Transcript Learn@Lunch with Professor Christopher Heeschen

Cancer & Embryos: the Yin and Yang of Stem Cells

11 July 2018


Welcoming remarks from Professor Rodney Phillips, Dean, UNSW Medicine


Ladies and gentleman, please do take a seat. Welcome. On behalf of the University of New South Wales, and in particular its Faculty of Medicine of which I'm privileged to be the Dean, my name is Rodney Phillips.


Before we start, I'd like to acknowledge the traditional custodians of this land in this place, particularly during NAIDOC Week when Australians I think quite rightly reflect on the colonial history of this country, and reflect, I think, quite carefully on the fact that the individuals who came to this place 50,000 years ago or more do have a very rightful place under the modern Australian sun. Regretfully, that rightful place under the modern Australian sun is not yet fully occupied in the minds of many of us, I think, here today.


It's my pleasure to welcome you to this series of lectures where we present some of the finest of our scientific leaders who present to us in language that I assume will be comprehensible, Chris, more or less. And it'll be an insight through these lectures as to the sort of quality of research that takes place at the University of New South Wales. I came back to Australia three years ago, and I was impressed with a number of things at the University of New South Wales, but the preclinical science, the basic science that addresses major problems in medicine, is an extraordinarily powerful grouping at the University of New South Wales. It's one of the great delights, to see it flourish.


Now, Chris Heeschen originally came to Australia to look at a job where he would work with me, but wisdom prevailed. Great wisdom, as it turns out, although he never worked with me before. But he stayed on in negotiations with us, and has come to us not as a administrator working in the mundane and grizzly catacombs of the faculty, but rather as a scientist. He holds senior rank in our scientists at the University of New South Wales. His training has been extensive throughout Europe and North America. We're very, very lucky to have him here. He works on a very fundamental biological concept, which is that there are cells persistent in the body of warm-blooded vertebrates like ourselves, which can form the basis of perpetuation of cellular lines and cellular elements in a constructive way. But regrettably, there are also cells that retain a precursor capacity to cause cancer. He's going to address some of the issues around the so-called stem cell. Chris, welcome, and thank you for your lecture.


3:45 Learn@Lunch presentation by Professor Chris Heeschen, UNSW Medicine

Thanks, Rodney. Thanks for all of you for coming. I will do my very best to make it simple but comprehensive at the same time. If you do have questions which really keep you away from understanding the big picture, just raise your hand ... Happy to answer them ... either during the talk or at the end.

So the yin and yang of stem cells, Rodney already addressed that there are obviously good things that have been going on with stem cells for a long time. I actually have been working on the sunny side of this concept, because I've been working on stem cells for cardiovascular regeneration during the first half of my career. But then I decided to look at the dark matters. And at that time ... That was about 10, 12 years ago ... the concept that there are stem cells driving not only tissue regeneration, but also driving cancer came up, originally from U.S., but we at that time in 2007 proved that concept also applies to pancreatic cancer. Well, pancreatic cancer that has been my passion since that.


So this is just a summary of what I'm going to talk about. I'm going to explain you very briefly what is a cancer stem cell, as opposed to a stem cell. We're going to show you some functional data showing how they are driving cancer progression and metastasis, the spread of cancer throughout the body. Obviously, our primary aim ... Because we are really a translational lab ... were to really bring things into the clinic. We are working very hard on eliminating those cells to prevent the relapse of the disease. So, just very basic, cancer is a disease of cells that have gone out of control. There are trillions of cells that we have in our body. At the end, it's just a number game, at what point of time this will go wrong, and some cells will acquire changes in their genome or epigenome that will make them go mad.


So this is a normal stem cell, what we have been working on previously. A normal stem cell is present in most tissues, particularly those that have a high turnover, like the liver or the intestine in the gut. So they give rise to what we call progenies. These are cells that then actually obtain the function for the organ, whereas the stem cell is just responsible for giving rise to those cells. And then those cells proliferate. So they divide one by one, and obviously this gives rise to an exponential number of cells. During that what we call mitosis or cell division, many things can go wrong.


Just to give you an idea of the number of cells we are dealing with on a daily basis, this is just what happened in the last minute in your body. 300 million red blood cells, the tiny red ones that are circling the blood, 12,000 million gut cells, so the gut has really a very, very high turnover, and skin cells, surprising not as many, 40,000 cells. This just refer to the last minute. So huge turnover of cells, which just illustrates you the risk that something can go wrong during all these cell divisions.


