Cells： from Robert Hooke to Cell Therapy - Professor Philippe Menasché

So good afternoon and welcome to the final session, the fourth of four absolutely amazing sessions, I'm sure you will agree, and the quality of the speakers has been unbelievable and we really owe it to the Royal Society for sponsoring this meeting and it's been a real privilege to work with my two co-chairs, Johan and Ian. So we now move on to what's called novel therapies and by novel therapies I'm not sure really what we mean but I think what we mean deep down is that these are therapies that don't just have an incremental change on patients' lives but actually are a step change and that's really what we're trying to achieve with cell and gene therapies. And our first speaker for the afternoon is a very good friend of mine, Professor Philippe Menachet who is from the Department of Cardiovascular Surgery at the Georges Pompidou Hospital in France. Could we please welcome Philippe to the podium. Thank you very much, it's a pleasure to have you here. Thank you for the privilege of attending this outstanding meeting. Cardiovascular diseases encompass a wide spectrum of pathologies and we'll concentrate on two of them which actually are the most common. Starting with acute myocardial infarction, I think that most of you are familiar with the common procedure which consists of harvesting bone marrow from the patient, isolating mononuclear cells and re-injecting the cells into the coronary artery which has been reopened by angioplasty and stent a couple of hours or days earlier. I think we have to be honest and objectively assess the outcomes which have been achieved so far and to do that probably the most relevant study is this recent meta-analysis which is interesting because based on individual and not pooled patient data. And we must admit that looking at the results it shows that this therapy, the way it's currently implemented, has failed to show any significant benefit. It's sad but it's life. It's even more true if we look at this other study which is also interesting because you can note that the benefits of stem cell therapy in acute MI settings, which have been occasionally reported by some studies, are no longer apparent if the results are assessed on MRI, not echo, and MRI is not operator dependent. It's more reliable and the benefits also have gone if you just focus on blinded trials. Is it surprising? Not so much. It's not so much surprising for two main reasons. The first is that some people have really rushed into clinical trials with very, very subtle and frail preclinical grounds. This is the seminal paper in which a given population of bone marrow stem cells was supposed to regenerate the infarcted marrow's heart. And I'd like you to look at the date, April 5, 2001. By the way, this paper has never been reproduced by others, the results. April 5. And this is the first clinical case, same year, August 24. So, just by sake of comparison, it took us 10 years of preclinical works before we did the first skeletal muscle transplantation in 2000 and again 10 years of preclinical work before moving to the first human embryonic stem cell trial that we are currently performing. So, that was probably one of the reasons why so many of these trials have failed, because of the lack of robust preclinical mechanistic data. The second reason is that over the past years there has been, fortunately, a striking decrease in the mortality rate of patients with acute MI, as you can see here on this registry. And so, it becomes terribly difficult now to demonstrate any additional benefit on top of these existing therapies. So, to finish this first section on acute MI, I think we can try to summarize by the conventional SWOT matrix. Stem cells, in the context of acute MI, have some strength, listed here, but they also have, as we have seen, a lot of limitations. Consequently, there are currently substantial threats over this therapy, and nevertheless, there might be some opportunities if we can address a couple of key questions we're going to talk about more in depth in the second part of the talk. And these opportunities consist of having a readily available off-the-shelf product which can be administered to probably well-selected patients, and more importantly, if we can improve on the delivery and retention of the cells. We should keep in mind that currently, with the trials that have been done so far, when you inject the cells down the coronary artery, at most 5% of them are going to remain in the myocardium. So, it's not that unexpected that the results have not met the spectacular expectations which were initially raised. Now, let's move now to chronic heart failure, which has been our major topic of interest. I'm a cardiac surgeon, so these are the kind of patients I'm dealing with. And so, chronic heart failure, as you know, is now a disease which affects an escalating number of patients worldwide, in particular because the mortality rate of acute infarct is being decreased, and so you have a greater number of patients at risk of developing late heart failure. And obviously, this has a major impact on costs, costs being primarily related to repeat hospitalizations of these patients. Now, once again, let's look at the most recent meta-analysis of stem cell trials in this context of chronic heart failure, and I think this is an interesting paper which has to be looked at very carefully, because if you just concentrate on the abstract, the signal is relatively positive, as you can see. Now, if you look at the paper more closely, it turns out that there are so many biases in the interpretation of these various trials that it becomes a little bit more difficult to interpret