Complications associated with lumbar drain placement for endovascular aortic repair

Objective We reviewed the complications associated with perioperative lumbar drain (LD) placement for endovascular aortic repair. Methods Patients who had undergone perioperative LD placement for endovascular repair of thoracic and thoracoabdominal aortic pathologies from 2010 to 2019 were reviewed. The primary endpoints were major and minor LD-associated complications. Complications that had resulted in neurological sequelae or had required an intervention or a delay in operation were defined as major. These included intracranial hemorrhage, symptomatic spinal hematoma, cerebrospinal fluid (CSF) leak requiring intervention, meningitis, retained catheter tip, arachnoiditis, and traumatic (or bloody) tap resulting in delayed operation. Minor complications were defined as a bloody tap without a delay in surgery, asymptomatic epidural hematoma, and CSF leak with no intervention required. Isolated headaches were recorded separately owing to the minimal clinical impact. Results A total of 309 LDs had been placed in 268 consecutive patients for 222 thoracic endovascular aortic repairs, 85 complex endovascular aortic repairs (EVARs; fenestrated branched EVAR/parallel grafting), and 2 EVARs (age, 65 ± 13 years; 71% male) for aortic pathology, including aneurysm (47%), dissection (49%), penetrating aortic ulcer (3%), and traumatic injury (0.6%). A dedicated neurosurgical team performed all LD procedures; most were performed by the same individual, with a technical success rate of 98%. Radiologic guidance was required in 3%. The reasons for unsuccessful placement were body habitus (n = 2) and severe spinal disease (n = 3). Most were placed prophylactically (96%). The overall complication rate was 8.1% (4.2% major and 3.9% minor). Major complications included spinal hematoma with paraplegia in 1 patient, intracranial hemorrhage in 2, meningitis in 2, arachnoiditis in 3, CSF leak requiring a blood patch in 3, bloody tap delaying the operation in 1, and a retained catheter tip in 1 patient. Patients who had undergone previous LD placement had experienced significantly more major LD-related complications (12.2% vs 3%; P = .019). The rate of total LD-associated complications did not differ between prophylactic and emergent therapeutic placements (8.1% vs 7.7%; P = 1.00) nor between major or minor complications. On multivariate analysis, previous LD placement and an overweight body mass index were the only independent predictors of major LD-related complications. Conclusions The complications associated with LD placement can be severe even when performed by a dedicated team. Previous LD placement and overweight body mass index were associated with a significantly greater risk of complications; however, emergent therapeutic placement was not. Although these risks are justified for therapeutic LD placement, the benefit of prophylactic LD placement to prevent paraplegia should be weighed against these serious complications.