greulich and Pyle atlas7 was used throughout puberty and complemented by the
method of Sauvegrain et al8 using the elbow, which is extremely valuable during the time of peak growth velocity from
11 to 13 years in girls and 13 to 15 years in boys.9,10 The
moment of triradiate cartilage closure was also noted during
this phase of growth. The Risser sign11 was used on the portion
of decelerating pubertal growth velocity until skeletal maturity.
Topographies of neutral, end, and apical vertebrae were
determined on standing posteroanterior spine radiographs, and
the angles of primary and secondary curves were measured
according to the Cobb method.12 There were 11 radiographs
per patient considered: 1 initial before puberty, 5 at 6 monthly
intervals during the first 2 years of pubertal growth (Risser 0),
and 5 from Risser 1 to 5. Two of the authors (Y.P.C. and
V.d.R.) who worked closely together took all serial radiographic measurements. The senior author (A.D.) realized initial
measurements. The 3 authors discussed the radiographic evolutions of all scoliosis cases to minimize intraobserver and interobserver errors. A curve progression was defined as an increase of ⱖ5°. Because of the retrospective nature or this
review, lateral spine radiographs could only be evaluated if
available to recognize tendencies on the sagittal plane deformity. These incomplete data were not considered for statistical
evaluation. The following frontal curve patterns were determined
according to the classifications of Lonstein13 and King et al.14
Major Thoracic and Minor Lumbar Curve Pattern. An
S-shaped curve that consists of a superior curve with the upperend vertebra at T4 or T5, a lower-end vertebra at T12, and an
apical vertebra at T8 or T9, and an inferior curve with the
upper-end vertebra at T12 and the lower-end vertebra at L4 or
L5. The apical vertebra is usually at L2. The thoracic curve is
larger and more structural than the lumbar curve. This curve
pattern corresponds to a King type II.
Major Lumbar and Minor Thoracic Curve Pattern. An
S-shaped curve in which the lumbar curve is larger and more
rigid. Its upper-end vertebra is usually at T10, its lower-end
vertebra at L4 or L5, and its apical vertebra mostly at L1. The
thoracic curve is more flexible and extends from T4 to T9 or
T10, with its apex at T6 or T7. This curve pattern may also fit
into the King type I category.
Single Thoracic Curve Pattern. The apex of this curve type
lies within the thoracic spine, usually at T8 or T9. The upperend vertebra is either T4, T5, or T6 and the lower-end vertebra
T11, T12, L1, or L2. The most common end vertebrae are T5
and T12. The great majority of these curves are convex to the
right. This curve pattern is also described as type III in the
classification by King et al.14
Single Thoracolumbar Curve Pattern With Thoracic
Predominance. This is a single curve pattern with its upper
end at T6, T7, or T8 and the lower-end vertebra at L3. The
apical vertebra is either T10, T11, or T12. The upper thoracic
and lower lumbar spine may show small compensatory curves,
which are usually completely flexible.
Single Thoracolumbar Curve Pattern With Lumbar
Predominance. This curve pattern is similar to the previous
one but lower. Its upper-end vertebra is at T8, T9, or T10 and
the lower-end vertebra at L3. The apical vertebra is at L1.
Single Major Thoracolumbar Curve Pattern. This is a long
thoracic curve with its upper end at T4, T5, or T6, and L4 tilts
into the curve at the lower end. The apical vertebra is usually at
T12. This curve pattern corresponds to King type IV.
Single Lumbar Curve Pattern. The upper-end vertebra of
this curve is at T11, T12, or L1 and the lower-end vertebra at
L4 or L5. The most common end vertebrae are at L1 and L4.
The apex is usually at L2.
Double Thoracic Curve Pattern. A left-upper and rightlower thoracic curve is typical of this pattern. The upper curve
has its apex at T3 or T4 and extends from T1, or T2–T5 or T6.
The lower curve has its apex within the thoracic spine and
extends from T5, or T6 –T11 or as low as L2. This curve pattern corresponds to type V curves according to the classification by King et al.14
Standing and sitting height measurements as well as the
determined Cobb angles were regularly referred to the pubertal
growth diagram5,10 at 6 monthly intervals. As shown in Figure
Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal Growth • Charles et al 1935
Figure 1. Pubertal growth diagram in girls and boys related to
annual growth velocities of sitting height and lower limb.