So usually this is a very controlled endeavour, so cells are in check by themselves. So they have an internal check, what we call a checkpoint. So whenever a cell goes through a cell cycle, which is the process for cell division, there's a check whether everything is okay with the genes and whatever comes out of that. But there's also an external check, which is the immune surveillance. So our immune system constantly checks the cells, whether there's anything going wrong. If there's anything going wrong, they eliminate the cells. And that is actually happening quite often. Just imagine, again, the numbers that we are talking about, quite often there's something going wrong, and the cell tries to repair itself, or it gets kicked out by the immune system. But if it's going wrong because cells have acquired whatever mutation, then there's a high chance that eventually there will be a cell that then escapes of this immune surveillance and then can start the process of what we call a transformation and become a cancer cell.


But a cancer cell is not like other cancer cells. There's a huge heterogeneity ... So that's a big topic nowadays in cancer research ... So they are cells that look different in a way that they do also express stemness factors such as NANOG... just to explain you where are the cancer cells. This is pancreatic cancer, by the way. These are the tumour cells here. What you see here, the more elongated cells, are actually the stroma cells. So this is kind of scar tissue. That's why cancer is also called a wound that never heals. So looking at these cancer cells, you'll see that some of them are brownish, and that is because they are staining for NANOG. NANOG is a transcription factor gene that is highly expressed in embryonic stem cells. So it's a embryonic stem cell transcription factor, which is really important for maintaining their stemness.


So for maintaining the stemness state, these cells need factors like NANOG. As you can see here, not all of the cells express NANOG rather only those that stain brownish, whereas all the other cells are not staining for NANOG. So this what we call functional heterogeneity in these cells. But obviously, just doing the staining doesn't prove that they are functionally different.


So why we're working on pancreatic cancer? So when I moved to my first independent position, I kind of thought, "Maybe I should work on a disease that is much more miserable." So came from ... I'm actually a MD, a medical doctor, by training, in cardiology. We've been working on cardiovascular regeneration... This is not so good… I've been working on cardiovascular regeneration to improve heart function in patient that have suffered from myocardial infarction, heart attack. But if you actually look at this curve, this is much worse than whatever you can see nowadays in patient with cardiovascular disease. So this a really a miserable outcome. So this is here showing a month. After two years, essentially all these patients with pancreatic cancers have died. So this was standard of care.


This is the chemotherapy that was standard for quite a while, until recently, when the addition of another chemotherapy, which is a bit of a nanoformulation, but just very slightly increases survival by two months. So this is what we call median survival, when half of the patients are still alive, is shifted by about two months. But what is important is actually if you look at the end of these curves, they're essentially more or less the same, because essentially still all these patients die. They die from relapse. So the disease comes back after there's initial response.


That's actually what we try to achieve with our research, to improve the long-term outcome by actually removing the stem cells. So essentially actually not removed by this therapy, and then eventually they lead to a relapse of the disease and then the patients succumb from the disease. But if you actually also eliminate the cancer stem cells, then the relapse will be prevented, and the patients would have an improved long-term outcome. So this is just obviously our long-term vision ... These are not real data ... but we have achieved this in mice already, and obviously the next step will be to translate this into the clinic.

So just to give you a brief introduction of what cancer stem cell is by definition. So reminiscent of a normal stem cell, there are inherent properties that define a stem cell. One of them is the indefinite self-renewal capacity. That means it can recapitulate itself or reproduce itself, and the same time give rise to these what we call progenies or daughter cells that then give rise to all these more differentiated cells. So these are the differentiated progenies, and this stem cell here stays there and can do this job indefinitely. It can give rise to the differentiated progenies, as I just mentioned. And what we use as a functional assay in the setting of cancer, is that they form tumours. So even if you take a single cell from this one here and put it into a mouse, it will eventually ... It takes a long time, obviously, because it's starting from a single cell ... it will eventually form a tumour. This will not be the case for this ... I'll show you some data in a minute. So there's a constant evolution in cancer.


So it's not just one cancer stem cell, rather than it's the evolution or process that eventually leads to cells that become mobile and can evade from the primary tumour and get into the bloodstream. This is supposed to illustrate the bloodstream. We have obviously lots of red blood cells, white blood cells, but also these cancer stem cells. We call them circulating cancer stem cells. These are the ones that actually can form or give rise to metastases. In the case of pancreatic cancer, this is mostly the liver.


So this is the evolution and the progression of the disease through the circulation circulating blood, and then eventually you have liver metastasis, which we call advanced pancreatic cancer. So how do we track those cells? It was quite interesting that this can be done using inherent features of those cells. This is called autofluorescence, which in the scientific world actually is ... is where I was suspicious that it's maybe an artefact, but we prove through many years of research, that this is actually real. So here you can see the cancer cells from the pancreas. And you can see that ... and so the blue is the nucleus, and the red is the cytoplasm, which is kind of the machinery of the cells, whereas in the nucleus, they have all the genetic information. You can see here these little green dots that are only present in a few of those cells. These cells are actually the cancer stem cells. Here's actually one that is currently dividing. If you look at this in high magnification, you can see that this is not in the nucleus, which is here, rather than this is in the cytoplasm.


