gms | German Medical Science

Objective: Surgery is accepted as a treatment for malignant hemispheric stroke (MHS) that is lifesaving but may result in moderate to severe disability. Since survival at the cost of disability may be acceptable to some but not to others, there remains a need to identify long-term outcome measures for MHS that can assist in surgical or non-surgical decision-making. Therefore, the aim of the present study was to identify prognostic variables for long-term outcome in patients undergoing decompressive hemicraniectomy (DHC) for treatment of MHS.

Methods: The present study included 97 patients with subtotal or total middle cerebral artery infarction, who underwent standardized DHC in our institution within 48 hours after stroke onset. After surgery, patients were transferred to the intensive care unit and a routine postoperative CT or MRI was performed within 24 hours. Intracranial pressure (ICP) was monitored and patients remained intubated / sedated until ICP was within normal ranges. The primary outcome measure was clinical outcome at 12 months according to the modified Rankin Scale (mRS) based on the initial infarct volume. Secondary clinical outcome measures included the time to treatment, estimated blood loss, postoperative ICP crisis, surgical complications and the duration of intubation. Secondary neuroimaging outcome measures included the involvement of vascular territories other than the MCA territory. All clinical and neuroimaging data was retrospectively reviewed and analyzed by two independent observers.

Results: At 12 months, overall mortality was 16% (16/97), while 64% (62/97) of all patients survived with severe (mRS 4-5) and 20% (19/97) with mild to moderate (mRS 2-3) disability. The Receiver Operating Characteristic (ROC) curve of the infarct volume mortality yielded a good classification result (AUC=0.77) and multivariate logistic regression was able to confirm infarct volume (*p=0.045) as a factor with significant association to mortality. Further, logistic regression analysis revealed a significant influence of the infarct volume on 12-month mRS (*p<0.001) with an effective power of 0.56 according to the Cohen classification. Post-hoc analysis with Bonferroni correction revealed significant lower infarct volumes in patients with mild to moderate disability (mRS 2-3) versus patients with severe disability or death (mRS 5-6) at 12 months. Additional univariate predictors of outcome were identified as involvement of the anterior (*p=0.01) and posterior cerebral artery (*p=0.001) territories, thalamic involvement (p=0.02), age>60 years (p=0.002), hypothermia (p=0.03), atrial fibrillation (p=0.01), and the duration of intubation (p<0.001).

Conclusion: In MHS, the initial volume of the infarction may help to weigh the consequences of DHC against the patients’ presumed individual expectation towards the benefit of surgery.