AACR 2014

Microenvironment of tumours key to cancer's progression

Dr Rakesh Jain - Harvard Medical School, Boston, USA

Can you tell us more about your work on re-engineering the tumour microenvironment?

Our hypothesis is that the tumour’s microenvironment is an important player in tumour progression and resistance to treatment. What we have learned in our work over the years is that the microenvironment of tumours is highly abnormal. Each component of this abnormal microenvironment facilitates resistance to treatment and our hypothesis is that if you can repair these abnormalities, or in engineering terms you would say if we can re-engineer the tumour microenvironment so it begins to resemble a more normal microenvironment, then perhaps we can improve various treatments in cancer patients.

Which components were different from usual?

A tumour is like an organ, like a tissue. It has cancer cells that drive the process but cancer cells cannot alone kill a patient until they can co-opt the host. These host components are known as microenvironment and they include blood vessels, lymphatic vessels and a variety of host cells, cells from the patient that are in the tumour, and they include fibroblasts, myofibroblasts, cells of the immune system, the resident cells like macrophages or transiting cells like T-cells. Each of these components are embedded in an extracellular matrix which is made out of collagen and hyaluronan and other proteins and other substances. Each of these components of the tumour microenvironment are abnormal and what we have been trying to do in our laboratory for the last two decades is try to repair each of these components and then test these concepts on patients. So we have done a number of clinical trials to test these ideas and they have held up so far, they are supportive of this concept.

How is it different than from within a normal tissue?

Let me step back a little bit. One of the major causes of tumour progression and resistance to treatment is hypoxia, low oxygen level or low pH in tumours. Tumours create this microenvironmental low pH and hypoxia by making the blood vessels of the tumour abnormal. Abnormality of the tumour vessels stems from two reasons: number one, tumour vessels are compressed by cancer cells, what is even more important is they are compressed by the matrix, both collagen and hyaluronan. Then the second reason why the tumour vessels are abnormal is because they’re leaky. So we had to develop different strategies for each of those. So, just to give you an idea, one of the most deadly human tumours is pancreatic ductal adenocarcinoma where the survival rate, five year survival rates, have not really budged in over thirty years, they’re around 5-6%. Now, what is remarkable about this tumour is only 5% of the cells in this tumour are cancer cells, 95% is the microenvironment. So that tells you the most deadly tumour is deadly because of not 5% cells but because of 95%. Of this 95% a good fraction of this is collagen and hyaluronan and what this does is it compresses tumour vessels. So 80% of the blood vessels in the pancreatic ductal adenocarcinoma are compressed, are squished. Therefore, the flow just doesn’t go through them. And if there is no blood flow in those areas it’s going to cause low oxygen levels. This hypoxia is going to have many different consequences – it’s causing immunosuppression, inflammation; it causes resistance to radiation therapy, a number of chemotherapeutic agents, causes resistance to immune therapy and the list goes on and on.

So we decided to alleviate hypoxia in tumours by decompressing blood vessels, by somehow opening them up. What we discovered is that we can do this by depleting collagen and hyaluronan from these tumours. So we’ve been looking at a number of strategies to do that. What I’ll talk about on Monday is one such strategy, that is to use widely prescribed anti-hypertensive drugs that are given to patients worldwide so they’re very safe. What they do is they indeed decrease collagen levels and hyaluronan levels in tumours and pancreatic cancer and they open up the blood vessels, they reduce hypoxia and you can deliver your drugs, you can deliver immune cells and radiation works better because oxygen is there. So this is one of the approaches for normalising the microenvironment, you’re repairing the matrix and you’re opening up vessels. So that’s the strategy.

Does the microenvironment precede the tumour?

It’s a very important question. I don’t think we have the definite answer to that question but at least my hypothesis it is there. We have done studies in the very early stages of carcinogenesis like during pre-neoplasia, and the two stages are called hyperplasia and dysplasia before frank neoplasia sets in. What we have found out is the microenvironment is already abnormal at that time, so the blood vessels begin to get squeezed at that time, the lymphatics begin to get squeezed at that time. Therefore the tumour cancer cells before they become frankly neoplastic they are already preparing the microenvironment, they’re preparing the niche where they’re going to grow.

Is there any comparison between tumour microenvironment and extracellular matrix during regenesis and stem cell migration?

Essentially if you look at a tumour it essentially recapitulates a lot of the embryonic programming, it just comes at the wrong time but it exploits that way. That’s why all the various transcription factors that are involved in embryogenesis also get activated during tumorigenesis. So your point is well taken. Some of those tumour elements, tumorous microenvironment phenocopies, some of the embryonic microenvironment. Except a lot of the proteins that come up are mutated, they don’t have the same structure, so to speak, or function as you would see in embryos.

And you’ve done quite a few trials?

This concept which I discuss about using anti-hypertensive drugs, there’s a trial going on started in the May of 2013 at the Mass. General Hospital being led by my colleague, Dr Ted Hong, where he’s combining… I’m not involved in this trial because of a conflicts of interest issue which I mentioned to you earlier. But my colleague, Dr Ted Hong, is leading this trial where he’s combining losartan, which is a generic anti-hypertensive drug, and he’s combining with standard of care chemotherapy known as FOLFIRINOX, it’s a four-drug combination, and he’s combining it with the losartan. Then he is giving radiation using proton beam and then seeing how it affects. It’s a very sophisticated trial design and it’s on the clinicaltrials.gov if you see. It’s really a pioneering trial.

