Abstract
Background: Scoliosis progression during adolescence is closely
related to patient maturity. Maturity has various indicators, including
chronological age, height and weight changes, and skeletal and sexual
maturation. It is not certain which of these indicators correlates most
strongly with scoliosis progression. The purpose of the present study was to
evaluate various maturity measurements and how they relate to scoliosis
progression.
Methods: Physically immature girls with idiopathic scoliosis were
evaluated every six months through their growth spurt with serial spinal
radiographs; hand skeletal ages; Oxford pelvic scores; Risser sign
determinations; height; weight; sexual staging; and serologic studies of the
levels of selected growth factors, estradiol, bone-specific alkaline
phosphatase, and osteocalcin. These measurements were then correlated with the
curve-acceleration phase.
Results: The period and pattern of curve acceleration began during
Risser stage 0 for all patients. Skeletal maturation scores derived with the
use of the Tanner-Whitehouse-III RUS method, particularly those for the
metacarpals and phalanges, were superior to all other indicators of maturity.
Regression of the scores provided good estimates of maturity relative to the
period of curve progression (Pearson r = 0.93). The initiation of this period
occurred simultaneously with digital changes from Tanner-Whitehouse-III stage
F to G. At this stage, curves also separated into rapid, moderate, and
low-acceleration patterns, with specific curve types in the rapid and
moderate-acceleration groups. The low-acceleration group was not confined to a
specific curve type.
Conclusions: The curve-acceleration phase separates curves into
various types of curve progression. The Tanner-Whitehouse-III RUS scores are
highly correlated with timing relative to the curve-acceleration phase and
provide better maturity determination and prognosis determination during
adolescence than the other parameters tested. Accurate skeletal maturity
determination should be used as the primary maturity measurement in girls with
idiopathic scoliosis.
Level of Evidence: Prognostic Level I. See Instructions
to Authors for a complete description of levels of evidence.
Scoliosis progression during adolescence is closely related to patient
maturity. The most commonly used maturity indicators in scoliosis are
chronological age, the Risser sign, and the age of menarche, each of which has
substantial variability relative to scoliosis progression. Maturity is
multidimensional and has various components, including chronological age,
height and weight changes, and skeletal and sexual maturation. Ideally, a
maturity measurement for scoliosis should be readily available to the
orthopaedic surgeon and it should correlate significantly with scoliosis
progression. The purpose of the present study was to evaluate several maturity
measurements for female children and adolescents with scoliosis. The maturity
indicators selected for analysis included height, weight, secondary sexual
characteristics, skeletal maturation, and chemical markers of growth, hormonal
development, and bone formation and resorption.
In this prospective, institutional review board-approved study, physically
immature girls with idiopathic scoliosis were prospectively recruited on a
sequential basis and participated following informed consent. Eligibility
criteria included female gender, premenarchal status, an age of eight to
twelve years, a diagnosis of idiopathic scoliosis, no evidence of neurological
abnormality, no previous spinal fusion, no abnormalities of maturation or
height (such as a lower extremity growth arrest or deficiency), no
abnormalities of the head or neck that would change height measurements, no
skeletal dysplasia or dwarfism, a Risser sign of 0 or 1, and a Tanner stage of
=3. All patients who were diagnosed under the age of eleven years had a
magnetic resonance imaging evaluation of the spine that revealed normal
findings except for the scoliosis. Bracing was used for eighteen patients and
was instituted according to generally accepted guidelines. Because further
curve progression could not be measured if a patient underwent surgery, any
patient who had surgery for the treatment of scoliosis (six patients as of the
time of study completion) was considered to have completed the study. The
girls were assessed every six months with the evaluation of standing and
sitting heights, arm span, weight, hand and spine radiographs, Tanner stage,
and serum chemistry studies, all of which are described in the following
sections.
Peak Height, Arm Span, and Weight Velocity Timing
Height was measured in centimeters according to the method of
Tanner1 on a
stadiometer with shoes and socks off. The child was encouraged to stretch out
to the maximum height with the application of gentle upward pressure under the
mastoid process and with the recorder watching to see that the heels were not
off the ground. Sitting height was similarly determined with use of a rigid
wooden stool, which was sufficiently high that the feet could not reach the
floor. Arm span was measured in centimeters from the ends of the middle finger
of one hand to the other with the arms maximally outstretched. Weight was
recorded in kilograms. The height velocity, sitting height velocity, arm span
velocity, and weight velocity were based on the number of centimeters of
growth or kilograms of weight increase every six months. Each height was
measured three times at each visit. The serial measurements were plotted, and
the age at the time of the peak height velocity was estimated from the
data2.
Skeletal Age
Skeletal age was determined on the basis of a posteroanterior radiograph of
the left hand and wrist. The radiographs were interpreted with use of the
Greulich and Pyle
atlas3 and the
Tanner and Whitehouse (Tanner-Whitehouse-III)
method4. The
Tanner-Whitehouse-III method involves specifically scoring the individual
bones with a letter that is correlated with a specific weighted score. The
scores are separate for the carpal bones (CARP) and the radius, ulna, and
small bones of the hand (RUS). The small bones consist of the metacar-pals and
phalanges. The modified Oxford score was evaluated on the basis of the visible
pelvic and proximal femoral ossification
centers5,6.
