Cancer is one of the leading causes of death worldwide. The development of novel high-throughput genome analysis technologies have enabled the detection of alterations in the genomes of tumor cells at single-base resolution and in ‘real-time’. Genome alterations (gene mutations, gene copy-number changes, genomic rearrangements, etc.) lead to irreversible changes in intracellular signal transduction pathways that the tumour cells become dependent upon. Understanding these genomic events as well as their impact on cellular signal transduction pathways is a crucial prerequisite to the development of novel therapeutic strategies for fighting cancer.

Lung Cancer Genomics

Our lab has a focus on analyzing understudied subtypes of lung cancer, including squamous-cell carcinoma and small-cell carcinoma (SCLC). We seek to identify genome alterations that permit selecting patients for personalized cancer therapy. Such alterations can enable treatment with higher efficacy and reduced toxicity.

We have recently initiated a large-scale sequencing efforts of SCLC with the goal of identifying novel and potentially clinically relevant mutations. We have performed an integrated genomic analysis of a collection of small-cell lung cancer including SNP array analyses, exome, transcriptome and genome sequencing. We have made novel observations and found focal amplifications of FGFR1 in 6% of SCLC cases, pointing to a potential therapeutic opportunity in these tumours (Peifer et al. Nat Genet. 2012). We are currently expanding these efforts to provide a comprehensive genomic encyclopedia of SCLC and of related tumors, e.g, large-cell neuroendocrine carcinomas or carcinoids. Our lab is also conducting functional experiments to characterize the functional properties of the mutations that we find.


Genomics of Resistance

Recent clinical successes in cancer genome-based personalized therapy have led to high response rates and improvements in progression-free survival. Unfortunately, though, resistance will ultimately occur, thereby limiting the overall efficacy of the treatment. Another major focus of our laboratory is therefore to elucidate the molecular mechanisms of cancer drug resistance with the goal of developing new therapies. To this end we employ genomic sequencing and functional genomics approaches to characterize tumor specimens obtained at the time of acquired resistance to targeted therapeutics.

Genomics of Drug Response

We have developed a functional genomics approach offering the possibility of characterizing the functional impact of many cancer genome alterations simultaneously. To this end, we have established a high-throughput pipeline for chemical perturbations in vitro (Sos et al., J Clin Invest, 2009; Sos et al., PNAS 2009). The centerpiece of this approach is a large collection of lung cancer cell lines whose genomes have been characterized in great depth. Importantly, these cell lines are genetically representative of primary, surgically resected lung tumors, thus allowing clinically relevant predictions. We perform large-scale compound and RNAi screening with the goal of identifying those cell lines that are particularly sensitive to a given perturbation. We employ computational prediction tools to identify genomic correlates of drug sensitivity. This approach offers the opportunity to generate both cell biology understanding and to identify genetically defined patient cohorts likely to benefit from the compounds tested at the same time. We have used this approach to find that tumor bearing alterations in the Ras-Raf signalling pathway are particularly sensitive to combined inhibition of the Pi3-kinase and Mek pathways (Sos et al., PNAS 2009).

Sos_PNAS  figure_2

Cancer Genome Diagnostics

Given the immediate relevance that several cancer genome alterations have in guiding molecularly informed therapy decision, sensitive analytical methods are needed to ensure accurate mutation diagnosis and therapeutic patient stratification. We work on developing sensitive methods for the parallel detection of somatic genome alterations in clinical cancer specimens. In this project, we work closely with the Lung Cancer Group Cologne (LCGC) and the Department of Pathology at the Center for Integrated Oncology ( Together, we have launched a new outreach initiative, Network Genomic Medicine (NGM) with the goal of providing accurate and comprehensive genomic diagnoses to non-academic hospitals as well as to oncologists and pneumologists in private practice. At this point, more than 3,000 patients are being diagnosed annually.

In order to provide sensitive and accurate genome diagnostics, we have developed a next-generation sequencing test and novel computational algorithms to detect all relevant genome alterations in parallel in a single-tube assay. We have optimized this assay as well as our tailored data analysis pipeline through multiple rounds of sequencing of clinical cancer specimens. This approach will be employed for routine clinical use within NGM before the end of 2013.