I am going to distil four and a half years of preclinical and clinical work into 30 minutes, so by definition my talk is going to be somewhat short of data, but I hope you take away some reasonably clear messages.
(Click on image for a larger version)
Approximately 80 per cent of individuals who are infected with hepatitis C virus (HCV) go on to become persistently infected. That has resulted in around 200 million infected individuals in the world, and around 200,000 in Australia.
Unfortunately, there are still around 12,000 to 16,000 new cases in Australia each year. Since we can only treat around 9000 or 10,000 per year, and as only 50 per cent of these people are cured, then we still have a net increase in the number of HCV carriers, both in Australia and in the world, every year. And that is pretty sobering.
Hepatitis C is a long-term disease – in fact, people are generally infected for life – and it is currently the leading single indicator for liver transplantation in the West.
There are very limited treatment options. The only licensed treatment for hepatitis C infection is combination therapy with interferon-alpha and ribavirin, either a six- or a 12-month course, depending on the genotype, but about 50 per cent of these people do recover.
(Click on image for a larger version)
You have seen slides similar to this in many talks, but I want to emphasise the potential of neutralising antibody to neutralise virus, as a protective mechanism, and clearance of infected cells by cell-mediated immune responses.
There is still a lot of discussion about whether neutralising antibody for hepatitis C is useful, but I guess all of us would like to see some effect of neutralising antibody in a potential vaccine. Equally, we would all be aware that clearance of infection will, really, generally be by cell-mediated immune responses.
So we recognised a number of years ago that the routes of these two different pathways, neutralising antibody and cell-mediated immune responses, had a common amplification pathway that required antigen-presenting cells – and in this particular case we will substitute dendritic cells (DCs) for these antigen-presenting cells – that were vital for amplification of both the humoral and the cell-mediated immune responses.
I want to say equally that cell-mediated immunity might be a double-edged sword in terms of hepatitis virus infection, in that although we probably recognise that the virus per se is non-cytolytic, we think it is equally likely that the hepatitis that results from virus infection is the result of the cell-mediated response trying to eliminate the virus-infected hepatocyte. So, intrinsic in any immunotherapy procedure or protocol there is a potential to generate hepatitis. That potential has influenced enormously the design of this trial and, in fact, the time that the trial has taken. I will come back to that point throughout the talk.
(Click on image for a larger version)
A very important point that I want to stress, following up on my point about inducing hepatitis, is that sustained viral responders to interferon treatment don’t generally show any spikes in their ALT (alanine aminotransferase). That is in direct contrast to treatment of hepatitis B carriers with interferon, and what that is telling us is that HCV is almost certainly eliminated in a non-cytolytic manner.
So we have the potential to induce a hepatitis, but in reality interferon treatment does not appear to induce a cytolytic response. Nevertheless, the ethics committees ruled that this was possible, and we had to act on their ruling , that there is that potential for hepatitis.
Equally, the viral load is decreased by several logs. In acute phase patients that is coincident with a T cell response, and that often happens without any rise in ALT. So these two points are telling us that it is very likely that the virus is eliminated in a non-cytolytic manner.
The rationale extended also to the fact that we figured we could load the DCs ex vivo much more efficiently than we could in vivo, and equally that immature DCs which express or process HCV proteins might induce tolerance. So we have to ensure ex vivo that we mature the DCs with strong maturation agents.
The objective was to concentrate interferon-alpha and -gamma producing cells in the liver that might eliminate the virus in a non-cytolytic manner without systemic side effects. The mathematical modellers will tell you that if you can actually get enough interferon into a patient, the patient will be cured of hepatitis C. The problem is that interferon is a very toxic molecule, and many of these patients have very serious side effects so we really just can’t get enough interferon into them. But we figured that if we could get the interferon produced locally in the liver, then it might eliminate the virus without those systemic side effects.