Individual markers for the Tanner-Whitehouse method and the modified Oxford
method were recorded.
Secondary Sexual Characteristics
Secondary sexual characteristic maturity was measured with use of the
method of Tanner7 by
a registered nurse (S.A.M.) who had been trained by a pediatric
endocrinologist (D.N.F.) in the accurate evaluation of secondary sexual
characteristic maturity. Menarche month and year were recorded.
Chemical Determination
Serum insulin-like growth factor-1 (IGF-1), insulin-like growth factor
binding protein-3 (IGFBP-3), dehydroepiandrosterone sulfate (DHEA-S),
estradiol, and osteocalcin levels were measured with use of enzyme-linked
immunosorbent assay kits (DSL-10-2800, DSL-10-6600, DSL-10-3500D, DSL-10-4300,
and DSL-10-7600; Diagnostic Systems Laboratories, Webster, Texas).
Bone-specific alkaline phosphatase was measured with the EIA kit (product
number 8012; Metra Biosystems, a division of Quidel, San Diego,
California).
Spinal Radiographs
Spinal radiographs consisted of standing posteroanterior radiographs, made
with the patient out of the brace, on a 36 × 91-cm (14 × 36-in)
cassette at a 2-m tube distance. The radiographs included the pelvis within
the field.
Curve Patterns
Because the system of Lenke et
al.8 is designed for
preoperative curve classification, a modified Lenke classification was
necessary for this mixture of curves in patients who progressed to surgery and
patients who did not progress to surgery and, therefore, did not have bending
radiographs. The modified Lenke system that was used in the present study
included six types of curves: type 1 (thoracic), type 2 (double thoracic),
type 3 (double major with the thoracic curve being larger than the
thoracolumbar or lumbar curve), type 4 (triple major), type 5 (thoracolumbar
or lumbar), and type 6 (double major with thoracolumbar or lumbar curve being
greater than the thoracic curve).
Determination of the Curve Acceleration Phase
The magnitude of the Cobb angles of the major curves was plotted over time.
It was quite obvious that the increasing curves had similar periods of rapid
acceleration, although at differing chronological ages. This acceleration was
termed the curve acceleration phase (CAP) and was determined by a change in
the rate of increase in curve size. For example, a curve accelerated if it
went from a 0.1° increase per month to a 1.0° increase per month.
Times before or after the curve acceleration phase were designated according
to the months before (negative) or after (positive) the curve acceleration
phase (e.g., as CAP —6 for six months before the initiation of the curve
acceleration phase and as CAP +12 for twelve months after the initiation of
the curve acceleration phase). Each of the various maturity measurements was
then compared with the curve acceleration phase. The average ages for various
maturity measurements was determined for the entire sample. Timing differences
for the maturity measurements were compared for individuals rather than for
group averages because of the individual variability.
Statistical Methods
Pearson correlation coefficients were computed for the curve acceleration
phase against each of the maturity measurement items. Linear regression was
used to fit the relationship between the RUS scores and the curve acceleration
phase. The program that was used was the PROC REG program in the SAS package
(version 9.1; SAS Institute, Cary, North Carolina). Curve-fitting was done
with use of a curve-fitting program (TableCurve 2-D, version 5.01; SPSS,
Chicago, Illinois).
Twenty-four girls were enrolled. Two patients were lost to follow-up very
early in the study and were not included in the data analysis. Another patient
was lost to follow-up after two years but had data of satisfactory quality to
be included in the analysis. Therefore, twenty-two patients were included in
the analysis. The average curve size was 23° (range, 11° to 43°)
at the time of enrollment and 45° (range, 13° to 82°) at the time
of the most recent evaluation.
Curve Patterns
Four curves changed patterns during the study: two curves changed from type
1 to type 3, one changed from type 2 to type 4, and one changed from type 5 to
type 6. At the completion of the study, there were eight type-1 curves, two
type-2 curves, four type-3 curves, one type-4 curve, five type-5 curves, and
two type-6 curves.
Curve Acceleration Phase
The curves demonstrated a marked change in behavior at the curve
acceleration phase (CAP 0) (Figs. 1-A and
1-B). There were three patterns of curve acceleration, most easily
characterized as low, moderate, and rapid. The average curve rate increase for
all three groups was 0.2° per month prior to the curve acceleration phase
and was not significantly different among the three groups, being 0.1°,
0.3°, and 0.1° per month for the low, moderate, and rapid-acceleration
groups, respectively. The average curve increase rate for all of the curves
was 0.9° per month following the curve acceleration phase, being 0.3°,
0.8°, and 1.6° per month in the low, moderate, and rapid-acceleration
groups, respectively. The curve patterns in the rapid-acceleration group
(seven patients) were all single thoracic curves or double-major curves with
the thoracic curve being larger than the lumbar or thoracolumbar curve (Lenke
type 1 or 3). All of the curves in the rapid-acceleration group had a final
magnitude >60°, except for one patient. This patient was lost to
follow-up at two years and had a 51° rapidly accelerating curve at CAP
+12. The curve patterns in the moderate-acceleration group (seven patients)
consisted entirely of double-thoracic, triple-major, thoracolumbar, and
double-major curves with the thoracolumbar or lumbar curve being larger than
the thoracic curve (Lenke types 2, 4, 5, and
6)8. These curves
also continued to worsen later in maturity, with the final curves ranging from
44° to 59°. The curves in the low-acceleration group (eight patients)
did not show a predilection for any curve type, and the final curve magnitudes
ranged from 13° to 35°.
