Indian Journal of Neurotrauma (IJNT)

55
Original article
2010, Vol. 7, No. 1, pp. 55-60

Unstable dorsolumbar fractures: A prospective

series of 94 cases
N Basheer MS, Deepak Gupta M Ch, GD Sathyarthi M Ch, Deepak Aggarwal M Ch,
S Sinha M Ch, BS Sharma M Ch, AK Mahapatra M Ch
Department of Neurosurgery, JPNA Trauma Centre, AIIMS,New Delhi.

Abstract: Dorsolumbar trauma is the most common cause of paraparesis or paraplegia. Optimal
goals of the management include establishment of a painless, balanced and stable spinal column with
vertebral fusion. We reviewed various types of fractures of dorsolumbar spine, their management
and outcome in a prospective study of 94 cases of dorsolumbar trauma managed surgically at our
centre (Jan 2008 – May 2009). All patients underwent complete neurological examination, CT and MR
imaging of the spine. Most common mode of injury was fall from height (66%). Majority of the
patients belonged to Frankel grade A (66 %). Eighteen patients (19 %) were operated using anterior
approach and rest by posterior approach. There was no deterioration in neurological status in any
of the patients while 26 patients had improvement. Three cases died out of which one was attributed
to associated head injury. We conclude that fall from height is the most common cause of dorsolumbar
fracture with majority affected belonging to young population and had significant deficits, thus
causing significant burden on the society. Surgical management is safe and helps in early mobilization
and rehabilitation, thus facilitating possible neurological recovery.
Keywords: dorsolumbar, fracture, paraplegia, transpedicular, rehabilitation

INTRODUCTION for the treatment of dorsolumbar burst fractures are

generally accepted.
Trauma to the thoracic and lumbar vertebra is one of
the most common cause of traumatic paraparesis or OBJECTIVE
paraplegia 1,2. These can occur with or without bowel-
Objective of the study was to review various types of
bladder involvement. Dorsolumbar fractures often cause
fractures of dorsolumbar spine, their management and
a neurologic deficit and present a significant economic
outcome managed by department of neurosurgery, at
burden to the family and society. Accepted methods of
Jai Prakash Narayan Apex Trauma Centre, AIIMS, New
treatment of dorsolumbar burst fractures include
Delhi from January 2008 till May 2009.
conservative therapy, posterior reduction and instru-
mentation, and anterior decompression and MATERIALS AND METHODS
instrumentation.
All surgically treated dorsolumbar fractures between
Early mobilization and rehabilitation is the most January 2008 and May 2009 were reviewed
important aim of the management1,3. Majority of the retrospectively. All patients had undergone complete
dorsolumbar fractures are unstable. Optimal goals of the neurological examination, computed tomography scan
management include establishment of a painless, and magnetic resonance imaging. Clinically Patients were
balanced and stable spinal column with fusion of least graded using Frankel classification of neurological deficits
number of vertebra4. The management of dorsolumbar pre- and postoperatively as follows:
fractures has been the subject of much controversy. There
exist different criteria for the choice of the management A. Absent motor and sensory function
based on the severity of kyphotic deformity, canal B. Sensation present, motor function absent
compromise, vertebral height loss, and neurologic
status1,4,5. To our knowledge, none of the existing criteria C. Sensation present, motor function active but not
useful (grade 2—3/5)
Address for correspondence:
Dr Deepak Kumar Gupta D. Sensation present, motor function active and
Department of Neurosurgery, AIIMS, New Delhi 110029 useful (grade 4/5)
Manuscript for consideration for best paper award at the 17th Annual E. Normal motor and sensory function
Conference of the Neurotrauma Society of India, 06-08 Aug 2009.

