Umeå University's logo

Objective: The association was studied of intracranial pressure (ICP) and cerebral perfusion pressure (CPP) on S-100B and neuron-specific enolase (NSE) in severe traumatic brain injury (sTBI). The relationship was explored between biomarkers, ICP, CPP, CT-scan classifications and the clinical outcome.

Materials and methods: Data were collected prospectively and consecutively in 48 patients with Glasgow Coma Scale score ≤ 8, age 15–70 years. NSE and S-100B were analysed during 5 consecutive days. The initial and follow-up CT-scans were classified according to the Marshall, Rotterdam and Morris-Marshall classifications. Outcome was evaluated with extended Glasgow outcome scale at 3 months.

Results: Maximal ICP and minimal CPP correlated with S-100B and NSE levels. Complex relations between biomarkers and CT classifications were observed. S-100B bulk release (AUC = 0.8333, p = 0.0009), and NSE at 72 hours (AUC = 0.8476, p = 0.0045) had the highest prediction power of mortality. Combining Morris-Marshall score and S-100B bulk release improved the prediction of clinical outcome (AUC = 0.8929, p = 0.0008).

Conclusion: Biomarker levels are associated with ICP and CPP and reflect different aspects of brain injury as evaluated by CT-scan. The biomarkers might predict mortality. There are several pitfalls influencing the interpretation of biomarker data in respect to ICP, CPP, CT-findings and clinical outcome.

The present study evaluates the types and dynamics of intracranial pathological changes in patients with severe traumatic brain injury (sTBI) who participated in a prospective, randomized, double-blinded study of add-on treatment with prostacyclin. Further, the changes of brain CT scan and their correlation to Glasgow Coma Scale score (GCS), maximal intracranial pressure (ICPmax), minimal cerebral perfusion pressure (CPPmin), and Glasgow Outcome Score (GOS) at 3, 6, and 12 months were studied. Forty-eight subjects with severe traumatic brain injury were treated according to an ICP-targeted therapy protocol based on the Lund concept with the addition of prostacyclin or placebo. The first available CT scans (CTi) and follow-up scans nearest to 24 h (CT24) were evaluated using the Marshall, Rotterdam, and Morris-Marshall classifications. There was a significant correlation of the initial Marshall, Rotterdam, Morris-Marshall classifications and GOS at 3 and 12 months. The CT24 Marshall classification did not significantly correlate to GOS while the Rotterdam and the Morris-Marshall classification did. The CTi Rotterdam classification predicted outcome evaluated as GOS at 3 and 12 months. Prostacyclin treatment did not influence the dynamic of tissue changes. The Rotterdam classification seems to be appropriate for describing the evolution of the injuries on the CT scans and contributes in predicting of outcome in patients treated with an ICP-targeted therapy. The Morris-Marshall classification can also be used for prognostication of outcome but it describes only the impact of traumatic subarachnoid hemorrhage (tSAH).

Objective: A decompressive hemicraniectomy is a potentially life-saving intervention following head trauma. Once performed patients are obliged to undergo a second procedure with cranioplasty. Two of the most commonly used materials are autologous bone and polymethylmethacrylate (PMMA). We have now evaluated complications following a cranioplasty using these materials. Materials and methods: During a 7-year period (2002-2008) 49 patients were operated with a decompressive craniectomy following head trauma. Patients received a cranioplasty consisting of autologous bone (30 patients, 61.2%) or PMMA (19 patients, 38.8%) and were followed at least 24 months. Patient data were collected retrospectively. Results: Twenty patients (20/49, 40.8%) experienced a complication that prompted a re-operation. There was a significantly higher rate of complications leading to a re-operation (53.3% vs. 21.1%, p = 0.03) and a shorter survival time of the cranioplasty (mean 48.1 +/- 7.8 vs. 79.5 +/- 9.0 months, p = 0.035) in patients with autologous bone compared to PMMA. Bone resorption and the presence of postoperative hematomas were significantly more common in patients with autologous bone. The material used for cranioplasty was the only variable that significantly correlated to the rate of complications. Conclusions: In our series we had a high percentage of patients needing re-operation due to complications following a cranioplasty. Though generally considered a straightforward procedure, complications and associated morbidity in patients undergoing cranioplasty should not be underestimated.