You have these very distinct vesicles that actually have a membrane, and there are something green, fluorescent in the centre. Funny enough, this is a vitamin. It's riboflavin. You may have noticed if you take a lot of vitamin B, you get this very strong yellow urine. The same is happening here. It's just amplified using our microscope. So there's a lot of riboflavin pumped into these little vesicles. They actually obviously not supposed to pump vitamin B in there, but this is a mechanism how they get rid of toxic substances. So many chemotherapy therapeutic agents are also transported through these transporters. This serves as a sink for these cells.


That has been working for us as a very powerful biomarker to track these cells. So it's an inherent feature. We don't have to add anything. We can just track these cells, and this can be used to sort the cells. This is what we call a cell sorter. And just by activating the cells with a laser, we can then decide, "This is a green cell and so that is a positive cell," and it becomes kind of disposed or deposed in this tube, whereas the other ones go into the other one. This is just for education purposes. So every dot that you see in this plot here is a cell. Cells can have this autofluorescence, so this would all be those here in the red square, whereas the negative, which is the vast majority of these cells, would be in this circle.


If you then use this cell sorter and sort the cells, and then you can do these what we call limiting dilution assays. So you take a decreasing number of cells, as I mentioned before, a single cell, and put them into a mouse, wait a long time, and then you see if the mouse forms a tumour or not. So these autofluorescent negative cells, even with 10,000 cells, there's hardly any tumour formation. Particularly if you use lower numbers, you don't see any tumour formation. Despite the fact that these are bona fide cancer cells, they don't form tumours in reasonable numbers of cells injected. These autofluorescent positive cells, they always form tumours. Even if you only inject a single cell, five out of 21 experiments came up with a tumour. So that's about a quarter of the cells were able ... Even though you just use a tiny single cell ... to form a tumour eventually. So there's a 200,000 fold enrichment.


So it's not black and white, because you do see some cancer formation if you use higher number of cells, but I'm going to illustrate why this is the case. Because obviously these are bona fide cancer cells, and if you use high numbers just by magnifying their numbers, they will always form a tumour if you inject them high numbers. But this is the decisive test here, where you have single cells, no tumours, and a quarter have tumours. They are resistant to chemotherapy, as I mentioned.


Gemcitabine is what we use for pancreatic cancer. So if you expose the cells to chemotherapy, you see that ... So these are the negative cells, which are the bulk of the cells. And these are the positive ones, which have a couple of percent usually, it varies from patient to patient. But if you treat these cells with chemotherapy, most of these cells actually die here, the negative ones. Whereas the positive ones see a relative enrichment because all those are gone, and it looks like they are relatively enriched. But they are simply surviving. So they don't necessarily expanding, but they are surviving this therapy.


So that's what the features are biologically as a cancer cell. But what is their regulatory mechanism? And that is really reminiscent of what we see in embryonic stem cells. An embryonic stem cell, which can be harvested obviously from a developing embryo, they have a core signalling network. I mentioned NANOG before. There are other factors like KLF4, SOX2, OCT4. This kind of creates a network, a signalling network that drives or keeps the cell in this stemness state. They obviously drive other factors that then determine how this cell behaves. So this is the embryonic stem cell state. But if you actually look at the cancer stem cells ... Which this is what we call a Western blot ... so it just shows the protein that you can find in the cell. And here showing you NANOG and SOX2. So these would be these two factors.


This would be the cancer stem cells that show a bend. That means they have a lot of protein for these two genes. Whereas the negative ones, which are the differentiated progenies, they don't have this bend. So it's highly expressed in the positive cells, and not present in the other. So they're really driving or riding on the stemness network you can find in embryonic stem cells. There's a tricky way we can track this in real time. I showed you this autofluorescence, but obviously that's not something we can use to further manipulate the cells. So we're actually using NANOG, this gene here, to drive a red colour. So this is just meant to show you which cell ... without doing this kind of Western blot here ... which cell is actually a stem cell, and lighting up in red. We call that a reporter.


So this red colour reports which cell is NANOG-positive. This is a ball of cells that is floating in a culture dish. There are about 50 cells. It looks quite nice and round because there's a lot of matrix, but these are actually 50 different cells. Some of them are highly red, so they are NANOG-positive cancer stem cells. We cannot only just show, "Here is the consensus," we can actually use then this gene to drive what we call a suicide gene. That is an interesting approach, where you can actually tell the cell, "Kill yourself, but only kill the red cells." So what we do here, is we link this to a TK ... Doesn't really matter what it is ... but TK converts something which is a drug but doesn't have any effect on all these cancer cells, but if a cancer cell has this TK, it will kill itself. So any red cell will have TK, and will die upon exposure to ganciclovir.


