Keywords transthoracic - kyphosis - Cobb’s angle - titanium mesh - transthoracic
Introduction
Due to high frequency of motor vehicle accidents and work place injuries, spine fractures
are common in today’s world. Following spinal injuries, there is 7% mortality rate.
The upper thoracic region (T1–T8) is rigid due to the rib cage, which provides stability.
The transition zone extends from T9 to L2, which is the transition between the rigid
and kyphotic upper thoracic spine and the flexible lordotic lumbar spine. This is
where most injuries occur.
The treatment of thoracolumbar fractures can be either conservative or surgical. There
were no advantages reported by Wood et al[1 ] for operative stabilization compared with nonoperative treatment in patient with
thoracolumbar burst fractures in neurologically intact patients. However, Siebenga
et al[2 ] reported significantly higher radiologic kyphosis and significantly higher pain
scores after nonoperative treatment.
Unstable thoracolumbar burst fractures may be treated with anterior, posterior, or
circumferential fusion. There are no clear-cut guidelines regarding surgical management
of unstable thoracolumbar fractures as the treatment protocol is decided taking into
account multiple factors like fracture pattern, canal compromise, the severity of
neurological injury, and surgeon’s experience. All these parameters affect the decision
regarding the surgical approach, with each parameter being an indication for one or
the other approach. Therefore, no consensus has been reached about the ideal treatment
approach.
Although there are several reports of thoracolumbar fractures, most studies are based
on only one surgical method or on internal fixation.[3 ] There is a paucity of the evidence-based guidelines for the treatment of these fractures
and, for a better comparison of surgical techniques, randomized controlled trials
are necessary.[4 ]
We present a prospective analysis of case series of patients who underwent either
anterior or posterior decompression or stabilization in our tertiary care center.
Our study was performed (1) to compare the neurological outcome of patients undergoing
surgery by either approaches and (2) to compare the radiological outcomes of anterior
and posterior operations.
Aim
To study the age and sex distribution of thoracolumbar spinal injuries in our patient
population.
To analyze the results of transpedicular decompression and anterior thoracic approach
in our patients with traumatic thoracolumbar spinal injuries in terms of:
Materials and Methods
This was a prospective study of patients with fractures of dorsolumbar spine conducted
at a tertiary care neurosurgical center from December 2011 to December 2013. Institutional
Review Board approval and informed consent were obtained.
Analysis of our surgical results was done as per the following:
Clinical parameters: Pre- and postoperative results according to American Spinal Injury
Association (ASIA) classification.
Radiological parameters: Pre- and postoperative radiological results in terms of spinal
deformity (kyphosis) was studied by Cobb’s angle.
Inclusion Criteria
For Anterior Approach
Significant anterior compression of the spinal cord (canal encroachment more than
50%) ± kyphotic deformity.
Absence of thoracic or abdominal pathologies that hinder the transthoracic or retroperitoneal
approaches.
Anterior and middle column disruption.
For Posterior Approach
Significant posterior compression of the spinal cord ± kyphotic deformity.
Patient condition not permitting lengthy anterior procedure.
Posterior column disruption.
Canal encroachment <50%.
Exclusion Criteria
Patients not willing to give consent.
Patients with age less than 18 years.
Patients with complete cord transaction.
Patients who had pathological fracture.
Clinical Assessment : A detailed history was obtained and examination was performed especially evaluating
the mode of trauma, ASIA grading, sensory level, and any spinal deformity. Full neurological
examination was done and documented repeatedly to look for and pick up neurological
deficits and deterioration.
Imaging : Plain X-rays in anteroposterior and lateral views were obtained to measure the angular
deformity both pre- and postoperatively; magnetic resonance imaging (MRI) or computed
tomography (CT) scans were obtained to further evaluate the important relationships
and instability of spine.