Correlation of the Actual Curve Acceleration Phase with the Maturity
Measurements
Table I shows the
correlation of the various maturity measurements with the actual curve
acceleration phase.
Chronological Age
The average age at the time of the curve acceleration phase (CAP 0) was
11.7 years. The correlation of the curve acceleration phase with chronological
age was good (r = 0.89), but the standard deviation was also relatively high
(ten months).
Peak Height Velocity Timing
Fourteen girls had identifiable growth peaks. Because the peaks could only
be identified retrospectively, it was expected that some of the recruited
patients would be beyond their peak growth. The average peak height velocity
was 10.5 ± 1.8 cm/yr across the range of ages (mean, 11.7 ± 0.9
years; range, 9.6 to 12.5 years).
Secondary Sexual Characteristics
The Tanner scores for both breast and pubic hair correlated well (r = 0.8)
with the curve acceleration phase, but the Tanner scores at the time of the
curve acceleration phase (CAP 0) ranged from 1 to 3 for breast and from 1 to 4
for pubic hair. The average menarche occurred at 13 ± 6 months (range,
three to twenty-seven months) after the curve acceleration phase (CAP 0).
Serum Chemistry Studies
IGF-1, IGFBP-3, and DHEA-S levels were significantly correlated with the
curve acceleration phase (p = 0.001), but these correlations were markedly
lower than those of the other maturity markers. Estradiol and osteocalcin
levels, while showing changes with maturation, were poorly correlated with the
curve acceleration phase stage. Bone-specific alkaline phosphatase peaked at
the time of the curve acceleration phase (CAP 0), but the range was too broad
to make it clinically useful (mean and standard deviation, 137 ± 29
U/L; range, 94 to 202 U/L).
Skeletal Maturity
All four methods of determining skeletal maturation (the
Tanner-Whitehouse-III method, the Greulich and Pyle method, the Oxford method,
and the Risser sign) correlated significantly with the curve acceleration
phase (p < 0.001). The weakest of these was the Risser sign. The curve
acceleration phase began during Risser 0 for all patients. The triradiate
cartilage was open in all but one patient at the time of the curve
acceleration phase (CAP 0). The Oxford method correlated well, but it was
skewed by the strength of the triradiate cartilage stage.
The Tanner-Whitehouse-III RUS method scores strongly correlated with the
curve acceleration phase (r2 = 0.87, r = 0.93, p < 0.001). The
grouping at CAP 0 was much tighter in association with the
Tanner-Whitehouse-III method than with the Greulich and Pyle method. Of the
individual bone components of the RUS score, the radius and ulna had the
lowest correlation with the curve acceleration phase. Therefore, we tested the
RUS system without the radius and ulna and termed this the digital skeletal
age (DSA). The correlation coefficient was also 0.93, but with a tighter fit
at the critical stage from CAP 0 to CAP +6 (with standard deviations of 14.84
and 5.13 at CAP 0 and CAP + 6, respectively, for the DSA score, compared with
43.93 and 22.96, respectively, for the RUS score). The relationships between
the DSA scores and curve magnitude demonstrated a marked curve acceleration
between DSA scores of 400 and 425 (Fig.
2). With use of the curve-fitting program, a curve relating the
DSA score to the curve acceleration phase was generated
(Fig. 3). The generated curve
equation is described in the Appendix. Radiographically, the timing of the
curve acceleration phase (CAP 0) corresponded with the change from a covered
(Tanner-Whitehouse-III stage F) to a capped (Tanner-Whitehouse-III stage G)
phalangeal epiphysis (Fig. 4).
The Tanner-Whitehouse-III carpal scores reached maturity about the time of the
curve acceleration phase (CAP 0) and were not as highly correlated as the RUS
scores with the curve acceleration phase.
Correlations with the Estimated Curve Acceleration Phase
Table II demonstrates the
relationships of estimated curve acceleration phase values (determined by
regression from the RUS scores) with other well recognized maturity
measurements. For type 1 and 3 curves, there appears to be a threshold of
30° at CAP +6 between the low and rapid acceleration groups. For the other
curve types, the threshold at CAP +6 between the moderate and low acceleration
groups is closer to 20°.
