Lung Cancer Model Systems
Lung cancer remains the leading cause of cancer-related mortality worldwide. Curing late stage disseminated lung cancer is challenging, as cell populations are highly heterogeneous both molecularly and histologically, and often genomically unstable. Despite promising progress in treatment strategies that harness anti-tumour immunity, therapeutic resistance limits clinical efficacy. A shift to early cancer detection and prevention is therefore necessary, and this requires an improved understanding of the biological basis of cancer initiation in its physiological niche.
We take a comprehensive approach to tackle challenges in the field of lung cancer research, progressing from mouse model systems of cancer to human disease cohorts. First, tumour-initiating events are identified in murine lung cancer cohorts, comprised of validated and putative driver combinations expressed in tissue progenitor cells. Second, we apply pathology-specific functional and immune profiling studies to identify drivers of malignancy. Third, we implement a mouse-to-patient comparative personalised medicine study, establishing the ability of cultured cells and tissue explants to reliably predict or validate in vivo drug sensitivities.
In previous years, our team has studied how phenotypic heterogeneity arises in non-small cell lung cancer (NSCLC). We generate somatic mouse models of lung cancer that combine germline conditional mutations with progenitor cell-targeted adenoviruses, permitting us to assess a variety of biological events contributing to lung cancer pathology and therapeutic response. Using preclinical NSCLC models representing the most common drivers of clinical disease, we showed that histopathology spectra are dependent on the tumour’s cell of origin. We further identified histopathology-selective gene expression profiles and immune microenvironments, as well as oncogenic signalling networks. We have applied this understanding to the building of diagnostic models, through activities part of FIMM’s individualised systems medicine Grand Challenge efforts, as well as European public-private collaborations in context of the IMI-PREDECT consortium project (www.predect.eu; 2011-2016; a project that set out to build sufficiently complex low-throughput explant models of cancer to ameliorate target validation). We used the NSCLC models to optimise a workflow for cultivation of precision-cut tumour slice explants, which, not surprisingly, was not an easy feat. Implementing such slice explants, we showed that response to combination treatment with signalling inhibitors corresponds with spatially-defined targeted pathway activities, and that combinatorial drug sensitivity aligns closely with histopathology type. Our findings underscore a generally underappreciated need to incorporate lesion-specific phenotypic heterogeneity in clinical settings.
Foreseeing rapid technological advances in personalised cancer diagnostics, we aim to implement our gained insights in the rational design of cancer treatment and intervention strategies. To achieve these goals, our research increasingly interfaces with biobanking and translational infrastructures available at FIMM, associated Life Science Biocentres, and the Helsinki University Hospital. We collaborate with pulmonary surgeons, clinicians and pathologists to implement personalised disease profiling on patient tumour samples. These efforts will permit us to evaluate whether the functional diagnostic elucidation of patient tumour-selective therapeutic vulnerabilities has the potential to benefit the health of locally-treated cancer patients.