What happens if you take these cells and put them into a mouse? Obviously, you get this very rapid tumour growth, so these are days. This is the tumour volume here. Within a couple of weeks, you get relatively big tumours, and that kind of are out of control. And then you treat them with this chemotherapy that I mentioned before, which is called gemcitabine. You see there's a bit of response because the tumours grow slower. They have a bit of a what we call stable disease, but then they progress and eventually become very large. So this is kind of the control scenario that we have in the clinic. We do have a bit of response, and then eventually tumours progress and the patients succumb from the disease. That's why I show you have initial response, but then eventually all the patients die.


What we were hoping for, is if you kill the cancer stem cells, you would actually kind of pull the plug from the tumour, and then the tumour would kind of implode. This is not what happened. If we treat the tumours with ganciclovir, then only the red cells will be killed, but the tumour’s actually continuing growing quite nicely, unfortunately for the mouse, of course. This just illustrates what's going on. So you have a tumour here. We have the red cells, which are the cancer stem cells, and you have outnumbering those cancer stem cells, the blue cells, are the more differentiated cells. If you now pull the plug and kill the red cells, what happens is that the tumour will still grow because you have lots of these blue cells that have high capacity to proliferate, and the tumour will grow. It has lost the capacity to form new tumours, but that's not necessarily what's relevant for the patient, because obviously they will die from this growing tumour.


So killing only the red cells will not stop the disease once the disease has started. If you kill these at the very early stage way before the cancer's diagnosed, this may have an impact, but usually, when the tumour is already growing, this does not have any impact. But interestingly, if we combine the killing of the cancer stem cells together with the killing of the differentiated cells using this chemotherapy, because this chemotherapy's very effective in killing the blue cells and this ganciclovir is very effective in killing the red cells, not much surprise, then everything is gone. And we really achieved cure in these mice, just by taking out, in addition to the bulk tumour cells, a few more cells using ganciclovir and thereby inducing the suicide of the small cancer stem cell population.


This is really interesting, because once the tumour’s growing and getting quite big and these cancer stem cells don't have much of an impact anymore, but you have to kill them to prevent relapse because eventually these tumour cells will take over. So it's important not only to kill the cancer stem cells, but to kill both, to really achieve this long-term cure. And this is a long time, 210 days for a mouse is an extremely long time. We cured the disease from these mice by combining actinotherapy with something that kills the cancer stem cells. So obviously this is not something we can do in a patient, but if we could identify a technology or a drug that actually kills the cancer stem cells in the patient, we may actually achieve a similar strong response and cure. So the key is now to find something which has a similar effect, which obviously is a challenge.


So how do we study cells from a patient? Every patient is different. Every tumour is different. There are various different cancer cells in the tumour, as I illustrated before. That's why we really have to study patient material. So usually where you get patient material is, the tumour from a patient is resected, and then you get a little piece, and then you can expand it in the laboratory. But in the case of pancreatic cancer, the minority of patients ... so most of the patients actually come with advanced disease where we don't get any tissue. So there are ways to get tissue, tiny tissue, using this biopsy technique, but it's tricky and obviously quite dangerous, so you don't want to do it just for research.


So what we have developed here is what we call a liquid biopsy. So we don't punch a needle into the tumour, we inject or use a needle to suck blood from the circulating blood. From this blood, we can then isolate these cancer cells because there are quite a few cancer cells in the circulation of these patients. This can be done in all these patients. We can repeat it, so we have a constant way of monitoring the disease in terms of their cancer cells. And then we can take them to the laboratory and either put them in mice, or put them where we now favour of course, put them in a dish, and create tumours in a dish. So these are quite complex structures that we can form in a dish.


How do we pull out these cells? It's really tricky, because cancer cells obviously are just out of control normal cells that you have in your body, so it's really difficult to differentiate this a cancer cell, and this is a normal cell. So we were looking for these what we call markers to pull out these cells. It was quite interesting, when we got in contact a team working on malaria vaccines in Copenhagen in Denmark in Europe, and they found that infected red blood cells, infected express on the surface something which is called VAR2CSA ... It's complicated abbreviation ... but it actually binds quite dramatically to these oncofetal CSA, which stands for chondroitin sulfate ... Doesn't really matter ... but it binds to the placenta of these pregnant women, and then causes foetal malaria, which is obviously a terrible disease.


