Complications of ventricular entry during craniotomy for brain tumor resection

OBJECTIVE Recent studies have demonstrated that periventricular tumor location is associated with poorer survival and that tumor location near the ventricle limits the extent of resection. This finding may relate to the perception that ventricular entry leads to further complications and thus surgeons may choose to perform less aggressive resection in these areas. However, there is little support for this view in the literature. This study seeks to determine whether ventricular entry is associated with more complications during craniotomy for brain tumor resection. METHODS A retrospective analysis of patients who underwent craniotomy for tumor resection at Henry Ford Hospital between January 2010 and November 2012 was conducted. A total of 183 cases were reviewed with attention to operative entry into the ventricular system, postoperative use of an external ventricular drain (EVD), subdural hematoma, hydrocephalus, and symptomatic intraventricular hemorrhage (IVH). RESULTS Patients in whom the ventricles were entered had significantly higher rates of any complication (46% vs 21%). Complications included development of subdural hygroma, subdural hematoma, intraventricular hemorrhage, subgaleal collection, wound infection, urinary tract infection/deep venous thrombosis, hydrocephalus, and ventriculoperitoneal (VP) shunt placement. Specifically, these patients had significantly higher rates of EVD placement (23% vs 1%, p < 0.001), hydrocephalus (6% vs 0%, p = 0.03), IVH (14% vs 0%, p < 0.001), infection (15% vs 5%, p = 0.04), and subgaleal collection (20% vs 4%, p < 0.001). It was also observed that VP shunt placement was only seen in cases of ventricular entry (11% vs 0%, p = 0.001) with 3 of 4 of these patients having a large ventricular entry (defined here as entry greater than a pinhole [< 3 mm] entry). Furthermore, in a subset of glioblastoma patients with and without ventricular entry, Kaplan-Meier estimates for survival demonstrated a median survival time of 329 days for ventricular entry compared with 522 days for patients with no ventricular entry (HR 1.13, 95% CI 0.65-1.96; p = 0.67). CONCLUSIONS There are more complications associated with ventricular entry during brain tumor resection than in nonviolated ventricular systems. Better strategies for management of periventricular tumor resection should be actively sought to improve resection and survival for these patients.