Trending: It’s true: Amazon has been profitable for nearly two years, even without AWS cloud windfall

One of the most popular sessions at the recent GeekWire Summit was a talk by Dr. Jim Olson, cancer researcher and serial entrepreneur, attending physician at Seattle Children’s Hospital, University of Washington professor, and member of Fred Hutchinson Cancer Research Center. He gave an inside view of the innovative approaches that he and his colleagues are taking in the battle against cancer.

The session was an update to a widely viewed TedX talk that Olson gave two years ago. As he told the audience, “Indeed, a lot has happened in just two years.”

Watch and listen to his GeekWire Summit talk above, and continue reading for an edited transcript, along with Olson’s slides and pictures.

Slide01Dr. Jim Olson: Everything that we do at Seattle Children’s Hospital and Fred Hutchinson Cancer Research Center is based on these kids. When I go into clinic each week I say to myself, “What am I going to be doing today that I never want to be doing again?” and “How can my lab help fix that?”

In 2004 we had a 16-year-old girl who came in with a brain tumor, and you can see on the left (below) there’s an abnormality in the MRI, and in the center you can see what the brain looked like to the surgeon.

Slide02Now if I gave you a scalpel and I asked you to cut out the tumor but leave the normal brain, where would you start and where would you stop? In this case the surgeon Rich Ellenbogen worked on this child for about sixteen hours, and in the end there was a thumb sized piece of cancer left behind. That day that we learned that, we decided that we’re going to find a way to make the cancer light up so that surgeons could see it in the future and distinguish the brain cancer from the normal brain.


Now, without going into great details, we turned to this little guy to solve the problem. This is the Israeli Death Stalker scorpion, and this particular scorpion makes a molecule called chlorotoxin that we learned went to brain cancer, but not to normal brain.

Slide04In fact what you see here on the left is a mouse that doesn’t have a brain tumor, and the right is a mouse that does have a brain tumor. The brain tumor is tiny, it’s like the size of a pencil lead, but you can see this molecule that we made by taking the chlorotoxin from the scorpion, and linking it to a little molecular flashlight called a near fluorescent dye. we could inject it into the tail vein of the mouse, it would go through the mouse, find the cancer cells and make them light up so we could see exactly where they were on almost a cell-by-cell basis. We imagined that this could someday be helpful for surgeons in human patients.

If we had discovered this for only for kids with brain tumors, it would be a real problem to find investors that would help us raise the hundreds of millions of dollars it takes to go through clinical trials to the FDA. So I went around the Fred Hutch, and I borrowed mice that had colon cancer, breast cancer, sarcoma, and prostate cancer, pancreatic cancer, and in the course of the next two weeks we learned that tumor paint lit up all of those different kinds of cancer.

Not just in mouse models of cancer, but in mice where human tumors were being grown in those mice. This was better than we could have ever imagined.


In the picture that you see with the red dots, everything that’s labeled with an N is a lymph node in a mouse that had prostate cancer, and every node that lit up with tumor paint, was shown to be packed with cancer cells. The nodes that didn’t light up with tumor paint didn’t have cancer cells in them. That little spot that’s labeled with an L is not a lymph node, but it’s a lymph channel and it’s a channel that connects one lymph node to the next and there are about 200 cancer cells traveling from one lymph node to the other. Tumor paint was sensitive enough to show light where those few cancer cells were, and this is 10,000 more times sensitive than an MRI scan, and the surgeons would be able to see it in real time.

We went from mice, instead of going directly from mice to humans, where there’s a 94% failure rate when drugs company do that, we decided to take a little detour through dogs.

Slide06We did this with families who brought their dogs in for cancer care, these are not research animals, these are family pets who had cancer, brought them into Washington State University Veterinarian School.

Slide07One of the kinds of cancers that dogs develop is sarcoma, and here you see the first dog that was injected with what is now the clinical product called BLZ-100, which is a molecule that we inject into the veins that goes through the body, it finds these cancers wherever they are and lights them up. You can see in this first case that the cancer lit up but the normal tissue did not light up. Here we are going from mouse, now into dogs with their own cancer rather than a research cancer.

Slide08This is a dog that has breast cancer — it’s called mammary carcinoma in a dog — and the surgeons knew about that big spot down in the bottom right, that was cancer, but really that’s all they were able to tell from the clinical exam, and the scans that were done in advance. This dog received a dose of tumor paint the day before surgery, and here is what the surgeons were able to see. Not only could they see the main tumor, but they could see additional areas of cancer that were not visible to the naked eye.

Slide09I think that all across America everyday, women are told they have clean margins, everything looks good, we’ll follow it with some scans, and then six to nine months later they start to get some bad news. I think it’s because the surgeons can’t always see exactly where the cancer is, and sometimes cancer isn’t contiguous. It jumps around into some spots a little ways away from the primary, and tumor paint is helping us see this, at least in the dogs.

Slide10This is a human breast cancer that’s growing in a mouse, and we’re using a new device that we developed in collaboration with the team at Cedars Sinai Hospital, and it turns out that these were surgeons developing the device, so we think it will be adopted by surgeons because it started with them, and they said, “This is a device that we think we can use.”