Choice of Approach
The decision on surgical approach is to a great extent based on “The Load Sharing
Classification.” The posterior approach is generally indicated for patients with score
of less than 6 points, whereas the anterior approach should be considered for patients
scoring more than 6 points. However, as mentioned earlier, this classification cannot
be the sole criteria for decision-making, as the surgical approach of choice also
depends on the surgeons’ familiarity with the surgical technique, and if anterior
surgery is not feasible due to the systemic condition of the patient or inadequate
technical facilities, extending posterior instrumentation and fusion length may be
the alternative option to restore coronal and sagittal stability, prevent recurrent
kyphosis, and promote fusion and post-reduction stability. Thus, factors such as an
anesthetic and surgical burden to the patient, morbidity, complication rates, costs,
and surgeon’s expertise should be taken into account in the choice of surgical approach.
The Thoracoabdominal Approach
It provides for the best exposure of T12–L2. Generally, resection of rib two levels
above the primary pathology is performed. Hence, resection of the 9th rib provides
the best window of access to T11–T12, and is accompanied by transthoracic approach,
whereas exposure of T12–L1 may be accomplished via a thoracoabdominal 10th rib approach.
The patients were placed in lateral decubitus position. A left-sided approach was
usually preferred to avoid inferior vena cava and the liver. An incision was made
from the lateral border of the par spinal musculature along the required rib to the
junction of rib and costal cartilage. The rib was removed. Peritoneum was dissected
off the inferior surface of diaphragm. The soft tissue was swept off the thoracic
and abdominal surfaces of the diaphragm, which was then incised circumferentially
leaving cuff of muscle attached to the chest wall. The crus of diaphragm was cut and
elevated off the spinal column.
Deaver retractor was used to retract the peritoneal sac anteromedially, and a large
rib retractor opens the intercostal space to reveal the thoracolumbar junction.
Complete discectomy at the level caudal and cephalad to the fracture was performed.
A large rongeur was used to remove the anterior cancellous portion of the vertebral
body. After the corpectomy was done, titanium mesh cage packed with corpected vertebral
body and rib was inserted. Bicortical fixation was performed in all constructs, which
were appropriately compressed to improve load sharing. Postoperatively, all patients
were managed in a Taylor brace for 3 months ([Figs. 1 ] and [2 ]).
Fig. 1 (A ,B ) Corpectomy of involved vertebra.
Fig. 2 Cage placement after distraction.
Posterior Approach
Twenty-eight patients were surgically treated with posterior segmental fixation and
posterolateral fusion using iliac crest bone graft. We performed pedicle screws fixation
at two levels above and one level below the fracture site in most of the patients.
In five patients, the lower level was extended to two levels below the fracture vertebra.
In most of the cases, transpedicular decompression was done whereas in three patients
fracture reduction was achieved by a combination of postural reduction, and by distraction
through ligamentotaxis. In two patients, mesh cage was placed after transpedicular
decompression at fractured vertebral site. For fusion, bone harvested from the decompression
site or iliac crests autograft was used.
Results were expressed as the mean ± the standard deviations.
Regardless of which approach was used, the goal of surgery was to decompress the canal,
correct the kyphosis, and stabilize the spine. After surgery, patients were mobilized
gradually while wearing lumbar orthoses. The patients underwent postoperative radiograph
before discharge. They were followed-up 1, 3, and 6 months postoperatively and then
annually. The radiographs were obtained with standing lateral views and were analyzed
with particular attention paid to the sagittal Cobb’s angles at the operation segments.
Result and Analysis
A total of 60 patients with traumatic spinal injuries were admitted during the study
period (December 2011–2013), of which 51 cases were finally selected and taken for
operations at our institution during this period for study proper as per the inclusion
and exclusion criteria laid down in the Material and Methods section of this study.
These 51 patients were selected and enrolled in the study proper after thorough consent
and were thoroughly studied throughout the study period of 24 months. During the study
period, 3 patients were lost in follow-up so we had only 48 patients remaining to
include in the study.
