Scientist Stories: Irv Weissman, Developing New Cancer Treatments

So as a prelude to my first slide, I'll say that in 1988, Mike McCune and I started a company called Systemics, the objective of which was to isolate human blood-forming stem cells and then to use them in therapies. We isolated them by 1991, and Tsukamoto was the lead scientist at the time, now my wife. And we showed that we could get them cancer-free, and in a clinical trial with Judy Jezuru in the back of the room, that we got much better outcomes with cancer-free stem cells to treat breast cancer patients who had metastatic breast cancer. We had programs to transplant them from donor to host, because Judy and I had shown that you can induce transplant tolerance with stem cells, that you can reverse the autoimmunity of juvenile diabetes and lupus. All of this was going on, and we even had a pretty significant program with anti-HIV based not on the Berlin patient because it was 1996. But by 2000, the company that bought us shut it all down with a business decision. And I want you all to remember that because that's essentially why in 2004, when Bob Klein and I were talking actually before 2004, we tried to find a way that we could take discoveries at academic institutions all the way through the first phases of clinical trials so there would be no risk, and therefore that companies would consider funding. So Judy, stand up. Judy is about to do a clinical trial at Stanford to treat children with severe combined immune deficiency for the first time with purified blood-forming stem cells. Good, thank you. So you can see here in this difficult diagram that we had discovered the way to isolate those 1 in 20,000 cells in blood-forming organs that are blood-forming stem cells called HSC. So that nobody else understands but us what it is. Blood-forming stem cells self-renew, but they also differentiate to make blood. So in 1998, we began the search for what cell in this series gives rise to human leukemia. And we found, first in mice and then in humans, that it wasn't at the level of the blood-forming stem cell but the first multipotent progenitor for most leukemias. And in 2004, you heard from Katrina, the blast crisis, the end stage of CML, she and I showed together, was at the level of another progenitor. In order to explain this, we finally first proposed, then demonstrated, that if this is the leukemia stem cell that self-renews, but is at the level of a cell that normally doesn't self-renew, we could go back and look at each blood-forming stem cell in the patient that's being treated and showed formally that each step that, that blood-forming stem cell has to have this gene or that gene go on leads from a normal stem cell to the leukemic stem cell. We then compared the leukemic stem cells' expression of genes with the normal progenitor in humans. And the first one that came out, which is hidden back here, is the gene CD47. We found that CD47 was a "don't eat me" signal. For the scavenger cells of the body called macrophages that normally contact and eat and then kill not only cancer cells but old red blood cells, old granulocytes, and so on. They clean up the mess, and after a heart attack, they clean up the, the damage that's left in the heart. But we found it on every human acute leukemia, a "don't eat me" signal. So we eventually figured out that might be important. And we realized that in that evolution from the normal blood-forming stem cell to the leukemia stem cell, many of the events would turn on programmed cell death and accompanying programmed cell removal, that is, an "eat me" signal. Every cancer, every single cancer, human cancer, overcomes programmed cell death by many pathways, but as I'm about to show you, overcome programmed cell removal by elaborating the "don't eat me" signal. So is it really a "don't eat me" signal for leukemias? We transferred the gene for CD47 into a leukemia and then incubated it with macrophages, or we just did a comparison. So this gray stuff here are the eating cells on a dish. If they don't have the "don't eat me" signal, they're all eaten and killed. If they have the "don't eat me" signal, none of them are eaten and killed, and they overgrow the macrophages. So we then made or obtained— first one we obtained, then we made— an antibody that blocked the "don't eat me" signal, one that you could put in the circulation. And here you're seeing human primary acute myelogenous leukemia growing up in the bone marrow of an immune-deficient mouse. And if anybody tells you you can get away with doing these things in a dish, they're full of it. You have to understand how these things work in a body, and it's better the body of a mouse than the body of a human. At least that's my feeling. So there's 4 months after Ravi Majetti and Mark Chow put human leukemias into the bodies of these mice, and just like in humans, they homed to the bone marrow, they grew in the bone marrow, then they spread spread from one bone marrow site to another. We then gave our blocking antibody for 2 weeks, and it cleared 90% of the human leukemia in the body of the mouse. The one example we couldn't clear— the macrophages are eating like mad, but they couldn't keep up with the tumor. So then we said, Well, actually, the antibody is blocking the "don't eat me" signal, and the leukemia, which has an "eat me" signal, is eaten because of that. And by the way, normal cells that temporarily express the "don't eat me" signal don't have the "eat me" signal, so even with our antibody, they are not removed. Normal blood-forming stem cells. Normal other stem cells, other tissue cells. But we knew that there was this antibody being used for successful treatment of non-Hodgkin's lymphoma called rituximab. And when we looked at how rituximab works, it actually has— when it binds to the lymphoma, on the back end of the antibody is an "eat me" signal. So we said, let's see if blocking a "don't eat me" signal will synergize with a plus "eat me" signal. So here's what happened in mice bearing a human primary follicular lymphoma. This happens to be a MYC and BCL-2, for those who are interested. If we give rituximab alone, you get a remission followed by a relapse. If we give our antibody for this very aggressive one, you get a remission followed by a relapse. Now, relapse is what occurs in most patients with non-Hodgkin's lymphoma with rituximab alone. But when we combine the two, we cure 90% of the mice having the human follicular lymphoma. This was just the first experiment. It repeats over and over and over again. So then we wondered, it's on leukemias, it's