Untitled

Gunshot injuries of the spine – a review of
49 cases managed at the Groote Schuur
Acute Spinal Cord Injury Unit
Registrar, Department of Orthopaedics, Stellenbosch University R. N. DUNN, M.B. CH.B., M.MED. (ORTH.), F.C.S. (S.A.) (ORTH.) Head of Spinal Services, Division of Orthopaedics, University of Cape Town dence of gunshot injuries. The Groote Schuur Trauma Unit sees between 70 and 120 gunshot injuries a month. Many ofthese involve the musculoskeletal system, and some the The Acute Spinal Injury Unit, relocated from Conradie spine. This is in stark contrast to the European experience.
Hospital to Groote Schuur Hospital in mid-2003, admit- Of 34 903 trauma cases seen in Scotland over 28 years there ted 162 patients in the first year of its existence. A large was only 1 case of gunshot injury of the spine.1 number of these injuries were the result of interpersonal The Acute Spinal Cord Injury Unit (ASCI), previously violence, particularly gunshot wounds.
based at Conradie Hospital, was relocated to Groote Schuur Aim. To review patients with gunshot injuries to the Hospital in 2003 as part of the government’s 2010 plan.
spine, with reference to neurological injury, associated This unit manages spinal cord injuries from the region on a injuries, need for surgery and complications.
referral basis. During the first year of opening, 162 patients Methods. A comprehensive database is maintained to were admitted with various spinal injuries. A large propor- collect data on all spinal injury admissions. These data, tion of these were gunshot injuries of the spine.
as well as case notes and X-rays, were reviewed for all A review of this subset of patients is presented.
gunshot spine patients admitted to the Acute SpinalInjury Unit over a year.
Forty-nine patients were identified. Thirty-eight were male and 11 female with an average age of 27.5 years(range 15 - 51 ± 8.53). The average stay in the acute All admissions to the ASCI unit are entered in a database.
Included in the data recorded is the mode of injury and the Results. The spinal injury was complete in 38 and neurological status. The latter is classified according to the incomplete in 8, with 3 having no neurological deficit.
American Spinal Injury Association (ASIA) scoring system, The level was cervical in 13, thoracic in 24 and lumbar in with motor function attributable to C5 - T1 and L2 - S1 12. Only 9 patients improved neurologically. The spine roots allocated a score between 1 and 5 based on the was considered stable in 43 cases. Stabilisation was Medical Research Council (MRC) power. This is done performed in the 6 unstable cases. The bullets were bilaterally. In addition sensation is scored in a similar fash- removed in 11 cases as they were in the canal.
ion on a provided form. The subsequent management, com- There were 55 significant associated injuries, viz. 14 plications and discharge neurological status are also haemo-pneumothoraces, 16 abdominal visceral injuries, 3 vascular injuries, 4 injuries of the brachial plexus and 3 Patients with gunshot injuries of the spine were identified of the oesophagus, 2 tracheal injuries, 1 soft palate from this database. Their case notes and X-rays were reviewed retrospectively, with a view to procedures per- Complications included 3 deaths and discitis in 3 formed, complications and neurological change. The cases, pneumonia in 6 and pressure sores in 6.
patients referred to the local rehabilitation service were Conclusion. Gunshot injuries of the spine are a preva- reviewed for further change in neurological status.
lent and resource-intensive cause of paralysis. There is A total of 49 patients with gunshot injuries of the spine a high incidence of permanent severe neurological were identified. Of these, 47 were from the ASCI service deficit, but usually the spine remains mechanically sta- and 2 from the adjacent private facility (University of Cape ble. Most of the management revolves around the asso- ciated injuries and consequences of the neurological All the injuries were from low-velocity weapons, as is typi- cal for our population. The most frequent calibre was 9 mm.
Of the 49 patients, 38 were male and 11 female. The aver- With the high level of crime in the Western Cape, the state age age was 27.5 (15 - 51 ± 8.53) years. The average in-hos- hospitals have witnessed a tremendous increase in the inci- pital stay was 30 (4 - 109 ± 28) days. Once the acute management was completed patients were discharged orreferred to the rehabilitation service depending on their neu-rological status. The rehabilitation service required patientsto be free of pressure sores before admittance. This led tosome of the prolonged hospital stays.
Neurological status
The anatomical level and neurological status of the spinalinjuries are tabulated in Table I. The vast majority werecomplete neurological injuries, with no incompletes in thethoracic group. Only 9 patients improved neurologically.
Three cervical completes regained 5 - 7 ASIA motor points.
