The influence of the type of intracranial lesion on the final outcome in a consecutive series of 277 severely head-injured patients was analyzed. Patients were studied with computerized tomography (CT) and underwent continuous measurement of intracranial pressure. They received identical treatment according to a standardized protocol. Outcome of patients with either epidural hematoma (38 cases), subdural hematoma (56 cases), brain contusion (87 cases), or diffuse brain damage (96 cases) was rather heterogeneous, and serial CT scanning allowed the authors to outline eight consistent anatomical patterns in the whole series which have stronger prognostic significance than the four major lesion categories mentioned above. Patients with pure extracerebral hematoma (19 cases), single brain contusion (45 cases), general brain swelling (41 cases), and normal CT scans (28 cases) had a significantly better outcome than patients developing acute hemispheric swelling after operation for a large extracerebral hematoma (27 cases), patients with multiple brain contusion, either unilateral or bilateral (74 cases), and patients with diffuse axonal injury (43 cases). These anatomical patterns are interesting because, in addition to having clinical and physiopathological significance, they provide useful prognostic information and facilitate improved therapeutic decision-making in severely head-injured patients.

The clinical, radiological, and histopathological features of 21 cases of angiographically occult intracranial vascular malformations (AOIVM's) are analyzed, and a review of 241 additional appropriately documented, histologically verified cases collected from the literature is presented. In all, there were 115 (43.8%) arteriovenous malformations, 82 (31.2%) cavernous angiomas, 26 (9.9%) venous angiomas, 10 (3.8%) cases of capillary telangiectasis, and 29 (11%) mixed or unclassified angiomas. The result of the analysis shows that there are no essential differences in the patterns of clinical presentation, the computerized tomography (CT) appearance, or the surgical prognosis among these pathological types of vascular malformations. Certain histological features common to all AOIVM's (such as the small caliber, the more or less complete thrombosis of the malformed vessels, and the changes induced in the surrounding brain tissue by repeated microhemorrhages) seem to determine the biological behavior of the anomaly rather than the predominant type of vessel involved. Thus, subdivision of AOIVM's into the four classical pathological types has little practical value. Most AOIVM's are visualized by the CT scan and show a rather typical appearance. Surgical removal, which prevents rebleeding and ameliorates or suppresses seizure activity, is usually easy to perform and represents the treatment of choice for patients with clinically symptomatic AOIVM's.

The clinical and computed tomographic (CT) findings in a series of 161 consecutive patients operated upon for postraumatic extradural hematoma are analyzed. Thirteen (8%) patients had delayed epidural hematoma formation. The overall mortality for the series was 12%, significantly lower than that observed during the prior "angiographic" period at the same unit (30%). Because all but 1 of the deaths occurred among the 66 patients unconscious at the time of operation (27% mortality in this subgroup), the authors sought differential factors between comatose and noncomatose patients at operation. There were no significant differences between these groups in age, sex, mechanism of injury, preoperative course of consciousness (lucid interval or not), or epidural hematoma location and shape. In contrast, significant differences were seen between the two subgroups in trauma-to-operation interval, hematoma volume, CT hematoma density (mixed low-high CT density vs. homogeneous hyperdensity), midline displacement, severity of associated intracranial lesions, and postoperative intracranial pressure (ICP). Patients comatose at operation usually evidenced a more rapid clinical deterioration (a shorter trauma-to-operation interval) and tended to have a large hematoma volume, a higher incidence of mixed CT density clot (hyperacute bleeding), more marked shift of midline structures, more severe associated lesions, and higher postoperative ICP levels.

The authors describe four cases of subependymoma studied with computed tomography (CT) and review 18 previously reported cases in an attempt to define the most characteristic CT presentation of this rare, benign tumor. Subependymoma usually appears as an isodense, or even hypodense, intraventricular tumor on plain CT scan and shows minimal or no enhancement in postcontrast studies. Differential diagnosis between subependymoma and the more malignant true ependymoma is difficult, particularly when the tumor occurs in the posterior fossa. Recognition of subependymoma should prompt the surgeon to attempt radical tumor removal because it can be achieved without sacrificing contiguous tissue and carries a good prognosis.

Mortality due to epidural hematoma is virtually restricted to patients who undergo surgery for that condition while in coma. The authors have analyzed the factors influencing the outcome of 64 patients who underwent epidural hematoma evacuation while in coma. These patients represented 41% of the 156 patients operated on for epidural hematoma at their centers after the introduction of computerized tomography (CT). Eighteen patients (28.1%) died, two (3.1%) became severely disabled, and 44 (68.8%) made a functional recovery. The mortality rate for the entire series was 12%, significantly lower than the 30% rate observed when only angiographic studies were available. A significant correlation was found between the final result and the mechanism of injury, the interval between trauma and surgery, the motor score at operation, the hematoma CT density (homogeneous vs. heterogeneous), and the hematoma volume. The patient's age, the course of consciousness before operation (whether there was a lucid interval), and the clot location did not correlate with the final outcome. The mortality rate was significantly higher in patients operated on within 6 hours or between 6 and 12 hours after injury than in those undergoing surgery 12 to 48 hours after injury. Compared with the patients operated on later, the patients undergoing surgery in the early period were, on the average, older and had more rapidly developing symptoms, more pupillary changes, lower motor scores at surgery, larger hematomas, a higher incidence of mixed CT density clots, more severe associated intracranial lesions, and higher postoperative intracranial pressure (ICP). The mechanism of trauma seems to influence the course of consciousness before and after surgery. Passengers injured in traffic accidents had a lower incidence of a lucid interval and longer postoperative coma than patients with low-speed trauma, suggesting more frequent association of diffuse white matter-shearing injury. The duration of postoperative coma correlated with the morbidity rate in survivors. Forty-eight patients (75%) had one or more associated intracranial lesions, and 70% of these required treatment for elevation of ICP after hematoma evacuation. An ICP of over 35 mm Hg strongly correlated with poor outcome; administration of high-dose barbiturates was the only effective means for lowering ICP in nine of 15 patients who developed severe intracranial hypertension after surgery. This study attempts to identify patients at greater risk for presenting postoperative complications and to define a strategy for control CT scanning and ICP monitoring.