Skeletal Maturity
While the stage of ossification of the iliac apophysis (the Risser sign) is
the most commonly used method of determining skeletal maturation in patients
with idiopathic scoliosis, it is the skeletal maturity measurement that is
most poorly correlated with the curve acceleration phase. The iliac apophysis
does not begin to show ossification (positive Risser sign) until an average of
eighteen months after the curve acceleration phase (CAP +18). This means that
most curve progression has occurred well before Risser stage 1 is evident. A
second problem with the Risser sign, identified by
Izumi9, is that the
appearance of the iliac apophysis on posteroanterior radiographs (which are
used to decrease breast irradiation) does not correlate well with that on
anteroposterior radiographs because of radiographic parallax of the x-ray beam
with the pelvic brim. The Risser sign is simple, is readily available, and
provides a general measurement of maturity, but it should not be used as a
primary method of maturity determination when more specific measurement of
skeletal maturity is required. The Oxford method uses other pelvic markers to
measure skeletal maturity but, since a hand radiograph provides more accurate
and useful information, routinely including the pelvis on spine radiographs
used for the evaluation of idiopathic scoliosis unnecessarily increases
radiation to the gonads.
The Greulich and Pyle atlas is widely used for assessing bone age and
skeletal maturity. The reference atlas is unfortunately quite widely spaced in
time during the critical CAP —6 to +6 period. Since few physicians have
access to the Tanner-Whitehouse-III manual, a simple way for clinicians to
assess maturity compared to the curve-acceleration phase is by using the
Greulich and Pyle atlas. Skeletal age 10 in the atlas corresponds to the
time-period before curve acceleration while skeletal age 11 demonstrates the
capping associated with rapid curve acceleration (CAP 0 to CAP +6). On the
basis of our data, the wrist scores (Tanner-Whitehouse-III CARP scores) are
not helpful relative to the curve acceleration phase. Therefore, when the
Greulich and Pyle atlas is used for the assessment of skeletal maturity in
patients with scoliosis, the orthopaedic surgeon should concentrate on the
phalangeal and metacarpal indicators.
Recently, Dimeglio et
al.10 described the
reliability of elbow secondary ossification centers as a maturity indicator
during the adolescent growth spurt. As we did not have this information for
the patients in the present study, we cannot comment on its utility, but this
method has the potential to reflect curve progression.
The Tanner-Whitehouse-III RUS scores strongly predicted the curve
acceleration phase. Figure 4
illustrates the stages surrounding the curve acceleration phase for the
digits. The Tanner-Whitehouse method is reasonably reliable, with trained
readers having maximum single-bone differences of one grade only 21% of the
time, with half of those differences being overestimated and half being
underestimated4. The
tight grouping of the DSA scores illustrates that the appearance of the
metacarpals and phalanges accurately reflects the curve acceleration phase.
Unfortunately, the method is fairly complex and time-consuming, and we are
currently investigating the reliability of several simplified modifications.
It may be that the Sauvegrain
method10 of using
the elbow or a simplified DSA measurement will provide an easier, but
accurate, determination of the curve-acceleration-phase stage.
While the current study was limited to girls with idiopathic scoliosis, the
study by
Duval-Beaupere11
demonstrated that curves progress during this period regardless of gender or
etiology. We strongly suspect that the stage of skeletal maturity will also be
a strong predictor of curve behavior in boys and in patients with other types
of scoliosis.
Peak Height Velocity Timing
The timing of the peak height growth
velocity11-21,
which is identified through serial height measurements and growth velocity
calculations, is highly prognostic in patients with idiopathic scoliosis. The
curve acceleration phase was correlated with the timing of the peak height
velocity in the fourteen girls with identifiable growth peaks. The correlation
was high but quite variable (standard deviation, nine months). We suspect that
this is an inherent problem in measuring serial heights and computing height
velocities. Despite accurate stadiometric measurements obtained by trained
personnel, height measurements, particularly in patients with a skeletal
shortening deformity, have substantial variability. There was also marked
variability in sitting heights and arm spans. It may be that peak height
velocity timing correlates more closely with the curve acceleration phase than
we were able to determine, but the variability indicates that it is a
difficult maturity measurement at best. Using the peak height velocity to
determine treatment is also difficult because it is only evident
retrospectively rather than concurrently with clinical decision-making.
Because the Tanner-Whitehouse-III skeletal maturity measurements correlated so
highly with curve progression, we recommend that peak height velocity timing
not be used as the primary method of determining maturity in patients with
scoliosis.
It was interesting that the curve acceleration phase correlated not with
the onset of the growth spurt but with the timing of the peak height velocity.
This finding may support the argument that instability resulting from rapid
growth22,23
is one important mechanism of scoliosis progression.
Serum Chemistry Studies
The specific serologic variables were selected because of their
relationship with maturation. The known relationship of each variable to the
adolescent growth spurt was described in our previous
work24. Briefly,
both growth hormone and the sex steroids, particularly estrogen, are necessary
for the normal adolescent growth spurt, which is the case for both boys and
girls. In girls, sexual maturity occurs in an orderly manner with adrenarche,
thelarche, and menarche. Adrenarche occurs about three years before the peak
height velocity and is reflected in the levels of circulating adrenal
hormones25,
including dehydroepiandrosterone sulfate (DHEA-S), whose elevation precedes
the peak height velocity and is well correlated with pubertal
stage26,27.
The gonadotropins luteinizing hormone and follicle stimulating hormone are too
pulsatile for clinical
use28, but
estradiol is not as variable and its levels correlate with pubertal
stage27,29-34.
Estrogen also appears to be the critical stimulus of the physis causing the
linear growth
spurt32,33,35-38.
