Ventricular Injury following Cranial Gunshot Wounds: Clinical Study

Military Medicine, Sep 2004 by Erdogan, Ersin, Izci, Yusuf, Gonul, Engin, Timurkaynak, Erdener

Gunshot wounds to the head are usually fatal injuries, despite all medical and surgical interventions. Ventricular injury is a poor prognostic factor for penetrating cranial gunshot wounds. Intraventricular hemorrhage and ventricular lacerations are the main components of such injuries. The incidence, management, and outcomes of cases of ventricular injury secondary to cranial gunshot wounds that were treated during a 9-year period at G�lhane Military Medical Academy were examined. The study group consisted of 67 consecutive patients who were admitted to the Department of Neurosurgery with the diagnosis of ventricular injury, with different penetration sites. The patients had been injured by either bullets or shrapnel. Surgical treatment was performed for all patients with ventricular injuries and 22 (32.8%) died. Ventricular injury in cranial gunshot wounds is a complex severe type of trauma that requires serious treatment. Early radiological diagnosis and accurate treatment frequently had lifesaving roles for these patients.

Introduction

The ventricular system of the brain is developmentally derived from the cavity of the neural tube. It consists of the lateral ventricles, the third ventricle, and the fourth ventricle. The ventricles are lined throughout with ependyma and are filled with cerebrospinal fluid (CSF).

Ventricular injury caused by penetrating craniocerebral gunshot wounds is a rare and usually life-threatening condition. Bullets and shrapnel are the most common agents of such injuries; they penetrate the walls of the ventricles after damaging the skull and adjacent cerebral tissues. Computed tomography (CT) has become the neuroradiological method of choice for the initial evaluation of patients suspected clinically of harboring a ventricular injury. Intraventricular blood, air, or foreign bodies detected on CT scans are the diagnostic features of a possible ventricular injury.

Methods

The study group consisted of 67 consecutive patients admitted to the Department of Neurosurgery with the diagnosis of ventricular injury resulting from penetrating craniocerebral gunshot wounds or missile injuries between 1992 and 2001. Data for all patients were analyzed retrospectively. Clinical, radiological, and surgical data were collected and investigated in detail. The initial presentation of each patient was assessed with the Glasgow Coma Scale (GCS) and results are summarized in Table I. Radiological investigation, including cranial X-rays and CT, was performed for all patients and noted the penetration and injury sites for each case (Fig. 1). We performed digital subtraction angiography for 17 patients with suspected traumatic aneurysms. These cases included those with wounding agents near important vascular regions, those with intracerebral hematomas evident on CT scans, and those in which an unexpected neurological event occurred. No aneurysm was identified. After the neurological and radiological examinations, patients underwent surgical treatment with various techniques. Primary suturing, craniectomy, debridement of necrotic tissues, evacuation of hematomas or foreign bodies, and durai repair were some of the surgical interventions performed for patients with craniocerebral injuries. In the postoperative period, close clinical and radiological follow-up monitoring was performed in the intensive care unit, and patients were assessed with the GCS. Postoperative complications such as CSF leaks, infections, hydrocephalus, and convulsions were detected and treated with either medical or surgical intervention. External ventricular drainage (EVD) was performed for patients with massive intraventricular hemorrhage (IVH) causing increased intracranial pressure and obstructive hydrocephalus. Cranial defects that occurred during the injury or surgery were repaired with methylmethacrylate or porous polyethylene, and good neurological and cosmetic results were obtained.

All surviving patients were monitored for 6 months to 9 years after discharge; the GCS scores and radiological findings were recorded. Both discharge and follow-up evaluations were performed in our department.

Results

We evaluated 67 consecutive patients with ventricular injuries who were treated in our department. Sixty-six (98.51%) were male and 1 (1.49%) was female. The mean age was 23.1 years (range, 19-45 years). The mean time between the injury and neurosurgical intervention was 2 hours. The GCS scores at admission were 8 to 10 for 25 patients (37.31%), 5 to 7 for 23 patients, 11 to 13 for 6 patients, and less than 5 for 13 patients. Shrapnel fragments were the most common wounding agents (n = 36, 54%), followed by bullets (n = 31, 46%). The CT findings for all patients were as follows: IVH for 62 patients (92.54%; Fig. 2), intraventricular foreign body (bone or metallic fragment) for 33 patients (49.25%; Fig. 3), and intraventricular air for 3 patients (4.48%). The lateral and third ventricles were the most common sites of hemorrhage, as observed for 59 (95.16%) of 62 patients. Blood in the fourth ventricle was detected on CT scans for only three patients (4.84%). The most common site of penetration was the frontal lobe, as noted for 26.4% of patients with ventricular injuries. After the first neurological and radiological examinations, primary suturing was performed for 13 patients (19.40%). Craniectomy, debridement of necrotic tissues, removal of foreign bodies, and durai repair were performed for 54 patients (80.60%). During the operations, we removed only palpable and visible foreign bodies (metallic or bone fragments) that were in the operative field. Bone and metallic fragments in deep or intraventricular locations were left in place (Fig. 3). The dura was closed for 54 patients who underwent operations with either primary closure (7 cases, 12.9%) or duraplasty (47 cases, 70.1%). All patients received third-generation cephalosporins (3 g/day) and metronidazole (7.5 mg/kg per day) for 14 days postoperatively. In addition, standard medication protocols, including antiedema, anticonvulsant, and analgesic agents, were followed. Seventeen patients (25.37%) whose CT scans showed extensive brain edema were intubated and connected to a mechanical ventilator. We monitored intracranial pressure closely for only 15 patients (22.38%), whose GCS scores were between 4 and 8; they were given antiedema therapy as required. Postoperatively, all patients were monitored with CT scans. EVD was performed, via Kocher's point, for 42 patients (62.69%) with intraventricular bleeding. Obstruction occurred and the intraventricular catheter was reopened for five patients. Ventriculoperitoneal shunts were required for two patients (2.98%) with obstructive hydrocephalus after the evacuation of IVH.