Prognostic Role of the Expression of Invasion-Related Molecules in Glioblastoma.

 Glioblastoma multiforme (GBM) is the most common malignant disease of the central nervous system. Its prognosis is unfavorable, and the median overall survival of patients is 16 to 24 months. The main cause of the poor survival data are the extensive invasion of cancer cells to the neighboring parenchyma, thus leading to inevitable local recurrence. The extracellular matrix (ECM) is a known factor in tumor invasion, and differences in the ECM of nontumor brain and glioblastoma has been proven.  In this research, 20 invasion-related expressions of ECM components were determined in 26 GBM flash-frozen samples using quantitative reverse transcription-polymerase chain reaction and proteomic measurements. Expression data were then set against the survival data of the patients.  Significant alterations between groups with different survival rates could not be established in the individual evaluation of the expression level of the selected molecules. However, statistical analysis of the expression pattern of invasion-related molecules revealed a correlation with prognosis. The positive predictive values of the messenger RNA (mRNA) and the proteomic expression studies were 0.85 and 0.89, respectively. The receiver operation characteristic value was 0.775 for the mRNA expression data and 0.875 for the protein expression data. Furthermore, a group of molecules, including brevican, cadherin-12, integrin β1, integrin α3, laminin α4, and laminin β1, that play a prominent role in invasion were identified.  Joint assessment of the expression of invasion-related molecules provides a specific invasion spectrum of the tumor that correlates with the survival of glioblastoma patients. Using statistical classifiers enables the adoption of an invasion spectrum as a considerably accurate prognostic factor while gaining predictive information on potential molecular oncotherapeutic targets at the same time.