This represents a single level of improvement, and probably Fig. 1. CT scan of comminuted body regarded as
indicates root escape rather than true cord recovery. One unstable.
complete patient regained 10 motor points and good sensoryrecovery.
Bullets in the canal
Of the 4 cervical incomplete injuries, 1 patient with no bony fracture recovered spontaneously. This was thought to The bullet was found to be in the canal in 12 cases – 1 cervi- be because of cord contusion as the bullet did not violate the cal, 4 thoracic and 7 lumbar. One patient absconded for canal or column. Another patient improved from a total of fear of retribution. The rest underwent surgery to remove 85 motor points to 98 at 3 months. This patient had a body fracture. None of those who recovered had violation of theircanals.
Associated injuries
There was no recovery in the thoracic group.
Three patients improved in the lumbar injury group. One There were many associated injuries, as listed in Table II.
patient improved from 70 to 88 motor points, then Of the 4 brachial plexus injuries, 3 were related to the cur- plateaued. This patient is now ambulatory with crutches.
rent injury and 1 to a previous gunshot wound. The latter Another had a 5-point improvement. Both of these were patient’s current injury was thoracic, and the brachial plexus incomplete injuries. Another patient judged on admission to injury was only explained when the X-ray revealed a previ- have a complete lumbar injury had return of distal motor ous bullet near the plexus. There was a high incidence of function after removal of the bullet.
associated chest trauma, with 14 cases having haemotho-races, pneumothoraces or lung contusions. These required Stability
intercostal drains to be inserted. There were 2 vertebralarterial injuries, one arteriovenous fistula requiring repair The injury involved the posterior elements in 25 cases, the and an axillary artery also requiring repair. Laparotomies body in 15 cases, and both in 2. There were no fractures in were performed in 9 patients for visceral injuries (Table III).
Our assessment of stability was based on the three-column system of Denis,2 as described for thoracolumbar fractures TABLE II. ASSOCIATED INJURIES (N)
from indirect forces. This was modified by the realisation Tracheal
that gunshot injuries are different to indirect forces, as they Oesophageal
are not associated with posterior column ligamentous Soft palate (Fig. 2)
injuries. Six patients were assessed as unstable. This was Brachial plexus
usually based on severe comminution of the anterior column Pneumo-/haemothorax
(Fig. 1), where it was felt that progressive kyphosis would Non-spinal fractures
occur. Three patients had cervical and 3 had thoracolumbar Vertebral artery
injuries. The cervical injuries were fused anteriorly by Axillary artery
means of an autogenous bone graft and plate, whereas the Liver, spleen, pancreas
thoracolumbar injury patients underwent posterior pedicle Diaphragm
TABLE I. NEUROLOGICAL STATUS PER
ANATOMICAL LEVEL.
TABLE III. ABDOMINAL SURGERY (N)
Splenectomy
Incomplete
Bowel repair
Complete
Loop colostomy
Hemicolectomy
improved
Renal stent
*Parentheses indicate number of patients with neurological improvement.
Nephrectomy
166 VOL 43, NO. 4, NOVEMBER 2005 SAJS
where if 2 columnsare involved the spineis considered to beunstable. In an indi-rect non-penetratingforce, one can extrap-olate that if there issevere anterior col-umn destruction,there is likely to beposterior ligamentousinjury. One then hasa clear indication formechanical stabilisa-tion. In gunshotinjuries of the spine itis possible for a singlecolumn to be injuredwithout associated lig-amentous injury.
Once there is an iso-lated body fracture itis often difficult todecide whether therewill be progressive Fig. 2. Gunshot C1 (transoral) with C1 arch fracture,
no neurology. The bullet was spontaneously spewed
out and the patient discharged.
Complications
spine, where there issupport from the tho- Six patients developed pneumonia, and 6 developed pressure sores. Only one deep-vein thrombosis was diagnosed. Three patients developed a pyogenic discitis (Fig. 3). Three deaths occurred and 1 patient developed postoperative sepsis.
Fig. 3. Magnetic resonance
image of subsequent discitis fol-
Discussion
lowing a gunshot injury of the
Gunshot injuries of the spine are endemic in the Western Cape. Management protocols have evolved as the incidence kyphosis is extremely functionally limiting in cases of quadri- has changed. Civilian low-velocity gunshot injuries are vastly or paraplegia, and we therefore proceed to anterior instru- different from the military experience which provided much mented fusion in these cases. Our conservative stance is of the earlier experience. As one would expect, survival is far echoed by others. Cornwell et al.6 reported on 141 thoracic likelier in the low-velocity group. Many facets of manage- spine gunshot fractures. Only 2 required surgical stabilisa- ment have been adopted from management of spinal injuries tion. Isiklar and Lindsey7 found that 10% were unstable in resulting from indirect trauma. Not all of these are appropri- their series. Three cases were cervical and 1 was lumbar.