1, the beginning of puberty is determined by an increase in
growth velocity, and the first 2 years represent the ascending
phase of pubertal growth. This period of highest growth velocity is characterized by a total increase of 7.5 cm in sitting height
and 7 cm in the lower limb segment for girls on average. The
respective values for boys are an increase of 8.5 cm in sitting
height and 8 cm in the lower limb.5 In addition to increasing
growth velocity, the beginning of this phase is generally marked
by the Tanner pubic stage 2 in boys and the breast stage 2 in
girls. Skeletal age at this time is around 11 years in girls and 13
years in boys. The end of this ascending pubertal growth phase
is marked by the fusion of all elbow epiphyses, which corresponds to 13 years of skeletal age in girls and 15 years in boys.
The Risser sign is still zero at this stage. This portion of growth
is split in 2 halves by the closure of the triradiate cartilage,
which occurs around 12 years of age in girls and 14 years of age
in boys.15
After elbow closure, mean remaining growth is 4.5 cm in
sitting height and 1.5 cm in the lower limb segment for boys
and girls. This descending phase of pubertal growth is characterized by a continuous growth velocity deceleration with an
early cessation of leg growth. The time from Risser 1, usually
13.5 years in girls and 15.5 years in boys, until skeletal maturity is clearly defined by 6-month sequences on the Greulich
and Pyle atlas.7 These skeletal age data complement the Risser
sign, which is more variable,16,17 and enable the mapping of
the patient’s skeletal maturity on the pubertal growth diagram.
Menarche mostly occurs at the beginning of this descending
growth phase, usually between 13 and 13.5 years, and corre-
sponds to Risser 1. Nevertheless, this indicator is less precise
when predicting remaining growth. Figure 2 illustrates an example of curve progression applied to the pubertal growth diagram.
The relationship between curve progression risk and gender, onset of scoliosis and the curve pattern were determined.
This risk was also calculated as a function of the primary
curve’s Cobb angle at onset of the pubertal growth spurt and of
the annual curve progression velocity during the first 2 years of
puberty. The phases of pubertal growth, in which the curves
essentially progressed, and the time of surgical intervention
were finally evaluated.
Statistical analysis was performed using BMDP software
(Statistical Solutions, Saugus, MA). The Fisher exact test was
used for 2-tailed qualitative data in the evaluation of curve
progression risk related to gender and the onset of scoliosis.
The Pearson ␹2 test was used for all other evaluations of qualitative risk factors. The significance level was set at P ⬍ 0.05.
Results
Age and Gender
The mean age at diagnosis of scoliosis in 205 patients
was 8 years and 2 months (range 5 years and 1 month to
10 years). These curves were divided into 52 juvenile I
with early onset before age 8 years and 153 juvenile II
with later onset from age 8 years or older. There were
163 girls and 42 boys. In the group of girls, 44 with
scoliosis were classified as having juvenile I and 119 as
having juvenile II. In the group of boys, 8 had juvenile I,
1936 Spine • Volume 31 • Number 17 • 2006
Figure 2. Slow curve progression under brace treatment until the end of growth.
and 34 had juvenile II. The ratio of boys to girls was
1:5.5 in the juvenile I group and 1:3.5 in the juvenile II
group.
A spinal arthrodesis was indicated in 99 of the 205
patients (48.3%). Of the 42 boys and 163 girls, 22
(52.4%) and 77 (47.2%), respectively, were operated
on. There was no significant difference in progression
risk toward surgery (curve more than 40° to 45°) between males and females (P ⫽ 0.2913). A spinal fusion
was performed in 27 of the 52 juvenile I curves (51.9%)
and in 70 of the 153 juvenile II curves (45.8%). These 2
groups did not present a significant difference concerning
the risk for surgery (P ⫽ 0.3271).
Curve Pattern
The frequencies of different curve patterns in 205 patients
and their respective percentage of curves, in which a spinal
arthrodesis was performed, are shown in Table 1. Primary
thoracic curve patterns of King types V, III, and II were at
highest risk for progression more than 40° to 45° and presented the highest rate of operated curves. Major lumbar
and minor thoracic curves showed a lower progression risk
than primary thoracic curves in the group of S-shaped curve
patterns. The prognosis for surgery decreased in single thoracolumbar curve patterns with a lower level of the apical
vertebra. Spinal fusion was not performed in any single
lumbar curve. These curves had a lower progression risk
than other curve patterns. Although sample sizes of some
curve patterns were quite small, these results proved to
be highly significant (P ⫽ 0.0001).
