Keywords
traumatic thoracolumbar fracture - spinal injuries - three-column fixation - transpedicular corpectomy - posterior only approach
Introduction
Thoracolumbar fracture is most common because it is uniquely positioned between rigid thoracic spine and mobile lumbar spine. Most authors agreed for surgical management of unstable thoracolumbar fracture, but choice of approach is still a matter of debate. Traditional posterior approach is preferred because of less complication rate and less complexity of surgery. Posterior approach has disadvantage of insufficient spinal canal decompression, less stable construct, and more chances of recurrent kyphosis. To overcome these disadvantages, anterior approach and later on combined anteroposterior approaches were developed. Access to thoracolumbar vertebral bodies or discs via traditional transthoracic or retroperitoneal approaches carries significant morbidity.[1]
[2] Additional posterior stabilization or decompression is often required, thus further increasing the risks of the operation.[3]
[4] Single-stage three-column fixation through posterior approach has been described in an attempt to simplify the surgical approach for circumferential decompression and simultaneous stabilization of the thoracolumbar spine.[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13] Many studies were done to compare anterior approach and combined anteroposterior approach with posterior approach but no study was done to compare combined anteroposterior decompression and stabilization through posterior only approach versus traditional posterior approach.
In the present study, we attempt to compare single stage three-column fixation through posterior approach with traditional posterior approach in terms of surgical, neurological, and functional outcome.
Method
Patients
This is retrospective study comprising 47 patients of traumatic thoracolumbar fracture admitted between September 2016 and January 2018 in neurosurgery ward of Dr. Ram Manohar Lohia Institute of Medical Sciences, Lucknow. Inclusion criteria include patients who were adult and diagnosed as a case of unstable thoracolumbar fracture, underwent either single stage three-column fixation through posterior approach or traditional posterior approach. Posterior approach includes short-segment fixation including fracture segment (SSF IFS) and long-segment fixation (LSF). Patients were undergone either of surgery depending on implant availability in the operating room.
Preoperative Assessment
Fracture was classified according to modified AO classification. Neurological assessment of all patients was done with ASIA impairment scale, assessment of pain was done with visual analogue scale (VAS), functional assessment was done with lower extremities functional score, and kyphosis was measured with Cobb's angle.
Operative Procedure
In the single-stage three-column fixation through posterior approach, under general anesthesia and prone position, midline longitudinal incision was taken. Bilateral facia and muscle were dissected till transverse process one level above and below the fracture segment. The lamina of the fractured vertebrae and the lamina of the cranial vertebrae were removed to expose the pedicles of the fractured vertebrae. Discectomies were performed one level above and one level below the fractured vertebrae to expose the end plates. The pedicles of the fractured vertebrae were exposed completely. Unilateral or bilateral pedicular resection was achieved based on anterior compression. We routinely performed only one side pedicular resection as optimum anterior decompression and cage placement can be achieved unilaterally. Through the resected pedicle, the posterior two-thirds of the fractured vertebra were totally removed with a high-speed drill and hand curettes. To protect the dura while drilling, a thin layer of cortex was left posteriorly which was broken with curved curette later on. Anterior decompression of the dura was achieved. Superior and inferior disc spaces were cleared with curved hand curettes and end plates prepared. Anterior part of the fractured vertebrae was kept intact, but a space for putting expandable cage was formed within the fractured vertebrae. Pedicular screws were introduced into the upper and lower vertebrae and distraction was applied to assist the entrance of the expandable cage. The expandable cage, filled with autologous bone graft, was inserted into the fractured vertebrae through the resected pedicle. The nerve root exiting through the inferior foramen was protected during placement of the cage. Cage was distracted so that it settles optimally on the end plates. The distraction of the cage was guided by preoperative anteroposterior and lateral fluoroscopic image. Finally, pedicular screws were loosened and fixed again in compression. The autologous cancellous bone chips were embedded over the decorticated posterolateral gutter to augment fusion ([Figs. 1]
[2]).
Fig. 1 (A) Posterior laminectomy with transpedicular anterior decompression of cord with placement of pedicle screw and opposite side rod fixation. Posterolateral gutter is made for distractable cage. (B) Distractable cage is placed.
