AACR 2016

Normal and neoplastic stem cells

Prof Irving L Weissman - Stanford University School of Medicine, Stanford, USA

We found early on, when we could compare blood forming stem cells and the leukaemia stem cells from patients that arise for their gene expression, that the leukaemia stem cells expressed a cell surface molecule that says, ‘Don’t eat me,’ to the macrophages of the innate immune system. So once we knew that that ‘Don’t eat me’ signal was there, and it wasn’t there on most normal cells, we wondered if we could block the signal with antibodies, which we did. So we took authentic human leukaemias and eventually lymphomas and all human cancers, every one expresses the ‘Don’t eat me’ signal. If you transplant them directly from the patient to the same organ in an immune deficient mouse, so you’re dealing with the authentic human tumour, then the antibody blocks the ‘Don’t eat me’ signal and the cells get eaten and the tumour usually goes away.

We wondered why this system came up and found that every cancer, every leukaemia, every lymphoma in its development from a normal stem cell of the tissue expressed an ‘Eat me’ signal before the ‘Don’t eat me’ signal came up. So you could imagine now you have cells that are expanding, millions of them, which have the ‘Eat me’ signal and they’re getting eaten as soon as they try to get out of their site of origin and go past macrophages.

So then what we did was found that the rare cell that had turned on the ‘Don’t eat me’ signal, which normal blood forming stem cells use when they go from one bone marrow part of the body to another part, they have to go past macrophages, that they put on the ‘Don’t eat me’ signal, then it shuts down. But sooner or later every developing cancer permanently puts on the ‘Don’t eat me’ signal.

We found that the major ‘Eat me’ signal was another surprise, it’s a protein that’s usually inside the cell and it’s chaperoning the development of protein signalling beyond the cell surface. But that protein called calreticulin, sorry about the name, is a protein that is not only a chaperone it has a binding activity for developing glycoproteins, carbohydrate stuck to proteins. When the cancer cell or a macrophage looking for cancer cells gets activated they break apart that protein called calreticulin, it gets secreted and then it binds to the cancer cell surface. One end of it binds very tightly, the other end is recognised by a receptor on the macrophage which eats anything the receptor engages. So now we’ve got the ‘Don’t eat me’ signal with its receptor, the ‘Eat me’ signal with its receptor and found pretty much that it works. So we receive funding from the California Institute of Regenerative Medicine, that’s from Proposition 71, the stem cells issue, we had learned about this field by first isolating blood forming then brain forming stem cells then isolating cancer stem cells from blood, brain and every tissue. So they were pleased that research on stem cells could lead to a cancer therapy. We obtained a grant of $20 million to form a team at Stanford to do what you would do in a biotech start-up, except at a university, and with that $20 million we made the antibody in a form that won’t be rejected or seen as foreign by humans, that’s called humanising the antibody.

We then showed that the antibody was extremely efficient at leading to the eating and killing of all human cancers we transplanted into the immune deficient mice – breast to breast, brain to brain, blood to blood. We then tested in a non-human primate that had exactly the same ‘Don’t eat me’ molecule distributed on normal cells exactly the same way and found how we could give the antibody safely. With that, and a number of other things, we filed an IND, initial new drug, application from the university, not a company, to the FDA for doing a trial with all solid tumours to see which one might be most sensitive. We filed at the same time in the United Kingdom to their FDA equivalent, the MHRA, to study leukaemia. Now they allowed us in the UK to do something we had hoped to do in the US and that was to take patients who are in a very end stage of their acute myelogenous leukaemia and give them a low dose the first week, twice, and then up the dose in the same patient the second week and up the dose the third week so that’s called intrapatient dose escalation. It means very few people would end up being guinea pigs to test the safety of the antibody, everybody has got some chance that if we get to a therapeutic level they’ll enjoy it, but we have the same safety requirements of a dose escalation to make sure something untoward doesn’t occur.

In both the leukaemia trial and the solid tumour trial we’re now pretty far into it and we’ve found that we can deliver the antibody safely and we’re approaching the doses that should be therapeutic. Along the way we said if the ‘Eat me’ signal is this calreticulin that’s bound to the surface, is it possible that known therapeutic anti-cancer antibodies like rituximab for lymphoma, Herceptin for breast cancer, something called cetuximab or panitumumab for pancreatic, colon, lung and brain cancer, if those antibodies were working by providing super ‘Eat me’ signals when they bound to the cancer cell surface and we found whatever else they do they do that. So that means that if you take one antibody that blocks the ‘Don’t eat me’ and you combine it with a super-strong ‘Eat me’ like rituximab we could cure human lymphomas growing in mice so fast that either antibody alone could lead to a remission followed by a relapse but not a cure and now we could cure them. So obviously we’re deeply interested in moving this forward.

In the past several months my university said enough of obligating the university to run clinical trials putting them in a situation they weren’t used to and they said, ‘We’re going to sell this to the highest bidder unless you form a group to carry it forward.’ So we did that in the last few months and we formed a company and now we’re trying to transition it to the company. I’ve formed companies before; in 1988 when we had isolated the blood forming stem cell we formed a company called SyStemix where we isolated human blood forming stem cells pure. Now at that time people were wanting to do therapy for breast cancer that had spread throughout the body. We knew that the therapies that were given, combination chemotherapies, were getting close to killing all the cancer cells in the body but they weren’t really there. So the practice had begun in the ‘90s of taking the patient’s bone marrow or mobilised blood, which contained blood forming stem cells, freezing it down, now treating the patient with as high a dose as you could give, hoping to kill the last cancer cell in the body but rescuing them with their own blood forming stem cells. Great idea except when we looked at the mobilised blood and the breast cancer they were contaminated with the patient’s own cancer cells. To make a long story short we purified the blood forming stem cells free of cancer cells. We treated many patients with that and we were moving on to more and more applications of it but my company was bought out, I won’t say in a hostile bid, but they were bought out by a large company which four years later shut it down. Everything.

Now, this will be the finish of this. Eighteen years later we went back and saw those breast cancer patients who had cancer spread throughout their body, hopeless in their possible outcome. One third were still alive eighteen years later and of those that got the unpurified mobilised blood 7% were alive. If that was a pill or a protein everybody would be taking it. So with the help of CIRM we went back and got all the rights to set up a clinic at Stanford where we’re purifying cancer free blood forming stem cells to repeat the metastatic breast cancer trial, 50,000 women a year in the US get to that stage, and the hard to treat lymphomas and myelomas, all with cancer free stem cells. Now, the two-thirds of the patients who redevelop their breast cancer at the end of the therapy had instead of hundreds of billions of cancer cells in the body, probably had a million or so or a hundred thousand. We know that our anti-CD47 anti-‘Don’t eat me’ signal can take care of that size of a tumour, probably Herceptin can also. So we expect to do trials, and I mentioned this in our talks, first to get rid of the cancer load in women who have no other hope, it’s a hospice or nothing else for them. Then following up to do the real clinical trials to see if our idea is correct that you can wipe up the rest of the breast cancer cells. So that’s pretty much what we do, we take stem cell biology, we understand how the stem cells work, we can trace, at least in the blood forming system, every step of mutation that takes a normal blood forming stem cell and anywhere from five to twenty steps later become a leukaemia stem cell. With that we discovered the ‘Don’t eat me’ signal and now we’re into therapies for all of it.