2010 San Antonio Breast Cancer Symposium, 8-12th December, USA
Interview with Dr Jeffrey Rosen (Baylor College of Medicine, Houston, USA)
Targeting of breast cancer tumour initiating cells
It’s the Brinker Award for basic research in breast cancer and they’ve been giving these for about eighteen years and they look at not just recent accomplishments but lifetime accomplishments in terms of breast cancer research. In our case we’ve worked normally on normal mammary gland development and moved that work on normal development now into more translational research in the clinic.
So we started about a decade ago now looking at stem cells in the normal mammary gland, trying to understand the stem cell hierarchy and understand how the different cell types originated in the mammary glands, similar to what people had done earlier on in the metabolic system where they looked at the different lineages in the metabolic system. That work on the metabolic system actually impacted leukaemias because the different leukaemias came from different cell types in the metabolic system. So similarly there are different subtypes of breast cancer and we had suggested, and other people now have confirmed, that the different subtypes may have different cells of origin, depending on the stem and progenitor cells in the mammary gland.
More importantly what we were interested in is whether these cells were what we call cancer stem cells or tumour initiating cells within the different types of breast cancer, were responsible for resistance to conventional therapies, both radiation and chemotherapy. So there is not only intratumoral heterogeneity between different breast cancer subtypes but within any given breast cancer there’s heterogeneity with the cells in that tumour. So the goal was to really see if there were subsets of cells or subpopulations of cells that were different from the bulk of the cells, that wouldn’t be a genetic change, that would be an epigenetic change that they had undergone, and whether if you treated with chemotherapy and radiation you could identify why those cells are resistant to treatment.
So we did this both in the clinic and in pre-clinical mouse models. In the clinic we have a population of women in Houston that about 1.5 million are uninsured. In fact, the state of Texas has the most uninsured people in terms of health insurance in the United States. So we have a county hospital that’s essentially the charity hospital in Houston which sees breast cancer patients who come in with an average breast cancer of about 9cm in size in contrast to the Methodist, the private hospital, and MD Anderson which see breast cancers of 1.8cm in size. So a big healthcare disparity a hundred yards each side of our building. These patients have such large primary tumours they usually have to be treated with what’s called neoadjuvant chemotherapy ahead of time.
And so this provides an opportunity where you can do vacuum assisted biopsies and our clinical colleague, Jenny Chang, was able to do this both pre-treatment and post-treatment. When we did that in the chemotherapy treated patients and then used FACS sorting to look at markers for stem cells that had been already used by other investigators, Mike Clarke and Max Wicha, we found that the proportion of these cells went up after chemotherapy. So the tumours were shrinking but there was actually an increase in these cells, they were resistant. We were actually able to grow them in sphere cultures as well and show that they were increased mammosphere cultures after chemotherapy.
So that was the clinical observation and then we developed some mouse models, genetically engineered mouse models where we can do similar kinds of things and show that they’re resistant to radiation, even up to sixth grade they’re not killed. We’ve looked at why they’re not killed. So they get double standard breaks - the DNA is broken by the radiation so it’s not that they’re resistant to the damage but they repair the damage much better. So they have what is called a DNA damage response. So if you look 48 hours after you give sixth grade radiation, the bulk of the tumour cells are damaged and they’re dying but these cells are repairing the damage and they’re resistant.
So we’ve studied that in trying to understand how we could sensitise those cells. What could we do to actually make them not repair the damage and actually be sensitive? We’ve done two kinds of experiments: the first is a nanomedicine experiment where my student discovered, by serendipity, that if the temperature went up from 37º to 42º that the cells now were sensitive to radiation. This was actually quite a funny story because she came in one day and her cells were dead and her glasses fogged up and she realised the temperature control on the incubator had been turned up by mistake and wasn’t working properly. So then we decided could we do this in vivo and we didn’t want to heat the whole mouse up. People have used hyperthermia, it’s actually one of the oldest treatments, I believe it goes back in papyrus to the Egyptians where they’d try to heat up the cancers to kill it. But rather than heating the whole body up, which has been done, which has effects on the immune system, we wanted to have targeted hyperthermia. So a group at Rice University had developed gold nanoshells, a derivative of Smalley and the buckyballs years ago, and people at MD Anderson were actually using these in some studies. So we collaborated with Sunil Krishnan and a group at MD Anderson where we could inject gold nanoparticles and they get concentrated in the tumours because of the leaky vasculature in tumours, it’s called the enhanced permeability effect. Then you can take a near infrared laser and heat them up very nicely.