Curve Degree at Onset of Pubertal Growth Spurt
At diagnosis, the primary curves measured 19° on average (range 5° to 68°). At the beginning of puberty, the
average value was 23° (range 8° to 64°). At the triradiate
cartilage closure, the average Cobb angle was 29° (range
10° to 85°). At the end of peak growth velocity, the end
of the ascending growth phase, the primary curves measured 33° on average (range 8° to 120°). The maximal
value during curve progression reached 41° on average
(range 10° to 135°) around Risser 3. Follow-up of curve
evolution showed that globally the scoliosis did not
Table 1. Curve Pattern Frequencies in 205 Patients and
Operated Percentage
Curve Pattern
Major thoracic-minor lumbar
(King type II)
Major lumbar-minor thoracic
(King type I)
Single thoracic (King type III)
Single lumbar
Single thoracolumbar-thoracic
predominance
Double thoracic (King type V)
Single thoracolumbar-lumbar
predominance
Single major thoracolumbar
(King type IV)
Pearson ␹2 test: P ⫽ 0.0001.
Frequency
(n ⫽ 205)
No. Operated Curves
per Pattern (%)
93
56 (60.2)
38
13 (34.2)
27
18
12
17 (63.0)
0 (0)
5 (41.7)
10
4
7 (70.0)
1 (25.0)
3
0 (0)
Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal Growth • Charles et al 1937
Table 2. Progression of Nonoperated Scoliosis
(n ⴝ 106/205)
Primary Curve Amplitude at
Onset of Puberty
Phase of Curve Progression
No progression
Ascending phase
Ascending and descending phase
Descending phase
ⱕ20°
(n ⫽ 92/105)
21° to 30°
(n ⫽ 14/56)
33.7%
28.3%
29.3%
8.7%
21.4%
28.6%
50.0%
0.0%
Pearson ␹2 test: P ⫽ 0.3335.
progress or only progressed slowly before entering the
phase of pubertal growth spurt. However, the range of
these values became much wider on the portion of ascending pubertal growth, which is caused by a high variability of curve progression depending on the initial angle and curve type, as well as the efficacy of brace
treatment.
To obtain a clearer idea of curve progression during
the different phases of pubertal growth, we classified the
scoliosis into 3 groups of primary curve amplitude at the
onset of puberty: ⱕ20°, 21° to 30°, and more than 30°.
In the section of amplitudes ⱕ20°, only 17 of 105 patients (16.2%) with scoliosis were operated on until the
end of growth. In the group of curves ranging from 21°
to 30°, 42 of 56 patients (75.0%) were treated surgically,
and in the section of curves, which were more than 30° at
the beginning of puberty, spinal arthrodesis was performed in all 40 patients (100%). Tables 2 and 3 show
the sections of essential curve progression throughout
pubertal growth. In the group of curves ⱕ20°, in which
spinal fusion was indicated, the majority progressed
slowly until the end of growth. Curves from 21° to 30°
all progressed during the first 2 years of the pubertal
growth spurt. Of 42 curves in this section, 18 (42.9%)
progressed rapidly during the ascending phase of pubertal growth. The other 24 curves (57.1%) began to
progress during the ascending phase at a slower velocity
but continued to progress during the descending phase as
long as surgery had not been previously indicated.
Although the curves ranging from 21° to 30° at the
beginning of puberty presented a global prognosis for
surgery of 75.0%, this group presented the highest variability in terms of annual curve progression rate, which
Table 3. Progression of Operated Scoliosis (n ⴝ 99/205)
Primary Curve Amplitude at Onset of Puberty
Phase of Curve
Progression
ⱕ20°
(n ⫽ 17/105)
21° to 30°
(n ⫽ 42/56)
⬎30°
(n ⫽ 40/40)
23.5%
64.7%
42.9%
57.1%
72.5%
27.5%
11.8%
0.0%
0.0%
Ascending phase
Ascending and
descending phase
Descending phase
Pearson ␹2 test: P ⫽ 0.0014.
makes prediction more difficult. In the curve section
more than 30°, 29 of 40 patients (72.5%) with scoliosis
had progression essentially during the ascending phase,
and spinal arthrodesis was indicated earlier. Nevertheless, the analyzed progression patterns for scoliosis requiring spinal fusion showed that all curves more than
20° at onset of puberty significantly increased during the
time of peak growth velocity (P ⫽ 0.0014).
