A translational study in pancreatic neuroendocrine neoplasms

Abstract

Pancreatic neuroendocrine neoplasms (PanNENs) are rare, highly heterogeneous tumours. There have been significant recent advances in our knowledge of genomic events underlying their pathogenesis. However, treatment decisions remain largely based on tumour stage and grade which is inadequate, the current classification paradigm failing to capture the significant heterogeneity in tumour biology. The first aim of this thesis was to establish a large registry for PanNENs and then clinically phenotype the patients included. The next aim was to develop a novel assay to subtype PanNENs, based on our previously derived PanNETassigner molecular subtypes, and to establish if the subtypes assigned were prognostic. The PanNEN registry and PanNETassigner assay were successfully developed. Clinical data and tissue samples from the registry were used to test, validate and refine this assay. The assay demonstrated that the metastasis-like primary-1 subtype (MLP-1) was associated with a poor prognosis. Novel therapeutic options are required in PanNENs and trials of immunotherapy are underway, although knowledge of the immune microenvironment in this disease is lacking. The last aim of this thesis was therefore to describe immune related gene expression across the PanNETassigner molecular subtypes. The poor prognosis MLP-1 subtype had an immune high phenotype, associated with hypoxic tumours and signalling within damage-associated molecular pattern pathways. The immune gene expression profile demonstrated generates the hypothesis that the MLP-1 subtype may be more amenable to an immunotherapeutic approach than other subtypes. Overall, this thesis demonstrates that molecular subtyping can be used to provide valuable additional information both regarding prognosis and the immune microenvironment in PanNENs. The assays and hypotheses developed here now require additional testing in pre-clinical mechanistic studies, larger cohorts of patients and prospective clinical trials. With such further validation the PanNETassigner subtypes may pave a way forward for a more personalised approach for patients with rare tumour type.