Abstract
Maximal safe resection of malignant tissue is associated with improved progression-free survival and better response to radiation and chemotherapy for patients with glioblastoma (GBM). 5-Aminolevulinic acid (5-ALA) is the current FDA-approved standard for intraoperative brain tumor visualization. Unfortunately, autofluorescence in diffuse areas and high fluorescence in dense tissues significantly limit discrimination at tumor margins. This study is the first to compare 5-ALA to an investigational new drug, panitumumab-IRDye800CW, in the same animal model. A patient-derived GBM xenograft model was established in 16 nude mice, which later received injections of 5-ALA, panitumumab-IRDye800CW, IRDye800CW, 5-ALA and IRDye800CW, or 5-ALA and panitumumab-IRDye800CW. Brains were prepared for multi-instrument fluorescence imaging, IHC, and quantitative analysis of tumor-to-background ratio (TBR) and tumor margin accuracy. Statistical analysis was compared with Wilcoxon rank-sum or paired t test. Panitumumab-IRDye800CW had a 30% higher comprehensive TBR compared with 5-ALA (P = 0.0079). SDs for core and margin regions of interest in 5-ALA-treated tissues were significantly higher than those found in panitumumab-IRDye800CW-treated tissues (P = 0.0240 and P = 0.0284, respectively). Panitumumab-IRDye800CW specificities for tumor core and margin were more than 10% higher than those of 5-ALA. Higher AUC for panitumumab-IRDye800CW indicated strong capability to discriminate between normal and malignant brain tissue when compared with 5-ALA. This work demonstrates that panitumumab-IRDye800CW shows potential as a targeting agent for fluorescence intraoperative detection of GBM. Improved margin definition and surgical resection using panitumumab-IRDye800 has the potential to improve surgical outcomes and survival in patients with GBM compared with 5-ALA.