Lower doses of estrogen, as during early puberty, stimulate growth, whereas
higher doses, as in later puberty, lead to growth cessation. Growth hormone
itself is too pulsatile to be useful in this setting, but its activity is
reflected in insulin-like growth factor-1 (IGF-1) levels, which are well
correlated with pubertal
stage30,39-45.
The IGF binding protein-3 (IGFBP-3) and the ratio of IGF-1 to IGFBP-3 are also
closely correlated with height
maturity42. With
rapid bone growth during the growth spurt, many bone markers have enormous
changes. Those that are most closely associated with pubertal stage include
bone-specific alkaline phosphatase and osteocalcin. Bone-specific alkaline
phosphatase secretion reflects bone formation, closely follows height
velocity29,46-51,
and has been proposed as a useful marker for scoliosis
progression48,49.
Osteocalcin also closely reflects the adolescent growth
spurt29,52-54.
In the present study, IGF-1 had the best correlation with the curve
acceleration phase of any of the serum chemistry variables but was not nearly
as good as other easier and less expensive markers. In a previous
study24, we found
that IGF-1 and estradiol were potentially good markers with which to identify
whether a patient was before or after the peak height velocity. However, when
correlated with the curve acceleration phase rather than the peak height
velocity, the results did not hold. We do not recommend any of these serum
markers for determining scoliosis maturity and prognosis.
Use of the Curve Acceleration Phase
Scoliosis progression occurs in four basic phases: infantile-rapid,
juvenile-slow, adolescent-rapid, and mature-slow. The present study focused on
changes that occur between the juvenile and mature phases. Prior to the curve
acceleration phase (i.e., prior to CAP 0), during the juvenile and
preadolescent slow phases, curves are at little risk of progression. As the
phalanges become capped as seen on the hand radiograph, the patient enters the
adolescent phase and curve progression accelerates. The curve acceleration
phase acts much like a prism separating curves into rapid, moderate, and low
rates of progression. Figure 5
illustrates the timing of the various maturity measurements and curve
progression relative to the curve acceleration phase. Adolescent curve
progression typically lasts about two to three years beyond the initiation of
the curve acceleration phase when girls are at CAP +24, which corresponds
approximately with Risser stage 4 and with <2 cm of growth remaining. There
is about a year of slow growth remaining at CAP +24, which we chose to term
the early mature phase. We did not attempt to measure curve progression in
this latter phase or beyond.
This leads to the obvious questions of why certain curves, specifically,
thoracic curves and double-major curves with a larger thoracic component
(Lenke type-1 and 3 curves), progress so rapidly at this stage; why double
thoracic curves, triple-major curves, thoracolumbar and lumbar curves, and
double-major curves with longer thoracolumbar or lumbar components (Lenke
type-2, 4, 5, and 6 curves) progress at an intermediate rate with later
progression; and why some curves do not progress substantially. We did observe
later progression in the moderate rate group, but this may not be real and may
simply reflect that the more severe, rapidly progressing curves underwent
surgery at earlier maturity levels. Another question is why the stage of
skeletal maturity correlates so highly with the curve acceleration phase,
particularly at CAP 0 to +6. Does this occur because skeletal maturation is a
better reflector of longitudinal growth with the spine becoming unstable from
rapid growth leading to curve progression, as suggested by Goldberg et
al.22,23?
Is it because the vertebral apophysis and growth plate is more susceptible to
Heuter-Volkmann compressive growth inhibition at this stage, or is some other
mechanism at work? Interestingly, two of the patients with the most rapid, and
ultimately the worst, curve progression had prolonged skeletal immaturity
during the curve acceleration phase but the sample size was too small to
determine whether this observation is a common cause of more severe curve
progression.
Use of the Estimated Curve Acceleration Phase
The actual curve acceleration phase is subject to the same limitation as
the peak height velocity in that it is only available retrospectively. The
estimated curve acceleration phase (eCAP) allows accurate maturity
determination during the critical period before and during early rapid curve
progression. It is available before any of the other common maturity markers.
The estimated curve acceleration phase can be calculated either with use of
the regression equation or from Figure
3. Unfortunately, the Tanner-Whitehouse-III method is fairly
complex. For a simple estimate, Table
II describes the appearance of the skeletal age radiograph at the
various stages. As the table also demonstrates, the various stages correspond
with other commonly considered maturity indicators. Unlike these other
indicators, the estimated curve acceleration phase does not require
determination of secondary sexual characteristics, serial height measurements,
or radiographs that include the gonads in the field. As such, this measurement
should provide a useful marker for additional studies of scoliosis prognosis,
bracing efficacy, and the crankshaft phenomenon and can assist physicians in
identifying the period of maximum risk of curve progression.