Curve Progression Velocity
Of 205 patients with scoliosis, 161 had progression during the ascending phase of pubertal growth. To determine the risk for surgery (curve more than 40° to 45°) by
assessing curve progression velocity during this time of
high growth velocity, 3 classes of annual Cobb angle
increase were compared: less than 6°, increase of 6° to
10°, and more than 10° per year. Table 4 shows that an
increase of ⱖ6° augments the risk for surgery significantly (P ⫽ 0.0001) compared to curves that progress
more slowly. An increase of more than 10° per year,
which could also be termed as 1° per month, represents
a 100% prognosis for surgery.
Timing of Spinal Fusion
Analysis of the time of spinal fusion related to curve
progression and pubertal growth showed that 42 of 99
patients (42.5%) with scoliosis had been operated on at
Risser zero, before entering the phase of decelerating
growth velocity. There were 10 patients who were operated on at Risser 1 (10.1%), 16 at Risser 2 (16.2%), 14 at
Risser 3 (14.1%), 8 at Risser 4 (8.1%), and 9 at Risser 5
(9.1%). Figure 3 shows the example of a patient who had
a primary thoracic curve of 35° at the onset of the pubertal growth spurt, which increased to 56° and was
fused posteriorly at Risser 2. This curve was reduced by
nearly 40%, and the remaining Cobb angle was 35° at
skeletal maturity. In contrast to this case, Figure 4 illustrates an example of early anterior thoracoscopic discectomy and vertebral growth arrest combined with a posterior instrumentation, allowing an almost complete
curve correction and a stable fusion until the end of
growth.
To determine if an early spinal fusion leads to a better
curve reduction than a later correction from the present
retrospective data, the percentage of primary curve correction (preoperative Cobb angle ⫺ Cobb angle at Risser
5) was calculated for each curve. Average values for all
Table 4. Velocity of Curves Progressing During
Ascending Phase (n ⴝ 161/205)
Annual Increase of Primary Curve
Nonoperated
Operated
⬍6°
(n ⫽ 58)
6° to 10°
(n ⫽ 54)
67.2%
32.8%
29.1%
70.9%
Pearson ␹2 test: P ⫽ 0.0001.
⬎10°
(n ⫽ 24)
0.0%
100%
1938 Spine • Volume 31 • Number 17 • 2006
Figure 3. Late operative indication.
subgroups at different stages of puberty from Risser 0 to
5 were then compared: 57.3% at Risser 0, 53.2% at
Risser 1, 48.2% at Risser 2, 46.2% at Risser 3, 47.7% at
Risser 4, and 47.0% at Risser 5.
Discussion
Remaining growth is a key factor in the progression of
idiopathic scoliosis with early onset like juvenile scoliosis. In most cases, the aim of orthotic treatment will be to
avoid spinal fusion at the end of growth. Nevertheless, a
curve progression can often be observed during the time
of the pubertal growth spurt, despite bracing, and a surgical intervention will likely be necessary.18 –20 The prediction of curve progression is relatively difficult to determine from the time when scoliosis occurs in younger
children until the beginning of puberty. Perdriolle and
Vidal21 stated that the main increase in juvenile scoliosis
occurs between the age of 6 years and the Tanner pubic
stage 2, followed by a lesser progression during the adolescent growth spurt. As previously shown by DuvalBeaupère et al2,22 and James,23 the results of the present
study confirmed that the main curve progression did not
occur during prepuberty but at the time of accelerated
growth velocity.
Furthermore, our study outlined that there was no
difference in progression risk between early or late onset
juvenile scoliosis occurring before or after the age of 8
years. Figure 5 illustrates how the spinal growth velocity
of thoracic and lumbar segments slows down and remains constant between ages 5 and 11 years in girls and
5 and 13 years in boys but then increases gradually after
this time,4,24 which may explain why juvenile scoliosis
essentially progresses during the period of pubertal
growth. Based on the fairly large number of patients with
juvenile scoliosis treated at our institution, we questioned the strategy of bracing patients over the entire
time of peak growth velocity and continuing this treatment during the descending phase of growth in some
cases. The rate of surgically treated patients in this cohort was relatively high (48%) and is almost comparable
to the results of Masso et al,25 who performed a spinal
fusion in 50% of 52 patients with juvenile scoliosis at a
mean age of 14 years.
