Hey Group,
I thought this article might be of interest to parents whose children
have progressive scoliosis but HOPEFULLY will resolve with casting.
Hey, once the curve resolves one has to consider pain issues,
right ?! ( NEVER EVER SAY NEVER :) )
A 25-YEAR FOLLOW-UP
Of 42 patients with resolving infantile idiopathic scoliosis, 34 were
followed up for more than 25 years. Twenty had been primarily treated
in a plaster bed and 14 by physiotherapy. The mean angle of the curve
at presentation was 17 deg and at follow-up it was 5 deg. No patient
had significant progression of the scoliosis during the growth spurt.
When adults few had back pain or an increased disability score and
there was no interference with work or social activities. The rib-
vertebra angle difference proved to be an essential radiological sign
when distinguishing resolving from progressive infantile idiopathic
scoliosis. There was no advantage of plaster over physiotherapy with
regard to either the time to resolution or the functional outcome.
Treatment of resolving infantile idiopathic scoliosis in a plaster
bed is therefore now outdated.
J Bone Joint Surg [Br] 2002;84-B: 1030-5.
Received 13 November 2001; Accepted after revision 3 April 2002
Infantile idiopathic scoliosis is a lateral curvature of the spine
with apical rotation and wedging, which may present before the age of
three years.1-3 In the 1930s, Harrenstein4 recognised that the course
of infantile scoliosis can be very variable. Some curves progress and
cause cardiopulmonary pathology 3-17 while others resolve or
disappear spontaneously within the first years of life.4,7,12,18-21
In 1951, James1 described this separate resolving type of infantile
scoliosis. Two theories have been proposed to explain its cause,
intrauterine moulding and postnatal external pressure on the spine.
The observation that plagiocephaly, the convex side of a curve, and
hip dysplasia all occur on the same side is evidence for the pressure-
moulding theory.6,17,19 The postnatal pressure theory is supported by
those cases in which scoliosis is not evident at birth.3,20,22,23
Some authors1,13,24 have divided resolving scoliosis into two types,
resolving and postural. The cause of resolving infantile idiopathic
scoliosis, however, remains unknown.
The review by Nachmenson25 and other long-term follow-up studies
during the last 30 years have proposed different treatments for
patients with idiopathic scoliosis.11,15,24,26-29 In order to achieve
sufficient numbers of patients many of these studies had poorly-
defines patient selection.16 Early-onset idiopathic scoliosis and
resolving idiopathic scoliosis have a different natural history from
adolescent-onset scoliosis, often making the groups of patients which
were investigated inhomogeneous.15,16,24,26,27,29 Varying forms of
treatment within these groups and the paucity of information do not
allow conclusions to be drawn. As yet, only short- and medium-term or
minor long-term observations concerning resolving infantile scoliosis
are available.2,5-7,9,10,12,17,19,21,30,31 In contrast to progressive
infantile, juvenile and adolescent idiopathic scoliosis, little is
known about its natural history at and beyond skeletal maturity.
While there are guidelines concerning the treatment of progressive
idiopathic scoliosis which have been verified in many clinical
studies, there is not a standardised concept for the management of
resolving idiopathic scoliosis. Treatment may be by a plaster shell
in and overcorrected position or by physiotherapy, and some authors
claim that no specific treatment is required.3,5,12,19-21,32,33 The
differing forms of treatment which are recommended probably reflect
the fact that it is difficult to differentitate between resolving and
progressive curves before the age of five years.2,21
Our aim was to describe the long-term outcome in patients with
resolving infantile idiopathic scoliosis, and particulary to consider
whether there is recurrence or progression of the deformity in
adolescence. We have also compared the correction which may be
achieved by either a plaster shell in the overcorrected position or
by physiotherapy.
Patients and Methods
Our study was based on a retrospective review of 73 patients with
infantile idiopathic scoliosis, who attended special clinics between
1964 and 1974 in two university hospitals. All the children had
developed a curvature before the age of three years. The mean age at
diagnosis was 15 months (2 to 36) Since differentiation of resolving
from progressive curves in the first years of life is difficult, we
classified the curves retrospectively according to the radiological
course during the first six years of life. The focus of the study was
the long-term observation and management of resolving infantile
idiopathic scoliosis, which was diagnosed in 42 of these patients.
Thirty-four (15 women and 19 men) were followed up for at least 25
years. Three had been lost to follow-up and for five the radiographs
were incomplete. These were therefore excluded from the study. The
mean follow-up was 28 years (25 to 34). Table I gives the details of
the patients.