Treatment Implications
The present study has a number of treatment implications. The first relates
to the efficacy of using braces for the treatment of scoliosis. To our
knowledge, all bracing studies to date have involved the use of the Risser
sign, chronological age, and menarche status to assess maturity, although none
of these maturity measurements correlate with curve progression nearly as well
as skeletal maturity measurements do. Because curve progression is determined
during Risser stage 0, more meaningful bracing evaluation will require
skeletal maturity stratification. Since various curves progress in different
fashions, it will also be important to stratify a curve's progression by its
rapidity during the curve acceleration phase. A Risser-stage-0, premenarchal
girl with a 30° thoracic curve may be at CAP 0 in the rapid-rate group or
at CAP +12 in the low-rate group. On the basis of the findings of the present
study, the prognosis differs for each of these patients: the first patient is
less likely to be managed successfully with bracing, and the second may not
require bracing. The numbers within the individual groups were too small to
allow us to make strong statements, but type-1 and type-3 curves of
=30° in patients at eCAP +6 all progressed markedly, whereas lesser
curves did not. Similarly, type-2, 4, 5, and 6 curves of >20°
progressed moderately, whereas lesser curves did not. Larger studies are
necessary to test these findings because it is possible that the bracing used
in the present study modified either the timing of the curve acceleration
phase or the behavior of the curves. However, because bracing is considered to
be the current standard of care, it was not considered ethical to withhold
treatment for this study. Whether the same results will hold for patients who
are not managed with bracing is unknown.
In summary, the curve acceleration phase (CAP) is a time of rapid change in
curve magnitude, with different behavior for different curve patterns. The
curve acceleration phase begins during a limited range of skeletal maturity,
and the curve progression pattern is quickly established as low, moderate, or
rapid. The present study identifies a strong correlation between skeletal
maturation according to the Tanner-Whitehouse-III method (particularly that of
the metacarpals and phalanges as reflected by the DSA score) and the curve
acceleration phase. Available before Risser stage 0, triradiate cartilage
closure, the peak height velocity, and stage-2 Tanner sexual characteristics
staging, the DSA score appears to provide a more predictive maturity
determination than the Risser sign, chronological age, serial height
determinations, secondary sexual characteristics, and several serum markers
that reflect various aspects of maturation. This method provides a useful
means of determining the period of maximum curve progression risk and appears
to be useful for determining the prognosis of differing curve types.
The equation that was generated for the relationship between the curve
acceleration phase (CAP) and the digital skeletal age (DSA) was
CAP=a+bDSA-cd|DSA-c|d-e
where a = 0.3479, b = 24.91, c = 406.04, d = 185.3, e = 0.4473 and
r2 = 0.90 (r = 0.95). ?
Tanner JM. Some main features of normal
growth in children. In: Zorab PA, editor. Scoliosis and growth:
proceedings of a third symposium held at the Institute of Diseases of the
Chest, Brompton Hospital, London, on 13th November, 1970. Edinburgh:
Churchill Livingstone; 1971. p
14-28.14
1971
Tanner JM, Davies PS. Clinical
longitudinal standards for height and height velocity for North American
children. J Pediatr.
1985;107:
317-29.107317
1985
[PubMed][CrossRef]
Gruelich WW, Pyle SI.
Radiographic atlas of skeletal development of the hand and wrist. 2nd
ed. Stanford, CA: Stanford University Press;
1959.
1959
Tanner JM. Assessment of skeletal
maturity and prediction of adult height (TW3 method). 3rd ed. London:
WB Saunders; 2001.
2001
Sanders JO, Herring JA, Browne RH.
Posterior arthrodesis and instrumentation in the immature (Risser-grade-0)
spine in idiopathic scoliosis. J Bone Joint Surg Am.
1995;77:
39-45.7739
1995
[PubMed]
Acheson RM. The Oxford method of
assessing skeletal maturity. Clin Orthop.
1957;10:
19-39.1019
1957
[PubMed]
Tanner JM. Growth and endocrinology of
the adolescent. In: Gardner LI, editor. Endocrine and genetic diseases
of childhood. 2nd ed. Philadelphia: WB Saunders; 1974. p
14.14
1974
Lenke LG, Betz RR, Harms J, Bridwell KH,
Clements DH, Lowe TG, Blanke K. Adolescent idiopathic scoliosis: a new
classification to determine extent of spinal arthrodesis. J Bone Joint
Surg Am. 2001;83:
1169-81.831169
2001
Izumi Y. The accuracy of Risser staging.
Spine. 1995;20:
1868-71.201868
1995
[PubMed][CrossRef]
Dimeglio A, Charles YP, Daures JP, de
Rosa V, Kabore B. Accuracy of the Sauvegrain method in determining skeletal
age during puberty. J Bone Joint Surg Am.
2005;87:
1689-96.871689
2005
[PubMed][CrossRef]
Duval-Beaupere G. Pathogenic
relationship between scoliosis and growth. In: Zorab PA, editor.
Scoliosis and growth: proceedings of a third symposium held at the
Institute of Diseases of the Chest, Brompton Hospital, London, on 13th
November, 1970. Edinburgh: Churchill Livingstone; 1971. p
58-64.58
1971
Burwell RG. The relationship between
scoliosis and growth. In: Zorab PA, editor. Scoliosis and growth:
proceedings of a third symposium held at the Institute of Diseases of the
Chest, Brompton Hospital, London, on 13th November, 1970. Edinburgh:
Churchill Livingstone; 1971. p
131-50.131
1971
Calvo IJ. Observations on the growth of
the female adolescent spine and its relation to scoliosis. Clin
Orthop. 1957;10:
40-7.1040
1957
Dimeglio A, Bonnel F. Le rachis
en croissance. Paris: Springer; 1990.
