Gunshot wounds to the spine: study of 246 patients

THESES

GUNSHOT WOUNDS TO THE SPINE: STUDY OF 246 PATIENTS (ABSTRACT)^*. THESIS. RECIFE, 2000.

HILDO ROCHA CIRNE DE AZEVEDO FILHO^**

The gunshot wounds of the spinal cord and cauda equina, surely, represent one of the most devastating calamities which can affect the human life. Very often, they cause irreversible neurological disturbances. In the civilian population, the specific literature is scarce of large retrospective series. Therefore, in the absence of extensive data, it is difficult to evaluate in depth the extension of the lesions, its complications and the role played by the several forms of treatment.

The objective of this thesis is to produce responses to several unsolved questions and to add information, which can improve the handling of the pathology.

This study was based on the analysis of 246 patients with gunshot wounds of the spinal cord and cauda equina, admitted to the Service of Neurosurgery, Hospital da Restauração, Recife, Brazil, between January, 1st, 1981 and June, 30th, 1998. Two hundred and thirty five (95.5%) patients were male and 11 (4.5%) were female. The age of the patients ranged from 10 to 65 years, with mean of 26.6 (SD=9.2). The median was 24 years. Twenty percent was less than 20 years old and 50% was younger than 25 years old. The neurological condition, by the time of the admission to the hospital, was judged according to the classification of Frankel ASIA/IMSOP. In group A were 147 (59.8%) patients; in group B, six (2.4%); in group C, 51 (20.8%); in group D, 35 (14.2%); and seven (2.8%) patients were allocated in group E. All patients had radiological investigation and 51 (20.7%) were submitted to computerised tomography of the affected vertebral segments. All patients have been injured by hand guns and 241 (98%) were hit by bullets of small caliber. The cervical spine was affected on 72 (29.3%) opportunities, the thoracic spine on 113 (45.9%) and the lumbar spine was injured on 61 (24.8%) occasions. The in-hospital time ranged from three to 182 days, with a mean of 28.3 (SD=23.5). The follow-up varied from three to 397 days, with a mean of 76.3 (SD=58.6).

One hundred and fifty seven (63.8%) patients were managed conservatively and 89 (36.2%) underwent surgical treatment. Thirty five (14.2%) patients died during the investigation. Urinary infection was present on 44.3%. Bedsores were detected on 35.9% and were significantly associated with the absence of motor activity (groups A & B). Laparotomy was carried out on 46 (18.7%) patients and was unrelated to mortality. Subcutaneous emphysema was evident on 18.7% and haemopnemothorax on 26.4% and both were unassociated with a higher mortality. Respiratory infection was diagnosed on 58 (23.6%) patients and was significantly related to the absence of motor power (groups A & B) and to a higher mortality rate. Pulmonary embolism was cause of death on 10 patients, in other words, on 4.1% of the population under study.

Amongst the 89 patients who were operated on, 15 (16.9%) had emergency surgery and 74 (83.1%) had elective operative procedures. All emergency surgeries were decompressive laminectomies. Amongst the elective surgical procedures, 61 (82.4%) were decompressive laminectomies, 12 (16.2%) constituted anterior approaches to the cervical spine for decompression and fusion, and on one (1,4%) opportunity the cervical spine was stabilised through a posterior approach. The percentages of surgical procedures related to the neurological status were as follow: 22.4% in group A; 83.3% in group B; 62.7% in group C and 54.3% in group D. No patient in group E was submitted to surgical treatment. The patients with incomplete lesions (groups B, C & D) had surgical procedures (60.9%) significantly higher than those with complete lesions (group A), which was 22.4%. On the other hand, group A patients had percentages of emergency interventions (27.3%) significantly more elevated than those with incomplete lesions (10.7%). The presence of the bullet inside the vertebral canal was significantly related to a higher rate of surgical and emergency procedures. Cauda equina injuries were significantly more operated on (72.3%) than spinal cord lesions (27.6%). Cerebrospinal fluid (CSF) fistula was present on 10 (4.1%) occasions and significantly associated with surgical procedures. Purulent meningitis was diagnosed on 12 (4.9%) patients and was significantly related to CSF leak, surgical treatment and emergency procedures. Thirty-one (21.1%) patients of the group A, two (33.3%) of the group B and two (3.9%) of the group C have died. Patients with complete lesions had significantly higher mortality rate than those with incomplete injuries (4.3%).

The mortality rate was 29.2% for affections of the cervical spine, 11.5% for the thoracic and only 1.6% for the lumbar spine aggressions. Included in the 211 patients who survived, 191 (90.5%) remained neurologically unchanged, 18 (8.6%) showed some degree of improvement and two (0.9%) patients deteriorated following decompressive laminectomies. The spinal cord injuries had improvement percentages of 8.5%, which was similar to the achievement of the cauda equina lesions (8.7%). The patients with metallic dust inside the vertebral canal demonstrated neurological improvement (13.3%) significantly better than those bearing the presence of the bullet inside the canal (3.1%). However, the radiological aspects of the metallic fragments did not correlate to the occurrence of neuropathic pain. Neuropathic pain was verified on 19.1% of the cauda equina lesions and on 9.5% of the spinal cord injuries, but the differences were not statistically significant. On the other hand, this type of pain was more evident when the lesion affected the thoraco-lumbar spinal cord than the cervical one. The outcome of the neuropathic pain was not significantly influenced by the surgical procedures. Amongst the 209 patients who survived and did not deteriorate following surgery, 10 (12.2%) out of 82 operated on, improved, as well as improved, 8 (6.3%), out of 127 conservatively managed. These differences did not reach statistical significance. All patients (12) submitted to emergency decompressive laminectomies did not change, while 10 (14.3%), out of 70 who had elective surgical procedures, showed some degree of neurological improvement. However, these differences were not statistically significant. Concerning cauda equina lesions, neurological progress was diagnosed on two (6.3%) out of 32 operated on and on two (15.4%) of the 13 conservatively managed, which differences did not obtain statistical significance. Studying the spinal cord injuries, unoperated patients and those treated by decompressive laminectomies, we observed that four (10.5%), out of 38 surgically treated, demonstrated neurological progress, while six (5.3%), out of 114 patients who did not undergo an operation, also showed some neurological improvement. These differences were not statistically significant.

Finally, according to the data extracted from this thesis, decompressive laminectomies were associated with a higher level of complications and did not influence positively the outcome of the patients. However, it is mandatory and we strongly recommend that a prospective multicentric study be organised with a larger population of patients. This will be, for sure, the only way we can better understand the whole complexity of this pathology, and therefore, be able to better manage and help these patients.

KEY WORDS: traumatic spinal cord lesion, gunshot wound.

** Address: Rua Apipucos 317/601, 52071-0000 Recife PE, Brasil.

*

Lesões da medula espinhal e cauda equina produzidas por projéteis de arma de fogo: estudo de 246 pacientes (Resumo). Tese apresentada à Faculdade de Ciências Médicas da Universidade de Pernambuco para Concurso de Professor Titular de Neurocirurgia.

Publication Dates

• Publication in this collection
  05 Oct 2001
• Date of issue
  Sept 2001