Radiological Evaluation
At the last follow-up, all 34 patients underwent radiological
evaluation (Fig. 1) with a standing anteroposterior (AP) view taken
on a 30 X 90 cm film. Observations and measurements were based on all
fulllength erect AP films of the spine made at the first visit and at
the final follow-up and any intervening radiographs. These were
interpreted and measurements made by one author (OD). The study
considered primary curves only. Lateral radiographs were often not
available and thus kyphosis, lordosis, and sacral inclination were
not studied. The Cobb34 angle was measured, and we compared the
course of short and long primary curves, setting the limit of
differentiation at nine spinal segments. We also determined the rib-
vertebra angle as described by Mehta,10 the rib-vertebra angle
difference (RVAD) at the apex of the primary curve and the
relationship between the rib head and the vertebral body (phase-I or
phase-II rib). In order to assess the predictive value of the RVAD in
the diagnosis of a resolving or progressive curve we retrospectively
measured it on the first available radiograph. Two of 31 patients
with progressive curves had incomplete radiographs and were excluded.
Methods of treatment. After the initial diagnosis of resolving
infantile scoliosis which was made at a mean age of seven months, 20
patients were treated in a plaster bed with the child supine and the
curve overcorrected in lateral flexion as was first described by
Harrenstein.4 It was retained day and night for at least 12 weeks. As
the children grew older or with improvement of the curve, it was used
only at night and physiotherapy was started.
Fourteen patients with infantile scoliosis were treated by
physiotherapy alone. Parents were advised to place the children in a
prone sleeping position. At that time the higher incidence of sudden
infant death syndrome in the prone position was not known.35
Back pain and work activity. The subjective clinical status was
assessed at final follow-up using two scores. We used a scheme
described by Weinstein et al15 which differentiates between the
frequency and incidence of spinal symptoms and whether
hospitalisation was required. The assessment of subjective disability
was based on the Oswestry Disability Index (ODI) version 1.0.(36)
Occupational activity was divided into three categories, heavy
(construction workers, farmers), medium (including household tasks)
and light work (office job).
Statistical Analysis
The independent t-test was used to analyse the differences in the
radiological data between the groups.
Results
Resolving scoliosis was diagnosed slightly more often in males (56%)
and was predominantly left-convex (56%) on the initial radiographs.
All 34 patients had single curves. The residual curve at final follow-
up was 10 deg in three (9%), with an overall mean of 5 deg (0 to 13).
None had significant progression of the scoliosis during the
adolescent growth spurt (Fig. 2). Treatment by surgery or with a
brace was not required. Although the curve was not clinically
apparent radiography had been performed in 16 patients (47%) during
adolescent growth.
Rib-vertebra angle difference (RVAD)
Figure 3 shows the RVAD. In cases in which it was
sensitivity was found to be 90% and specificity 85% in the diagnosis
of a progressive scoliosis. In all cases of resolving scoliosis the
RVAD decreased with skeletal growth. In none of these patients were
the ribs found to be in phase II. However, in 17 of the 29 patients
(59%) with a progressive curve there was an overlap of the head of
the apical rib on the convex side on the upper corner of the
corresponding vertebral body on the initial radiograph. In a further
seven progressive cases (24%) the rib-vertebra relationship converted
from phase I to phase II during the following six months.
Pattern of curves. Patients with resolving infantile idiopathic
scoliosis had either a C-type thoracic or a thoracolumbar single
curve at the initial presentation. In six infants flexion radiographs
taken within the first year after birth were available; five showed a
fixed lateral curve. In distinguishing between short and long C-
shaped curves we set the limit at nine spinal segments, measured from
the apical to the distal vertebral body of the major curve. Only four
patients had short major curves.
Degree of Initial Curve
The mean angle of the curve at presentation for those with resolving
scoliosis was 17 deg (5 to 36); 26 patients (77%) had a curve of 30
deg. At the final follow-up the mean decrease was 12 deg (-29 to 3).
In five of eight patients, in whom the initial curve was
there was a virtually complete (20 deg it was 4 deg (0 to 8).
Methods of treatment. The median treatment time for a resolving
deformity was three years (nine months to 15 years). Both groups
showed most improvement during the first four years of life (Fig. 2).
There was no significant difference in the time to resolution between
the two treatment groups. At the final follow-up the lateral curve
had decreased by a mean of 11 deg (-26 to +4; SD 8.0) in group I and
by a mean of 13 deg (-29 to 0; SD 9.9) in group II. There was no
significant difference between the groups with regard to either the
decrease in the curve during the period of observation or the
persistent curve at the time of followup.