1990
Duval-Beaupere G. Les reperes de
maturation dans la surveillance des scolioses. [Maturation indices in the
surveillance of scoliosis]. Rev Chir Reparatrice Appar Mot.
1970;56:
59-76.5659
1970
Heine J, Reher H. Die progredienz der
unbehandelten skoliose nach poliomyelitis bis wachstumsabschluB. [Progress of
untreated scoliosis following poliomyelitis up to the end of the growth-period
(author's transl)]. Z Orthop Ihre Grenzgeb1975;113:
876-9.113876
1975
[PubMed]
Hopf C, Sandt E, Heine J. Die
progredienz unbehandelter idiopathischer skoiosen im rontgenbild. [The
progression of untreated idiopathic scoliosis in the xray image].
Rofo. 1989;151:
311-6.151311
1989
[PubMed]
Little DG, Song KM, Katz D, Herring JA.
Relationship of peak height velocity to other maturity indicators in
idiopathic scoliosis in girls. J Bone Joint Surg Am.
2000;82:
685-93.82685
2000
[PubMed]
Song KM, Little DG. Peak height velocity
as a maturity indicator for males with idiopathic scoliosis. J Pediatr
Orthop. 2000;20:
286-8.20286
2000
[CrossRef]
Loncar-Dusek M, Pecina M, Prebeg Z. A
longitudinal study of growth velocity and development of secondary gender
characteristics versus onset of idiopathic scoliosis. Clin Orthop Relat
Res. 1991;270:
278-82.270278
1991
Risser JC. The Iliac apophysis; an
invaluable sign in the management of scoliosis. Clin Orthop.
1958;11:
111-9.11111
1958
[PubMed]
Goldberg CJ, Fogarty EE, Moore DP,
Dowling FE. Scoliosis and developmental theory: adolescent idiopathic
scoliosis. Spine. 1997;22:
2228-38.222228
1997
[PubMed][CrossRef]
Goldberg CJ, Dowling FE, Fogarty EE,
Moore DP. Adolescent idiopathic scoliosis as developmental instability.
Genetica. 1995;96:
247-55.96247
1995
[PubMed][CrossRef]
Sanders JO, Browne RH, Cooney TE,
Finegold DN, McConnell SJ, Margraf SA. Correlates of the peak height velocity
(PHV) in girls with idiopathic scoliosis. Spine.
2006;31:
2289-95.312289
2006
[PubMed][CrossRef]
Rotteveel J, de Ridder C, Schoute E,
Delemarre-van de Waal HA. Androstenedione, dehydroepiandrosterone sulfate, and
estradiol levels throughout female puberty: relation to height velocity.
Horm Res. 1997;48:
263-7.48263
1997
[PubMed][CrossRef]
Lashansky G, Saenger P, Fishman K,
Gautier T, Mayes D, Berg G, Di Martino-Nardi J, Reiter E. Normative data for
adrenal steroidogenesis in a healthy pediatric population: age- and
sex-related changes after adrenocorticotropin stimulation. J Clin
Endocrinol Metab. 1991;73:
674-86.73674
1991
[CrossRef]
Reiter EO, Fuldauer VG, Root AW.
Secretin of the adrenal androgen, dehydroepiandrosterone sulfate, during
normal infancy, childhood, and adolescence, in sick infants, and in children
with endocrinologic abnormalities. J Pediatr.
1977;90:
766-70.90766
1977
[PubMed][CrossRef]
Kasa-Vubu JZ, Padmanabhan V, Kletter GB,
Brown MB, Reiter EO, Sizonenko PC, Beitins IZ. Serum bioactive luteinizing and
follicle-stimulating hormone concentrations in girls increase during puberty.
Pediatr Res. 1993;34:
829-33.34829
1993
[PubMed]
Blumsohn A, Hannon RA, Wrate R, Barton
J, al-Dehaimi AW, Colwell A, Eastell R. Biochemical markers of bone turnover
in girls during puberty. Clin Endocrinol (Oxf).
1994;40:
663-70.40663
1994
[PubMed][CrossRef]
Grumbach MM. Onset of puberty. In:
Berenberg SR, editor. Puberty, biologic and social components.
Leiden: HE Stenfert Kroese; 1975. p
1-21.1
1975
Grumbach MM. The neuroendocrinology of
human puberty revisited. Horm Res.
2002;57 Suppl 2:
2-14.572
2002
[PubMed][CrossRef]
Grumbach MM. Estrogen, bone, growth and
sex: a sea change in conventional wisdom. J Pediatr Endocrinol
Metab. 2000;13 Suppl 6:
1439-55.131439
2000
Grumbach MM. The role of estrogen in the
male and female: evidence from mutations in synthesis and action. Horm
Res. 2000;53 Suppl 3:
23-4.5323
2000
[CrossRef]
Reiter EO. Neuroendocrine control
processes. Pubertal onset and progression. J Adolesc Health
Care. 1987;8:
479-91.8479
1987
[CrossRef]
Nilsson O, Chrysis D, Pajulo O, Boman A,
Holst M, Rubinstein J, Martin Ritzen E, Savendahl L. Localization of estrogen
receptors-alpha and -beta and androgen receptor in the human growth plate at
different pubertal stages. J Endocrinol.