Mannherz et al26 recommended an orthosis in 32 of
43 patients; 40% had progression and required an operation despite bracing until skeletal maturity. Figueiredo
and James27 treated 56% of 98 patients with juvenile
scoliosis surgically. It has been their policy to indicate
spinal fusion in all progressive thoracic curves at about
the age of 10 years to avoid the inconvenience of wearing
a brace for a long and uncertain period of time. We
usually wait at least if the curve progresses during the
Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal Growth • Charles et al 1939
Figure 4. Early spinal fusion during ascending phase of pubertal growth.
first year of the pubertal growth spurt. To recognize
curves that are likely to progress and predict the prognosis for surgery, we determined significant risk factors that
led us to anticipate spinal arthrodesis and avoid too long
orthotic treatment.
Primary thoracic curves, especially King types V, III,
and II, proved to be the most likely to progress more than
40° to 45° and were characterized by a high rate of sur-
Figure 5. Annual growth velocity
of thoracic and lumbar spine.
Modified with permission from
Springer; 1990.4
gical correction in this cohort. According to the results of
Lonstein and Carlson,1 double thoracic curves (King V)
present the highest tendency to progress. Their apical
vertebrae are at high levels, and, therefore, these curves
are difficult to control in a brace. We also agree with the
findings of Robinson and McMaster,28 who showed that
single thoracic curves (King III) as well as primary thoracic curves with secondary development of a lumbar
1940 Spine • Volume 31 • Number 17 • 2006
curve (King II) present a high risk of progression. The
apical vertebra of these curves is usually located at the
eighth thoracic level. Single thoracolumbar curves with
an apical vertebra at or around T10 still present a relatively high progression risk, especially if the trunk is
shifted toward the convexity of the curve. Primary and
single lumbar curves have a more benign prognosis. Sagittal plane radiographs were not statistically evaluated in
the present study because of incomplete data, but we also
believe that a hypokyphotic component of the thoracic
segment as well as an increased vertebral rotation must
be considered as negative prognostic factors,22,26,28,29
which are part of the 3-dimensional spinal deformity.
Lonstein and Carlson1 analyzed the evolution of idiopathic scoliosis in general, and showed that the incidence
of curve progression was correlated with curve magnitude as well as the patient’s chronologic age and the
Risser sign. Nevertheless, juvenile idiopathic scoliosis
needs to be considered independently of adolescent idiopathic scoliosis. Duval-Beaupère et al2 pointed out that
skeletal age assessment as well as the determination of
secondary sexual characteristics were valuable parameters for evaluating remaining growth and the progression
risk of scoliosis.
The best method for detecting the beginning of the
pubertal growth spurt is to measure the child’s standing
and sitting height regularly at 6 monthly intervals, and
match these data with skeletal ages from the left hand
and wrist7 and the left elbow.8,9 The ascending growth
phase, also known as peak height velocity,3 starts around
11 years of skeletal age in girls and 13 years in boys, and
is characterized by a gradual increase of the spinal
growth rate over a period of 2 years. Mean increase in
sitting height is about 3.5 cm per year in girls and 4 cm
per year in boys.
The present study showed that curve magnitude
groups juvenile scoliosis significantly in terms of progression risk at this particular point of increasing growth
velocity. Curves exceeding 30° at the beginning of the
pubertal growth spurt progress rapidly more than 40° to
45° and present a 100% prognosis for surgical correction. If the risk factor of a primary thoracic curve is
associated, we usually wait until the curve has proved to
progress during the first year of puberty before indicating
spinal fusion. Curves ranging from 21° to 30° are at a
75% risk and present a higher variability of curve progression. All curves in this section, which will progress
and require spinal fusion at some point, are characterized by a significant increase in curve magnitude during
the 2 years of peak growth velocity. Therefore, the third
parameter of annual curve progression velocity is useful
in predicting prognosis and making an appropriate therapeutic decision as soon as possible. A Cobb angle increase of 6° to 10° per year represents a prognosis for
surgery around 70%, an increase more than 10° per year,
or 1° per month, represents a 100% risk. Usually, these
curves up to 30° at onset of puberty need to be observed
until the second year of pubertal growth. For clear de-
tection of curve progression, we recommend mapping
successive radiographs on the pubertal growth diagram
as shown in Figure 1, which makes the follow-up easy.