But as you can already tell from the name here, oncofetal, onco, it also binds to cancer cells. So there was a coincidence that we found that this VAR2CSA also binds very effectively to these cancer cells. It doesn't bind to any normal cells. So these are kind of sections from normal tissue, brain, liver, intestine. You don't see any staining, so it's all kind of white. This is a specific protein that Mother Nature has developed to bind to the placenta, but that also binds to cancer cells. We have been making use of that to actually bind then the cancer cells in the blood. It looks complicated, but at the end what's important for you is that we can pull out a thousand cells roughly from the cancer patients. So we just take 60 ml of blood, which is not too much ... It's something they can bear ... and we get about a thousand cells, which is enormous amount, if you think about it.


And then you see here the cancer cells in green that we have pulled out in this particular patient. The red ones here are white blood cells. These are normal cells. But the green ones are the ones that we are interested in. This is a high magnification of these cells. This is a cancer cell pull out from a patient. These are actually three cells. You may wonder about the funny shape here. This is actually because they're reusing magnetic beads which are coated with this protein, the VAR2, and they bind to the cells, or to this cluster in this case. And then with a strong magnet, we can pull them out from the blood.


You can't only just pull them out, you can obviously now have all the variety of tests you can do with these cells. Here just shows a very simple one. So we have a target, which we use for immunotherapy, which is called RAY11, and then we can test whether these cells that are cancer cells in the circulation of this patient express RAY11, so that this immunotherapy against this target would be effective in this patient. Other patients are negative for this target, so those would not be good candidates for this kind of therapy. So this is a bit of a precision medicine approach that we are trying to develop.


So this is interesting in a way that we have what I told you was patient that have the resectable pancreatic cancer ... So this is a tumour that is confined in the pancreas and can be resected by surgery ... they have a lot of cells in the circulation already. Whereas the patient that have metastatic disease, so that the cancer has already spread, for example, into the liver, they have a greater variety of cells. But in general, there are not many more. So even in earlier stage of disease, we find a lot of cells. We are now looking into whether we can find those cells even in earlier stage of disease, because the key for cancer treatment is to diagnose patients as early as possible.


And this is just an example here. This is a biomarker that people use for tracking pancreatic cancer, CN99. You can see that it is elevated in red here in very few patients. Most of the patients actually have normal values. But all of these patients have increased circulating tumour cells. Our assay, the cell-based assay is much more sensitive than for example this biomarker. So we hope that we can find and track these cells at a much earlier stage, and therefore develop a biomarker we can use for early detection of cancer, but also for monitoring cancer.


What is interesting is, if you look at these cells that you can pull out from the blood, they don't all look alike. They are very different, if you just look at the size. This is a white blood cell here. But the green ones here are the cancer cells. And here you have a huge cell, whereas this one is much smaller, not much bigger, actually, than the white blood cell. This is what we call heterogeneity of the cancer cells. We are dissecting this heterogeneity by picking these cells as a single cell event. So it's, again, important to really look at individual cells, and compare them side by side.


So what we use for that? This is a robot. Here's a cell that you're interested in. These are cells you're not interested in because they are all not cancer cells. Only this cell here expresses the cancer marker. And then you take a robot and you pick the cell, and then you can put it into a little tube, and then it's gone from your slide where you picked it, but now it's in the tube, and you can analyse it. If you do that hundreds and hundreds of times ... And it takes forever, of course ... then you get kind of profiles for each of them.


So this what we call a surface marker profile. We're not looking at genes. We're just looking at what's on the surface of these cells, because we want to identify them. Surface markers are always good for identification of cells. So then we ask the computer, "Can you cluster them? Can you identify cells that are different from the others?" So while those cells here ... So each of these circles here represents a cell ... so they're mostly here similar, very difficult to distinguish from each other, but these are really different. And what was the defining marker that made these different from the other ones? Well, it's that they express a receptor which is called CC04. It's just a receptor that we have been studying previously. We were curious to see whether those cells may be cancer stem cells, as opposed to those here are more differentiated cancer cells that are also shed from the tumour into the blood, but don't have this specific function.


This was reassuring because these CC04-positive cells express SOX2. If you remember the four factors, SOX2 was one of them. And there are many others that are highly expressed in these positive cells, as opposed to the remaining cells. So these are, again, suggesting that these cells are different, despite the fact that they're all in the circulation, but these cells here express high SOX2 levels. And then again, this decisive experiment is to take these cells that you pull out from patient, a single cell, and put it into a mouse. We did that with the positive and the negative cells. This is tumour formation here. In about 70% of the cases, if we take these positive signal cells, put it into a mouse, we do see a tumour, whereas we never see a tumour if we inject any of the other cells. So these are the cells that are really dangerous for the patients, that can cause metastasis.


Obviously, the efficiency in a patient is not 70%. So just because a cell is in the circulation, doesn't mean it will cause a metastasis, a spread lesion, rather than it's a very complicated process for the cell then to enter into a new space, and then escape the immune system, et cetera. But these have the potential. And these are the ones that we should track because these are the really dangerous cells, whereas the other ones probably have a neglectable level of tumorigenicity.