Age Distribution
Among our 48 patients, youngest one was of 20 years and eldest of 58 years. Most of
the patients (41.66%) were in the age group of 30 to 40 years. The mean age of our
patient population was 38.33 years ([Table 1 ]).
Table 1
Age-wise distribution of patients with traumatic thoracic/thoracolumbar fracture
Age (in years)
No. of patients (n = 48)
Note: In our study, there were 40 (83.33%) male and 8 (16.66%) female patients.
20–30
8 (16.66%)
31–40
20 (41.66%)
41–50
16 (33.33%)
>51
4 (8.33%)
All the patients in the study populations were thoroughly assessed clinically and
grading of completeness of spinal cord injuries was assessed as per ASIA grading of
all patients. Patients with ASIA grade E were not included in the study ([Tables 2 ] and [3 ]).
Table 2
Distribution of patients according to American Spinal Injury Association (ASIA) grade
of spinal injury
Grade
Completeness of spinal cord injury
No. of patients (n = 48)
Abbreviation: M, motor.
A
Complete (M—, S—)
4 (8.33%)
B
Incomplete (S+, M–)
10 (20.83%)
C
Incomplete (M < 3)
20 (41.66%)
D
Incomplete (M > 3)
14 (29.16%)
E
Normal (S+, M+)
Not included
Table 3
Distribution of patients having urinary incontinence according to American Spinal
Injury Association (ASIA) grade
ASIA grade
No. of patients (n = 25)
A
4 (16%)
B
8 (32%)
C
8 (32%)
D
5 (20%)
Among patients who were decided to be treated by anterior approach, out of 20 patients,
4 were having grade B, 8 grade C, and 8 patients had ASIA grade D.
Among patients operated by posterior approach, out of 28 patients, 4 patients had
grade A, 6 patients had grade B, 12 patients had grade C, and 6 patients had grade
D.
Preoperative Classification
Preoperative Classification
Patients were classified preoperatively as per the data provided in [Tables 4 ] and [5 ] .
Table 4
Subclassification of patients falling in different American Spinal Injury Association
(ASIA) grades according to operation done
ASIA grade
Anterior approach (n = 20)
Posterior approach (n = 28)
A
0
4 (14.2%)
B
4 (20%)
6 (21.42%)
C
8 (40%)
12 (42.85%)
D
8 (40%)
6 (21.42%)
Table 5
Classification of patients on the basis of mobility in anterior and posterior approaches
preoperatively
Type of patients
Anterior approach (n = 20)
Posterior approach (n = 28)
Bed ridden
13 (65%)
22 (78.57%)
Walks with support
4 (20%)
3 (10.71%)
Walks without support
3 (15%)
3 (10.71%)
PostoperativeSurgical Outcome
PostoperativeSurgical Outcome
Postoperatively, clinical improvement was again assessed by ASIA grading system before
discharge, at 6 weeks, 3 months, 6 months, 9 months, 12 months, and then every 6-monthly
follow-up. The best grade of the patient achieved in follow-up period was taken as
the final ASIA grade. Range of follow-up was from 1 month to 20 months, with a mean
of 7.4 months.
In Anterior Approach
Two patients did not improve, that is, out of four patients having grade B, two remained
to have grade B, one patient moved to grade C, and one patient to grade D.
Five patients having grade C developed power > 4 and three patients having grade D
developed power > 4. One patient having grade B developed power > 4 and was able to
walk with support. So, in total, nine patients were able to walk with support. Rest
eight patients could walk without support.
Out of 20 patients, 3 (15%) patients were bed ridden, 9 (45%) persons could walk with
support, and 8 (40%) without support.
In Posterior Approach
Out of four patients having grade A, three remained in grade A while one improved
to grade C with power of 2. Among six patients with grade B, two were able to walk
with support and four patients improved to grade C. Among 12 patients having grade
C, 5 were able to walk with support and 2 without support. In grade D, out of six
patients, one patient was not able to walk as he had power of 3, while three patients
were able to walk with support and two patients without support.
