Prediction of clinical outcome in subacute subarachnoid hemorrhage using diffusion tensor imaging

Abstract

Objective: Clinical outcome in nontraumatic subarachnoid hemorrhage (SAH) is multifactorial and difficult to predict. Diffusion tensor imaging (DTI) findings are a prognostic marker in some diseases such as traumatic brain injury. The authors hypothesized that DTI parameters measured in the subacute phase of SAH can be associated with a poor clinical outcome.

Methods: Diffusion tensor imaging was prospectively performed in 54 patients at 8–10 days after nontraumatic SAH. Logistic regression analysis was performed to evaluate the association of fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values with a poor clinical outcome (modified Rankin Scale score ≥ 3) at 3 months.

Results: At 8–10 days post-SAH, after adjusting for other variables associated with a poor outcome, an increased ADC at the frontal centrum semiovale was associated with a poor prognosis (OR estimate 1.29, 95% CI 1.04–1.60, p = 0.020). Moreover, an increase of 0.1 in the FA value at the corpus callosum at 8–10 days after SAH corresponded to 66% lower odds of having a poor outcome (p = 0.002).

Conclusions: Decreased FA and increased ADC values in specific brain regions were independently associated with a poor clinical outcome after SAH. This preliminary exploratory study supports a potential role for DTI in predicting the outcome of SAH.