A review of our patients’ charts has outlined that spinal fusion was performed too late in some cases, which
permitted only partial curve reduction, as in Figure 3.
Based on the results of this study, this patient presented
significant risk factors of curve progression that could
help predict the prognosis: a primary thoracic curve and
a Cobb angle of 35° at the onset of the pubertal growth
spurt. An earlier intervention could have been more advantageous, allowing easier curve reduction. Nevertheless, we could not show that an earlier intervention systematically leads to higher curve correction, as shown in
Figure 4. It is difficult to determine from the present retrospective data if an early spinal fusion leads to a better
curve reduction than a later correction because curve
magnitudes and curve patterns varied in subgroups of
patients at different stages of skeletal maturity. Furthermore, surgical indications consisted of an anterior growth
arrest and posterior arthrodesis for immature spines, or a
single posterior instrumentation for still relatively supple
curves, or an anterior release and posterior fusion for structural curves. The calculated global results for each subgroup show only a tendency of higher correction in younger
patients. If an early spinal fusion leads to a better long-term
curve correction would need to be verified on prospective
data for comparable curve patterns and magnitudes in patients at different stages of skeletal maturity.
The risk factors presented (i.e., curve pattern, curve
degree at onset of puberty, and curve progression velocity) represent useful parameters in clinical practice,
which now allow us to evaluate progression risk in a
more precise manner in conjunction with clinical parameters, such as balance of the trunk. If these factors indicate a high risk of curve progression, surgery might be
indicated earlier, during the ascending phase of pubertal
growth, which might probably be preferable to obtain
better correction. In these patients at Risser zero, especially those with open triradiate cartilage, an anterior
spinal growth arrest is necessary to prevent a crankshaft
phenomenon.30 –36 We usually perform an anterior endoscopic anular release and discectomy37 before posterior instrumentation. This protocol seems appropriate
for the early correction of primary thoracic curves. Anterior instrumentation would be an alternative procedure, permitting growth arrest and curve correction
through a single procedure.
However, we only indicate this type of surgery, which
was not performed in the present cohort, in major thoracolumbar curves. An anterior and posterior fusion at
the beginning of pubertal growth represents only a minor
sacrifice in terms of sitting height. As illustrated in Figure
6, the average remaining spinal growth is about 3.6 cm
for the thoracic segment and 2.1 cm for the lumbar segment in girls with a skeletal age of 11 years. For boys, the
analog values at 13 years of skeletal age are 3.9 cm of
remaining growth for the thoracic spine and 2.3 cm for
Progression Risk of Idiopathic Juvenile Scoliosis During Pubertal Growth • Charles et al 1941
Figure 6. Remaining growth of
thoracic and lumbar spinal segments related to remaining sitting height until skeletal maturity.
Modified with permission from
Springer; 1990.4
the lumbar spine. Two years later, at the end of peak
growth velocity, remaining growth of the T1–T12 segment is about 1.5 cm in girls and 1.6 cm in boys, and the
L1–L5 segment has 0.9 cm of growth left in boys and
girls.4,24 These values show that a perivertebral arthrodesis does not significantly shorten sitting height if it is
performed during the ascending phase of pubertal
growth. Furthermore, a certain amount of sitting height
is lost in a curved spine. An early curve correction permits the distraction of the spine while it is still relatively
supple and its fusion in a straight position. Therefore, the
sitting height deficit appears to be a relative problem.
An attractive alternative to spinal arthrodesis of the
growing patient with open triradiate cartilage could be the
vertebral body stapling procedure, which seems to offer
promising results for curves less than 50°.38 If the presented
risk factors indicate a high progression risk (i.e., a primary
thoracic curve more than 30° at the beginning of puberty),
vertebral body stapling could allow a straight growth of the
spine and avoid a perivertebral arthrodesis.
There were 3 significant predictive factors of curve
progression determined in this study: curve pattern,
Cobb angle at onset of the pubertal growth spurt, and
curve progression velocity during the first 2 years of puberty. The best strategy for the follow-up and treatment
of idiopathic juvenile scoliosis is to measure the child
regularly to determine the beginning of peak growth velocity, which is the most critical period for curve progression. Skeletal age data and the assessment of secondary
sexual characteristics are helpful as additional parameters that allow precise mapping of the patient on the
pubertal growth diagram. Curve progression and risk
factors in the evolution of scoliosis are then easily recognizable if all successive spine radiographs are placed in a
panoramic view related to the pubertal growth diagram.
