The molecular medicine programme essentially was generated as a platform to facilitate the interaction between the clinicians and translational and basic scientists. Within that programme my group is specifically involved in ovarian cancer and also in some studies on tumour vasculature, namely the ability of a tumour to recruit its own blood vessels, which is a key step towards the growth of any solid tumours and also provides the highways to tumour cells to disseminate to distant organs. This is, of course, a crucial step because, as it is widely known, the real lethal step in any solid cancer is not the primary tumour in most cases rather than metastasis.

So in that particular aspect my group has been focussed on an unexpected player in the field, and when I say unexpected I mean that we work on a class of molecules that were discovered in the central nervous system where they act as players in brain development and plasticity. However, we actually found out that specifically one of these molecules, which is called L1, is actually also found in the vasculature of tumour tissue but very rarely in the vasculature of normal tissues which suggests to us that, first of all, that this can be a biomarker of pathological vessels. This is really important because there are no clear biomarkers that can discriminate between pathological and normal counterparts. But more importantly this allowed us to ask the question – is this just a marker or is it functionally contributing to tumour vascularisation and, more in general, to tumour development? This is actually the question we addressed using a mouse tumour model in which we deleted, specifically, this molecule, L1, from the vessels and thus the question what would happen to a tumour in this type of context? In this particular case we used a mouse model of pancreatic cancer but now we know that this is actually a model that can also be extended to other different tumour types. However, going back to the pancreatic model, what we found is that the ablation of the molecule from the tumour vasculature is actually sufficient to dramatically reduce the growth of the tumour, not only the growth of the tumour itself but also the number of vessels that any tumour is able to recruit, indicating that this L1 within the vascular compartment is actually a key requisite for endothelial cells to be recruited by the tumours and form the vasculature of the tumour itself.

One additional observation that we have made is actually that L1 in endothelial cells, so in the vessels, is not only involved in the number of vessels but also in their maturation state. In other words, tumour vessels are known to be pretty immature and this is what makes them difficult to attack from a therapeutic point of view. It also makes them more permeable which also allows, for example, aggressive tumour cells to enter the circulation and disseminate into distant organs. We found out that actually L1 within the endothelial cells of tumour vessels plays a causal role in facilitating and enhancing the permeability of these vessels so facilitating the metastatic dissemination of a tumour. And to complete the story, we actually asked the question can we actually interfere with this process? We generated some antibodies that specifically inactivate L1 in the vascular compartment and this turned out to be a suitable strategy to first of all decrease the number of tumour vessels which is already a result on its own because this is what the whole anti-angiogenic field is looking for. But, more importantly, this was also a way to restore a normal mature phenotype in tumour associated vessels in order to interfere with the L1 induced permeability of the vessel themselves. We now think that this actually can be a strategy to test, first of all in pre-clinical models but then hopefully also in a more clinical setting, as a strategy to decrease tumour vascularisation and also to help the restoration of the normal phenotype in a tumour vasculature. Because what people think is that restoring a normal phenotype in the vessels, so making them less permeable, less leaky, would also help in the classical chemotherapy to systemically administer to reach the tumour tissue, especially the central core of solid tumours, and to get hopefully a better response of the tumour in a therapeutic perspective.

And you’re starting to look at human tissue?

Of course, after being encouraged by the results in the pre-clinical models we are now, first of all, looking in a wide variety of human tumours to see whether this aberrant expression of L1 in the vascular compartment is a common thing in tumours, and it looks like that’s the case. Of course, this makes L1 a general biomarker for tumour vasculature in the first place but, most importantly, it makes it a potential target for therapeutic strategies aiming at the tumour vasculature. One added value of this potential strategy is that also many tumours express L1 not only in the vasculature but also in the tumour cells where it has been proposed also by us, for example, in ovarian cancer to functionally contribute to aggressiveness. So we think targeting L1 could have a dual beneficial effect on the vascular compartment but also on the tumour cell malignancy itself.