So the aims were to alter the equilibrium between the virus and the immune response in patients to result either in a reduction in the viral load or in viral clearance. The hypothesis that we worked on was that DCs from patients that are loaded and matured ex vivo with HCV lipopeptides – and I will come back to the lipopeptides in a minute – followed by autologous transfusion will prime naïve T cells or stimulate existing HCV-specific immune responses.
Perhaps we should at least mention the word ‘naïve’, because I think if I was to do this trial now I probably wouldn’t be as naïve as I was six years ago when I started this!
(Click on image for a larger version)
The trial process is shown here. The patients go for leukopheresis, performed at the Peter McCallum Institute, in Melbourne. The cells are then subjected to CliniMACS. That is a CD14 monocyte purification procedure, using this fancy machine from MiltenyiBiotec to generate CD14 monocytes. These are cultured ex vivo and remain immature DCs which are then pulsed with either antigen or, in this case, lipopeptides to produce mature antigen-loaded DCs, which are then reinfused back to the patient. I will come back to many of these points a bit later.
(Click on image for a larger version)
The trial was actually approved by the TGA (Therapeutic Goods Administration). There is a bit of history as to why it went to the TGA, but we felt, given the potential for the hepatitis, that it was useful to have the TGA approve it. So it is, after all, a phase I dose-escalation trial approved by the TGA as a CTX (clinical trial exemption). For those of you who are not really familiar with this: the TGA approve only about two of these per annum – not that there is any reduction or number limitations; it is simply difficult to get CTX-approval.
The trial uses lipopeptide-pulsed, human monocyte DCs which are cultured in serum-free medium. The serum-free medium is absolutely vital. The TGA are not particularly happy about culturing cells for clinical trials in fetal calf serum, and equally we know that serum from HCV patients often contains very high levels of IL-10. We figured that the cytokines in HCV-positive serum are likely to influence the outcome of the DC culture more than anything else, so we have spent some time, as I will show you in a minute, culturing these cells in serum-free medium.
The inclusion criteria were, first, that patients had to have failed interferon-based therapy. That was imposed on us by the TGA. Given that 50 per cent of patients will recover if we treat them with interferon, then the TGA insisted that these patients have a first line of treatment that might cure them. So we had no choice; we had to use patients who had failed therapy.
These are generally genotype 1 infected people and this of course is more than six months, but many of these patients have been infected for many, many years.
Ages were 18 to 60; the patients were male or female.
The lipopeptides we use are HLA A2 restricted, so we had to perform HLA typing on each patient.
They should have had a biopsy, and most of these people did have a biopsy because at least one of the criteria for previous treatment by interferon was to have a liver biopsy.
And they had to use effective birth control, although for the life of me I can’t decide why.
The endpoints were product safety – it was a phase I trial. We did want to see if there was an increase in HCV-specific cell-mediated responses, and if there was any reduction in the viral load.
(Click on image for a larger version)
We talked a moment ago about the outcomes. The routine monitoring was done in the hospital, and as you can see here it really covered most things. Those checks were done on a daily basis for four days after the infusions, and then on a weekly basis for six weeks. So you can see that the monitoring was pretty full on.
(Click on image for a larger version)
I want to spend a minute on the culture in the serum-free medium. The top part of this slide shows that, relative to culture of these cells in RPMI and 10 per cent fetal calf serum, we actually had a higher proportion of cells recovered in the serum-free medium than we did from the RPMI and 10 per cent fetal calf serum. That was very encouraging.
Equally, comparing the phenotype in the RPMI-cultured cells with those cultured in the serum-free medium, you can see very clearly that they are actually very similar. In fact, the serum-free medium cultured cells actually show some better evidence of being immature DCs than the cells cultured in the RPMI and 10 percent fetal calf serum.
Here are some flow charts showing phenotypes of immature DCs versus matured DCs on a daily basis, and you can see that they are actually very similar.
(Click on image for a larger version)
Here are the HCV-specific CTL peptides. These were made under GMP (Good Manufacturing Practice) conditions by David Jackson, in his lab at Melbourne University. In case any of you think that GMP is straightforward, I can tell you (and Dave will tell you equally) that it is not. It is very strict and we had to comply.