Primary thoracic curves exceeding 30° at the beginning
of puberty are likely to progress rapidly from the first
year of peak growth velocity. Therefore, surgical correction should be anticipated as soon as they reach values of
40° to 45°. Curves 21° to 30° are also at a high risk for
progress but present a higher variability. The parameter
of curve progression velocity is useful for the evaluation
of these curves, which should be observed at least until
the second year of the pubertal growth spurt. With the
use of the pubertal growth diagram and knowledge of
the presented risk factors, the prognosis of juvenile scoliosis can be assessed more precisely, and surgical treatment might be proposed earlier.
Key Points
● The first 2 years of pubertal growth, 11–13 years
in girls and 13–15 years in boys, are decisive in the
progression of idiopathic juvenile scoliosis. Nearly
90% of all operated curves progressed essentially
during this phase of peak growth velocity.
● Curves more than 30° at onset of the pubertal
growth spurt increase rapidly and present a 100%
prognosis for surgery (curve more than 40° to 45°).
Curves 21° to 30° also present a 75% progression
risk and need careful follow-up.
● An annual curve progression velocity of 6° to 10°
during the pubertal growth spurt represents a prognosis of around 70% for spinal fusion. An increase
of 1° per month represents a 100% risk.
● Primary thoracic curves (King types V, III, and II)
present a higher progression risk than primary
lumbar or thoracolumbar curves.
● Spinal fusion might be indicated early in primary
thoracic curves more than 30° at the beginning of
puberty if they proved to progress during the first
year of pubertal growth. Curve pattern and curve
progression velocity are useful additional parameters to predict which curves of 21° to 30° are likely
to progress more than 40° to 45° during the first 2
years of puberty and will necessitate surgical
correction.
1942 Spine • Volume 31 • Number 17 • 2006
References
1. Lonstein JE, Carlson JM. The prediction of curve progression in untreated
idiopathic scoliosis during growth. J Bone Joint Surg 1984;66-A:1061–71.
2. Duval-Beaupère G, Dubousset J, Queneau P. Pour une théorie unique de
l’évolution des scolioses. Presse Med 1970;78:1141– 6.
3. Little DG, Song KM, Katz D, et al. Relationship of peak height velocity to
other maturity indicators in idiopathic scoliosis in girls. J Bone Joint Surg
2000;82-A:685–93.
4. Diméglio A, Bonnel F. Le Rachis en Croissance. Paris, France: Springer;
1990.
5. Diméglio A. Growth in pediatric orthopaedics. J Pediatr Orthop 2001;21:
549 –55.
6. Tanner JM, Whitehouse RH. Clinical longitudinal standards for height,
weight, height velocity and weight velocity and the stages of puberty. Arch
Dis Child 1976;51:170 –9.
7. Greulich WW, Pyle SI. Radiographic Atlas of Skeletal Development of the
Hand and Wrist. 2nd ed. Stanford, CA: Stanford University Press; 1959.
8. Sauvegrain J, Nahm H, Bronstein N. Etude de la maturation osseuse du
coude. Ann Radiol 1962;5:542–50.
9. Diméglio A, Charles YP, Daures JP, et al. Accuracy of the Sauvegrain method
in determining skeletal age during puberty. J Bone Joint Surg 2005;87-A:
1689 –96.
10. Diméglio A. Growth in pediatric orthopaedics. In: Morrissy T, Weinstein SL,
eds. Lovell & Winter’s Pediatric Orthopaedics. Vol. 1. 5th ed. Philadelphia,
PA: Lippincott Williams & Wilkins; 2000:33– 62.
11. Risser JC. The iliac apophysis: An invaluable sign in the management of
scoliosis. Clin Orthop Relat Res 1958;11:111–9.
12. Cobb JR. The problem of the primary curve. J Bone Joint Surg 1960;42-A:
1413–25.
13. Lonstein JE. Idiopathic scoliosis. In: Lonstein JE, Winter RB, Bradford DS, et
al, eds. Moe’s Textbook of Scoliosis and Other Spinal Deformities. 3rd ed.
Philadelphia, PA: Saunders; 1995:219 –56.
14. King HA, Moe JH, Bradford DS, et al. The selection of fusion levels in
thoracic idiopathic scoliosis. J Bone Joint Surg 1983;65-A:1302–13.