So, out of 28 patients, 14 (50%) patients were bed ridden, 10 (35.71%) patients were
able to walk with support, and 4 (14.28%) patients without support.
Follow-Up Kyphotic Angle
Our patient population postoperatively was subjected to serial radiographs. Initially
on third postoperative day check digital X-rays, both anteroposterior and lateral
views were taken. Out of 48, only 43 came for regular follow-ups postoperatively,
that is, 18 in anterior group and 25 in posterior group. They were subjected to serial
radiographs. CT and MRI scans were obtained in selected cases where digital X-rays
were not informative or contradictory.
Kyphotic angle changes were seen in 16 patients out of 18 in anterior group and 20
patients in posterior group out of 25.
Out of 18 patients in the anterior group, 2 showed no change in angle, 2 showed increase
in kyphotic angle (deterioration), and 14 showed reduction in kyphotic angle of 1
to 10 degrees (improvement)·with mean improvement of 4.07 degrees.
In the posterior group, out of 25 patients, there was no change in kyphotic angle
in 5 patients, 7 patients showed improvement of 1 to 8 degrees (reduction in kyphotic
angle), whereas 13 patients showed deterioration of 1 to 12 degrees. The mean improvement
was 2.14 degrees in 7 patients and mean deterioration was 4.92 degrees ([Figs. 3 ]
[4 ]
[5 ]).
Fig. 3 Patients showing changes in kyphotic angle after operation through anterior approach.
Fig. 4 Number of patients showing changes in kyphotic angle in posterior approach group.
Fig. 5 Patients showing changes in urinary incontinence before and after operation through
anterior approach or posterior approach.
Patients with Urinary Incontinence
In the anterior group, out of 20 patients, 8 patients had urinary continence before
operation. After operation, 3 patients improved and rest 5 continued to have incontinence.
In the posterior group, out of 28 patients, 17 had urinary incontinence. Only 2 patients
improved and rest 15 patients continued to have urinary incontinence ([Table 6 ]).
Table 6
Urinary inconsistency after surgery
Incontinence
Posterior group (N = 17)
Anterior group (N = 8)
Corrected
2 (11.76%)
3 (37.5%)
Not corrected
15 (88.23%)
5 (62.5%)
Statistical Analysis of Clinical Assessment and Results
On comparing the two groups preoperatively after applying chi-square test, the p -value was 0.56, which is > 0.05. Hence, there was no statistically significant difference
between the two groups ([Table 7 ]).
Table 7
Preoperative neurological status of patients
Status
Posterior group
Anterior group
χ2 test
Bed ridden
22
13
χ2 (2) = 1.16
p = 0.56
Walks with support
3
4
Walks without support
3
3
Total
28
20
Postoperatively
In the postoperative group, the p -value is 0.025, which is < 0.05. Hence, the result is statistically significant after
using chi-square test. There was statistically significant improvement in the postoperative
group in anterior approach as compared with posterior approach ([Table 8 ]).
Table 8
Postoperative neurological status of patients
Anterior Approach
Posterior approach
Total no. of patients
Bed ridden
3
7.08
(2.35)
14
9.92
(1.68)
17
Walks with support
9
7.92
(0.15)
10
11.08
(0.11)
19
Walks without support
8
5.00
(1.80)
4
7.00
(1.29)
12
20
28
48
Statistical Analysis of Kyphotic Angle
Comparing the kyphotic angles in two groups, p -value in deteriorated group is 0.014, which is < 0.05, so the result was statistically
significant. The posterior approach group showed statistically significant deterioration
in kyphotic angle whereas on comparing the improvement in kyphotic angle between anterior
and posterior groups, p -value was found to be 0.004 (<0.05). So, there was statistically significant improvement
in kyphotic angle in anterior approach group ([Table 9 ]).