Removal of bullets remains controversial. Concerns High-dose steroid administration for indirect, non-pene- regarding lead toxicity (plumbism) have been reported.8,9 The trating trauma has been promoted by Bracken and Shephad3 incidence is rare and should probably not be used as a reason in the first 8 hours following indirect spinal cord injury.
to remove all bullets. Bullets in disc spaces and joints are Despite this being of doubtful clinical benefit in this non- more likely to release heavy metals. Scuderi et al.10 found penetrating trauma group, some still administer steroids to only 12 cases of bullets in disc spaces over a 24-year period patients with gunshot wound injuries of the spine. Heary et among 238 gunshot injuries of the spine. Only 1 of these 12 al.4 and Levy et al.5 have shown that there is no benefit in developed clinical signs of lead toxicity. They recommend terms of neurological recovery in penetrating injuries. If one that rather than imperative bullet removal, signs of lead toxi- considers the risks of infection and immune compromise, it is illogical to administer steroids if there is no evidence of Bullets in the canal are more of a concern. Basic science neurological benefit. We therefore regard administration of research has shown that of the heavy metals likely to be in steroids in the gunshot spine scenario as inappropriate.
bullets, copper is the most toxic to the cord in the animal The issue of spinal stability in mechanical terms is a diffi- model.11 Bono and Heary12 reviewed the topic well and com- cult one, as there is no good classification for gunshot mented that firstly one should ‘do no harm’. Removal of the injuries, making interpretation somewhat subjective. Much bullet did appear to alter the rate and incidence of neurologi- of our knowledge and experience are based on indirect cal recovery, but there was an increased incidence of infec- injuries. Many base this on the 3-column system of Denis,2 Waters and Adkins13 reported that bullet removal did not Generally the spine is mechanically stable and the neurologi- alter infection rates or sensory recovery. However, in the cal status static. Management revolves largely around the cauda equina region they found that removal may increase associated injuries and supportive care of the paraplegia.
the neurological recovery. This was also observed in our Steroid administration is not indicated in these injuries.
cohort. Of course one cannot be sure this was not simply a Although bullets should not be removed routinely, there is a matter of passage of time rather than the actual intervention.
case for removal if they are in the canal, especially if the The current policy of our unit is to remove the round after a few days, to allow the dura to seal and inflammation to Because of the high incidence of associated injuries and resolve. When retrieving bullets from the canal it is impor- permanent neurological deficit, gunshot injuries of the spine tant to have recent X-rays available and imaging in theatre.
The bullet may move. In our series there was a case of entryat L3, but the bullet was removed from S1 posteriorly.
There are case reports that confirm this ‘wandering’ of thebullet.14,15 This tends to be in a caudal direction because of Connell RA, Graham CA, Munro PT. Is spinal immobilisation neces-sary for all patients sustaining isolated penetrating trauma? Injury 2003; widening of the canal from T10 downwards.
34: 912-914.
Decompression has frequently been performed in an Denis F. The three columns of the spine and its significance in the clas- attempt to improve neurological outcome. The results of sification of acute thoracolumbar spine injuries. Spine 1983; 8: 817-831.
intervention must be seen against natural history, where Bracken M, Shephad M. Administration of methylprednisolone for 24or 48 hours or tirilazad mesylate for 48 hours in the treatment of ASCI.
spontaneous improvement may occur. This may be an addi- Results of the 3rd National Acute Spinal Cord Injury Randomized tional root level, as seen in our study, or more significant Controlled Trial, NASCIS. JAMA 1997; 277: 1597-1604.
recovery in incomplete injuries where a degree of spinal cord Heary RF, Vaccaro AR, Mesa JJ, et al. Steroids and gunshot wounds to
the spine. Neurosurgery 1997; 41: 576-583.
contusion may have occurred. In our study no patient with Levy ML, Gans W, Wijesinghe HS, SooHoo WE, Adkins RH, cervical or thoracic canal violation showed recovery. In Stillerman CB. Use of methylprednisolone as an adjunct in the manage- those who recovered, it would appear that the recovery was ment of patients with penetrating spinal cord injury: outcome analysis.
Neurosurgery 1996; 39: 1141-1148.
because of resolving cord contusion. These cases were body Cornwell EE 3rd, Chang DC, Bonar JP, et al. Thoracolumbar immobi- fractures and probably cord injuries owing to the local energy lization for trauma patients with torso gunshot wounds: is it necessary? Arch Surg 2001; 136: 324-327.