15. Diméglio A. Pubertal peak, triradiate cartilage and apophysis of the great
trochanter. Paper presented at: Annual Meeting of the Pediatric Orthopaedic
Society of North America; May 1– 4, 2000; Vancouver, Canada.
16. Little DG, Sussman M. The Risser sign: A critical analysis. J Pediatr Orthop
1994;14:569 –75.
17. Shuren N, Kasser JR, Emans J, et al. Reevaluation of the use of the Risser sign
in idiopathic scoliosis. Spine 1992;17:359 – 61.
18. Mener G, Rigault P, Pouliquen JC, et al. Résultats du traitement orthopédique dans les scolioses de l’enfant de moins de 7 ans. A propos de 75 cas.
Rev Chir Orthop Reparatrice Appar Mot 1986;72:355– 66.
19. Morin C. Traitement des scolioses idiopathiques chez l’enfant en période de
croissance. Bull Acad Natl Med 1999;183:731–5.
20. Dobbs MB, Weinstein SL. Infantile and juvenile scoliosis. Orthop Clin North
Am 1999;30:331– 41.
21. Perdriolle R, Vidal J. Thoracic idiopathic scoliosis curve evolution and prognosis. Spine 1985;10:785–91.
22. Duval-Beaupère G. Mesure de gibbosité et d’angle couché comme facteur
pronostique des scolioses mineures. Acta Orthop Belg 1992;58(suppl 1):
26 –32.
23. James JIP. Idiopathic scoliosis, the prognosis, diagnosis and operative indications related to curve patterns and the age of onset. J Bone Joint Surg
1954;36-B:36 – 49.
24. Diméglio A. Growth of the spine before age 5 years. J Pediatr Orthop 1993;
1:102–7.
25. Masso PD, Meeropol E, Lennon E. Juvenile-onset scoliosis followed up to
adulthood: Orthopaedic and functional outcomes. J Pediatr Orthop 2002;
22:279 – 84.
26. Mannherz RE, Betz RR, Clancy M, et al. Juvenile idiopathic scoliosis followed to skeletal maturity. Spine 1988;13:1087–90.
27. Figueiredo UM, James JIP. Juvenile idiopathic scoliosis. J Bone Joint Surg
1981;63-B:61– 6.
28. Robinson CM, McMaster MJ. Juvenile idiopathic scoliosis. Curve patterns
and prognosis in one hundred and nine patients. J Bone Joint Surg 1996;
78-A:1140 – 8.
29. Tolo VT, Gillespie R. The characteristics of juvenile idiopathic scoliosis and
results of its treatment. J Bone Joint Surg 1978;60-B:181– 8.
30. Dubousset J, Herring JA, Shufflebarger H. The crankshaft phenomenon.
J Pediatr Orthop 1989;9:541–50.
31. Hefti FL, McMaster MJ. The effect of the adolescent growth spurt on early
posterior spinal fusion in infantile and juvenile idiopathic scoliosis. J Bone
Joint Surg 1983;65-B:247–54.
32. Sanders JO, Herring JA, Browne RH. Posterior arthrodesis and instrumentation in the immature (Risser grade-0) spine in idiopathic scoliosis. J Bone
Joint Surg 1995;77-A:39 – 45.
33. Shufflebarger HL, Clark CE. Prevention of the crankshaft phenomenon.
Spine 1991;16(suppl 8):409 –11.
34. Lapinsky AS, Richards BS. Preventing the crankshaft phenomenon by combining anterior fusion with posterior instrumentation. Does it work? Spine
1995;20:1392– 8.
35. Sanders JO, Little DG, Richards BS. Prediction of the crankshaft phenomenon by peak height velocity. Spine 1997;22:1352–7.
36. Roberto RF, Lonstein JE, Winter RB, et al. Curve progression in Risser stage
0 or 1 patients after posterior spinal fusion for idiopathic scoliosis. J Pediatr
Orthop 1997;17:718 –25.
37. Lenke LG. Anterior endoscopic discectomy and fusion for adolescent idiopathic scoliosis. Spine 2003;20(suppl 15):36 – 43.
38. Betz RR, D’Andrea LP, Mulcahey MJ, et al. Vertebral body stapling procedure for the treatment of scoliosis in the growing child. Clin Orthop Relat
Res 2005;434:55– 60.