2003;177:
319-26.177319
2003
[PubMed][CrossRef]
Cutler GB Jr. The role of estrogen in
bone growth and maturation during childhood and adolescence. J Steroid
Biochem Mol Biol. 1997;61:
141-4.61141
1997
[CrossRef]
Styne DM. The regulation of pubertal
growth. Horm Res.
2003;60(Suppl 1):
22-6.6022
2003
[PubMed][CrossRef]
Lee PA, Witchel SF. The influence of
estrogen on growth. Curr Opin Pediatr.
1997;9:
431-6.9431
1997
[PubMed][CrossRef]
Bala RM, Lopatka J, Leung A, McCoy E,
McArthur RG. Serum immunoreactive somatomedin levels in normal adults,
pregnant women at term, children at various ages, and children with
constitutionally delayed growth. J Clin Endocrinol Metab.
1981;52:
508-12.52508
1981
[PubMed][CrossRef]
Grumbach MM, Styne DM. Puberty:
ontogeny, neuroendocrinology, physiology, and disorders. In: Williams RH,
Wilson JD, Foster DW, editors. Williams' textbook of endocrinology. 8th
ed. Philadelphia: WB Saunders; 1992. p
1139-221.1139
1992
Hesse V, Jahreis G, Schambach H, Vogel
H, Vilser C, Seewald HJ, Borner A, Deichl A. Insulin-like growth factor I
correlations to changes of the hormonal status in puberty and age. Exp
Clin Endocrinol. 1994;102:
289-98.102289
1994
[CrossRef]
Juul A, Dalgaard P, Blum WF, Bang P,
Hall K, Michaelsen KF, Muller J, Skakkebaek NE. Serum levels of insulin-like
growth factor (IGF)-binding protein-3 (IGFBP-3) in healthy infants, children,
and adolescents: the relation to IGF-I, IGF-II, IGFBP-1, IGFBP-2, age, sex,
body mass index, and pubertal maturation. J Clin Endocrinol
Metab. 1995;80:
2534-42.802534
1995
[CrossRef]
Rosenfield RI, Furlanetto R, Bock D.
Relationship of somatomedin-C concentrations to pubertal changes. J
Pediatr. 1983;103:
723-8.103723
1983
[CrossRef]
Delemarre-van de Waal HA, van Coeverden
SC, Rotteveel J. Hormonal determinants of pubertal growth. J Pediatr
Endocrinol Metab. 2001;14 Suppl 6:
1521-6.141521
2001
Rotteveel J, Schoute E, Delemarre-van de
Waal HA. Serum procollagen I carboxyterminal propeptide (PICP) levels through
puberty: relation to height velocity and serum hormone levels. Acta
Paediatr. 1997;86:
143-7.86143
1997
[CrossRef]
Crofton PM, Stirling HF, Kelnar CJ. Bone
alkaline phosphatase and height velocity in short normal children undergoing
growth-promoting treatments: longitudinal study. Clin Chem.
1995;41:
672-8.41672
1995
[PubMed]
Crofton PM, Stirling HF, Schonau E,
Kelnar CJ. Bone alkaline phosphatase and collagen markers as early predictors
of height velocity response to growth-promoting treatments in short normal
children. Clin Endocrinol (Oxf).
1996;44:
385-94.44385
1996
[PubMed][CrossRef]
Izumi Y. Alkaline phosphatase as a
biochemical maturity index in female adolescence. Spine.
1993;18:
2257-60.182257
1993
[PubMed][CrossRef]
Willner S, Johnell O. Study of
biochemical and hormonal data in idiopathic scoliosis in girls. Arch
Orthop Trauma Surg. 1981;98:
251-5.98251
1981
[CrossRef]
Zorab PA. Total hydroxyproline excretion
in scoliosis. In: Zorab PA, editor. Proceedings of a second symposium
on scoliosis: causation. Edinburgh: E and S Livingston;
1968.
1968
Zorab PA, Clark S, Harrison A, Seel JR.
Hydroxyproline excretion and height velocity in adolescent boys. Arch
Dis Child. 1970;45:
763-5.45763
1970
[CrossRef]
Johansen JS, Giwercman A, Hartwell D,
Nielsen CT, Price PA, Christiansen C, Skakkebaek NE. Serum bone Gla-protein as
a marker of bone growth in children and adolescents: correlation with age,
height, serum insulin-like growth factor I, and serum testosterone. J
Clin Endocrinol Metab. 1988;67:
273-8.67273
1988
[CrossRef]
Johansen JS, Jensen SB, Riis BJ,
Rasmussen L, Zachmann M, Christiansen C. Serum bone Gla protein: a potential
marker of growth hormone (GH) deficiency and the response to GH therapy.
J Clin Endocrinol Metab.
1990;71:
122-6.71122
1990
[PubMed][CrossRef]
Vihervuori E, Turpeinen M, Siimes MA,
Koistinen H, Sorva R. Collagen formation and degradation increase during
growth hormone therapy in children. Bone.
1997;20:
133-8.20133
1997
[PubMed][CrossRef]