the outcomes in a meaningful fashion. So, nevertheless, I think that all these studies are extremely instructive if we can draw appropriately some lessons, and if we can incorporate these lessons and make them sort of building blocks for the next generation of clinical trials, and that's what I would like to discuss with you now, concentrating on five major messages which have come from this accumulated both experimental and clinical experience with stem cells. The first very important thing, I think, is what could be called a mechanistic shift, because, as you will see, this has major practical implications which could really change the picture of stem cells in the context. Initially, I think we have been very simplistic and very naive, in particular us, the surgeons, but that's by nature, because we thought that transplanting cells within an infarct myocardium, scarred areas, would allow the cells to physically, structurally replace dead cardiomyocytes, that they would couple electromechanically with the host cardiomyocytes, and consequently that they would improve function. If we objectively look at the literature published so far, it seems that, at least for the time being, it's wishful thinking. Conversely, those cells definitely have positive effects, and there is now increasing and compelling evidence that they primarily act by releasing factors which are going to harness endogenous reaper pathways. This is a nice illustration of the concept. In this particular study, they took adipose stromal cells, they reprogrammed the cells to make them iPS, and then they re-differentiated these iPS into endothelial cells. This is now the most important and the most interesting part of the study. This is rather, I would say, classical, and what you can see here is that injecting these endothelial cells derived from iPS, they had an improvement in function, which, again, is fairly conventional, and note, please, that at the same time point, all the cells have gone. The interesting part of the study is that, at the completion of the experiments, they took out the hearts from both controlled and infarcted animals, and looked at the expression of various genes, and were able to show that just by injecting the cells, they were able to upregulate a wide array of genes involved in the prevention of apoptosis and in the stimulation of angiogenesis. This is one piece of evidence that probably one primary mechanism of action, and we'll see the practical implications it has, one of the primary mechanisms of action is really that the factors released by the cells activate some endogenous reparative pathways, and I prefer reparative regeneration. The second message relates to the nature of the cells. I don't have anything to sell, so although we are using the embryonic stem cells so far, if tomorrow there is another cell type which turns out to be more effective for the patients, I will switch without any problem. But if we admit that the cells are acting primarily through paracrine mechanisms, and then the question comes, are they all equal? Can we use any cell type? Because many of these cells actually secrete the same kind of factors. It doesn't seem that it's the case. If we look, the head-to-head comparative studies which have been published so far, assessing cardiac committed cells versus non-cardiac committed cells, and regardless of the endpoint which is looked at, you see that the best outcomes are consistently associated with the cardiac committed cells, which to some extent makes sense. You want to repair the heart, rather give heart cells rather than non-heart cells. So then, what are the potential sources of these cardiac committed cells which could be used clinically? Well, the first source is the secret positive cells, which can be harvested from the right atrium at the time of coronary surgery and then re-injected into the coronary arteries a couple of weeks or months later. The second source is the so-called cardiospheres, which are pieces of right ventricular tissue that you can harvest with an intermyocardial biopsy, exactly what we do for monitoring rejection in our transplant patients. And these are agglomerates of heterogeneous cell populations comprising mesenchymal stem cells, fibroblasts, endothelial cells, maybe cardiac stem cells. But this is cardiac cells. And finally, you have this other approach consisting of taking bone marrow cells, but engineering the cells by exposing them to a cocktail of cardiopoietic factors, leading to the overexpression of some cardiac transcription factors in these mesenchymal stem cells. So they're no longer naive MACs, but say cardiac-like MACs. And I show you these three cell sources because each of them has been the subject of phase one or phase two clinical trials, which say encouraging results, which now need to be validated by larger phase three trials, some of which are already ongoing. Our approach has been slightly different. We also wanted to use cardiac-committed cells, but we started from the beginning. That is, from the pluripotent stem cells, which, because of this pluripotentiality, can be driven in any direction towards any lineage, including the cardiac lineage. And we were able then, as others, to generate these beating clusters of cardiomyocytes, which form the basis of the cell population that we have now started to use clinically in an ongoing clinical trial. Incidentally, the cells that we are currently transplanting are not beating major cardiomyocytes. They are early cardiac progenitors derived from human embryonic stem cells. A third important point, which has to be revisited in light of the paracrine hypothesis, is retention versus survival. There are different approaches