A major problem in pediatric neuro-oncology is lack of insight into underlying tumor biology, making treatment with targeted therapy challenging for this group of patients. To enable targeted therapy choice, understanding of underlying signal transduction pathway activity is required, as well as availability of appropriate diagnostic tests to predict therapy response. We have developed a novel method to identify activity of the ER, AR, HH, Wnt, TGFbeta, NFκB, and PI3K signal transduction pathways in an individual tissue sample. Our method is based on measuring mRNA levels of the pathway transcription factor target genes and interpreting these expression levels using a knowledge-based Bayesian network computational model which infers the probability of activity of the respective transcription factor (Verhaegh et al, Cancer Res. 2014). Public patient sample datasets from the GEO database were used to investigate pathway activation in pediatric brain tumors. In normal brain tissue from medulla, cerebellum, cerebral cortex and thalamus (GSE50161), pathway analysis showed no significant oncogenic pathway activation.

Observed pathway activity in medulloblastoma was strongly correlated with known pathway-activating mutations, which was used as initial validation of these models for use in pediatric brain tumor diagnostics. In ependymoma (GSE66354), three distinct subtypes were analyzed: infratentorial posterior fossa A (PFA), PFB and supratentorial. These tumors are known to have a subtype-dependent prognosis; mainstay of therapy is surgery and radiation. Tumor driving pathway identification in pediatric ependymoma using mutation analysis is usually not helpful because epigenetic mechanisms prevail (Mack et al, Brain Pathology 2013).

Using our pathway activity analysis, we find remarkable differences between different types of ependymoma with respect to signal transduction pathway activity. For PFA we find high NFκB and Wnt pathway activity. The dominant NFκB activity is in agreement with the described inflammatory phenotype. On the other hand, for PFB we find low NFκB and Wnt pathway activity. Finally, for supratentorial ependymoma we find almost all samples to have an active Hedgehog and Wnt pathway. We also analyzed multiple other pediatric tumors for pathway activity, for which data will be presented.

The above pathway activity results provide highly interesting leads for targeted therapy selection, and we expect that such pathway analysis will have clinical utility in a neoadjuvant setting to reduce tumor size prior to surgery, and possibly also in an adjuvant setting to prevent recurrences.

Citation Format: Laurent Holtzer, Wim Verhaegh, Anja Van de Stolpe. Elucidating pediatric brain tumor pathophysiology by assessing signal transduction pathway activation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 705. doi:10.1158/1538-7445.AM2017-705