We used six CTL epitopes that are highly conserved and HLA-A2 restricted. You can see that three of them are from the core protein, which of all the HCV proteins is the most highly conserved, followed by one in NS3 and two in NS4B. I would urge you to pay attention to those for the moment, because we will come back to them a bit later in the talk.
(Click on image for a larger version)
Some of you may think this dose schedule is a bit strange. We had planned to do 12 patients. We ran out of money and time, and in the end we did only six patients. One of the reasons why we ran out of money and time is that the ethics people imposed a four-week interval between patients 1, 2, 3 and 4. So, essentially, we sat around for half of the year when we did this, making sure the patients didn’t actually develop a hepatitis.
In part that arose because we went to the TGA and the ethics people about the same time as that horrendous problem in London. So this is a direct response to that, but if the TGA tell us to jump, we ask how high.
We had a single dose, what they called a low dose, an escalating dose, double, medium and high. This schedule was imposed on us by the TGA. I am not sure it is particularly scientific, but that is the way it worked.
Unfortunately – and this again is one of the problems of taking something into the clinic – when we came to patient 6, we were unable to get that level of cell recovery, and so, essentially, we duplicated patients 5 and 6 with the dose for patient 2. I will come back to that point in a minute.
(Click on image for a larger version)
This slide gives you an idea of how difficult it is to work in the ‘clean room’.
You see here the ex vivo schematic. We pulsed the immature DCs with the lipopeptides to result in DC maturation for transfusion back to the patients. We hoped that the DCs would then migrate to the lymph node and generate HCV-specific CTLs that in turn migrate to the liver to target the infected cell.
These people are in the clean-room GMP facility for about eight hours at a time. The room is under positive pressure. There are bacteriological culture plates open on the bench throughout that time, and if any of these criteria are not met then the product fails.
Some of these people, as you would well know, have probably been intravenous drug users. One of the people we put on the leukopheresis machine simply had their veins collapse and we were unable to generate a product. Not only does that delay us in time but it costs an enormous amount of money to do that.
(Click on image for a larger version)
Before we got to any patients we did three dry runs. I would just like to focus on the three dry runs for the moment. Shown first here is the CD14 apheresis product – this is the proportion of CD14 monocytes in the product – from the dry run normal individuals. Let me rephrase that. Since that first individual was me, we will call it a ‘healthy’ rather than a ‘normal’ individual. But as you can see I had a reasonably high proportion of CD14 monocytes. The second dry run was with David Jackson, who had a similar high number. The third was with one of the people from the microbiology department at Melbourne University.
You can see, however, that when we get into the patients the variability is much higher. It is something we really could not control. And again I would have thought really carefully about doing this, had I known that we would have this huge degree of variability.
The same was true of patients 4–6. (I have only put up the first three patients.)
(Click on image for a larger version)
The criteria within the red lines here are something that we had to meet for the TGA. CD14 post-CliniMACS had to be more than 85 per cent. You can see that we had always greater than 90 per cent CD14 monocytes to start the DC culture. The viability post-culture (pre-infusion) had to be higher than 75 per cent, and it was always very high. And the CD14 pre-infusion had to be lower than 10 per cent, and again always was.
The two criteria at the bottom were not criteria that we had to meet but we certainly would like them to be higher than 75 per cent for HLA-DR. We seldom achieved that. And for CD86 again it was highly variable. But these were not breakers, whereas the criteria on lines 2–5 certainly were.
(Click on image for a larger version)
So we had six patients. We have finished the trial now. The data are still falling out, but all of the patients had some side effects: a mild headache, some mild flu-like symptoms, shivers and nausea that were really quite common in all of them, a mild erythema at the ID site (the cells were injected both IV and ID) and some other side effects that were probably unlikely to be caused by the infusion. But there were no severe adverse events. All of these adverse events or side effects went to a data safety monitoring board, and it was only after those people went through them that we could then go on to the next patient. So it was very strict and well controlled.