for delivering stem cells to the heart. Unfortunately, the least invasive are the least effective. In other words, if you give the cells intravenously, none of them is going to reach the heart. So, the most effective approach is the direct approach, that is, giving the cells directly in the myocardium, either from the inside by an endoventricular catheter or from the outside, which we do in surgery, that is by transepicardial injections. But even with this direct approach, as you can see here, by, say, one week, you only have 15% of the cells which remain on board. And on top of that, 90%, not to say 100% of them, are going to die within the next weeks. You see here an interesting comparison of different cell types, the bone marrow cell, the myoblasts, MSCs, fibroblasts, and you see that regardless of the cell type, if you look at the bioluminescence signal, all the signals go down, down, down, and you do not have any cell left after, say, one month. We know that there are different causes for the cell death, which has generated a lot of studies trying to empower stem cells. But although these approaches have been occasionally successful in the lab, none of them is really easy to translate into the clinics. But a basic question now is, should we really strive for keeping the cells alive for a very long period of time, or should we not rather concentrate on optimizing early retention to fully exploit the paracrine effects? And this line of reasoning is supported by this interesting study in which the authors compare two groups, MSCs conventionally injected in the myocardium, or MSCs encapsulated in a gel. The issue here is not the gel. The issue is that if there is this biomaterial, not unexpectedly, you see that by three days you have a much greater retention of the cells, which is quite expected. The interesting thing also is that by one week all the cells have gone with or without the gel, but by three days there were more cells in the encapsulated group. Now, what happens one month later, the best functional outcomes are seen in the group in which retention was the highest after three days. In other words, with the limitations of the study, it has been enough for the cells to be present in large amounts initially after the procedure to apparently trigger reaper mechanisms which have then remained self-sustained over time, accounting for the benefit at a time where the trigger cells had disappeared. So that's the reason why we strongly believe that efforts should be put on the optimization of early retention rather than maybe trying to keep those cells alive forever, which once again I think is wishful thinking, with a caveat of maybe the Parkinson's disease, where we have shown very impressive data showing the long-term survival of some of these neurons, but maybe the brain behaves differently as the heart, which beats. So how can we try to improve retention? Well, in a very simple way, by designing more elaborated delivery devices such as this one, or probably still more importantly, by combining cells with biomaterials. Biomaterials which can be injected in a liquid form in which polymerize in situ, improving the retention of cells, or biomaterials that can be used under the form of a patch, which is the approach that we have adopted now, the patch which will be delivered onto the epicardium of the disease area in the heart. This is just an illustration of that comparison of injection versus the patch with or without a biomaterial, and you see that regardless of the material itself, alginate or chitosan, and regardless of whether you use the injection or the patch, you get a much better retention of the cells compared with the controls. So, in addition to improving retention, biomimetic or bio-inspired materials have a lot of advantages, and I think that, and the question was raised this morning, I think that one of the major lessons we have learned over the past years is that if we want to optimize the effects of cells, we probably have to combine them with some form of scaffolding, which will help keeping them alive, in particular because the biomaterial may allow the cells to secrete their own matrix and to remain anchored to a matrix, which is critical for their survival. You inject the cells in a suspended fashion, they have lost attachment one to the other, they have lost attachment to a matrix, and this sets the stage for them to die. Just to show how these advantages of biomaterials translate into functional improvements, this is a head-to-head comparison we did a couple of years ago, comparing conventional injections with the epicardial delivery of a patch, and you see that the functional results here speak for themselves. So there are different materials so far that can be used, and I'm not going to detail all of them because it's not one size fits all. I mean, the biomaterial has to be tailored to the type of cells that you are using, and different characteristics have to be taken into consideration. In our own experience, fibrin has turned out to be a nice material to incorporate the embryonic stem cell derived cardiac progenitors, and this is the last preclinical study we had to do before having permission to move to the clinical trial, where in a blinded fashion we compared control animals with animals receiving a cell-free fibrin patch versus the cell-loaded patch, and you see that even the fibrin by itself has some effect, obviously loading the patch with the cells dramatically improves the efficacy of the procedure. Now another important message alluded to by Professor Ritescu this morning is the requirement for an off-the-shelf product. Otologous cells have been the first one to be used. They're not that easy to use