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010

56 N Basheer, Deepak Gupta, GD Sathyarthi, D Aggarwal, S Sinha, SS Kale, BS Sharma, AK Mahapatra

Preoperative plain radio- graphs with sagittal and axial

computed tomography (CT) scans were reviewed regarding
three separate characteristics of the fracture site.
1) The amount of vertebral body actually comminuted
by the injury, as best seen in sagittal CT, or lateral
plain X-ray.
2) The apposition of the fracture fragments, as best
seen on axial CT cuts through the fracture site.
3) The fractures were categorized to 5 main groups
according to mode of injury and pathomorphological
uniformity6:
• 1. Compression Fracture
• 2. Burst Fractures
Fig 2: A case of fracture L1, pre op x ray, CT Scan and MRI
• 3. Flexion Distraction Injuries
• 4. Fracture Dislocations
• 5. Spondyloptosis
All patients were operated and fixation with fusion
was done. Post operatively patients underwent either
X-ray or CT and were examined for construct placement,
spinal canal decompression and kyphosis correction. All
of them were mobilized using dorsolumbar brace from
the second postoperative day. Regular physiotherapy was
done in all patients after the surgery (Figs 2–4).
Analysis was done using SPSS 13 for windows.
RESULTS
We analyzed all surgically treated dorsolumbar fractures Fig 3: Post operative CT
treated by neurosurgery department of this tertiary care scan of the same patient
trauma centre from January 2008 till May 2009. showing transpedicular
rod and screw fixation
Epidemiology: Our patients ranged from 10yrs to 65
yrs (Mean ± SD 31.5± 12.15) with 71 males and 23
females. Most importantly majority of them belonged
to age group of 20- 39 (65%) (Fig 1). Most common
mode of injury was fall from height (66 %). Road traffic

Fig 1: Distribution of age (n = 94) Fig 4: A case of L4-L5 spondyloptosis

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010

 Unstable Dorsolumbar spinal trauma: a single institutional experience of 94 cases at level I apex trauma centre 57

accidents accounted for 26 out of which almost all were patients developed sepsis with multiorgan dysfunction
due to either because of running over by the vehicle or syndrome. Three patients died in perioperative period,
due to high speed acceleration deceleration injury .One two of them had severe chest infection with sepsis &
patient had fracture due to gunshot injury. other one had associated head injury. Four patients had
construct failure with screw pullout and had to be
Clinical Profile: Majority of our patient (67%) belonged
operated again.
to Frankel grade A, i.e., complete motor and sensory
loss below the injury. Only 20 % of the patients had Early Outcome: Immediate outcome was measured using
some useful motor sensory function. Sixty five patients Frankel grading on post operative day 7. Mean Frankel
(68%) had bowel bladder involvement. Majority of the grading on post operative day 7 was 2.2 ± 1.6 compared
patients (90%) had some neurological deficit. Twenty to pre operative mean of 1.93 ± 1.4. None of the patient
nine patients (32%) had associated systemic injuries (16 had neurological detioration. Overall 23 patients had
out of 35 in high velocity injury compared to 13 out of improvement with 12 of them had significant motor
59 low velocity injury) with 13 patients having associated improvement. Out of 75 patients belonged to Frankel
head injury. Twenty two patients had multiple level A- C, 14 patients improved compared to 9 out of 9 in
vertebral fractures, (34% and 8 % in high and low velocity Frankel grade D. But there was no statistical significance
group respectively). for Frankel grade in predicting the improvement of the
patient (Table 2). The surprising findings were 3 patients
Radiology: Most common vertebra involved was L1
of Frankel grade A also showed improvement in their
vertebral body (36 patients). Common type of fractures
neurological status.
were compression and burst fractures both accounting
for 26 each (Table 1). There were 7 cases of Fourteen out of 32 (43 %) patients who were operated
spondyloptosis. On MR imaging 28 patients (29.8%) within 7 days of injury showed improvement compared
showed complete cord transection. Another 43 (45%) to 9 out of 62 who were operated after 7 days of injury
showed cord signal changes with 23 (24.3%) showing (Table 3 ). This figures show that early surgery is beneficial
normal cord. in neurological recovery of these patients. But the P
value was not statistically significant (p = 0.16).Our study
Surgical Management: Mean duration from injury to
also showed that there was no correlation of outcome
surgery was 11 ± 5.6 days. Majority of them (76) were
with with the surgical approach (anterior or posterior )
approached posteriorly .Eighteen patients were operated
preoperative methyl prednisolone, other systemic injury
through anterolateral approach. Seven patients were
or hypotension (Tables 4 & 5).
operated through minimally invasive percutaneous screw
and rod placement. Short segment fixation was done in Final Outcome: Final outcome was measured at 3 months
61 patients where as in 33 patients long segment fixation follow up or at last follow up visit. Mean follow up was
was done. Synthetic graft was used in 42 patients and 2.3 ± 1.2 with range of 1 to 6 months. Out of 94 patients
autologous bone graft was used in remaining 45 patients 34 patients had a follow up of 3 months. Out of 34
for fusion. Both anterior and posterior column fusion patients, 21 (61.6%) showed some improvement,
was done in 20 patients.
Table 2: Correlation of prognosis with
Post operative Complications: Most common Frankel grade on admission
complication was chest infection (14%). Wound Frankel Grade Improved Same P value
infection was present in 5 patients (6 %). One patient A (n = 63) 3 (4.7%) 60 0.126
had unilateral vision loss due to prone position. Three
B/C(n=12) 11(91.6%) 1
Table 1 D (n = 9) 9 (100%) 0
E (n = 10) 0 10
Compression 26 (27.6%)
Flexion distraction 17 (18.1%) Table 3: Correlation with timing of surgery
Fracture dislocation 16 (17%) Duration before surgery <7 days (N=32) >7 days (N=62) P value
Burst 28 (29.8%) improvement 14 (43%) 9 (14.5%) 0.16
Spondyloptosis 7 (7.3%) Same 18 53