Table 9
Statistical analysis of kyphotic angle
Kyphotic angle
Anterior group (N
1 = 18)
Posterior group (N
2 = 25)
Z -test or p -value
Same
2 (11.1%)
5 (20%)
Z = 0.36
p = 0.718
Deteriorated
2 (11.1%)
13 (52%)
Z = 2.45
p = 0.014
Improved
14 (77.8%)
7 (28%)
Z = 2.91
p = 0.004
Statistical Analysis of Urinary Incontinence
On comparing the two groups by chi-square test, no statistical difference was found
(p > 0.05) regarding improvement in urinary incontinence during the follow-up period
([Table 10 ]).
Table 10
Urinary incontinence
Incontinence
Posterior group
Anterior group
χ2 test
Corrected
2
3
χ2 (2) = 0.93 (Yates corrected)
p = 0.33
Not corrected
15
5
Total
17
8
Complications
Three patients in the posterior approach group developed superficial skin infections,
which were managed conservatively by appropriate antibiotic treatment. Three patients
in the anterolateral approach group required prolonged chest tube after surgery due
to opening of the pleura cavity during the operation. Two patients who underwent anterolateral
approaches developed low thoracic dermatomal pain from intercostal neuralgia that
was improved with pain medication. One patient had an aortic injury during exposure
that was successfully repaired by cardio-thoracic surgeon and the patient had an uneventful
course. There were no other complications such as neural structure injury and hardware
failure.
Discussion
Almost two-thirds of all traumatic TL spine injuries occur at the TL junction, most
frequently at L1, followed by T12. From an anatomical standpoint, the ideal treatment
of unstable TL junction fractures should consist of complete kyphosis correction and
optimal spinal canal decompression in case of a neurological deficit. The treatment
of TL junction burst fractures remains very controversial for several reasons. Anterolateral
decompression in comparison to posterior decompression allows direct decompression
of ventral osseous and offering superior canal clearance as compared with the posterior
approach (ligamentotaxis, posterolateral decompression techniques).
Surgical Outcome in Terms of Motor Functions
Robert and Kumar[5 ] usually perform posterior surgery on cases with ASIA class A spinal cord injuries.
The extent of instrumentation is usually two or three levels above and two levels
below.
Cases with partial neural deficits are ideal candidates for anterior decompression
as they have the greatest chance for neurological recovery. In their experience, blood
loss, neurological outcomes, and overall management of morbidity and mortality have
been very low in either approach. The most important factor determining the operative
approach is the individual surgeon’s own outcomes, which lead to the best neurological
and functional outcomes. In the overwhelming majority of patients, successful decompression
and stabilization/fusion can be accomplished by either approach.
Belanger et al[6 ] reported that even in cases of long-standing compression, anterior decompression
can result in modest improvements in neurological function. The results of surgical
decompression (SD) in 59 patients with neurological deficits secondary to thoracic
or lumbar fractures were evaluated at a mean of 3.7 years after injury. The purpose
was to determine whether SD could be correlated with subsequent neurological outcome.
Follow-up neurological evaluations showed a greater neurological improvement in the
20 patients who were treated with anterior spinal SD as compared with the 39 patients
who received posterior or lateral SD (88% vs. 64%).
Bradford and McBride[7 ] reported an average 25.9% of residual canal compromise following posterior surgery
compared with less than 1% after anterior decompression.
Humphries et al[8 ] were of the opinion that anterior grafting favored fusion because of better blood
supply and bone contact, and the fact that the graft is under compression.
Outcome in Relation to Kyphotic (Cobb’s) Angle
In Ghanayem and Zdeblick[9 ] study, 10 of 12 patients who underwent anterior arthrodesis using the Z plate thoracolumbar plating system after a one-stage anterolateral decompression
and reduction procedure for burst fractures from T9 to L3 maintained their postoperative
sagittal alignment or a significant portion of their kyphosis reduction. Two patients
with severe kyphotic deformities greater than 50 degrees lost 10 degrees and 20 degrees
of their reduction, respectively, at last follow-up.
