Background: In modern clinical neuro-oncology, the pathologic diagnoses are very challenging, creating significant clinical confusion and affecting therapeutic decisions and prognosis.
Methods: TP53 and PTEN gene sequences were analyzed, and microarray expression profiling was also performed. The authors investigated whether gene expression profiling, coupled with class prediction methodology, could be used to determine the prognosis of gliomatosis cerebri in a more consistent manner than standard pathology.
Results: The authors reported the results of a molecular study in 59 cases of gliomatosis cerebri, correlating these results with prognosis. The well-known prognostic factors of gliomas (ie, age, Karnofsky performance status, histology [grade 2 vs 3], and contrast enhancement) were found to be predictive of response or outcome in only a percentage of patients but not in all patients. The authors identified a 23-gene signature that was able to predict patient prognosis with microarray gene expression profiling. With the aim of producing a prognosis tool that is useful in clinical investigation, the authors studied the expression of this 23-gene signature by real-time quantitative polymerase chain reaction. Real-time expression values relative to these 23 gene features were used to build a prediction method able to distinguish patients with a good prognosis (those more likely to be responsive to therapy) from patients with a poor prognosis (those less likely to be responsive to therapy).
Conclusions: The results of the current study demonstrated not only a strong association between gene expression patterns and patient survival, but also a robust replicability of these gene expression-based predictors.