Risk of surgical site infection in 401 consecutive patients with glioblastoma with and without carmustine wafer implantation

Objectives: Patients with glioblastoma (GBM) have an inherently shortened survival because of their disease. It has been recently shown that carmustine wafers in addition to other therapies (surgery, temozolomide, and radiation) can further extend survival. There is concern, however, that these therapies may increase infection risk. The goals of this study were to calculate the incidence of postoperative infection, evaluate if carmustine wafers changes the risk of infection and identify factors independently associated with an infection following GBM surgery. Methods: All patients who underwent non-biopsy, surgical resection of an intracranial GBM from 2007 to 2011 at a single institution were retrospectively reviewed. Stepwise multivariate proportional hazards regression analysis was used to identify factors associated with infection, including the use of carmustine wafers. Variables with P < 0.05 were considered statistically significant. Results: Four hundred and one patients underwent resection of an intracranial GBM during the reviewed period, and 21 (5%) patients developed an infection at a median time of 40 [28-286] days following surgery. The incidence of infection was not higher in patients who had carmustine wafers, and this remained true in multivariate analyses to account for differences in treatment cohorts. The factors that remained significantly associated with an increased risk of infection were prior surgery [RR (95% CI); 2.026 (1.473-4.428), P = 0.01], diabetes mellitus [RR (95% CI); 6.090 (1.380-9.354)], P = 0.02], and increasing duration of hospital stay [RR (95% CI); 1.048 (1.006-1.078); P = 0.02], where the greatest risk occurred with hospital stays > 5 days [RR (95% CI); 3.904 (1.003-11.620), P = 0.05]. Discussion: These findings may help guide treatment regimens aimed at minimizing infection for patients with GBM.