There was conversion from single C-curves to double S curves in three
of the 20 children, who had been treated in a plaster bed (15%). In
group I we also found a change in the direction of the primary curve
with a slight permanent overcorrection in four of the patients (20%).
No similar complications were found in group II.
Back Pain And Work Activity
At the final follow-up 13 patients (38%) claimed not to have suffered
from back pain. One had been hospitalised and six (18%) had visited a
physician with back pain at some time (Table II). Figure 4 shows the
ODI for resolving curves at the final follow-up and compares it with
normative data of healthy control subjects and patients with
adolescent idiopathic scoliosis.37 There are no significant
differences between patients treated in a plaster bed (group I) and
those treated by physiotherapy (p = 0.46).
At the final follow-up only two patients (6%) were unemployed. None
had retired prematurely or was unable to work because of spinal
symptoms. Twelve patients (35%) assessed their occupation as light
work, 14 (41%) as medium strenuous and eight (24%) as heavy work.
Only two patients (6%) claimed that they were restricted in their
occupational or recreational activities because of spinal symptoms.
The occupations were not significantly different between the
treatment groups.
Discussion
The reported incidence of resolving infantile idiopathic scoliosis
varies between 17%(38) and 92%(19) of all patients with infantile
idiopathic scoliosis.5,7,10,17 The true incidence is clearly unknown
since many minor and resolving curves are not diagnosed;(21) 58% of
our patients had resolving curves.
There are few studies which deal with early-onset resolving scoliosis
from the time of diagnosis to skeletal maturity. Most authors observe
patients only until or just after the adolescent growth
phase.5,9,21,30,38 At the time of the final follow-up most of our
patients had a residual curve of
Radiological Findings
Rib-vertebra angle difference (RVAD). Our results underline previous
studies2,5,7,10,21 which have shown that the RVAD is an essential
radiological sign in distinguishing between resolving and progressive
curves. In all resolving cases the RVAD decreased during the period
of observation. In none of these patients were the ribs found to be
in phase II. This confirms the conclusions of Mehta10 that the
diagnosis of the rib head in phase II is a definitive sign of a
progressive scoliosis. When the rib head is in phase I, the RVAD is
even more important since it determines the prognosis. This
diagnostic tool has a marked influence on the treatment of infantile
scoliosis.
Pattern Of Curves
The most common pattern of resolving infantile idiopathic scoliosis
is the long single C-type lateral thoracic or thoracolumbar curve
with a slight rib hump on the convex side. In contrast to progressive
infantile and juvenile idiopathic scoliosis, double primary curves
are rare.10 Five of six patients who had lateral lexion radiographs
within the first year had a fixed lateral curve. Although
statistically not significant this suggests that resolving infantile
idiopathic scoliosis can be associated with structural changes which
may be identified radiologically.
Degree Of The Initial Curve
In our study most resolving infantile idiopathic curves had a Cobb
angle of 30 deg. The most pronounced curve which resolved was of 36
deg. Curves of 40 deg have been shown to resolve
spontaneously.7,10,21 The Cobb angle at the final follow-up was not
related to the severity of the curve at presentation.
Back Pain And Work Activity
According to the ODI patients with a resolving infantile idiopathic
scoliosis graded their disability as minimal. Few patients sought
advice for back pain and only one required hospitalisation. Almost
identical ODI values were found for the normal population in the meta-
analysis of Roland and Fairbank,37 who also showed that patients with
adolescent idiopathic scoliosis had slightly higher values. Patients
with a resolving infantile idiopathic scoliosis, when compared with
the healthy population, do not have increased back pain or disability
scores, and the variety of occupations pursued is not limited. The
treatment of resolving infantile idiopathic scoliosis had no effect
on disability at skeletal maturity.
Treatment
The treatment of early-onset resolving scoliosis by means of a
plaster bed which was common practice until recently, now seems
incomprehensible. Although it was accepted that infantile idiopathic
scoliosis may resolve spontaneously, the fear of missing the
progressive form led surgeons to advise this form of treatment for
all infantile idiopathic curves. More recently, numerous authors have
proposed that resolving idiopathic scoliosis should be treated by
early physiotherapy.3,5,12,21,23,33 We believe that once a resolving
idiopathic curve has been identified no further treatment is
required.
Critical Aspects
Although the importance of long-term reviews for a better
understanding of the natural history of spinal deformities is well
established,1,10,16,19,38 the design of retrospective studies remains
controversial. The inevitable initial selection of patients, flaws
resulting from irretrievable patients and insufficient documentation
justify some scepticism when interpreting the results. Our
observations may not represent the normal incidence found in an
unselected group of patients. However, there will be few prospective
studies with a follow-up of more than 25 years.