Era of Hope 2011 Plenary Sessions

Session Title: Closing Session – Bringing New Ideas into Reality

Title of Presentation: Developing a Prophylactic Breast Cancer Vaccine

Dennis Slamon, MD, PhD, University of California, Los Angeles: The next speaker will be addressing some of the issues that I think are relative to the objectives of the Artemis Program in terms of a vaccine and the role that the immune system plays in breast cancer for both therapy as well as a potential for prevention. So I want to introduce Stephen Johnston who is from Arizona State who will talk about that topic.

Stephen Albert Johnston, PhD, Arizona State University: Thanks Dennis. So I got a note from Fran that said keep to 12 minutes so this will be conceptual and it’s—and it’ll be to try and present a new sort of view on what kind of vaccines we might be making than our original objective.

Now I’m stressing that the goal of our project has—has always been to make a prophylactic, a preventive cancer vaccine. And the—the concept was—is that—is that we—we advance the treatment of breast cancer to the point that hopefully you didn't even know you had it. It was just like an infectious disease. You can receive a vaccine and you don’t get the cancer, and coming from the infectious disease area that this seemed like a conceivable type of project.

But what I’m going to tell you about is that we’ve made some progress in the area but we’ve also decided that the—we may need to open some more fronts on how we develop this vaccine. I mean as Fran knows I’m a student of the Apollo Project and whenever you wanted to get something done for everything that needed to get done in the Apollo Project they didn't open just one option to get—accomplish that—that aspect of the project. They opened up three or four different ways to get that done because they had to make sure one of them worked and you couldn’t predict which.

Now when we started this project we thought that it had two components. One was the—that we would be engaged in; one was make progress on the technical challenges of what goes into the vaccine; how do you make the vaccine? The other one that we were also just as sure of that we had to do was convince other peoples to engage in this project because it was going to take too many different areas of expertise and perspectives to just count on our own capability and—and fortunately we’ve seen that happen since the Artemis Project has started and we’re very pleased to see that.

The third one that—that actually came up because of conversations around this with people in the DoD Program was that we realized that we had to develop a diagnostic for early detection to accompany the clinical trial for this vaccine. This happened because I explained to Leslie Bernstein how I thought we would conduct a clinical trial for a prophylactic cancer vaccine and after I helped her back off the floor and into her seat she explained to me why we needed to think about this differently. And—and from that conversation I realized we have to have a very simple effective mechanism to determine early breast cancer and we’ve opened up a whole new front on technology development for that. And I don’t have time to tell you about it but it’s very promising.

Now as to the components of the vaccine we thought that there—in our mind at the time and playing off of the infectious disease area is that you had to have the antigens to go in there which we thought was going to be the biggest problem; the adjuvant, you had to have a delivery system. We assumed that targets would be the tumor itself but as we’ve seen in this meeting that may not be the only target that you could go after for a vaccine.

We knew we would have to be concerned about autoimmune disease; I actually don’t think that’s a big problem but—but people that aren't immunologists do and then we did realize the clinical—the trial—a trial for a prophylactic cancer vaccine would be very challenging and probably would demand as much innovation as developing the vaccine itself.

So just a couple of definitions; the—what we’re talking about again is a preventive vaccine, a prophylactic vaccine versus a therapeutic vaccine. And almost all the work in the cancer field has been on a therapeutic vaccine. Almost all the work in infectious disease has been on a prophylactic vaccine. And—and in my mind the definition of the two is that a therapeutic is a vaccination with the antigens after the tumor presents the antigens. And a prophylactic is vaccination with those antigens before the tumor presents those antigens. So that’s our simple functional definition.

Now as I said, almost all the work in the area of cancer has been on therapeutic vaccines. And they inherently it’s—I mean in the area they point to Provenge and as—as a great stride in this area; I—you know I’m not so excited about it but it—the therapeutic vaccines have inherent problems with them and they’re becoming—some of them may be solved but—but there are some that may not be.

First of all you have a—as Peter Lee would point out you have a simple numbers problem. You have a lot of tumor cells and you have only so many T-cells that you can mobilize against them. By the time that you’re vaccinating, the tumor is already there. The tumor has already had the chance to evolve stealthy mechanisms which they do readily. And then you have as—as has become more and more apparent the tumor cells themselves have very clever ways to immune-suppress your own immune system through T-regs and other cells that you’ve heard about.

