Metastasis is the final stage of cancer progression and is responsible for more than 90% of cancer related deaths, claiming millions of lives worldwide. The most formidable challenge in the treatment of metastatic cancer is the emergence of resistance to current therapies. Informed by clinical data and using functional in vivo and in vitro models our group seeks to understand the mechanistic determinants for cancer progression, defined by metastatic competence and resistance to therapies. Both rely on programs that ensure the cancer cells’ survival during periods of enhanced cellular stress, limited survival cues and attacks by the immune system. Although fascinating and clinically important, mechanistic insights into these programs are still lacking. Our ultimate goal is to enhance the efficacy of current anti-cancer therapies using combinatorial treatment regimens and find new therapeutic avenues to fight metastatic cancer.

The treatment of metastatic cancer has undergone a paradigm shift in the last couple of years. The identification of specific “driver mutations” in tumors holds the promise of tailored, mechanism-based treatment strategies, which are commonly referred to as ‘targeted therapies’. These new generations of targeted therapies can achieve tumor control for several months and have replaced unspecific cytotoxic chemotherapies for many cancer types. However, durable treatment responses are rare, due to the emergence of aggressive, drug-resistant clones that drive relapse and rapidly form new metastases (Figure 1, below). As a result, cure rates and long-term survival rates of metastatic patients treated with targeted therapies remain disappointingly low.

To better understand the clinical emergence of resistant cells our work focuses on the poorly understood events during tumor regression.


By combining the power of experimental model systems, in situ gene expression profiling techniques, and computational analysis, we recently discovered that targeted therapy with kinase inhibitors induces a complex network of secreted signals in drug-stressed melanoma and lung adenocarcinoma cells (Obenauf A et al., Nature, 2015). This response, termed therapy-induced secretome (TIS), does not only enhance the survival of drug-sensitive cells, but also stimulates the proliferation, invasion, and metastasis of drug-resistant clones that are lurking in the background of the regressing tumors (Figure 2a-c).

We also found that the regressing tumors in our animal models act as potent ‘magnets’ to attract drug-resistant cells from the circulation. This process, termed self-seeding, could add an additional layer of complexity to the treatment and relapse of patients on targeted therapy (Figure 2d). Our findings establish a general mechanism by which drug-stressed tumor cells can aggravate cancer progression and it could – at least partly – explain why targeted therapies rarely lead to complete tumor regression. By dissecting the processes that lead to this phenotype in melanoma, we identified that the TIS consists of a vast number of signals that are capable of activating multiple signaling pathways, including an important survival and proliferation pathway (PI3K/AKT/mTOR pathway) (Figure 2e). The addition of AKT/PI3K/mTOR inhibitors blunted the outgrowth and metastasis of resistant cancer cells in animal models. These experiments suggest that this drug combination is a potential strategy to delay tumor relapse in patients. Even more importantly, our study started to expose the significant changes in tumors treated with targeted therapies. These changes are still largely unexplored and are expected to have a major influence on the efficacy of other drugs, such as immunotherapies, when given in combinations.