These are all evil cells. So we show that they're really terrible cells that can cause cancer even at the single cell level, but we want to eliminate them, of course. We are working on inhibiting various signalling pathways. Well, that's obviously quite technical, so I just want to give you a notion of how we can use actually our immune system to target these cells. So what we do here is which we take the T cells, which is a immune cell, from the patient.


So while we pull out the cancer cells, we also pull out the T cells from an individual patient. And then we can manipulate these T cells, these immune cells, and make them chimeric antigen receptor. So we express a receptor on these cells that are specific for the cancer. So while a T cell in a cancer patient is what we call is exhausted and doesn't really have a lot of activity against the cancer cells and that's why the cancer actually grew, we take a virus and that virus carries this receptor, which we call the chimeric antigen receptor, into these T cells, and then they become CAR T cells. And then this receptor combined to this cancer antigen.


The key, of course, is that this antigen, this binding structure here, is only expressed on a cancer cell, because this cell now will kill everything that has this antigen on the surface. That has been very, very difficult. It has been very successful in liquid cancers like leukemia, where there is a structure what's called CD19, which is not specific, but it's only relevant in numbers on these leukemia cells. And that has now been approved for treatment in the clinic. But for solid cancers ... Five minutes ... it has been much more difficult to find those because, again, if this is expressed on a normal cell, this normal cell will also be killed. And if it's a dispensable cell that may be acceptable, but if it's an important cell, this will be very dangerous.


But we have identified an antigen, which I unfortunately cannot disclose, but it is a very exciting antigen that is only expressed on the cancer cells. And now those cells will interfere or will bind to this antigen on these cancer cells, including the stem cell compartment, and then will kind of secrete their toxic factors and can kill these cells.


So this is a natural process that we are using. We're just educating the cells to bind to those cells because without this modification, these cells would not kill the cancer cells, including the cancer stem cells. This is how they look like. This is a cancer cell. These are the T cells, the modified T cells, the CAR T cells. This is in action. So you see this cancer cell here is attacked by this T cell, and then eventually will be killed. There are various types. This is just illustrating here the conventional one, where we just express this receptor and then it binds to the tumour cell. This will be a uncontrolled killer. It will kill any cell that has this antigen, and it's very difficult to stop this. That's why there has been also some fatalities using these conventional CAR T cells.


But we have been now working and developing a system which we call switchable CAR T cell. This CAR T cell is also a modified T cell. It has the structure here. But this structure doesn't identify the tumour. It actually identifies ... This comes up ... a switch. This switch is something which links the CAR T cell with the cancer cell. If the switch is absent, nothing. It's unarmed, and nothing will happen. If you infuse, inject the switch into the patient, they will get armed, and they will take the action. But these have a very short half-life, so within a few hours, these little fragments are gone. So whenever you stop the infusion, the system is off. So this is very different from this conventional one, where you cannot control activity, whereas here you can switch it off just by stopping the infusion. So this is a very powerful system, and it's controllable. It's also very, very effective, as I'll show you in a second.


What I mentioned before was the precision medicine. You want to identify patients that have high expression of our target. So here, you see that essentially all of the cells are positive, 98%. Whereas in the other patient, there were cells that actually negative for the target. So you will choose this cell, this patient for this kind of therapy. This is how it looks like. You have a mouse. This false colour here illustrates the tumour at an earlier stage. And here becomes a high burden disease, and then the mice have to come down.


With the unarmed CAR T cells, so this is when the switch is not infused, you can that the tumour progresses quite massively, actually. But then you have the armed CAR T cells, where you actually infuse the switch on a daily basis. And then, nothing. So this is day 63. There's no detectable disease. And then we obviously wait. There's no more treatment. It's only treated for three weeks. On day 235, the disease is still gone. So these are cured mice. So we actually kept them even longer, and then eventually we looked at their pancreata, the pancreas, and there was not a single cancer cell. So they had all been wiped out by these weapons.


That's what we are very excited about, and we would like to get that into the clinic. It is a bit difficult because the switch in the CAR T cells have to be made in a way that they can actually be used in the clinic. So this is experimental, what we do so far. Just to bring them into what we call a clinical grade takes probably two years and a lot of money. But we have applied for funding for that, and hopefully we will have that ready by 2020 for clinical trial, which again, we will run as a precision immunotherapy trial, where we select patients based on the circulating tumour cells, identify the ones that have high expression of this target, and then we can also monitor the treatment using these circulating cells. And then we'll see if that works in patients as well as it did in the mice.

With that, this is my team from in London. It has now been diminished, of course, because we moved to Randwick. We're building actually two campuses, one's in Liverpool at the Ingham Institute, and one is in Randwick on the main campus of UNSW. We are recruiting heavily. Hopefully we're going to be up and running soon. So thanks for the attention. Looking very much forward to your questions.