Fig. 2 Combined decompression and stabilization through posterior only approach: (A) Preoperative magnetic resonance imaging, (B) preoperative X-ray, and (C) postoperative X-ray.
For traditional pedicle screw fixation, pedicle screw was placed one level above and one level below including fracture segment and posterior decompression was done in SSF IFS method ([Fig. 3]). In LSF method, pedicle screw was placed two levels above and two levels below the fracture segment ([Fig. 4]).
Fig. 3 Short-segment fixation including fracture segment (SSF IFS): (A) Preoperative magnetic resonance imaging, (B) preoperative X-ray, and (C) postoperative SSF IFS.
Fig. 4 Posterior approach long-segment fixation (LSF): (A) Preoperative magnetic resonance imaging, (B) preoperative X-ray, and (C) postoperative LSF.
Duration of surgery, intraoperative average blood loss is noted in each surgery.
Follow-Up
All patients were discharged on seventh day after stich removal. Postop X-ray at 1 week and 3 months was done. Postoperatively, AIS, VAS, LEFS, and kyphotic angle was noted in all patients at seventh day and 3 months. To decrease bias in neurological and functional assessment, parameters were not measured in immediate postoperative days.
Data Analysis
Patient data were analyzed using STATA-12 Software. Unaired student t-test and Kruskal–Wallis nonparametric test were used for statistical analysis.
Results
Three patients had injury due to fall from height, while 17 patients had road traffic accident ([Table 1]). Out of 47 patients, 33 patients were operated through traditional posterior approach and 13 patients were operated through single stage three-column fixation through posterior only approach. In traditional approach, 21 patients were operated using SSF IFS method and 12 patients were operated using LSF.
Table 1
Patient-related parameters
|
No of patient (%)
|
|
Abbreviations: FFH, fall from height; RTA, road traffic accident.
|
|
Age (in years)
|
|
10–20
|
10 (21.3)
|
|
20–30
|
10 (21.3)
|
|
30–40
|
8 (17.0)
|
|
40–50
|
7 (14.9)
|
|
>50
|
12 (25.5
|
|
Sex
|
|
Male (M)
|
40 (85.1)
|
|
Female (F)
|
7 (14.9)
|
|
# Level
|
|
D11
|
1 (2.1)
|
|
D12
|
19 (40.43)
|
|
D12, L1
|
2 (4.26)
|
|
L1
|
22 (46.81)
|
|
L2
|
3 (6.4)
|
|
Mode of injury
|
|
FFH
|
30 (63.81)
|
|
RTA
|
17 (36.81)
|
|
Type of # (AO classification)
|
|
A3
|
10 (21.28)
|
|
A4
|
21 (44.7)
|
|
B2
|
4 (8.6)
|
|
B3
|
6 (12.8)
|
|
C
|
6 (12.8)
|
The mean age of patients treated through traditional approach is 38 ± 2.6 (standard deviation [SD]) years, whereas it is 32.9 ± 3.0 (SD) years in single-stage three-column fixation approach (combined). Out of 33 patients, those operated with traditional approach, 30 patients are male and 3 patients are female, whereas out of 14 patients operated with combined approach, 10 patients are male and 4 patients are female ([Table 2]). Overall L1 vertebrae fracture is most common in our study ([Table 1]). In traditional approach, most commonly patient with L1 fracture is operated, that is, 17 out of 33 patients, whereas patient with D12 level fracture is most commonly operated with three-column fixation approach, that is, 7 out of 14 patients ([Table 2]). Most common fracture type is AO class A4 in both approaches followed by AO class A3 in traditional approach and AO class B3 and C in combined approach ([Table 2]).