Another concern that we’ve had because from the beginning we said that if we’re going to develop a preventive vaccine for cancer it has to be for the whole world; it can't just be for people that can afford to get something in the developed world. And so the—if—if you follow that logic through it has to—this—at least for almost all of the things that have come through the therapeutic vaccine possibilities now, the cost is going to be high. And so that was—that was another reason that we didn't want to explore the therapeutic.

On the—in contrast, the prophylactic vaccine by definition if it’s—if you’ve already pre-armed the immune system, you have a low number of cells—tumor cells if it’s caught early, they don’t have time to evolve, so all these stealthy mechanisms go by the wayside. You don’t have immune suppression and the important thing is—is that it could be simple and inexpensive. Vaccination is the most cost effective medical intervention that has ever been by hundreds to a 1,000-fold over everything else we’ve done. And so if we could mimic what they’ve done in infectious disease and in cancer it—it means it could be a—have an impact for the whole world and not just the developed world.

So there’s been two approaches or two ideas of what kind of antigens would go into this vaccine. One is self-antigens and you’ve heard talks about MUC and HER and things like this. The advantage with those—that they’re easy to find; the disadvantage is that because they’re self-antigens what you’re left with in your immune system are low affinity T-cells; you sort of start with this—with the immune deck stacked against you.

The other way is to use neo-antigens. These are things that your immune system has not seen before that the tumor could present. And the—these are things like translocations, new mutations, insertions, deletions that make neo-proteins, proteins that you’re—or peptides that your immune system hasn’t seen. The disadvantage of these is that they’re hard to find. The advantage is that when they’re there they can—you can get high affinity T-cells for them.

Our view, our simplistic view when we started this project was that the way we would make the breast cancer vaccine is that we would look in a lot of tumors, find out neo-antigens which we were going to focus on that occurred in each tumor and then find ones that occurred across all tumors to the point that we could add up enough of them that they would have 100% coverage—that is that there would be a—10 or 15 or 100 components in this vaccine and any tumor that showed its head would present one or more of those that you’ve already armed the immune system by—by that vaccination so it would go after that tumor. That was the simplistic concept.

It—it was simplistic and we thought that the DNA sequencing effort that was going on would just make this easier and easier for us because people would be sequencing all these tumors and we just looked through what turned out to be mostly their garbage. Most of the people doing all the DNA sequencing of tumors throw away the information that we need to make the vaccine and we’ve had to look through the garbage cans to find it which isn't always easy.

But we could find them but the—it turned out as you see—heard Joe Gray and others mention that from the overlap of these DNA modifications from tumor to tumor is relatively small. So only a few percent of tumors present any specific translocation or insertion/deletion and only a few percent of those actually would make an antigen that would be useful in the vaccine.

So that was disappointing. What was—what was encouraging though is that we discovered that tumors do—start to get very trashy once they start developing. They start taking RNA pieces and fusing them together and they do them in repeatable ways, reproducible ways. So they start creating what we call frame shifts and other things that are—when they take a piece of gene one and they put it up to—against the piece of gene two those two pieces aren't supposed to be together. And when they get translated they make a piece of junk peptide and that peptide is new to the immune system.

And that could be used as a source of vaccine. So we collected lots of these from screening mammalian human breast tumors and thought we had the armament to go make the vaccine—that we had the components. So we started testing them in—first in therapeutic models where we inject a tumor and see if these things will protect. Using the—the antigens in mice that we had found also in human, about 20% of the antigens we find in human tumors are repeated exactly in mouse tumors, so we could test them there. And they worked beautifully in therapeutic tumors—vaccination, works great in mice.

And then we went to—we set up a bunch of mouse models for spontaneous tumors and it didn't work very well. And in fact, when we vaccinated with our control which was our—just our adjuvant, we actually saw protection from the adjuvant. But we didn't see any increased protection from adding these antigens. We never had seen that before in the therapeutic models.

