To The Editor:
We read with interest the article entitled "Operative Compared with
Nonoperative Treatment of a Thoracolumbar Burst Fracture without Neurological
Deficit. A Prospective, Randomized Study" (2003; 85:773-81), by Wood et
al.
The authors speak about "stable" burst fractures of the
thoracolumbar junction. The inclusion criteria included the absence of a
posterior column lesion (except for a laminar fracture, which was neither an
exclusionary criterion nor a contraindication for nonoperative treatment). The
first definition categorizes the fractures as burst type-A3 fractures
according to the AO classification system1, which is widely used. A
laminar fracture, however, indicates a posterior column lesion, thereby
categorizing the fracture as type B1 according to the AO
classification1. According to the criteria described by
Denis2 and Louis and Goutallier3, a type-A3 burst
fracture, by definition, represents an unstable lesion as two of three columns
are involved and a type-B1 fracture represents a highly unstable lesion as
three columns are involved. Therefore, we think that it is incorrect to speak
about a "stable" burst fracture.
Furthermore, we fear that comparing operative treatment with nonoperative
treatment on the basis of the follow-up data for forty-seven (89%) of
fifty-three patients is misleading. We do not believe that it is advisable to
test for significant differences between two small groups of patients because
of the possibility of inducing incorrect conclusions.
The study is referred to as a long-term investigation, but the mean
duration of follow-up was only forty-four months. We think that this interval
represents only short or intermediate-term follow-up and does not allow one to
conclude how the patients will fare over a longer period. Indeed, it is known
that residual kyphosis induces back pain after ten years or
more4-6. Therefore, we think that it is wrong to state that one
treatment or the other provides no major long-term advantage.
Furthermore, the operative group was composed of an inhomogeneous group of
patients who underwent anterior or posterior procedures, which cannot be
compared. The authors did not indicate how many patients presented with
multiple-level fractures, which were not considered to be an exclusion
criterion. How did such patients fare compared with those who had single-level
lesions? Can a patient who has operative treatment of a multiple-level injury
be compared with a patient who has nonoperative treatment of a single-level
type-A3 burst fracture? The authors provided no information regarding
osteoporosis, preexisting back pain, or discopathies.
Finally, there are many surgical alternatives for the treatment of a spinal
fracture. Therefore, the title should have been more precise concerning the
type of surgical treatment, in this case spondylodesis. Indeed, the authors
did not refer to any other form of treatment, such as cages or
vertebroplasty7.
While the idea of comparing operative treatment with nonoperative treatment
is interesting, we doubt that the conclusions of the authors can be shared. We
would hesitate to treat burst fractures with nonoperative means on the basis
of this one study, even though we are aware of the fact that there is not yet
a unanimous consensus regarding a standard surgical treatment for such
lesions.
We appreciate the letter by Drs. Wettstein and Mouhsine. As to their first
concern regarding the inclusion of laminar fractures, the original article by
Magerl et al.1
plainly indicates that a laminar fracture is part of a type-A compression
injury and does not necessarily constitute a type-B distraction injury.
Although Louis may have thought that a laminar fracture is an unstable injury,
the current thinking is that, in the presence of an intact posterior column
and intact ligaments and facet joints, such a fracture is not an unstable
injury. Finally, the concept that the involvement of two of three columns
represents an unstable lesion has been safely discarded in recent
literature.
As to the second point, we agree that the numbers of patients in each group
were less than ideal, but this is reflected on the first page, where the level
of evidence of the study is addressed by The Journal.
The third point with respect to the mean follow-up of forty-four months
representing only short or intermediate-term results may be applicable. How
these patients will fare as the years go by remains to be seen. There is in
fact no definition of what constitutes short, intermediate, and long-term
follow-up, so whether this study represents intermediate-term or long-term
results remains debatable. It was our judgment that four years of follow-up
could typify a long-term study. In fact, in the nearly five years since these
data were collected, no new problems have arisen within our study group and
all patients have continued to be followed at this center.
The fourth point, regarding the inclusion of both anterior and posterior
procedures, may have some merit. We are currently in the process of comparing
two groups of study patients (those treated with an anterior approach and
those treated with a posterior approach) and have found no major differences
between the two; thus, we think that it was reasonable to include the two
groups together within the operative segment. We believe that it is fair to
compare a five-level, four-disc spine fusion with a three-level, two-disc
anterior reconstruction, principally because the caudalmost vertebral body
remains the same. The levels of extension occur in the thoracolumbar and lower
thoracic spine and do not put strategic motion segments within the lumber
spine at any further risk. We did not specifically address osteoporosis or
discopathies; however, the patients' preexisting visual analog pain scales
reflecting existing low-back pain can be seen in the tables cited in the
study.
As to the last point, it is widely accepted at this point in time that
vertebroplasty is not a viable treatment option for burst fractures associated
with canal compromise. It is indeed true that a cage could be placed
anteriorly following a subtotal corpectomy; however, we did not employ such a
regimen. If the authors wish to pursue such a study in which such treatment is
compared with the other surgical options, we would read the results with great
interest.
In summary, we stand by our conclusion that nonoperative treatment remains
a viable alternative. It certainly has been shown in numerous previous studies
that nonoperative treatment is a viable option for stable burst fractures, and
our study adds further data to this literature.
Magerl F, Aebi M, Gertzbein SD, Harms
J, Nazarian S. A comprehensive classification of thoracic and lumbar
injuries. Eur Spine J.1994;3:
184-201.3184
1994
[PubMed][CrossRef]
Denis F. The three-column spine
and its significance in the classification of acute thoracolumbar spine
injuries. Spine.1983;8:
817-31.8817
1983
[PubMed][CrossRef]
Louis R, Goutallier D. Fractures
instables du rachis (symposium). Rev Chir Orthop.1977;63:
415-81.63415
1977
Oner FC, van Gils AP, Faber JA, Dhert
WJ, Verbout AJ. Some complications of common treatment schemes of
thoracolumbar spine fractures can be predicted with magnetic resonance
imaging: prospective study of 53 patients with 71 fractures.
Spine.2002;27:
629-36.27629
2002
[PubMed][CrossRef]
Wu SS, Hwa SY, Lin LC, Pai WM, Chen
PQ, Au MK. Management of rigid posttraumatic kyphosis.
Spine.1996;21:
2260-6.212260
1996
[PubMed][CrossRef]
Karjalainen M, Aho AJ, Katevuo K.
Painful spine after stable fractures of the thoracic and lumbar spine. What
benefit from the use of extension brace? Ann Chir
Gynaecol.1991;80:
45-8.8045
1991
Nakano M, Hirano N, Matsuura K,
Watanabe H, Kitagawa H, Ishihara H, Kawaguchi Y. Percutaneous
transpedicular vertebroplasty with calcium phosphate cement in the treatment
of osteoporotic vertebral compression and burst fractures. J
Neurosurg.2002;97(3 Suppl):
287-93.97287
2002
[CrossRef]