Table 2
Patient-related parameters according to surgical approach
|
Characteristics
|
Surgical approach
|
p-Value
|
|
Traditional
|
Combined
|
|
Abbreviations: AO, Association for the Study of Internal Fixation; SE, standard error
|
|
Mean age ± SE (in years)
|
41.0 ± 2.6 years
|
22.9 ± 3.0 years
|
p < 0.001
|
|
Sex (M/F)
|
M = 30, F = 3
|
M = 10, F = 4
|
p = 0.086
|
|
Fracture level
|
|
D11
|
1
|
0
|
|
|
D12
|
12
|
7
|
|
|
D12, L1
|
2
|
0
|
|
|
L1
|
17
|
5
|
|
|
L2
|
1
|
2
|
|
|
Fracture type (AO class)
|
|
A3
|
10
|
0
|
|
|
A4
|
12
|
9
|
|
|
B2
|
3
|
1
|
|
|
B3
|
4
|
2
|
|
|
C
|
4
|
2
|
|
Changes in neurological and functional status of in two groups were compared at preoperative state, 1 week and 3 months postoperatively ([Table 3]). The median change in AIS score at third month observed in traditional approach is 1 (IQR 0.1) and that observed in combined approach is 1.04 (interquartile range [IQR] 0.1). This difference was found to be statistically nonsignificant (p = 0.991). (Kruskal–Wallis test nonparametric test) The median change in LEFS score at third month observed in traditional approach is 34 (IQR 22.46) and that observed in combined approach is 38 (IQR 32.50) ([Table 3]). This difference was found to be statistically nonsignificant (p = 0.561) (Kruskal–Wallis test nonparametric test).
Table 3
Changes in neurological and functional outcome
|
Preoperative
|
Postoperative at 1 week
|
Postoperative at 3 Months
|
|
Traditional
|
Combined
|
Traditional
|
Combined
|
Traditional
|
Combined
|
|
Abbreviations: AIS, Asia Impairment Scale; ASIA, American Spinal Injury Association; LEFS, Lower Extremity Functional Scale.
|
|
AIS (ASIA impairment scale)
|
|
A
|
2
|
0
|
1
|
0
|
2
|
1
|
|
B
|
10
|
7
|
7
|
5
|
4
|
2
|
|
C
|
6
|
4
|
10
|
5
|
5
|
4
|
|
D
|
14
|
3
|
14
|
4
|
13
|
6
|
|
E
|
1
|
0
|
1
|
0
|
9
|
1
|
|
VAS
|
|
1–3 (mild pain)
|
0
|
0
|
26
|
12
|
33
|
14
|
|
3–7 (moderate pain)
|
1
|
0
|
7
|
2
|
0
|
0
|
|
7–10 (severe pain)
|
32
|
14
|
0
|
0
|
0
|
0
|
|
LEFS (changes in score)
|
|
<10
|
|
|
10
|
6
|
2
|
1
|
|
10–20
|
|
|
18
|
4
|
4
|
2
|
|
20–30
|
|
|
5
|
4
|
8
|
0
|
|
30–40
|
|
|
0
|
0
|
7
|
6
|
|
>40
|
|
|
0
|
0
|
12
|
5
|
The mean VAS score at third month in subjects treated with traditional approach was 1.8 ± 0.61 whereas in subjects treated with combined approach the mean VAS score was 1 ± 0.39 ([Table 4]). The difference in VAS scores observed at third month was found to be statistically significant (p < 0.001).
Table 4
VAS score and Kyphosis correction observed at 3 months after surgery
|
Traditional posterior
|
Combined posterior only
|
p-Value
|
|
Abbreviations: SD, standard deviation; VAS, visual analogue scale.
|
|
VAS score (mean ± SD)
|
1.85 ± 0.61
|
1.0 ± 0.39
|
0.0001
|
|
Kyphotic correction at 3 months in degree (mean ± SD)
|
11.73 ± 3.60
|
15.36 ± 4.25
|
0.0045
|
In traditional surgeries, there was an observed correction of 11.73° ± 3.6° in kyphosis, while in combined surgeries the observed correction in kyphosis was 15.36° ± 4.25° ([Table 4]). This difference was found to be statistically significant (p = 0.004) (unpaired student t-test).
In traditional approach, there was an observed average blood loss of 108 ± 27.32 mL, while in combined approach the observed average blood loss was 263 ± 40.84 mL ([Table 5]). This difference was found to be statistically significant (p = 0.001).
