Abstract 3174: Colorectal cancer specific host response signatures achieved by next generation sequencing and multiplex RT-qPCR

Abstract

Abstract Background: colorectal cancer (CRC) develops over a period of several years and is often preceded by adenoma formation. Detection rates for colorectal adenoma and early CRC, however, are unsatisfactory due to low compliance towards invasive screening procedures such as colonoscopy. There is a large unmet screening need calling for an accurate, non invasive and cost effective test to screen for early neoplastic and preneoplastic colorectal lesions. We developed a screening test aimed at detecting precancerous lesions and CRC at early stages, based on a multigene expression profiling on peripheral blood mononuclear cells (PBMC). Methods: 92 colonoscopy screened subjects served as a training set. Colonoscopy revealed 21 patients had CRC, 30 adenomas larger than 1cm and 41 had no detectable lesions. 16 mL of blood was collected from each individual and PBMC were purified using Vacutainer® CPT tubes (Becton Dickinson). Total RNA was extracted and a whole genome expression analysis by digital gene expression TAG profiling (Illumina) was performed on a subset of these 92 subjects. Deep sequencing generated on average 1.5 million reads per sample and they matched to 20288 different transcripts. Different univariate and multivariate statistical methods were then applied in order to find genes differentially expressed between control, adenoma and CRC (pvalue <0.05). 45 candidate genes discovered with TAG profiling were then analytically validated using a multiplex RT qPCR methodology. All candidate genes were normalized with a stable normalization strategy in order to obtain comparable results across assays. An additional set of 57 candidate genes was also added based on literature review and validated using the multiplex RT qPCR methodology. Multiple statistical methods were applied on the normalized PCR data and 43 biomarkers, with significant p-value (<0.01) for most of the methods, were selected. Two distinct molecular signatures were derived from the normalized biomarker combinations based on penalized logistic. Results: a validation set was defined using random resampling (bootstrapping method), leading to the separation of individuals without lesion from those with CRC (Se 70%, Sp 94%, AUC 0.89) and from those with adenoma larger than 1cm (Se 67%, Sp 87%, AUC 0.80). In addition, as a test set, the organ and disease specificity of these signatures was confirmed by means of patients with other cancer types and inflammatory bowel diseases. Conclusion: Next generation sequencing is a powerful research tool to define specific signatures for precancerous and early cancerous lesions when coupled to multiplex RT qPCR and normalized with specific reference genes. A prospective, multi-centric, pivotal study is underway in order to validate these results in a larger cohort. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3174. doi:10.1158/1538-7445.AM2011-3174