This is for pancreatic cancer?

The ductal adenocarcinoma which are the majority of pancreatic cancers and the ones where we haven’t made much progress.

How is it different than from within a normal tissue?

Oxygenation also helps with radiation in addition to chemotherapy so this is the rationale. So this is just one… But again, as I said to you, one approach to repair the vessels is through opening them up and 25% of human tumours, 25%, have compressed vessels like this. So this concept, if it works in pancreatic ductal adenocarcinoma it’s going to have… I can tell you the implications of that, you can deduce them. So we should know the results and, of course while this trial is going on our laboratory is also developing better anti-hypertensive drugs that are even more effective. So let’s see what happens, hopefully we’ll know the answers for that in the next several years.

But, as I said to you, that’s only one component of normality, the abnormal matrix. The blood vessels in the tumour are also leaky and if you’re watering your lawn and suddenly if you put a leak in the hose or the pipe what will happen to the water? It will begin to shut down, it will not flow as rapidly. That’s exactly what happens with tumours. The tumour blood vessels are leaky and this leakiness makes the blood-flow quite sluggish in tumours so again it creates hypoxia. So what we proposed in 2001 is a very controversial idea that let’s use antiangiogenic drugs to repair this leakiness. You can imagine it was not very well received because at that time the whole concept of antiangiogenic therapy was to starve the tumour, to simply shut off the blood supply and I was trying to propose the opposite – let’s fix the blood supply, let’s improve the profusion in that tumour. Then once you do that then hypoxia would go down, pH will begin to move towards normal and the same thing as I said earlier, by opening out vessels, the decompression, the same kind of advantages will come in. The tumour immuno-suppressor microenvironment would become immuno-stimulatory, drugs will get in there or begin to work better, radiation will work better, immune therapy will work better and the list goes on and on. So now we tested this first in animal models. I proposed this idea in 2001, so thirteen years ago. We did a number of preclinical studies to test this concept and indeed that was happening if you used a lower dose of antiangiogenic therapy. So then we did the very first clinical trial at Mass. General, led by my colleague Dr Chris Willett in rectal carcinoma patients using bevacizumab or Avastin, whichever word you’re comfortable with. We found out, indeed, the blood vessels were getting repaired, normalised with Avastin.

But once we finished that work it led to the next set of questions – when does normalisation come, set in, when does it end? Does it really benefit patients? So we went back to the bench-side and began to work with brain tumours in mice. We answered all those questions. We found out that normalisation begins in a day but unfortunately lasts only five days in mice. But the good news is if you give radiation during these five days the outcome is far superior compared to if you give it before or after the normalisation window. So, again, this raised the next set of questions – what about patients. With my colleague, Dr Tracy Batchelor, who is the head of neo-oncology at Mass. General Hospital, we initiated a number of clinical trials in glioblastoma, in brain tumours. As we had anticipated there was a window of normalisation in patients, it’s about a month, not as long as we would like it. In patients we would like it to be six months because that’s a total course of chemoradiation therapy but nevertheless the most exciting finding came when we found out in our very first glioblastoma trail that had 30 patients that in 7 out of 30 patients blood perfusion actually went up after you give antiangiogenic therapy. Now remember, until that point the whole field is thinking antiangiogenic therapy is starving tumours and guess what? These 7 patients, the ones where perfusion went up, these patients lived six months longer median survival compared to where the perfusion went down or remained stable. So when we finished that trial that was very gratifying that now we know how to select cancer patients who would actually benefit. We also know not to give this drug to 23 other patients because these drugs are toxic and expensive.

But, again, with one trial you cannot deduce so we said, ‘OK, let’s do another trial.’ Dr Batchelor completed a second trial, this one on newly diagnosed glioblastoma patients and this trial had 40 patients in it. In 20 patients out of 40 blood perfusion went up, went up after antiangiogenic therapy.

Did you look at the effect on blood vessels?

Yes, of course. We did this with MRI. So then what we found out is that these 20 patients had a median survival which was about nine months longer than the patients where the perfusion went down or remained stationary. So now we have two trials where we can show this in GBM. One could say, alright, maybe it’s specific to GBM. So there are two unpublished studies right now, but the abstracts were presented at ASCO meetings, where the same thing has been seen with bevacizumab in breast cancer and non-small cell lung cancer. So those studies have not been published yet so they are unpublished data but they have been presented at ASCO. So it looks like this principle of decreasing hypoxia by repairing blood vessels seems to make sense.

How does that fit in with them being antiangiogenic?

Originally, so the anti-VEGF or bevacizumab, which is an antibody against VEGF, VEGF was originally discovered at Harvard, not as VEGF but as VPF by Dr Harold Dvorak; he called it vascular permeability factor. So the effect of VEGF is a double-edged sword: it makes blood vessels leaky which reduces perfusion but VEGF is also a survival factor for endothelial cells. So if you remove just a little bit of VEGF, you repair the leak, you improve the perfusion. If you give too much VEGF you’re taking away the growth factor endothelial cells need to survive and then what happens is you shut down the blood flow. That was the original idea. So the idea is what we need to do is give just a judicious dose, personalise the dose. You cannot give the same dose to every patient. The dose may need to be titrated based on imaging and those are some of the questions Dr Tracy Batchelor and other oncologists at Massachusetts General Hospital are right now answering and asking in different trials.

The other thing we are trying to do is figure out what can we add to bevacizumab so that the normalisation window is longer and more stable. So we have some leads on that and we are working on that.