Table 5
Average blood loss and duration of surgery
|
Traditional posterior
|
Combined posterior only
|
p-Value
|
|
Blood loss
|
108 ± 27.32 mL
|
263 ± 40.84 mL
|
0.0001
|
|
Duration of surgery
|
120.3 ± 25.43 minutes
|
150 ± 13.01 minutes
|
0.0002
|
Likewise, in traditional approach, there was an observed duration of surgery 120.3 ± 25.43 minutes, while in combined approach the observed duration of surgery was 150 ± 13.01 minutes ([Table 5]). This difference was found to be statistically significant (p = 0.002).
Intraoperatively, dural tear was found in two patients which was repaired successfully. No other complications related to operative site infection or implant failure were reported in immediate postop period and at 3 months follow-up.
Discussion
The approach which gives maximum benefit to the patient of traumatic thoracolumbar fracture is still the matter of debate. Several research article and meta-analysis were published, but no study favors one over other. Few studies were done for comparing combined approach versus posterior approach. Till date no study was done to compare combined anteroposterior decompression with stabilization through posterior only approach versus traditional posterior decompression with stabilization.[14]
[15]
[16]
[17]
[18]
A meta-analysis was done by Oprel et al to compare combined anteroposterior surgery versus posterior surgery for thoracolumbar fractures to identify better treatment.[18] They concluded that a small significantly higher kyphotic correction and improvement in vertebral height observed for the combined anteroposterior group is cancelled out by more blood loss, longer operation time, longer hospital stay, higher costs, and a possible higher intra- and postoperative complication rate requiring reoperation. The surgeons’ choices regarding the operative approach are biased: worse cases tended to undergo the combined anteroposterior approach.
A prospective multicenter study done by Knop et al and a retrospective study done by Been and Bouma to compare combined approach versus posterior approach show similar results.[15]
[19] In these studies, combined method showed comparative high blood loss, longer operation time, longer hospital stay, higher costs, and a possible higher intra-and postoperative complication rate because of anterior approach. In our study, these complications and morbidity mostly associated with anterior approach are not found because combined decompression and stabilization were done through posterior only approach.
Our study shows kyphosis correction at 3 months is more in three-column fixation than traditional posterior approach. It shows that on 3-month follow-up examination, loss of correction is smaller in three-column fixation than traditional posterior approach as seen in previous studies on anterior and combined approach. Four case series on combined decompression and stabilization through posterior only approaches in patients of traumatic thoracolumbar fracture concluded that significant correction in kyphosis and nonsignificant loss of kyphotic correction.[10]
[11]
[12]
[13]
Postoperatively, pain on VAS score was significantly improved in both approaches, but it is significantly reduced in combined decompression through posterior only approach at 3 months than traditional posterior approach.
Our study showed that both traditional posterior approach and combined decompression and stabilizations through posterior only approach were equally effective when the neurological (AIS), functional (LEFS) parameters were compared. There were no differences in the neurological or functional outcome as concluded in other studies.
We found lower blood loss and shorter operation time in traditional posterior approach surgery than the three-column fixation through posterior only approach. This is expected because of additional procedure in the combined approach.
Anterior neurodecompression still remains the practice in many institutions, as experimental data have shown that early neurodecompression is crucial for regeneration of nerve tissue.[9] Our results showed that the choice of three-column fixation through posterior only approach or posterior approach for decompression did not influence the neurological or functional outcome; however, neurological and functional improvements were seen in each patient. This might be because more severely injured and higher AIS grade patients tend to get operated through combined decompression method thorough posterior only approach.
The limitations of the study were the relatively small study size and unusual allocation of the patients into groups. Patients were divided into two trial groups according to the implants available in the operating theater at the time of the patient's admission. The lack of funds was the reason why implants for combined surgery (expandable cage) were rarely available. On the other hand, these circumstances favor traditional posterior approach fixation, which is preferred technique in our institution.
The next step should be a prospective randomized controlled trial of longer follow-up with proper supply of needed implants.
Conclusion
Combined anteroposterior decompression and stabilization through posterior only approach is convenient for complete decompression of cord, stabilization, and restoration of vertebral height, and there is statistically significant kyphotic correction, pain relief (VAS) but there is no statistically significant neurological and functional outcome than traditional posterior approach. Most neurosurgeons are familiar to posterior approach; hence, it should be used in unstable thoracolumbar fracture whenever needed, while avoiding various dreaded complication of combined approach.
