Molecular profiling of intestinal cancers Towards clinical implementation

Gastrointestinal (GI) cancers are molecularly heterogeneous, resulting in pronounced intertumoral differences in disease progression and treatment response. The diversity of GI malignancies complicates estimation of prognosis and optimal timing and selection of treatment regimens for individual patients. In recent years, it has become clear that gene expression (transcriptomic) profiling of GI tumors reflects key elements of tumor heterogeneity, and as such, is capable of stratifying patients based on molecular tumor properties. Often, these molecular tumor subgroups contain prognostic and predictive value. The Consensus Molecular Subtypes (CMSs) of colorectal cancer (CRC) are an example of such a taxonomy. Although biologically highly informative, clinical usage of transcriptomic taxonomies remains challenging, due to unclear clinical consequence, high costs, the required expertise and technical limitations. As an illustration, tumor classification on low-quality RNA-data from formalin-fixed paraffin-embedded (FFPE) tissue samples, which are widely used in the clinic, remains complicated.
In this thesis, we set out to further explore and expand the clinical potential of transcriptomic subtyping of GI malignancies. We did so by optimally making use of the available literature regarding transcriptomic tumor stratification and clinical management, by performing in-depth molecular characterization of GI malignancies, and by studying large retrospective cohorts of patients. First, we focused on CRC, which is a commonly occurring and well-studied example of a GI malignancy with clinically relevant transcriptomic subgroups (the CMSs), and used this cancer type as a source of inspiration for studies on small intestinal adenocarcinomas (SIAs), a rare cancer type for which molecular profiling efforts are scarce.