Brain metastases are secondary cancers that travel from a primary location of cancer, like lung cancer, breast cancer or melanoma. These cancers seed the brain even after treatment of the primary cancer and cause a second cancer in the brain. These cancers are notoriously difficult to detect and to model or to try to develop scientific systems to portray accurately what happens in human patients. They are also very difficult to treat. Currently if you have a brain metastasis the median survivorship is 2-5 months and the most common treatment is palliative radiotherapy and there are very few targeted therapies for brain metastases.
The three most common sites of brain metastases are, first of all, from lung cancer, secondly from breast cancer and third from melanoma or skin cancer. We have chosen to profile lung to brain metastases as they are the most common type of brain metastasis.
Let’s get into the details then - the therapeutic targeting at the pre-metastatic stage.
Yes. I mentioned it’s very difficult to track the metastatic process in a human patient because if you can imagine how many steps are involved in the metastatic cascade, first the cell has to break off from the primary tumour, then it has to invade into the bloodstream, then it has to circulate and overcome all the barriers of actually getting through the circulation and then the cell has to somehow invade into the secondary site, in this case the brain. That means these cells have to cross the blood-brain barrier and then they form what we call micrometastases, tiny little spots, and then finally a full blown macro metastasis which produces symptoms and is then detected and very often surgery is done if possible and, if not, palliative radiation therapy.
So you can imagine with all of those different stages at the human patient level we’re not able to detect most of them. Most of the time patients show up with a full-blown brain metastasis. So I felt it necessary in my lab programme to develop an accurate model that recapitulates every stage of the metastatic cascade. So we take cells from patient derived brain metastases, those patients who have had surgery, and we dissociate those cells and then inject them back into immunocompromised mice. This is like an assay or a model system where we can read-out what the human cells will do in what we call an in vivo environment or in an actual other mammal. What happens with these cells is when we inject them into the lung, remember that’s their primary site of origin, we can actually track through time every stage as the cells make their way to the brain because we’ve labelled the cells with a green fluorescent protein. With mice that can be sacrificed at different stages of time you can then detect where the human cells have gone and what are they doing.
So we were able to detect a stage that we were quite excited about. What happened is we injected cells into the mouse lungs and the majority of mice would die of lung cancer burden because tumours would form in the lung before they would have any visible signs of brain metastasis. So that means if we were to do an MRI on the mouse then the brain would be clean. If we were to even remove the brain of the mouse post-mortem the brain would look normal. So we would think, ‘Oh, we didn’t capture the metastatic stage, we missed it.’ But, remember, we tagged those cells with GFP, green fluorescent protein, so what that means is we can take the entire mouse brain, which to all appearances is normal, and we can dissociate it. What we found was we could recover a very rare fraction of GFP positive cells and there was maybe only 1% got into the brain and they were not detectable by any usual method. So we called this stage the pre-metastatic stage because this must be the stage in the metastatic cascade where the cells have just crossed the blood-brain barrier, they’ve just invaded into the brain but they haven’t even yet formed a micrometastasis that would be detectable.
Remember, the most important thing about this model is we will never be able to detect that stage in a human patient and yet, if we know what those original invaders are, if we know the nature of those few cells that first cross into the brain, perhaps we can target them and then block or prevent brain metastases from even forming.
What is being done to further characterise this invasive percent?
We’ve recovered those cells back and now we’ve profiled them. So we’ve looked at their gene expression profile, we understand the genes that they express, and now we can do something called connectivity mapping which means we can take those gene signatures and input them into a big computer programme that tells us which drugs would then be able to target that particular gene signature. Now, of course, we have a nice model, an experimental model, where we can test those drugs. In fact, we found one drug in our recently published Cancer Research paper from this year, we found one drug that is actually able to prevent those cells from crossing into the brain.
So we treat mice with this drug, which is called apomorphine. If we treat mice with apomorphine then the percentage of GFP positive cells that get into the brain after our assay is zero. So this is a proof of principle experiment, we don’t know if treating human patients with apomorphine would actually prevent their brain metastases but theoretically as a proof of principle this is a first step to try to understand that if we can profile someone’s original lung cancer and then somehow have these predictive biomarkers that tell us, maybe this pre-metastatic signature will inform us about whether you’re at risk of getting brain metastases or not. Furthermore, if we can profile those cells we can tell you you should be treated preventatively with this drug so that you’ll never get brain metastases.
So we’re hoping that this model will provide a new therapeutic paradigm where we can start thinking about not treating necessarily the full-blown brain metastasis – when the horse is already out of the barn, there’s no point in closing the door now – but can we get ahead of the metastasis curve with a preventative therapy.