Background We report a retrospective analysis of a two-level, variable-volume epidural blood patch (EBP) technique for the treatment of spontaneous intracranial hypotension (SIH).

Methods Ninety-four patients with SIH underwent EBPs. Injectate volume was variable and guided by the onset of back pain, radiculopathy or symptoms referable to the EBP. Patients were a responder if no invasive treatment was necessary for SIH symptoms after the first EBP. Baseline patient characteristics, including presenting symptoms and imaging findings, were compared between responders and non-responders. Adverse events were recorded.

Results The mean first EBP volume was 45.3±23.2 (range 4–124 mL). The responder rate was 28.7% after the first EBP, improving to 41.5% and 46.8% after a second and third EBP, respectively. Baseline characteristics were similar between groups, except for the greater prevalence of subdural fluid collections in the responder group (48% vs 18%, p=0.003). Two severe complications occurred: transient bilateral paraplegia and cauda equina syndrome from arachnoiditis.

Conclusions Our protocol can be used to treat SIH but carries risks that require meticulous attention. EBP volumes were variable across patients, demonstrating the variability in the elastance of the epidural space. Rigorous monitoring for neurological symptoms is necessary throughout the procedure to limit complications and determine the endpoint of the EBP. We advocate for rigorous confirmation of epidural placement of the EBP with contrast-imaging studies to limit intrathecal placement of blood, which can lead to arachnoiditis. Randomized controlled trials are needed to determine the safety and efficacy of large-volume EBPs.