in the real world, particularly because of the tremendous variability in cell function between one patient and another, not to speak about the complexity of logistics and cost, and most of these hurdles can be overcome by the use of an allogenic product coming from a well-qualified bank and which can be readily available whenever you need it. So the industry has understood that, and this is an interesting representation of the advantages versus disadvantages of otologous allergenic cells, clearly showing that if you are treating a very limited number of patients you can afford otologous cells, but if you're targeting wide populations of patients with heart failure, then definitely you probably need an allergenic source. Now the downside of the allergenic cell obviously is rejection, but once again in light of the paracrine hypothesis, maybe this has to be revisited. Why? Well, maybe it's just enough for the cells to reside for a couple of days or weeks, in which case the issue is no longer to avoid rejection by a lifelong immunosuppression, but just to delay rejection by a couple of weeks. This is an interesting illustration of that. A comparison of syngenic, equivalent to otologous, versus allergenic and syngenic cardiospheres. One week after transplantation you see equivalent engraftment rates, and not unexpectedly three weeks later all the allergenic cells have gone because of rejection, that is three weeks after the transplantation. Now what happens six months later, exactly similar functional outcomes. So once again it has been enough for the cells to remain there for a couple of weeks. I don't know how long they should be there, but in this particular case three weeks and the other study three days, but what it means is that a transient period during which the cells are present seems to be sufficient for them to exert their paracrite effect and trigger some form of cardioprotection. And in this case then it becomes more acceptable from a clinical standpoint to give this very sick heart failure patients a transient immunosuppression. And in this setting an increasing interest is paid to some form of induced tolerance. I must say that in our current protocol we're just giving conventional immunosuppression with very low doses just for one month, but induction of tolerance is an interesting approach because it might allow the cells to be tolerated for a couple of weeks, maybe months, and that might be enough. And just to illustrate that, this is an unreadable slide just depicting the number of allergenic stem cell trials currently recorded on the official website in the context of heart failure. And finally the question comes if really the cells act in a primarily paracrine fashion by releasing factors, why don't we give the cells? Why do we give the cells? Why don't we give the factors? And this question obviously is now very timely appropriate and the idea here is to try to duplicate the effect of the cells on the endogenous reaper pathway. So one approach is to use acellular biomaterials, biomaterials which by virtue of their physical or chemical properties could induce this form of activation of endogenous reaper pathways. Many experimental studies tend to support the concept. We must admit that the first clinical study has not been that positive. This is to my knowledge the first clinical study in the context of heart failure in which a polymer has been used without cells and indeed you see that in terms of function there was no benefit. Maybe it has to do with the dosing or with the polymer, but this is the results we have so far. So the other approach is to use biological compounds supposed to recapitulate the paracrine effects of the cells. One option is to use growth factors and this is a paper which has raised a lot of publicity recently in which the authors have incorporated a growth factor into a collagen patch and were able to show a reduction in fibrosis, the appearance of new cardiomyocytes in the infarcted heart and this translated into an improvement in function. So this is one growth factor among many others that could be used. This one is supposed to use through an activation of the epicardial cells which could re-enter a developmental program and become again new cardiomyocytes. Micro RNAs are another option and this is one example with one single micro RNA turned out to be cardioprotective since knockout animals did not show any kind of protection which could be re-established by a of this micro RNA. Our current approach is rather to focus on extracellular vesicles and reason for that is number one that it seems maybe a little bit over simplistic to think that the effects of the cells which are very complex can be duplicated by a single factor targeting a single pathway. A second reason is that extracellular vesicles shuttle a very wide array of biomolecules in particular micro RNAs and you know extracellular vesicles include both the exosomes and the extracellular microparticles. It's a different size primarily but all these extracellular vesicles normally secreted by cells in particular on the conditions of stress shuttle a wide variety of micro RNAs known to play on different key signaling pathways. This is a recent interesting study showing that if you take conventional fibroblasts and expose them to extracellular vesicles coming from cardiac cells you can really change the phenotype of the fibroblasts and make them anti-fibrotic and pro-angiogenic. This is another interesting example where you see the cascade of events following an infarct with different changes in micro RNA expression. Now if you transplant skeletal muscle cells in this context you completely reverse the up or down regulation of these micro