(Click on image for a larger version)
To come to the bottom line: I am going to show you a couple of slides that show the ELISPOT results against a series of HCV peptides. CTLs 1 through 6 are the lipopeptides or the peptides that were in the cellular vaccine, and the other peptide pools represented the complete polyprotein of hepatitis C.
In general, all of the patients had a baseline in the ELISPOT of about 20 spots per million. Many of you probably don’t know that the frequency of HCV-specific T cells in HCV carriers is extremely low, and it is not unusual to have a low background like this.
Two weeks post-infusion there was nothing particularly encouraging to report. But although patient 1 received CTLs 1, 2 and 3 from the core in the vaccine, and although those don’t appear to react directly, at six weeks post-infusion the core peptide pool which clearly contains some of these sequences was clearly higher, much higher than in this patient, as it was with E2 and NS4B.
Unfortunately, as you can see from this slide, at 12 weeks post-infusion those responses were not sustained. Again that is pretty disappointing, but I think it is clear that we did have some responses, even though they were not sustained.
But equally, although we didn’t put an NS5B peptide in the cellular vaccine, there was a clear response to NS5B, and that has actually been consistent in a number of patients.
(Click on image for a larger version)
I will show you patient 2. The baseline again is about 20, maybe a little higher. At two weeks post-infusion 1 we see a response to NS5A in this time, remembering that NS5A was not in the cellular vaccine either. There is a good response to one of the peptides that was in the vaccine, and that is reflected equally in the peptide pool. So there is a lot of consistency in this.
You can see that it is also a temporary effect. By the time we get to 12 weeks post-infusion we are virtually back to background.
(Click on image for a larger version)
I don’t want to go through all six patients with this. What I want to do now is to reflect on the highest ELISPOT responses. Bear in mind that the background is usually around 20. Patient 1 had responses to a couple of peptides that were around 40 or 50, and to another peptide that was 150. Patient 3 had similar. And patient 2 again. So we are seeing some useful responses in each of the patients.
(Click on image for a larger version)
And again remembering that the background is about 20, here are the responses in patient 4, patient 5 and patient 6.
(Click on image for a larger version)
Now, some of you will say that this guy is a virologist and he has absolutely no idea how to do an ELISPOT, and a couple of years ago I would probably have agreed with you. But I have to tell you that these ELISPOTs were performed using a procedure that was licensed by Rose Ffrench through NATA (National Association of Testing Authorities). And when we did flu responses in each of the patients, the flu responses were usually pretty standard throughout. So I don’t think there is anything wrong with the protocol; it is simply that the HCV responses do vary considerably.
I will just summarise that. The infusions showed only some minor side effects but no serious adverse events. Fortunately, there was no increase in the ALT levels. There were de novo responses to core, E2 and NS5, and clearly that showed evidence of epitope spreading. The responses were clearly transient. And, although I haven’t shown the data, there was absolutely no effect on the viral load. Again that was really quite disappointing.
But of the aims of the study – that is, to show safety and to show some immune responses – I think we have done two out of three and I guess we learned a lot. We could probably design a study that would improve on this if any of us had the strength and the will to take it on. We are still trying to decide if we should do that.
(Click on image for a larger version)
This was funded from the NIH (National Institutes of Health) with a previous study funded by NHMRC (National Health and Medical Research Council). Shown here are the chief investigators, who did some of the work, but most of the people in the second column did all of the work – the people from my own lab, Kathryn Jones, Phil Latour, Rose Ffrench who was incredibly helpful, David Woollard and Li Shuo; and from David Jackson’s lab at the University of Melbourne, Emily Ericksson, Irene Dinatale, Georgia Deliyannis and Brendan Chua.
The Peter McCallum people were incredibly useful, in fact seminal, to the study because they did most of the GMP stuff at the Peter McCallum: Maureen Loudovaris, Jude Moloney and Kerrie Stokes. We have the people at the Alfred Hospital, where Jo Mitchell was the nurse study coordinator. And MiltenyiBiotec were incredibly helpful through the help that Paula Stoddart gave us.