So we could do that where we rapidly heated up just in twenty minutes and then cooled down while we were doing the radiation and we found we could sensitise these cells. What we learned from those studies is interesting. We learned that looking at total tumour size and shrinkage is not a good indication of whether your fifty cells, as I’ve mentioned, and that’s the RECIST criteria for clinical trials – you look at tumour shrinkage. In fact these cells could have increased actually while you’re doing this kind of treatment, probably because you’re killing the bulk of the cells but also there’s an inflammatory response that probably stimulates these cells as well when you’re giving chemotherapy radiation.
Then the other thing we noticed is that if we took the cells that survived and transplanted them, the ones that grew back out after radiation alone actually were more aggressive looking and worse, they probably accumulated additional mutations and they actually were more metaplastic, while the cells that grew out after the hyperthermia and the radiation actually looked more differentiated. So this led us to suggest that what we need to do is not just treat and look at tumour size but look at what’s left over after you’re done with the treatment to try to see how efficacious the treatment is.
This is a really interesting approach, it’s very useful for a local disease where you can use the laser to heat but it’s not very good, obviously, for metastatic disease at distant sites. So it may be actually developed into clinical trials looking at locally recurrent breast cancer or possibly inflammatory breast cancer where it’s a local disease. But we need systemic treatments if we’re going to treat metastatic disease.
In the course of our studies looking at gene profiling on these tumour initiating cells we found that there were alterations, as I said, in the DNA damage pathways. There were also alterations in the P10/AKT and WNT pathways and so we confirmed that at the protein level, not just at the messenger RNA level, and then we actually set up an experiment to give an AKT inhibitor to the mice orally, it’s an orally active drug called Perifosine that actually inhibits AKT as well as downstream TOR1 and TOR2. Then we gave them radiation and we were able to show that that inhibited the AKT activation and also the downstream WNT pathway and actually sensitised these cells.
The other thing we did, which I think is going to be important, is we put in a pathway reporter and not just using surface markers to mark these cells but we actually put in a WNT pathway reporter with an antivirus and we could actually show that those cells that had the active WNT signalling were the ones that had the tumour initiation capability. That’s interesting because, again, studies in the normal gland have shown that the canonical WNT signalling is important for self-renewal of the stem cells by [Rollason] and colleagues.
So we went back and looked at the DNA damage response and when we blocked with the AKT inhibitor now, the tumour initiating cells looked just like the bulk of the cells and they were actually being killed as well. So that’s again generated a fair amount of excitement and there’s a number of clinical trials going on now with various AKT inhibitors to use them in combination possibly with chemotherapy to see if they can sensitise. Max Wicha did similar studies with the same drug using Doxorubicin, a chemotherapy agent, and showed similar results in some human xenografts. So the studies we’ve done in the genetically engineered mouse models have been replicated in human xenografts as well.
What clinical trials are underway?
There are some trials already with the gold nanoparticles for head and neck cancer that are being done in the Texas Medical Centre. The actual trial I mentioned is being written up as a proposal to get approval for the nanoparticles. The AKT I believe has already been started now at a few institutions to look at its effect.
Our goal is that we think that as single agents by themselves these may not be very useful but if you combine them with the standard of care, whether it be radiation or chemotherapy, it will be much more effective. We have to do that, anyway, in the clinic because you have to give the standard of care, you can’t not do that. So we think that’s the way to go and that we should test these first in the neoadjuvant setting just to see if we can show that we’re hitting the right targets and that they’re actually performing as predicted and then they could be moved into the adjuvant setting.
What’s the likely impact in the clinic?
The nanoparticles, the hyperthermia, has gotten people excited again but I think the difficulty there is for primary breast cancer, most times you’re treating it with surgery so it really has very little benefit right now. Unless you have local recurrence it’s not operable or inflammatory breast cancer that’s not operable.
It’s going to take a while for metastatic disease until they develop methods of activating these particles and targeting them. It’s the same problems we’ve had with gene therapy – we know what the defects are but we have to get to the right cells to actually be able to target.
The AKT inhibitors, as I said, there are a number of trials, clearly the PA3 kinase AKT pathway is one of the more mutated pathways in all cancers and there are a number of trials on these inhibitors going on now which hopefully will be promising. We’re going to need to do personalised medicine and treat patients individually to know what sort of mutations that they have and it’s going to be important to select the right patients for these trials as well. If you use it on a total group of patients and treat all breast cancers the same you probably won’t see very much of an effect but if you pick the right group of patients that have the right genetics, I think that you might see dramatic effects.