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010

58 N Basheer, Deepak Gupta, GD Sathyarthi, D Aggarwal, S Sinha, SS Kale, BS Sharma, AK Mahapatra

Table 4: Comparison of anterior and posterior approach & outcome Table 8: Other factors in prognosis: final outcome
Improved Same P value
Anterior Posterior P value
Approach (n=18) Approach (n=76) Age <20 (n=5) 3 (60%) 2 (40%)
20-40 (n=19) 12 (63%) 7 (37%) 0.663
Improvement 5 (27.78 %) 18 (23.7 %) 0.672(NS) > 40 (n=10) 6 (60%) 4 (40%)
Same 13 58 MRI
Worse 0 0 Cord Change (n=27) 16 (59%) 1 1 (41%) 0.09
No change (n= 7) 5 (71.4%) 2 (29.61%)
Table 5: Other prognostic factors Single fracture (n=21) 1 7 (80.1%) 5 8 (19.86%) 0.74
Improved Same P value Multiple fracture (n=13) 4 (30.73%) 1 3 (69.26%)
Age < 20 (n=13) 3 (23.07%) 10 (77%) Mechanism of injury
20-40 (n=60) 14 (24.2%) 46 (75.7%) 0.763 Low velocity (n=23) 1 4 (60.86%) 9 (39.13%) 0.85
> 40 (n=21) 6 (28.5%) 15 (71%) High velocity (n=11) 7 (63.65%) 4 (36.31%)
MRI Systemic injury (n=10) 4 (40%) 6 (60%) 0.43
Cord Change (n=71) 6 (8.4%) 65 (91%) 0.089 Methyl prednisolone (n=7) 3 (42.85%) 4 (58.15%) 0.69
No change (n= 23) 17 (73.9%) 6 (26.1%)
Single fracture (n=77) 19 (24.6%) 58 (75.4%) 0.912
Multiple fracture (n=17) 4 (23.53%) 13 (76.5%) DISCUSSION
Mechanism of injury Dorsolumbar fractures account for the most common
Low velocity (n=59) 14 (23.72%) 45 (76.3%)
High velocity (n=35) 9 (25.71%) 26 (74.28%) 0.84 cause of traumatic paraplegia. Most of the affected belong
to the productive age group, thus having a major
Systemic injury (n=29) 6 (20.6%) 23 (79.3% 0.231
economic burden on the society. The aim of treatment
Hypotension (n=12) 3 (25%) 9 (75%) 0.169
is restoration of function of the patient by creating a
Methyl prednisolone(n=30) 7 (23.3%) 23(76.7%) 0.693 healing environment to allow a stable pain free spinal
16 (47.5 %) of them showing significant motor column, with the minimal risk to the patient7,8,9.
improvement. None of the patient had deterioration. The management of fractures in the thoracolumbar
The only factor which was significant in deciding the region is a controversial subject. Disadvantages of
outcome was pre operative Frankel score, with almost conservative treatment include deterioration in
all patients who had preoperative frankel of C, D or E neurological status in 17% of the patients, progressive
showing improvement (P Value.032). None of the other kyphotic deformity in 20%, persistent backache,
factors such as age, time of surgery, pre operative decubitus ulcer and deep venous thrombosis. Most of
hypotension, or other associated injury showed significant these complications can be avoided by early mobilization
association with final outcome. (See Tables 6, 7 & 8) and decreased hospital stay by early surgery 7,9,10. Patients
which we operated ,majority had severe neurological
Majority of the patients (27) were still voiding through deficits, with our pre op mean Frankel score being 1.93
urinary catheter. Bedsore was present in 6 (17.6%) ± 1.4. This is very low compared to all the studies
patients. conducted previously as shown in the table. This could
Table 6: Frankel grade and final outcome explain the lower percentage of neurological
Frenkel Grade Improved Same P value improvement compared to others in our study (See
Table 9). As shown with other studies pre operative
A (n=13) 5 (38.4%) 8 0.032
Frankel score was the single most important factor
B/C(n=10) 9 (90 %) 1 deciding the neurological outcome 5,6. Even though
D(n=7) 7(100%) 0 statistically insignificant, greater fraction of patients
E(n=4) 0 4 Table 9: Management of dorsolumbar spine trauma
in the literature
Table 7: Correlation with Timing of surgery: final outcome
Study (n) Mean Frankel Improved Worse
Duration before surgery < 7 days > 7 days P grade
(N= 11) (N=23) value Patrick w. Hitchon etal 8 63 3.7 ± 1.1 23 2
improvement 8 (72 %) 13 (56.5 %) 0.417 Mohammad F. Butt et al 11
50 2.2± 1.21 24 0
Same 3 10 Present study 2009 94 1.93 ± 1.4 23 0