42.45 Q&A with audience

Rodney Phillips: Thank you very much, Chris. It was a tour de force of this fascinating topic. The picture here gives us no indication as to why you left sunny London and came to this place, but we can only guess.

Chris Heeschen: Yes. There's currently a heatwave at UK.

Rodney Phillips: We have some time for questions. We will take some. I have one, if I might start with, though. For a long time when I was young, there was a doctrine that it was going to be very difficult to mount an immune response, natural or acquired, against cancer, because cancers were composed of self-proteins. There was no foreignness about a cancer cell. Obviously, that paradigm has changed a little bit. Could you explain to us how likely it is that a given cancer would in some way or other have its own exclusive oncoantigen, as it were?

Chris Heeschen: Yes. That's obviously a very good question. Why not? I mean, people have been struggling finding these specific antigens. We were just probably lucky to find it. It's actually something which is also found in normal cells, it's just not on the surface. And that is obviously ... I mean, there's a caveat that the cells will just not expose it on the surface anymore, it will internalize it, it will recall it, and then the treatment will not work. The switch technology has the advantage that we can switch from different antigens. Obviously for now, we hope that this one is going to make an impact. We will be able to monitor if there are cells popping up that don't have that on the surface anymore. I mean, that's always a risk. I mean, cancer cells are extremely smart. We haven't seen this kind of internalization in our animal experiments. It doesn't mean that it's not going to happen in a human being, of course. But it gives a good hope that it's very specific. It doesn't seem to be internalized and like the Golden Shield mode. So I think there's good reason to believe in that working.

Speaker 1: Has the role of viruses been considered in changing stem cells to cancer stem cells?

Chris Heeschen: Yes. I mean, I work in pancreatic cancer, and there's no evidence for virus being involved in this particular cancer. It was interesting that the microbiome seems to have quite an impact ... That's the bacterial load in the gut which also crawls up into the pancreas ... kind of has an impact. So the environment, not necessarily viruses, but our general environment that we are exposed to seems to have an impact on developing a cancer.

Speaker 2: I remember back in the '80s when we were doing blood cancers. We spoke about T cells being the magic bullet in the '80s. I'm glad that it's come to this stage. It's pretty close to cure. But what my question is this, you're looking at a pancreatic cancer tumour, which is just as deadly as the brain tumour. There are two questions I want to ask you.

Speaker 2: One question, would brain tumour cells would have the same kind of mechanism of attack, because it recurs very quickly after you incise it because what you're saying here is the remitting stem cells, cancer cells, proliferate again, right? That's why it causes a very quick recurrence of the disease. The second question is, what you think low-dose chemotherapy and high-dose intravenous vitamin C would have any effect on this kind of tumours?

Chris Heeschen: That's actually an important question as well. We actually have kind of a clinical trial network where we investigate the administration of very high-dose vitamin C and chemotherapy, particularly platinum therapies. So this causes oxidative stress and it could lead to kill off cancers. We have tested that in our laboratory. I mean, the clinical trial was planned to go ahead anyway, but we tested that in our laboratory. We found that the cancer stem cells are resistant to this kind of ROS or oxidative species inducing therapy. We will see. Our prediction would be that this would not be effective. But I said this trial was planned anyway, so it moved ahead. They have a very strong antioxidative capacity, so they can fight these oxidative species. And that makes them more resistant to this kind of therapy. We will see how that works.

We're working on glioblastoma as well. It's not our main topic, but we have been studying of those cells. We actually find that autofluorescence feature in those cells as well. They show the same response and the same resistance, and then relapse obviously starting from the cancer stem cells as well. So it seems to apply to that cancer as well. We have colleagues in U.S. actually that also work on CAR T cell therapy for glioblastoma. They're actually using HER2, which is mostly used in breast cancer, but there are HER2-positive glioblastomas as well.

Speaker 3: Thank you. Is it likely that that technique could be relatively easily transferred to other solid tumours?

Chris Heeschen: Yes. The target that we have identified is actually what we call a universal target, so we see that on many solid tumours that we have tested. I mean, we've found it on all of them so far. So that would give hope that you could then apply it. We will start with pancreatic cancer. It's a very challenging disease. There's a lot of medical need because there's no effective therapy. So it's a good start, I think. I mean, it's not going to be an easy task because you have the tumour stroma that I mentioned at the beginning. So you have your cancer cells, and when we model this in the mouse because we're using the human cells, the tumour microenvironment is not to the same standard as what you have in human. So we have some concerns that the cells may have some difficulties reaching those tumours.

