To The Editor:
We read "Intraobserver and Interobserver Reliability of the Classification
of Thoracic Adolescent Idiopathic Scoliosis" (80-A, 1097-1106, Aug.
1998), by Lenke et al. The purpose of that study was to determine
the interobserver and intraobserver reliability of the King classification
of thoracic adolescent idiopathic scoliosis. King et al.3 classified thoracic curves in a paper
entitled "The Selection of Fusion Levels in Thoracic Idiopathic
Scoliosis." Since the King classification system deals solely with
thoracic scoliosis, we fail to understand why, in order to consider the
reliability of the King classification, parameters such as the "lumbar"
curve, the "thoracolumbar" curve, the "double-major" curve, and others
were introduced in the trial. In our opinion, only thoracic curves
should have been examined.
Furthermore, on the basis of the work by Barr et al.1, Ibrahim et al.2,
Lenke et al.6, Lonstein7, and Richards8,
there is great terminological confusion between a "double-major"
curve and the King type-I curve, and we believe that the term "double-major"
curve should not have been included in the study.
Finally, in the 1994 study by King5,
there is no mention of the criteria that are used to distinguish
type-II and type-III curves. The difference between these two types
of curves, which is related to whether the lumbar curve crosses
the midline, was established by King et al. in 19833 and 19884.
Angel M. Hidalgo-Ovejero, M.D.
Serafín García-Mata, M.D.
Manuel Martinez-Grande, M.D.
Tomás Izco-Cabezò¬ª M.D.
Corresponding author: Angel M. Hidalgo-Ovejero, M.D.
Avenida Pio XII - 16 Escalera Derecha 3°C
31008 Pamplona, Spain
L. G. Lenke, R. R. Betz, K. H. Bridwell,
D. H. Clements, J. Harms, T. G. Lowe,
and H. L. Shufflebarger reply:
We appreciate the comments made by Hidalgo-Ovejero et al., as
they point out some of the reasons for performing this study.
The first point made by Hidalgo-Ovejero et al. is that we should
have evaluated only thoracic curves in our study. In fact, all twenty-seven sets
of radiographs demonstrated an operative main thoracic curve. Thoracic
curves do not exist in isolation but in the context of structural
or nonstructural curves, above and below, that warrant evaluation
and possible treatment. None of the cases that we reviewed had an
isolated thoracolumbar or lumbar curve. In essence, we did look
at curves that were similar to those described by King et al. in
19833.
The second point is a valid criticism of our use of the term
double-major as a possible classification choice for the reviewers
in our study. We offered this option because it is a commonly used
classification term that occasionally is used to refer to King type-I
and type-II curves, as noted by Barr et al.1.
We consider a double-major curve to be one in which both the thoracic
and the thoracolumbar or lumbar components are considered structural and
must be included in the arthrodesis. There certainly are many curves
in which the thoracic and lumbar components have nearly equal Cobb angles,
rotation, and degrees of deviation from the midline that also may
satisfy the criteria for a King type-II curve but in which both
the thoracic and lumbar components must be instrumented and arthrodesed6. We wonder if it is still appropriate
to call such a curve a King type-II curve, which implies that selective
thoracic arthrodesis may be performed. Alternatively, perhaps it
should be called a King type-I curve even if it does not meet the
criteria, just because such curves are more likely to be treated
as true double-major curves. Double structural thoracic and thoracolumbar
or lumbar curves are not well categorized by the King system or
any other system. Additional work is needed to appropriately classify
such curves and to adequately evaluate treatment protocols2.
The last comment is related to the distinction between King type-II
and type-III curves. It is true that in their 1983 article3, King et al. stated that type-III
curves do not cross the midline. However, as we mentioned in our
study, we do not know exactly how much of the lumbar curve has to
cross the midline in order for it to be classified as a King type-II
curve. As shown in Figure 3 of our report, there are many lumbar
curves in which the center sacral line barely touches the apex of
the lumbar curve and there is contralateral rotation of the lumbar spine
compared with the rotation of the thoracic spine. It appears there
is an intermediary curve between King types II and III that has
been a challenging curve to classify. It also appears that a more
strict definition of the exact part of the lumbar apical body that
must cross the midline is necessary in order to distinguish between King
type-II and type-III curves in those borderline cases, and it may
be necessary to create a separate category.
We appreciate the comments of these astute readers, and we believe
that they further emphasize the need for a more critical evaluation
of the classification of scoliosis in order to obtain a comprehensive,
reliable, and practical system for universal use. The Scoliosis
Research Society is aware of this need and has commissioned an ad
hoc committee on the classification of idiopathic scoliosis
in order to attempt to address these and related concerns.
Lawrence G. Lenke, M.D.
Randal R. Betz, M.D.
Keith H. Bridwell, M.D.
David H. Clements, M.D.
Jürgen Harms, M.D.
Thomas G. Lowe, M.D.
Harry L. Shufflebarger, M.D.
Corresponding author: Lawrence G. Lenke, M.D.
Department of Orthopaedic Surgery
Spinal Deformity Service, Washington University
One Barnes-Jewish Hospital Plaza
Suite 11300, West Pavilion
St. Louis, Missouri 631101.