Factors predicting response to the first epidural blood patch in spontaneous intracranial hypotension

Spontaneous intracranial hypotension (SIH) is usually treated with an epidural blood patch (EBP), but response rates vary and there are no clear outcome predictors. Wu et al. identify specific brain and spinal imaging measures which, together with EBP blood volume, can be used to predict targeted first-EBP response in SIH. Spontaneous intracranial hypotension results from cerebrospinal fluid leakage. Currently, the treatment of choice for spontaneous intracranial hypotension is the epidural blood patch, which has a variable response rate and no clear outcome predictors. This study aimed to identify predictors for response rate of a first targeted epidural blood patch in patients with spontaneous intracranial hypotension. We reviewed cases of patients with spontaneous intracranial hypotension who received targeted epidural blood patch at our hospital between 1 January 2007 and 1 July 2014. The outcome measure was first epidural blood patch response. We analysed demographics, clinical manifestations, neuroimaging findings (non-contrast heavily T-2-weighted magnetic resonance myelography and brain magnetic resonance imaging), and blood volume as potential outcome predictors. Significant predictors were tested and a decision tree was used to construct a predictive model. In total, 150 patients with spontaneous intracranial hypotension were included for final analyses. Their overall first targeted epidural blood patch response rate was 58.7%. Among patients with a greater injected blood volume (>= 22.5 versus <22.5 ml), the response rate was higher (67.9% versus 47.0%, P = 0.01). In brain and spinal magnetic resonance imaging studies, significant predictors included anterior epidural cerebrospinal fluid collection length (<8 versus >= 8 segments; 72.5% versus 37.3%, odds ratio = 4.4, 95% confidence interval: 2.2-8.9, P < 0.001) and midbrain-pons angle (>= 40 degrees versus <40 degrees; 71.3% versus 37.5%, odds ratio = 4.1, 95% confidence interval 2.1-8.3, P < 0.001). Decision tree analyses showed that patients with anterior epidural CSF collection involving <8 segments and an injected blood volume >= 22.5 ml had an 80.0% response rate. Patients with anterior epidural cerebrospinal fluid collection involving >= 8 segments and a midbrain-pons angle <40 degrees had a 21.2% response rate. These three variables predicted first epidural blood patch response in 71.3% of patients. Brain and spinal neuroimaging findings and epidural blood patch blood volume can be used to predict targeted first epidural blood patch response in patients with spontaneous intracranial hypotension.