NETwork! is a new translational programme purpose-built to conduct deep genomic analysis of NETs within an integrated scientific and clinical practice.
A clinical and ethical framework was established to support, interpret and return genomic analyses, including a clinically-annotated national registry of NETs coupled to a NET-specific MDM to harmonise variability and facilitate tissue collection. Genomic analyses of the first 61 pNETs are reported here. Hybridisation capture DNA sequencing (578 cancer associated genes; >750 x coverage) and microarray mRNA expression analysis were conducted for all pNETs, and 12 tumours underwent additional whole genome sequencing, RNAseq, methylation and miRNA expression analysis. Clinical, pathological and genomic data were compared using a customised bioinformatic platform.
pNETs carry relatively few mutations, their genomic landscape dominated by large scale copy number, epigenetic and gene expression changes. Mutations were found in 75 genes associated with cancer, with 64 exclusive to single tumours. Apart from recurrent mutations in MEN1 (39%) and ATRX (7%), the driver genomic changes in pancreatic NETs were highly tumour-specific, including somatic mutations in: FANCA, APC, BRCA2, PTEN, EGFR, MDM4, MSH2 and VHL. Mutations were seen in ten additional genes not traditionally associated with cancer. There was striking aneuploidy eg, 18% of pNETs showed an identical and undescribed pattern of LOH of the same ten whole chromosomes. 12 tumours had gene expression profiles of immune activation. Therapeutic choice was suggested using single biomarkers e.g., FANCA, MSH2 but further informed by multi-level genomics eg, in one case the impact of a PTEN SNV was negated by LOH in downstream mTOR thus reducing pathway activity, whereas another case showed mTOR hypomethylation and and expression changes consistent with pathway activation.
Deep genomics of carefully annotated NETs revealed new insights into NET tumorigenesis, and enables rational but unexpected therapeutic choice to be applied in clinical trial. Genomic variability despite clinical homogeneity argues for multi-level sequencing of metastatic NETs.
Clinical trial identification
Legal entity responsible for the study
University of Auckland
Translational Medicine Trust
All authors have declared no conflicts of interest.