Discussion
Question: Eric, can you give a general commentary on the utility of using peripheral blood? When you have got an infection like hepatitis C, all the action is going on in the liver. A lot of cells presumably are recruited into the liver, doing their thing there. You are measuring in peripheral blood with ELISPOT a reflection of what is going on where the action should be. You don’t want them circulating in the blood, presumably; you want them in the liver. Can you give us a feeling in this field – I know it is the art of the possible, you measure what you can, but the more we look at this, is there any indication that we are actually missing an analysis of the immune response where it should be going on?
Eric Gowans: I think you’re absolutely right. There have been a number of studies that look at the T cell responses in liver biopsy samples, and there is certainly a higher frequency in the liver, relative to the peripheral blood. And you are absolutely right equally that to try and get liver biopsy samples from this group would be impossible. So we do what we can, and that was the best we could do, I think.
Question: I wondered about T-regs. In HPV I know some of the flares are associated with a decrease in the number of circulating T-regs. I don’t know about HCV, but I know that in some chronic infections they do play a role. Did you think about taking them out of your T cell assays because then that would allow the T-effectors to have an effect?
Eric Gowans: I think that would add a level of complexity in a clinical trial that, at least for us at the moment, would be unlikely to get through. We are actually very much aware of T-regs. We have someone in the lab who is a T-reg expert, and we had some very exciting data which is still being analysed. But I think to build that into a clinical trial at the moment is probably a little premature. You need to bear in mind that this project was written six years ago, and it takes a long time for a clinical trial to come through – as you know.
Question: Did you measure whether the DCs make interferon-alpha in response to that PAM-CYS ligated peptide, which was one of your early objectives?
Eric Gowans: We haven’t done that yet with the patients. We are still doing a lot of the data. Many of the cells have been frozen and will be done in one batch towards the end of the study. But what we have done has shown that these lipopeptide-pulsed DCs can stimulate interferon-gamma from PBMCs (peripheral blood mononuclear cells) from autologous patients.
Question: What is the current status of these patients? Have you made it worse, or is there any improvement in these patients? Can you update that?
Eric Gowans: As far as I am aware, there has been no clinical change. There is certainly no change in the viral load, much to our disappointment. I guess we will check them again at three-month intervals, but at this stage we doubt very much if there will be any decrease in viral load.
Question: Have you ever checked the SCV (small cell variant) itself, to see whether the SCV-derived antigen was captured by the Kupffer cells?
Eric Gowans: Oh yes, we have shown that the DCs take up the peptide…
Question (cont.): The DC or the Kuppfer cell? We first presented the acute liver injury mouse model, using the antigen-specific Th1 transfer system. In that system the antigens were captured with Kuppfer cells, and just transferring the antigen-specific Th1 cells induced acute liver injuries. So I just wondered: if you stimulate host Th1 immunity by vaccination of DCs pulsed with HCV antigen, if the host is a very Th1 type host, with so much increase of Th1 immunity the liver will either reject the virus-infected liver cell itself but also react with the Cooper cells, and the Cooper cells produce large amounts of TNF and finally induce acute liver injury. So I am just afraid of that.
Eric Gowans: We certainly didn’t see any increase in ALT. What we did note, though, was an increase in responses to epitopes that were not present in the cellular vaccine that does suggest that there was some cross-presentation of viral antigens that were present in infected hepatocytes.
Question (cont.): So this is not existing in the native HCV antigen, you mean?
Eric Gowans: Yes.
Question (cont.): Ah, that’s okay.
Eric Gowans: The responses to NS5A and NS5B could only have resulted from cross-presentation of viral antigens present in hepatocytes. At least, that’s my interpretation of it.Visit these websites
The Sir Mark Oliphant Conference Series is supported by the Australian Government under the International Science Linkages Program.
© 2008 Sir Mark Oliphant Conferences.
Website design acidgreen
