My lab is interested mainly in breast cancer research and more specifically we’re focussing on the question of targeted therapy and in how far, actually, targeted therapy is not working or leaves residual cancer cells.

That is a long-standing question in the field because it’s hard in patient tissue to get a hold on these residual cells; you don’t really identify them.

So we try with our model system to get this substrate and to be able to study that and then map back to the animals and also to patient samples.

So that’s in the ballpark what we are trying to do.

Are you working in vitro?

We are working actually with a combined in vitro/in vivo model.

So we are taking from the mouse model where we can induce breast cancer at a given time point in adult stage.

So we can take the exact same cells and put them into an in vitro system, in an in vitro 3D collagen Matrigel system and develop in parallel to the mouse, almost like in an isogenic way.

So we have the mice with nine mammary glands left over and one mammary gland we take, we digest the cells out and we establish our in vitro system in parallel.

So we can really look at the effect of oncogene induction and treatment of these tumours in vitro in parallel to the mouse in vivo, so almost in an isogenic way.

What are your findings so far?

We are in early days but what we could characterise right now is this substrate of surviving cells that will go into relapse later on.

By this trick of having this in vitro system we can get a good look at gene expression patterns in these cells and at the exact differences of those dormant cells as compared to normal breast cells.

That gave a good first idea on what might be going wrong in these cells, that they will come back as a relapse later on.

This is a quite important question also for breast cancer patients in general because the percentage of patients that relapse after five year survival or even longer, ten year survivors, in breast cancer specifically is very high.

So this is a pertinent question in the field.

So to get the profile of those cells and get an idea on what’s wrong in metabolism, what’s wrong in their DNA structure and so on and so forth, just the ability to do that is a first big step for us.

Now, we were able to drive this a bit further in the question of deregulated lipid metabolism which we found.

This seems to create a rather dangerous state in these cells that leads to deregulations and better fatty oxidation in all these kind of pathways that at the end of the day will result in an increase in reactive oxygen species or other dangerous products of the cells that are accumulating there.

We have first data now that suggests that this kind of accumulation… so these cells are hanging on, if you want, they’re not dying, they’re still there but they accumulate damage.

They get adduct formation on their DNA; they seem to have not the capacity to deal with that.

So either they have downregulated DNA repair mechanisms or they’re just overwhelmed.

That results, and that was the hypothesis, that that will result into point mutation and drive of the relapse.

Now, what we did and what we recently got the data in is we did take a forward approach and try to suppress the reactive oxygen species at that state and could delay, actually, relapse.

This sounds probably a little bit trivial but what was the real surprising thing for us is that in the absence of an oncogenic signal, so these resting cells, do have features that you would expect to have in the presence of oncogenic signal, that’s well known.

But they seem to carry this over and that seems to make them prone to get into relapse.

Are these cells spatially separate from the ones that responded to the therapy?

What we do see is a massive reduction of the tumour volume in response to applying… we call it the best possible therapy because what we are doing in our animal models is we simply switch off the oncogenes.

So this is, if you want, the best possible therapy thinkable.

But the bad news is that you still get relapses in animals so that’s the bad news for the patients, I would say.

So even if you have the best possible treatment the probability to get relapse is pretty high.

But, to your question, 95% of the cells or even more are cleared.

So this principle of oncogene dependence is working.

So if you inhibit the original signal of the oncogene you will get a lot of apoptosis in the tumour and you will actually almost rebuild the mammary tree again.

Functional mammary tree, by the way, is so these animals are able to nurse their pups.

However, whoever is surviving, and that’s a good question because that we cannot really answer since we cannot really trace them in the animal model yet.

We’re working on basically introducing markers, actually.

But that’s probably the difference to the in vitro system where we do see a surviving rim of cells.

So also there we have, the same way as in animal models, a large percentage of cells clearing but we don’t know the exact order of the surviving cells.

We do believe we have a heterogeneous rim which will be pretty much the same as in the animal so that we cannot say right now.

But, again, we do have a heterogeneous population of survivors and that’s also the next step we want to invest in, to try to separate those surviving cells in different pots – how many populations do we get and which one is the population that is mostly prone to get into relapse again?