So that's why we will combine it with what we call a stroma targeting therapy. So we'll at the same time take apart the stroma of these tumours, which is a double-edged sword because you can actually make it worse, but at the same time you hit the cancer cells then with these very effective CAR T cells, I think we have a very good chance that it will be for the good of the patient. We are testing this obviously now in the mouse. So this is all in human cells, but we have not developed a mouse CAR T cell and study this target in a fully immunocompetent mouse where we have all the features of human disease. We don't have the variety of human disease in the mouse because the mouse is made by genetic manipulation, so it's quite homogenous. But as a model system, this works very well to study what are the resistance mechanism, what are the barriers and hurdles to get these CAR T cells into the tumour. This will be a combination therapy.

Speaker 4: All right. Could you make any comment on what predisposes a human to pancreatic cancer? Is there any trigger mechanisms that you could comment about?

Chris Heeschen: Yes. Pancreatic cancer is one of the cancers that doesn't have a clear correlation with risk factors, but it has some association with whatever causes what we call a chronic inflammatory state in the pancreas. So this is alcohol, excessive alcohol use, diabetes as well, smoking. But in general, cancer is caused by a chronic inflammatory state. So whenever there's remodelling, things kind of being repaired all the time, that's when things are happening. So when you're smoking, obviously you're causing constant stress and injury to your lungs. And there's something equivalent, but not to the same extent, of course, for the pancreas. So chronic pancreatitis, which is the chronic inflammation of the pancreas, is a very strong risk factor for pancreatic cancer. So whenever you have all these inflammatory cells invading, not only doing good, but doing all this repair mechanism that can cause cells to transform and become cancerous. So there's always the good and bad of inflammation. Inflammation obviously is needed to remove these transformed cells, but at the same time it also causes or increases the likelihood that there will be these transforming events.

Rodney Phillips: Chris, one the cell types you showed us had a receptor on the surface, which I think is a chemokine receptor. Is there any evidence that the natural chemokines are actually driving proliferation of those particular cells?

Chris Heeschen: Yes.

Rodney Phillips: Is that a potential blocking function? If you block the chemokine, could you block some of the drive for proliferation?

Chris Heeschen: So you're referring to the CXCR4, which is the name of this receptor. The chemokine that is binding to it is stroma derived factor, or CXCL12. And that is produced in very high amounts in the liver, also the lung. The cells are traveling from the pancreas through the portal vein to the liver, and that's when they find, "Oh, here's a lot of CBF1." And so they like that environment, and that's why they obviously reside there and can expand. You can actually block that process. It's a double-edged sword because obviously CXCR4 is used for other cells as well.

Rodney Phillips: Does it cause a degree of immunosuppression if you do actually block that chemokine effect?

Chris Heeschen: Yes. Yeah.


Rodney Phillips: Is it a dangerous level of immunosuppression?

Chris Heeschen: It wouldn't be my first choice of therapy. For us, it serves as biomarker. You can inhibit the spread of the cells. But mostly, I mean, when the patients literally come into the clinic and they have this cancer spread already anyway, so I don't think that's a therapeutic angle because you have to inhibit their expansion. But the CXCR4 inhibition doesn't inhibit cell proliferation.

Rodney Phillips: Could I just press you a bit harder on the ... The first issue that crops up, obviously, is this, are there always going to be useful antigens? The second issue that seems to me fundamental in plausibility or otherwise, is that with a leukaemia, you can imagine the infused T cells getting into the right environment where the leukaemia is because of, as we call them, they're liquid tumours. But if you look at solid tumours in reality in the surgical theatre, plausibility of those T cells getting to every cell in a very large tumour mass seems to me quite low.

Chris Heeschen: Well, it worked surprisingly well in the mouse. Yes. I mean, that's a well taken point. We'll see how that works in humans eventually. But as I already mentioned, we have to combine it with something which makes the tumour more accessible. I mean, pancreatic cancers are extremely stiff tumours. In some of them, 90% of the cells are actually stroma cells. And then there are immune suppressor cells, so cells that counteract immune activity. So all these elements need to be addressed to some extent. I mean, these are really killer machines, so we think they're much more potent than the endogenous T cells that you have. But there will be a level of resistance, and that's why we think this combination hopefully will address that issue. But it's a well taken point. I mean, in liquid tumours, they're floating around. I mean, these cells are exposed to whatever you infuse, whereas in solid cancer, this is not the same.

Rodney Phillips: Right. I think that I'm getting a signal from the back that it's time, but I'm sure Chris would be willing to answer your questions, should you wish to stay behind.

Chris Heeschen: Thanks a lot.

Rodney Phillips: On behalf of the University of New South Wales and the Faculty of Medicine, I'd like to thank Chris Heeschen very much for his outstandingly clear and massively excellent tour de force in this topic area. I would invite you in the future to come to our lecture series.