The point here is that if you go back to that first slide that I showed you of the brain and ask where is the tissue? Where is the cancer? And I gave you a scalpel. In this case, would you be able to take out the tumor, and leave the normal tissue behind?

We think that most of you in this room actually could do that, and it shows the difference between operating with tumor paint and without tumor paint, and our hope is that someday surgeons would never be able to believe that they used to do surgery without it for cancer patients.


Here’s an example of a scan that none of us want to see when we’re in clinic. This is a brainstem giloma, this is a cancer that is right between the brain and the rest of the body. The question is, even if you lit this up with tumor paint, you could never do an operation here. It would disconnect the brain from the rest of the body. In the twenty five years that I’ve been taking care of kids with brain cancer, I’ve never had a single child with this kind of cancer survive.


It’s really hard when I go into the rooms and I meet a child that is something like this little boy, Kyle, and I have to tell him and his parents that this is a tumor that’s more than likely going to take their life within the coming year. We decided that we were going to go after not only this disease, but other diseases that we currently consider incurable by taking advantage of this whole new platform with drugs that we were learning about.


Everything I told you about with tumor paint came from the Israel death stalker scorpion, but it turns out that every plant, and every animal, needs to make drugs just to get through their everyday lives. The scorpions need it to paralyze their prey. These sunflowers need to make drugs that protect them from insects just eating them. This looks like breakfast to an insect, it’s bright yellow, it’s full of nutrition, why wouldn’t they just go through and devastate this whole field of sunflowers? But there’s no pesticide on these sunflowers, these sunflowers are making their own trypsin inhibitor which blocks the enzyme that bugs would squirt on them to break them down.

It turns out that when I looked into this that the trypsin inhibitor that is secreted by the sunflowers, follows the same molecular rules as the molecule that was made by the scorpions. It turns out that this is a category of drugs that are called knottin peptides or mini proteins, because they actually literally tie themselves in a molecular knot. The nice thing about that is that knot makes them resistant to stomach acid, to the enzymes in blood that would take a protein drug and make you think that it’s nutrition and try to turn it into food for your body and break it down. These are really beautiful scaffolds for human drugs if you just look at them through a broad perspective.

Slide18These are made by many different kinds of plants and animals throughout the world, I’ll give you just one example. The second panel there is the brazian fruit in Africa. Everybody may or may not know, you might not even want to think about it, why is fruit sweet? Fruit is sweet because they make sugar, and it’s sweet because animals will want to eat them and then spread their seeds. The brazian plant actually instead of making sugar, which takes an enormous amount of energy from each plant, they make one little knottin that binds to our taste buds a thousand times stronger than sugar, and tricks the mouths of animals into believing that they’re eating sugar, and by that it gets its seed spread. That’s just one example of how nature has made these drugs to protect themselves, or to foster future generations of the same plant or animal.


It turns out that these were first described about 23 years ago, and when we started the project I looked into why had the pharmaceutical companies failed to make more progress when the first knottins were described twenty three years ago? We came up with this list, I’m not going to go through a science lesson here, but one example is these fold poorly in E. coli and yeast. Yeast and bacteria are the workhorse of the biotech industry and the pharmaceutical industry if you’re working on protein drugs. If you can’t make these things in those organisms, then I can understand why they wouldn’t make a $100 million investment into trying to figure out a different solution.

Colin Correnti and Chris Mehlin

We came up with this list of reasons that they had failed, and then with the gift from the grandmother of one of my patients, I was able to hire these two guys, Chris Mehlin (right) and Colin Correnti. Chris had been the head of molecular biology at Amgen here in town, Colin had been a student, and the day before I hired him he defended his thesis, and he had figured out a way to trick human kidney cells into making these little knotted proteins, for an entirely different reason. It was just by chance that I happened to go to his thesis defense, and he had already taken another job at another biotech company out in Woodinville, and I asked him, “Well what are you going to be doing out there?” He said, “I’ll be doing crystallography to help them discover this new drug.” I said, “Well, what do you want to do in 10 years?” He said, “Well I’d like to be a CEO.” I said, “Well, how many times have you seen somebody’s crystallography bitch turn into a CEO in the next 10 years?” With that I had him hooked, and he came over, and in the next six months he and Chris resolve five of those six problems.


In the past couple years we’ve made significant progress on the sixth one, and I’ll show you some of that data today that nobody else has seen.


At the time that we started this project there were about 300 knottins identified by all the scientists in the world. There were about six thousand that were identified in the data base as possible being knottins. I recruited a guy from Oracle, a computational guy, he wrote a python program that crawled through all of the genomic data bases in the world, and in a day in a half he had identified two hundred and six thousand additional drugs that are made by plants and animals that fall into this category. These are now blueprints for what we do.


We went from being able to make about twelve candidates a year, to now being able to make about ten thousand in three weeks, and we actually surpassed that recently where we made 30,000 in six days.


This is what these little knottin proteins look like, you can see that they’re tied in a knot and all the active groups, the charges and the groups that make them interactive for the proteins, are on the surface.


