29 

Does Decompressive Craniectomy Really Improve Outcome After Head Injury?

Brian T. Andrews

BRIEF ANSWER

Background

From an evidence-based perspective, decompressive craniectomy has adequate class II support in recent literature to support its use as a level II recommendation for the treatment of refractory elevations in ICP after head injury. In terms of improving clinical outcome, it should be considered a level III recommendation because the only sound data that are available are limited to prospective or retrospective case series that often differ in the exact clinical time point at which decompression was performed. This limitation of the data makes it difficult to determine accurately the effect of decompression on clinical outcome. There exist no reliable prospective randomized controlled trials of this technique. Historical perspective gives us little useful information because decompressive craniectomy was generally applied in the past only as a last resort when all other options had been exhausted, and thus for patients beyond any reasonable hope for a functional recovery.

Literature Review

Physiologic Effects

In terms of ICP, persistent elevations of ICP after severe head injury are associated with poor outcome (class II data).¹ Miller et al² found that head-injured patients with uncontrollable elevations of ICP beyond 20 mmHg had a mortality rate of 100%, whereas those with ICP increases above 20 mmHg who responded to therapy had a mortality rate of 29%. Normal ICP throughout a patient’s course was associated with a mortality rate of only 14% (class II data).

Decompression is an effective means of lowering elevated ICP. Yoo et al³ measured ICP before and after bilateral craniectomy, as well as before and after the dura was opened, in 20 patients with massive brain swelling (class II data). The mean ICP decreased to 50.2% of the initial ICP after craniectomy alone and decreased an additional 34.5% after durotomy; the final pressure was reduced by an average of nearly 85%. Kunze et al⁴ showed that, in 28 patients with an initial mean ICP of 41.7 ± 10.1 mmHg, postoperative ICP was reduced to 20.6 ± 15.3 mmHg (class III data). Polin et al⁵ also noted a highly significant (p =.0003) reduction in ICP in 35 patients who underwent bifrontal decompression for diffuse traumatic cerebral swelling; postoperative reduction was also significant when compared with ICP values at 48 to 72 hours postinjury of weighted control patients from the Traumatic Coma Data Bank (class II data).

In a computed tomographic (CT) analysis, Münch et al⁶ have shown that unilateral decompressive craniectomy for patients with acute subdural hematomas with severe mass effect significantly decreases midline shift and improves visibility of the mesencephalic cisterns (class III data).

In the study described above, Yoo et al3 reported that in 20 patients with massive diffuse brain swelling and elevated ICP treated with bilateral decompression, the overall mortality rate was 20% (class II data). Fully 69% of survivors (11 of 16 patients) had a good outcome, defined as a Glasgow Outcome Scale (GOS)⁷ score of 4 or 5. Importantly, persistent elevation of ICP above 35 mmHg during the first 24 hours after decompression correlated with 100% mortality.

In the report by Polin et al,⁵ bifrontal craniectomy was performed in 35 patients with ” malignant posttraumatic cerebral edema“ (class II data). The overall mortality rate was 23%, with 37% of patients achieving a good outcome or moderate disability (GOS 4 or 5); children had a higher rate of favorable outcome (44%). Patients who underwent surgery more than 48 hours after injury had worse outcomes than those who were decompressed within 48 hours.

Unilateral decompressive craniectomy has also been reported to be beneficial. In Kunze et al’s⁴ retrospective review of 28 patients with uncontrollable elevation of ICP, 17 of the 28 were treated with unilateral craniectomy for swelling that was primarily located in one hemisphere (class III data). Overall mortality was 11%, and 54% of patients had a good outcome (GOS 4 or 5) after 1 year. These authors found that patients under 30 years of age had particularly good outcomes and low mortality.

Guerra et al⁸ prospectively treated 57 patients with unilateral (31 patients) or bilateral (26 patients) decompression (class II data). Overall mortality was 19%, and GOS scores of 4 or 5 were achieved by 58% of patients. These authors found that the Glasgow Coma Scale (GCS) score on the first day after trauma and the mean ICP were most predictive of outcome. Of interest, however, Münch et al⁶ found that when unilateral decompression was performed for patients with a unilateral intracranial lesion and midline shift greater than 1 cm, the clinical results were no better than those reported in the Traumatic Coma Data Bank for similar patients treated without decompression (class III data).⁹

A recent study in pediatric patients concluded that early decompressive craniectomy was associated with better control of ICP and with better functional outcome in children with traumatic brain injury.¹⁰ However, management protocols continually evolved and changed during the course of the study, including goals for cerebral perfusion pressure, use of hyperventilation and hypothermia, and administration of intravenous fluids. The decompressive craniectomies consisted only of limited bitemporal decompression, without opening of the dura in most cases. Other problems included the small number of patients (n = 27), the protracted time course of the study (7 years), and the method of obtaining outcome assessments. For these reasons, despite its laudable efforts to prospectively randomize patients, this trial may be regarded at best as a class II study.

Some authors have limited the use of decompression to patients younger than 50 years of age because of the generally poor prognosis of older trauma patients.^8,11 The author’s experience, however, is that older patients may respond reasonably well to decompression, and therefore no specific age of exclusion for this technique is imposed.

Complications

Decompressive craniectomy has been associated with various complications (class II and class III data).^3–5,8 Leakage of cerebrospinal fluid (CSF) may occur after the craniectomy,and a small percentage of patients have difficulty with scalp breakdown at the bone edge after ICP normalizes and the scalp recedes. Wound infections may occur after reimplantation of the bone flap.³ Delayed subdural hygromas at the site of decompression may require drainage and/or ventricular or lumbar shunting.

Surgical Technique

Decompression

Decompressive craniectomy can range from a simple subtemporal decompression (unilateral or bilateral) to subtotal calvariectomy.^3–5,8,12 Several authors describe large bifrontal craniectomies, particularly for patients with refractory elevations of ICP due to diffuse cerebral swelling.^3,5 The author prefers a large, unilateral craniectomy to include the frontal, parietal, and temporal bones, centered over the area of maximal brain injury as identified on the most recent preoperative CT scan. The bone flap should extend at least to or beyond the margins of injury or swelling and down to the floor of the middle fossa. This extension of the inferior margin of the craniectomy to the temporal skull base has been shown by Münch et al6 to result in optimal decompression of the perimesencephalic cisterns (class III data).