Now I—this made us step back and rethink the system. Why did that happen? And we think we know why now. Because in order to get a prophylactic vaccine to work it—this is the way we envisioned it from here to here. You start the immunization; you get a strong immune response. That strong immune response basically the number of high affinity T-cells has to get to a certain point. If that’s high enough at the time the tumor is making—and this is the number of tumor cells going on, and at that point that those tumor cells get to this point and they present your neo-antigen you should have those—those T-cells ready to outnumber those tumor cells and kill them because they’re presenting these neo-antigens. It seemed like that should work.

Well there’s another factor in the system that we probably didn't consider. And that is that normal cells have a quality control system so that they get rid of these aberrant peptides very quickly. They’re very efficient at doing that, so they have a QC system. And tumor cells we know out here, they don’t have a good QC system. It’s gone basically; they’re presenting all sorts of junks that normal cells don’t do. The problem is this decline in that QC system has to be to a point that it’s presenting these antigens for these cells to kill those tumors and we—what we probably don’t have is that we don’t have the immune system expanded enough or the antigens on this side of this line here enough that we can get the prophylactic vaccine to work in a spontaneous tumor model.

So what—we think this is correctable and we’re going on with that, but there was another interesting observation and—and it made us step back and that was we had always thought that—that these same antigens that we had because we knew they worked in therapeutic models that there was another life for them and that’s this kind of scheme that we—we’ve always presented to people is that eventually you might think of having a new phase of a personal therapeutic vaccine where somebody has a tumor. It’s detected very early. You can go in; you’re going to have to get some of that material, biopsy it, and then sequence it. And this is where this revolution in sequencing could potentially find all those neo-antigens that were personal to you, the ones that can create the high affinity T-cells.

Then you go make a vaccine out of those and you go back and vaccinate yourself; well you vaccinate yourself with your own antigens and—and in that way you—you still—you have all those liabilities of it being a therapeutic vaccine but now you’re vaccinating with your own—and focusing your own high affinity T-cells that are remaining on your—on your system. And if you could accompany that with some sort of lower regulation of the—of the—of the TH1, these regulatory cells that suppress your immune system then you might be able to make this work. So that—that’s possible and the technology is getting that way.

However, there was a funny thing that happened on the way to the forum on those—all these experiments that took us a year and a half to do where we didn't see any effect of our—our antigens. These mice generally developed two to three tumors in this mammary tumor model. Although we didn't see any reduction in fatality of the mice, the mice that received the antigens didn't—didn't develop a second tumor. So they delve out the primary; the primary grew, eventually killed the mice, but where they got one of our antigens they didn't develop a secondary tumor.

And this suggests another kind of vaccine that we—we’d like to pursue and we’re starting to pursue this now and that is a vaccine against metastasis. So the idea is this; we take those same neo-antigens that we have. We vaccinate; you get a low level response and but when you use—you still may get a primary tumor but you have this low level response, and it turns out that primary tumor actually boosts that immune response. We’ve seen this in mice. This is what happens. So what happens is you may have to remove that primary tumor but you won't get a metastasis if—if the same thing happens in humans as we’ve observed in mice.

Now that has a—a potential advantage; as you know most fatalities don’t happen because of the primary tumor. It’s because of metastasis. The tumor would—your vaccine would essentially be latent; it wouldn’t be doing anything. It would just be there at a low level and then it would be primed by the tumor existence, the occurrence of the tumor itself and that low level of—of the vaccine actually relieves some of the safety concerns people might have about having these vaccines in the first place.

So I think the bottom line is—is that yes, we’re—I think that the prophylactic vaccine is doable but it may have other forms that it could take and we’ll be exploring them and we hope other people explore all of them. And in final comment I say that you know I—Fran mentioned the 2020 Project; I’m an avid supporter of that project. I think that—that is right on line in what we should be doing. And in the academic community there’s frankly a lot of resistance or hesitation about being enthusiastic about doing that and I think they’re wrong. And for money reasons—and Fran knows some of them but one of the things that—the reason I support it is because I see pathways to get there. And I think the prophylactic vaccine approach may be one of the best pathways to get to that goal.

By the way; the Era of Hope and historians generally announcing era is about 20 to 30 years, so this era should be over of hope pretty soon, about the time of 2020.