Various authors have reported on decompression in this Isiklar ZU, Lindsey RW. Low-velocity civilian gunshot wounds of the
spine. Orthopedics 1997; 20: 967-972.
scenario. Kahraman et al.16 reported on 106 patients, where Linden MA, Manton WI, Stewart RM, Thal ER, Feit H. Lead poison- 60% were operated on. There was similar recovery in both ing from retained bullets. Pathogenesis, diagnosis, and management.
the surgically managed group and the conservatively man- Ann Surg 1982; 195: 305-313.
Grogan DP, Bucholz RW. Acute lead intoxication from a bullet in an aged group. The study by Benzel et al.17 showed root intervertebral disc space. A case report. J Bone Joint Surg Am 1981; 63:
improvement in complete injuries as opposed to conserva- tively managed cases. The incomplete cord injury and cauda Scuderi GJ, Vaccaro AR, Fitzhenry LN, Greenberg S, Eismont F. Long-term clinical manifestations of retained bullet fragments within the inter- equina injury groups showed improvement irrespective of vertebral disk space. J Spinal Disord Tech 2004; 17: 108-111.
decompression. This has been echoed by other authors.18,19 Tindel NL, Marcillo AE, Tay BK, Bunge RP, Eismont FJ. The effect of Stauffer et al.20 reported 19% iatrogenic instability in their surgically implanted bullet fragments on the spinal cord in a rabbitmodel. J Bone Joint Surg Am 2001; 83-A: 884-890. laminectomy group, which highlights a risk associated with Bono CM, Heary RF. Gunshot wounds of the spine. Spine J 2004; 4:
posterior decompression. Our policy is not to decompress unless there is neurological deterioration subsequent to Waters RL, Adkins RH. The effects of removal of bullet fragments injury, with compression demonstrated on imaging.
retained in the spinal canal. A collaborative study by the National Spinal
Cord Injury Model Systems. Spine 1991; 16: 934-939.
The associated injuries are a major factor in the manage- Gupta S, Senger RL. Wandering intraspinal bullet. Br J Neurosurg 1999; ment of these patients. Transvisceral injuries are a particular 13: 606-607.
concern because of possible infection of the spine. This Oktem IS, Selcuklu A, Kurtsoy A, Kavuncu IA, Pasaoglu A. Migration seems to be a less frequent problem than expected. Kumar of bullet in the spinal canal: a case report. Surg Neurol 1995; 44: 548-
550.
et al.21 reported on 31 cases (13 transcolonic injuries) treated Kahraman S, Gonul E, Kayali H, et al. Retrospective analysis of spinal with antibiotics for 2 - 43 days. None of their cases devel- missile injuries. Neurosurg Rev 2004; 27(1): 42-45.
oped a vertebral osteitis, and they recommend conservative Benzel EC, Hadden TA, Coleman JE. Civilian gunshot wounds to the
spinal cord and cauda equina. Neurosurgery 1987; 20: 281-285.
treatment of the spine. Kihtir et al.22 reviewed 21 transperi- Hammoud MA, Haddad FS, Moufarrij NA. Spinal cord missile injuries toneal gunshot injuries, including 5 transcolonic injuries.
during the Lebanese civil war. Surg Neurol 1995; 43: 432-437.
There were no vertebral infections. Roffi et al.23 followed up Kupcha PC, An HS, Cotler JM. Gunshot wounds to the cervical spine.
42 patients with 51 visceral perforations. These included 14 Spine 1990; 15: 1058-1063.
Stauffer ES, Wood RW, Kelly EG. Gunshot wounds of the spine: the colonic and 15 small-bowel injuries. They used antibiotic effects of laminectomy. J Bone Joint Surg Am 1979; 61: 389-392.
cover and reported 3 spinal infections. They concluded that Kumar A, Wood GW 2nd, Whittle AP. Low-velocity gunshot injuries of early bullet removal did not seem to be helpful.
the spine with abdominal viscus trauma. J Orthop Trauma 1998; 12:
514-517.
Kihtir T, Ivatury RR, Simon R, Stahl WM. Management of transperi- Conclusion
toneal gunshot wounds of the spine. J Trauma 1991; 31: 1579-1583.
Roffi RP, Waters RL, Adkins RH. Gunshot wounds to the spine associ- Gunshot injuries of the spine are devastating to the individu- ated with a perforated viscus. Spine 1989; 14: 808-811.
al in terms of the severe neurological consequences.
168 VOL 43, NO. 4, NOVEMBER 2005 SAJS

Source: http://www.spinesurgery.co.za/Attachments/SAJS%20Nov%2005%20GSW%20Spine.pdf