Abstract 3421: Comparing gene expression of matched FFPE colorectal cancer samples measured by Nanostring nCounter® platform and Affymetrix GeneChip® Human Transcriptome Array platform

Abstract

Abstract OBJECTIVE: Due to the lack of clinically annotated fresh frozen tumor samples, defining molecular features that can predict the recurrence of colorectal cancer (CRC) for stage II or stage III patients remains challenging in cancer research. Most available clinical samples are Formalin Fixed and Paraffin Embedded (FFPE). Nanostring nCounter® and Affymetrix GeneChip® Human Transcriptome Array (HTA) are two latest platforms that enable gene expression profiling of FFPE samples. In this study, we evaluated these two platforms for validating the prognostic gene expression signature based on FFPE-derived CRC samples. METHODS: We designed a custom nCounter® codeset based on elements from multiple published prognostic gene signatures for CRC based on frozen tissues and used this platform to measure the gene expression of FFPE-derived CRC tissues passing strict quality control measures. Then, we measured the gene expression of 42 matched FFPE-derived samples on the HTA platform. Gene expression data for matched CRC samples measured by the two platforms were compared. Finally, we compared the previously published prognostic subtype assignments based on gene expression data measured by the two platforms. RESULTS: In total, 42 pairs of matched CRC samples including 31 pairs of technical replicates and 11 pairs of biological replicates measured by both platforms passed the quality control. Our results showed moderate positive correlation of gene expression between matched samples measured by both platforms, which is not correlated with the RNA quality. When comparing the gene expression of patients with metastatic recurrence to that of patients with no recurrence over a 7.4-year mean follow up, we identified 5 differentially expressed genes (SYT17, LRIG1, PTPRJ, BSG, PTEN) in common (p < 0.05) based on both platforms. Subtype analysis assigned Zhu et al.'s subtypes (Zhu, J. et al., PLOS One 2013) to 29 pairs of matched technical replicates measured by both platforms with posterior probability greater than 50%. Overall, we found that the gene expression patterns visualized in the HTA dataset were more consistent within defined subtypes as compared to those of the nCounter dataset and the HTA-based assignments are better supported by the corresponding survival estimates. CONCLUSION: We found a moderate positive correlation between matched samples across both HTA and nCounter platforms, however the gene-by-gene correlations were weak. Both platforms support 5 potential biomarkers of poor prognosis. Discovery of CRC subtypes was more definitive using data from the HTA platform as compared to the nCounter® platform. Further comparisons of matched FFPE-derived CRC tissues across these platforms can lead to the discovery of additional prognostic biomarkers and increased confidence in the optimal platform for expanded clinical trial studies. Citation Format: Jing Zhu, Natasha G. Deane, Keeli B. Lewis, Mary K. Washington, Xi Chen, Robert D. Beauchamp. Comparing gene expression of matched FFPE colorectal cancer samples measured by Nanostring nCounter® platform and Affymetrix GeneChip® Human Transcriptome Array platform. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 3421. doi:10.1158/1538-7445.AM2015-3421