Predictive markers for response to adjuvant procarbazine, lomustine (CCNU) and vincristine (PCV) in anaplastic oligodendrogliomas and oligoastrocytomas
Session type: Parallel sessions
High throughput ‘omics' data may accelerate the identification of predictive response markers by screening for genes, methylation sites or genetic changes that are associated with clinical outcome. However, most clinical trial samples are fixed in formalin and embedded in paraffin (FFPE) which results in heavily degraded and chemically modified DNA and RNA. To demonstrate that expression profiling is feasible using FFPE tissues, we performed analysis of 55 paired snap frozen (FF)-FFPE samples. We show that differences in mRNA expression levels are retained in FFPE samples. Similarly, we performed paired FF-FFPE analysis using the Infinium HumanMethylation450 beadchip. Our data show that the reproducibility between replicates was high and independent of time in paraffin (>15 years). Expression and methylation profiling therefore is feasible on RNA/DNA isolated from FFPE samples.
We then performed gene expression and methylation profiling samples from the EORTC-26951 clinical trial. This trial examined the effects of adjuvant procarbazine, CCNU and vincristine (PCV) chemotherapy in anaplastic oligodendrogliomas and oligoastrocytomas and demonstrates that adjuvant PCV chemotherapy improves overall survival. However, a subset of patients clearly benefitted more from adjuvant PCV than others. Gene expression profiling demonstrates that intrinsic glioma subtypes (i.e. subtypes with similar gene expression profile) are highly prognostic in EORTC26951 and improve outcome prediction when combined with other prognostic factors. Our data also show that tumours assigned to a specific intrinsic subtype, IGS-9, benefit from adjuvant PCV. Methylation profiling demonstrates that the CpG island methylator phenotyp (CIMP) status is prognostic for overall survival in the EORTC26951 clinical trial and that methylation on specific sites within the MGMT promoter are correlated to benefit to PCV chemotherapy. In an exploratory analysis, we have also identified 259 CpG sites associated with treatment benefit in this study. Our data demonstrate the power of using high-throughput ‘omics' in identifying patients that benefit from treatment.