on lymphomas, is it on any other cancers? So interacting with Jonathan Barak and the group in obstetrics and gynecology, we obtained human ovarian cancers. And funded largely by the Ludwig Foundation and Joe and Laurie Lakup, we began research on ovarian cancer. So here we transferred from the peritoneal cavity of a woman in the clinic with spread ovarian cancer, the ovarian cancer cells into the peritoneal cavity of an immune-deficient mouse. And here's how it grew in the mouse if we gave control antibody treatments. And here, which we waited for the cancer to take and grow in the peritoneal cavity, and then began treating with the antibody, And it not only prevented its growth, it cured all of the animals in this particular experiment. So what else? So talking with Mike Clark, my colleague who I recruited to do cancer stem cell research at Stanford, we had— we got human primary breast cancers, and they grow in the mammary gland of immune-deficient mice. So we put again the cancer cells into the mammary gland, and then we waited 2 weeks, a month, and then treated them either with the antibody that blocks the "don't eat me" signal or the control antibody. And again, most but not all of the breast cancers were cured. If they were very large, we could only hold them back; we couldn't cure them. So we need to be able to use things like cancer-free stem cells to get the women with metastatic breast cancer to the point that we can then treat with the antibody. That's obvious and logical. Here's one that was a total shock. Primary brain cancer that kills almost everybody fast—glioblastoma multiforme. Every glioblastoma that Sam Cheshire and Sid Mitra obtained from the surgery group and we put into the brains of immune-deficient mice 3 to 4 months later, it is growing like mad. And again, if you treat it then with antibody to CD47, even though you put it here, it clears the brain tumor in almost all cases. Same is true for other brain cancers. Over the last 6 years, by lots of interactions with our surgical and medical oncologists, we've gotten all of these cancers— in fact, more than I'm showing you now— and every one of them responds to the anti-CD47. All of them we prevent metastases, and the others we have static, and we have never found a resistant one. So obviously, this Use of a mouse antibody would be good if we could get it to humans. That's where CIRM comes in. So we applied to CIRM for a $20 million disease team so that we could take it at Stanford forward as far as possible in order to lead to eventually treatments. And we had the— and have the full intention to do the Phase 1 and 2 trials before Stanford we advised Stanford to license it to an entity that could commercialize it. We want to know, will there be any aberrant toxicities? Will it work in human cancer? So CIRM funded us, and we are moving toward our first clinical trial. In the first $20 million phase, we've proven that not only it works, but the antibody that we made that we humanized, made it so it won't be rejected in patients. Works. We've gone through the toxicity studies. Mild toxicity, all of which we can overcome in non-human primates. Because CIRM made an arrangement with the Medical Research Council in the UK, we have grants that are side by side And just to give you the names of the people there, Perez-Villas and Alan Burnett, just to give you an idea of the power of their clinical trials unit, because they have a single-payer— and I'm not a communist now, you guys— but because they have a single-payer, they organized themselves to be clinical trialists 20 years ago. In the first 3 months of our association, with that group in Oxford. We got 574 independent acute myelogenous leukemia samples, which we've tested. All of them are CD47 positive. The subset that we've tested by putting in animals, we can again treat with our antibody. Now, of those 574 patients, many will come to relapse and will be in our phase 1 trial. And we have the unique circumstance that we have the leukemia frozen away, we can have it growing in animals, and we treat the patients. So we are moving to a phase 1 or phase 2 trial simultaneously in Oxford and at Stanford. Oxford, we are now funded by Cancer Research UK for solid tumors Medical Research Council for leukemias and lymphomas. And here we're about, we hope, to be funded by CIRM, but we haven't got the answer yet. So we could do it all in England, you know. And then Brits would be the first people treated. This isn't a threat, but I'm just— it's an observation. So we are moving to clinical trials, all comers, all cancers. Now I don't want to fool you into believing that we know this is going to cure all cancers or any cancers. We're at the risk phase now where you put the antibodies in people who have these cancers and you look for toxicity. That's the part of a phase 1 study. But it's agreed amongst all the investigators, both sides of the Atlantic, that all patients who come to the study will have measurable tumors, so that even in the phase 1 study, as we do increased dose and then repeat dose at the level they can take, we'll know whether it's likely to work in breast cancer only or lymphoma only, or— good and bad news— all cancers. Good because there's a hope, Bad because somehow all those people who tell us you gotta focus will be unhappy if we don't focus. But I want to remind all of you that these are diseases that happen to everybody, and if you focus and decide one versus the other, there's that short window of time that people had to be treated that they wouldn't be treated because you decided for whatever reason, financial or otherwise, to focus. So CIRM has allowed us to carry forward something that I know by testing the NIH will not fund. So it can only be funded in this way. You can only carry it forward this way, and you can only then say to the public and Stanford University, okay, it's time to license it to somebody who's now going to take it through the Phase 2 trials required to get it rapidly to trials. And I'll say to my friends in the audience who are venture capitalists that I think we took something that had a pre-money valuation of $10 million when we discovered it 4 years ago, just 4 years ago, to something whose pre-money valuation, as my friend Mike Longacre likes to say, in the hundreds of millions. And that means Stanford and CIRM will benefit much more from the recovery of the funds if commercialism was your only intent. But of course, for most of us who don't have to deal with those business decisions, what we really want to do is get the therapy out to everybody. Thank you.