
After tumor resection, radiotherapy (60 Gy) and chemotherapy with TMZ in accordance with the Stupp protocol were performed. Then, an Ommaya reservoir was placed under the scalp for obtaining CSF samples in the cavity. If the ventricles were opened, they were closed as much as possible with Gelfoam ® (Pfizer Japan, Inc., Tokyo, Japan) and fibrin glue. Eight BCNU wafers were placed to closely cover the surface of the tumor resection cavity in each case and secured with Surgicel absorbable hemostat (Surgicel) (Johnson & Johnson, Tokyo, Japan) to prevent the wafers from floating up. After the tumor was resected, BCNU wafers were implanted on the surface of the tumor bed. All patients underwent craniotomy for tumor resection. Informed consent was obtained from all individual participants enrolled in the study after the risk of the surgical procedure and the potential risks of microsurgery and chemoradiotherapy, including implantation of BCNU wafers (Gliadel ®) were explained. Among them, 3 patients (2 GBM and 1 AMG3) were assigned to the control group in this study. Materials and MethodsĪ total of 18 patients whose histologies were verified as malignant gliomas (14 GBMs, 4 astrocytoma IDH mutants, grade 3 (AMG3)) who were treated according to the same treatment protocol except for implantation of BCNU wafers at the Department of Neurosurgery of Ehime University Hospital between November 2014 and December 2016 were enrolled. These studies provide crucial information for obtaining optimal efficacy in the treatment with BCNU wafers in GBM. In the present study, we examined gene expression of CD44 in the tumor tissues of GBM to identify the features of specific phenotypes of GBM. We have previously reported that imaging features of GBM on MRI and Met-PET well correlated with the expression level of CD44 in the tumor tissues of GBM. In addition, to investigate a phenotype of GBM showing resistance to high-dose BCNU, we examined imaging features of GBM on magnetic resonance imaging (MRI) and 11C-methionine (Met)-positron emission tomography (PET) and analyzed the correlation between imaging features and anti-tumor effect of high-dose BCNU. We also investigated whether a high concentration of BCNU that was locally released into the area adjacent to the tumor resection cavity affected the pattern of tumor recurrence.

Here, to investigate whether BCNU wafers are effective for controlling tumor progression, whether anti-tumor effects of BCNU wafers depend on the accumulating concentration of BCNU, and what type of tumors show a resistance to BCNU wafers, we measured the BCNU concentration in the resection cavity up to 30 days after implantation of wafers by assessing cerebral spinal fluid (CSF) samples and analyzed the relationship between the total concentrations of BCNU and survival times in patients with GBM. However, a high concentration of BCNU may have potential to provide some survival benefit for less-invasive type GBM. Total BCNU concentrations did not correlate with tumor progression or survival. No differences in AUC all were seen between local and non-local recurrence groups. A total of 9 patients experienced recurrence, with 6 local, 2 diffuse, and 1 disseminated patterns. MRI revealed that all patients in the ER group had highly invasive GBMs, whereas all patients in the LN group had less-invasive GBMs. AUC all tended to be lower in the ER group than in the LN group, but the difference was not significant. GBM patients were classified into two groups: early recurrence (ER) and late or no recurrence (LN), using median progression-free survival as the cut-off. The BCNU concentration was maximal 1 h postoperatively, rapidly decreased within 24 h, and remained relatively high for 7 days. The area under the curve (AUC) all was calculated from BCNU concentration curves, and the relationships between AUC all and survival, tumor phenotypes on MRI, and recurrence patterns were analyzed. BCNU concentrations in the tumor resection cavity were measured for 30 days postoperatively. BCNU wafers (Gliadel ®) were implanted with an Ommaya device in 15 patients, including 12 patients with GBM.


We investigated the accumulating high-dose effects of BCNU released from the wafers on the survival of GBM patients by measuring BCNU concentration in the resection cavity of GBM over time. The effectiveness of carmustine (BCNU) wafers on local recurrence of glioblastoma (GBM) remains contentious.