Hammad and Mohamed[10 ] in their study found that in anterior approach the kyphotic angle improved from
the mean angle 8.8 degrees (range from–6 to 20 degrees) before surgery to the mean
angle–1 degree (range from–10 to 4 degrees) after surgery and final follow-up. A significant
difference was seen between kyphotic angle preoperatively and at final follow-up (p = 0.0001) in all cases.
Changes in Bladder Symptom
Chiu and Liao[11 ] in their study of pure conus medullaris syndrome with L1 burst fracture found that
despite surgical treatment regardless of approach, only one-half of the patients regained
normal bladder and sexual function.
Elsawaf[12 ] in their study showed that in the posterior approach group, 18 patients had urinary
symptoms in the form of urine retention. With continuous follow-up, 11 patients showed
progressive improvement and all of them were not using a urinary catheter anymore.
On the other hand, the anterior group showed 16 patients of acute urinary dysfunction:
13 of them had complete improvement in the problem by the follow-up period; the other
3 patients had persistent urinary dysfunction but were in need of only intermittent
catheterization with bladder distension. His follow-up duration was 4.9 years.
We, however, could not find any significant difference in the postoperative improvement
in bladder symptoms in either approach or the two approaches were similar in outcome
as far as bladder symptoms were considered.
But here also we would like to emphasize the limitation of short follow-up, small
sample size, and proper bladder exercises. In the long-term follow-up, the results
may vary.
Analysis of our clinical outcome demonstrates that neurological improvement was documented
in our patients regardless of which approach was used. The results were better with
the anterior approach since ~80% of the patients in the anterior study group were
able to walk, either with or without support. Radiological results in our study show
that 77.77% showed improvement (reduction) in kyphotic angle whereas in the posterior
group only 28% showed improvement in kyphotic angle. Currently, guidelines based on
randomized trials for the treatment of thoracolumbar burst fractures are unavailable.
In terms of urinary incontinence, there was no statistical improvement in the anterior
group as compared with posterior group or vice versa. Verlaan et al[13 ] conducted a literature review of 132 papers, involving 5,748 patients with thoracic
and lumbar fractures treated with posterior, anterolateral, or combined approaches.
They concluded that evidence-based guidelines for the treatment of these fractures
were absent and suggested that, for a better comparison of surgical techniques, randomized
controlled trials were necessary. We agree with their conclusions, but we need to
consider several parameters before we chose the selection of treatment options. This
selection should be based on clinical and radiological findings such as deformity
angulation, residual canal diameter, and vertebral body height including neurological
deficit.
Conclusion
With all the available material and methods and after analyzing all the results and
statistical values, this study concluded that:
Traumatic thoracic/thoracolumbar injuries (41.66%) are more common in persons of productive
age group (31–40 years) with male preponderance (M > F, 83.33%).
More persons were able to walk with or without support in anterior group (85%) whereas
in posterior group the improvement was less (50%).
There is statistically significant difference in change of kyphotic angle in long-term
follow-up. In the anterior group, 77.77% showed improvement (reduction) in kyphotic
angle whereas in posterior group only 28% showed improvement in kyphotic angle.
In terms of urinary incontinence, there was no statistical improvement in anterior
group as compared with posterior group or vice versa, with only 37.5% showing improvement
in anterior group as compared with posterior group (11.76%).
In conclusion, it can be said that the anterior approach provides a better outcome
in terms of improvement in motor and sensory function, as well as more decrease in
kyphotic deformity as compared with posterior approach. But there seems to be no significant
difference in terms of autonomic improvement in form of bladder continence.
There were certain limitations in our study. The sample size was not very big; moreover,
the long-term follow-up may have some impact on the kyphotic angulation regardless
of approaches.
