Prognostic and Predictive Markers in Glioma and Other Neuroepithelial Tumors

Work to date on the identification of predictive and prognostic markers in brain tumors provide reason for both caution and optimism with respect to improvements in the diagnosis and treatment of patients with these tumors. It is sobering to consider that the multitude of efforts into the identification of prognostic markers in gliomas has led to development so far of only one molecular test that is relatively common: 1p/19q testing in oligodendrogliomas. More effort is required for the identification of markers that distinguish patients in a way that influences therapeutic options. In addition, although prognostic factors such as the RTOG-RPA classification are relatively powerful with respect to patient stratification, one goal should be to include molecular markers into this RPA classification to more precisely predict outcome. Finally, with the advent of newer targeted therapies, new predictive markers specific to a particular targeted agent need to be identified, perhaps to determine whether an individual patient is likely to respond to that therapy. There is, however, some cause for optimism in that the tools for high throughput molecular marker analysis markers will allow efficient screening for new prognostic factors on the DNA, RNA, epigenetic, and possibly protein levels. These technologies will lead to multi-marker profiles, which, given the heterogeneity and complexity of cancer, are likely to be more robust with respect to outcome prediction than prognostic classes defined by the status of a single marker. Some of these technologies are allowing, or in the future, will allow, the use of analytes from formalin-fixed paraffin-embedded tissue, which represents an important practical consideration especially when contemplating screening of tissue from large multi-institutional clinical trials. It can be argued that the best way to move forward will be a situation where future clinical trials require not only a slide for diagnostic conformation, but also tissue to identify molecular predictors using these new tools. In this way, progress toward truly personalized medicine, whereby treatment is tailored to the molecular signature of the tumor, may be possible.