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010

 Unstable Dorsolumbar spinal trauma: a single institutional experience of 94 cases at level I apex trauma centre 59

operated within 7 days showed improvement compared Spine Concept in Acute Spinal Trauma.
to those who where operated after 7 days. As shown by Clin Orthop 1984;189:65–76.
other studies there was no correlation of outcome with 3. Denis F. The Three Column Spine and Its Significance in the
the different surgical approaches6. Classification of Acute Thoracolumbar Spinal Injuries.
Spine 1983;8:817–31.
The study showed that almost all patients with
4. Panjabi MM, Oxland TR, Kifune M, Arand M, Wen L, Chen
incomplete cord injury improved with time. only factor A. Validity of the Three-Column Theory of Thoracolumbar
which was significant in deciding the outcome was pre Fractures. A Biomechanic Investigation.
operative frankel score, with almost all patients who had Spine 1995;20:1122–7.
preoperative frankel of C, D or E showing improvement 5. Alvine GF, Swain JM, Asher MA, Burton DC. Treatment of
(P Value.032) as shown by other studies as well6,7. Even thoracolumbar burst fractures with variable screw placement
in complete cord injury the incidence of complication or Isola instrumentation and arthrodesis: Case series and
due to immobilization of patients were reduced literature review.
dramatically with improved quality of life. Thus early J Spinal Disord Tech 2004; 17:251–64.
surgery can help in rehabilitation and mobilization of 6. Gertzbein SD. Spine update. Classification of thoracic and
patients thus preventing the complications such as lumbar fractures.
decubitus ulcers, chest infection deep venous thrombosis Spine 1994; 19:626–8.
etc. this also makes patients independent on others for 7. Lemons VR, Wagner FC, Montesano PX. Management of
their daily activities12. thoracolumbar fractures with accompanying neurological
injury.
CONCLUSION Neurosurgery 1992; 30:667–71.

Fall from height is the most common cause of 8. Hitchon PW, Torner JC, Haddad SF, Follett KA. Management
dorsolumbar fracture with majority affected belonging options in thoracolumbar burst fractures.
Surg Neurol 1998; 49: 619–27.
to young population and presenting with significant
deficits, thus causing significant burden on the society. 9. Rechtine GR II, Cahill D, Chrin AM. Treatment of
Surgical management is safe and helps in early thoracolumbar trauma: comparison of complications of
operative versus nonoperative treatment.
mobilization and rehabilitation, thus facilitating possible
J Spinal Disord 1999; 12:406–9.
neurological recovery. Long term follow up of patients
are awaited. 10. Hitchon PW, Torner JC. Recumbency in thoracolumbar
fractures.
REFERENCES Neurosurg Clin N Am 1997; 8:509–17.

1. Hassan Dashti, Haw Chou Lee, Eldin E Karaikovic, Robert 11. Mohammad F. Butt , Munir Farooq, Bashir Mir, et al.
W. Gaines Jr. Decision making in dorsolumbar fractures. Management of unstable thoracolumbar spinal injuries by
Neurology India 2005; 53:Issue 4. posterior short segment spinal fixation.
International Orthopaedics 2007; 31: 259-64.
2. Denis F. Spinal Instability As Defined by the Three-Column

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010

60 N Basheer, Deepak Gupta, GD Sathyarthi, D Aggarwal, S Sinha, SS Kale, BS Sharma, AK Mahapatra

Indian Journal of Neurotrauma (IJNT), Vol. 7, No. 1, 2010