RNAs and end up with a limitation of apoptosis. And the third argument for us to go in this direction is the really encouraging data we have collected from our first head-to-head comparison of cells versus the vesicles. This is the summary of these results. It's a mouse model of chronic heart failure. We gave the cardiac progenitors derived from human embryonic stem cells exactly the ones I'm currently transplanting in the clinics and we compared that where the extracellular vesicles derived from the same batch of cells and surprisingly to some extent the functional outcomes were completely equivalent. We then looked at gene expression and found that both cells and vesicles expressed the same set of genes quite different from those in the control group and once the genes were clustered into biological pathway it turned out that most of them were acting on events leading to some form of cardioprotection. So it's clear that we're still far from clinical applications but if it turned out that this hypothesis was validated that might completely change the picture by streamlining the process and allowing to get closer to a drug-like paradigm with the potential applicability of the technique to a greater number of patients. So to conclude I think first we have to select the right patient. Heart failure is a continuum and not all patients obviously are candidates. Probably the good candidates are those with severe heart failure which have exhausted conventional therapies but not end-stage patients in whom the only option remains heart transplantation. The second thing is that we should be careful about the fact that their colleagues in other areas are also making substantial progresses and you just have here a couple of examples. So I don't think that in the context of heart failure stem cells are going to be a panacea, a miraculous treatment but at most and that would be already a good thing it could be one option in addition to the others. The third is a surgical message. My experience as a surgeon is that the only techniques which work are the simple ones. If it becomes too complicated and restricted to a club of happy fews you can just forget it. Maybe you could get a nice paper in science or any kind of these journals but it will be difficult to translate it into a clinically useful therapy. And just to illustrate that this is the protocol that we have followed for generating our cardiac progenitors. We set up, we scaled up the cells, undifferentiated obviously, set up two banks, the master and the working. Then we induced cardiac specification, isolated a marker, a surface marker which allowed to identify cells which had lost pluripotency so that we could sort the cells, end up with a purified population of cardiac progenitors and incorporate these cells into the fibrin patch. If the protocol has been allowed to be translatable into the clinics it's exclusively because we use a drug BMP2 for inducing cardiac differentiation. We use immunomagnetic sorting which has been in use in hematology for years for purification and we use a surgical glue for making the patch and that were the prerequisites for making it something clinically acceptable and something that could be implemented into the patients. Then once again do not expect miracles. Some people are skeptical because the improvement in ejection fractions in most studies is say between five and seven percent. Five, seven percent is not negligible. It might allow a patient who is lying in his bed to walk around the block which is not too bad. But I just would like to draw your attention on these drugs. These drugs are life-saving drugs, no question about that. If you look back at the very early studies, the early functional sub-studies of these drugs in which their functional effects were assessed, you will note that the improvement in ejection fraction was not that dramatic. So the fact that we do not have you know spectacular increases in ejection fraction doesn't mean that these cells might not be therapeutically effective. And finally I think it's critically important to honestly report what works and which doesn't work. And to illustrate that this is a very distressing slide I'm going to show you. This is a study coming from the UK by the way in which the authors have scrutinized most of the stem cell trials published so far and have identified what they call methodological inconsistencies. For example 15 patients and among the 15 patients you have 12 male and nine women. These sort of things. So they've really looked at all the inconsistencies across all the studies and then they correlated the number of inconsistencies with the outcome. The distressing result is that as you are going into an increasing number of inconsistencies the outcome is more and more positive. But unfortunately if you look at the five trials in which there was no inconsistency there was no benefit. Once again it's sad it's life and it's better to acknowledge that than to try to oversell a product whatever it is. So in conclusion I think that really myocardial infarction is not clearly an unmet medical need for stem cells in contrast to heart failure. I think that really what we have learned is probably we should try to target the use of cardiac committed cells combined with some form of tissue engineering whatever it is and that we really have to carefully look at the possibility of using cell mimetics to try to recapitulate the effects of the cells once again in the perspective of a greater simplicity in the implementation of the procedure and consequently with the possibility of having a greater number of patients drawing benefit from it. Once again I'd like to thank you for the invitation and your patience.