Biological context

Lung cancer is the leading cause of cancer-related death worldwide. Small cell lung cancer (SCLC) is an aggressive subset of lung cancer, representing 10 to 15% of newly diagnosed lung cancers per year and characterized by rapid-onset symptoms due to high tumor growth rate. Even if recently, chemotherapy combination with immune checkpoint inhibitors showed substantial improvement in two phase III trials, the prognosis remains very poor with a median overall survival of about one year. SCLC shows a high propensity for metastatic spreading and therefore generates a higher number of circulating tumor cells (CTC) relative to other tumor types. Recent studies have demonstrated that CTCs isolated from patients diagnosed with SCLC can be deeply characterized on the genomic level and present a potential prognosis significance.

We set up a clinical study to collect blood from SCLC cancer patients treated at the Rennes University Hospital. We developed a versatile and easy-to-use workflow for CTCs detection, count and isolation from whole blood samples. Phenotypic analysis and genomic analysis of isolated cells confirm their tumor lineage. Generation of CTC-Derived Xenografts (CDX) in immunocompromised mice support the tumorigenic properties of CTCs. Further genomic analysis of CTCs and matched tumor biopsies from four patients reveal commonly found genomic alterations and biological pathways impaired in SCLC (Ricordel et al. biorxiv).

Bioinformatics background

Our goal is now to analyze the genomic alterations of a cohort of more than 20 patients and to associate these alterations with clinical characteristics in order to identify potential biomarkers. To this end, we have developed a workflow implemented in a conda environment to automate the analysis of the Whole Exome Sequencing (WXS) data produced for this project, from the pre-processing of the fastq files to the filtration and prioritization of the variants (SNVs). We added tools (CNVKit, FACETS) to this workflow to identify somatic copy number variants (CNVs) and to infer ploidy, purity and mutational burden of each sample. Based on the genomic profile of a sample, we can analyze the intra-tumor heterogeneity and deduce the number of clones that constitute the tumor population as well as the evolutionary relationships between them (molecular phylogeny). To do so, we have integrated the tools Pyclone-VI and PhyloWGS in our workflow.

Bioinformatics objectives

The overall objective of the internship is to perform the final analysis of our cohort. The main objective is to finalize the evaluation of tools allowing to (i) determine the Copy Number profile of a sample, (ii) identify clonal subpopulations of a tumor by integrating other tools (PALIMPSEST, SciClone) and/or, (iii) produce evolutionary trees based on this clonal deconvolution (MEDICC, SCHISM, PhyloWGS, TuMult). The evaluation of the tools will be based on their capacity to identify genomic variations (biomarkers) and associated potential drug targets linked with clinical characteristics (mainly resistance and recurrence/metastasis). A secondary objective is to correlate the final analysis with the clinical data.