This is what we do with them. This is a base protein, and you can see how we can computationally change it, and turn it into different shapes and different charges, each of which could interact with the proteins that cause Alzheimer disease, autism, cancer, and interfere with those processes. This is really a way to address diseases that can’t be addressed by the little tiny molecules that are like 20 times smaller than these, that are made by Merck and Pfizer and the other drug companies.


When we design these on the computer we just put in the sequence of them, put it in the computer, the guys in the lab give them my credit card, and this guy shows up with a Fed Ex envelope that has 10,000 genes in it the following Tuesday.


We take those genes and we put them into a modified HIV virus that can’t cause HIV but it can deliver genes, and it spits them into human embryonic kidney cells that we’ve immortalized, and those human embryonic kidney cells become churning out drugs overnight while we’re home sleeping, and in the morning we come back, take it off the shaker, put it onto a machine that purifies them, and by the end of the day we have these new drug candidates.



We just ordered our first robot to make these, that robot will be able to make about two thousand individual proteins a month, right now we’re making about a ten a week. You’ll see the ramp up that we get by bringing robotics into this.


Again, in addition to making these things, we have a therapeutic team that focuses on these diseases that affect kids. We’ve identified one new target for glioblastoma, one of the most challenging kind of brain cancer, that on the top you see one glioblastoma cell that’s dividing into two, and all of the DNA is perfectly dividing and pretty soon there will be two cells, but when we hit that target the cell explodes instead. It’s called mitotic catastrophe, and that’s what we’re working toward is these new drugs that will actually make cancer cells explode and disintegrate, rather than turning into two daughter cells. We’re making really nice progress on this.


One of the things that happened at a lot of conferences like this when I would give the talk is people would say, “If you can deliver light to the cancer, why can’t you just deliver the drugs to the cancer?” The reason is that the chlorotoxin that we use from the scorpion, in addition to going to go the cancer, a lot of it also went to the liver, and to the spleen. If we put a toxin on it we’d wipe those two organs out, and we didn’t want to do that.

A few years ago we had a fundraiser over at Fred Hutch and the Kathi Goertzen Foundation sponsored it, and they raised about a hundred thousand dollars that night, and we said, “We want to do something unique with this money to honor Kathy,” so we started the Kathi Goertzen  Library of 100 of these molecules that we would share freely with any scientist in the world. I said to the team, “Since we’re making these anyways, let’s put some fluorescent lights on them, put them into mice that have cancer, and see where some of these go, do they go to interesting places?” It turns out purely by serendipity the third molecule we made went to cancer, but it didn’t go to the liver and to the spleen. Our Holy Grail, we saw it in the third molecule that we made.


It turns out that this molecule, anybody have any ideas where it came from? The grasshopper. The grasshopper is teaching us now, and we have, with our computational abilities, made thousands of variants with this molecule, and we’re honing in on the ones that we think would be best able to deliver the drugs right to the cancer. We hired we first chemist last week. In fact, 2-and-a-half years ago when we started this project we had two scientists working on it, right now we just hired our thirty second scientists, and I just interviewed the 33rd this morning.


One of the things that we wanted to do is just see where these things go. Different plants and different animals have reasons to get drugs to different parts of our body, and instead of fighting against nature we thought we would learn what nature is doing, and then work with it. We made twenty eight proteins that came from scorpions and spiders, and we found three more that went into the brain like chlorotoxin. These molecules can be used as a foundation for Schizophrenia, for Alzheimer’s disease, autism, any disease that effects the bring that would be difficult to treat with the small molecules.


We found one of these molecules that accumulates in the fluid around the brain, and if you look closely at that little tail that’s coming out of the black spot in the top panel in the center, those are actually the new nerve cells that are being formed in the brain, and so this drug is able to act with specifically the new nerve cells that are coming out of the nerve stem cells. You can imagine for nerve generative disease just like Alzheimer disease, that this would be a really important advancement. Again, you guys are the first seeing these slides in public.


How many of you have had a joint that started bugging you in your knee, or foot? Yeah, look at all of the hands that go up now. I ask you a science question and no hands.

Unexpectedly, of those molecules that we tested from the twenty eight scorpion and spider molecules, we found quite a few of them that go specifically to the cartilage in every disk, that’s in our backbone, in the places where our ribs join our breast bone, in all the joints in our hips, our shoulders, every single joint in the body, and look at how specifically and how high these go to those joints.


Imagine that we can take a pain-relieving medicine, like a steroid, and just attach it to this, and deliver it right to the place where people are having pain, and skip those injections that work for four or five days, and don’t effect the other joints, and you can’t get another one for six months. Skip the side effects that would come by just taking a steroid systemically, and having to deal with all those side effects. You can imagine that a pharmaceutical company doing a partner with us, for what would be a blockbuster drug for osteoarthritis, could fuel a lot of the work that we do on pediatric brain cancers, and other rare diseases at the same time. This is very much aligned with our overall mission.


This is one of the buildings at the Fred Hutch Cancer Center, and uniquely they gave us an additional four thousand square feet, that whole top row of windows is our new production facility for these proteins. They recognize that we’re doing something nobody else in the world can do, and ultimately we want to build a world class program where scientists from all over the world can come and spend three weeks, or three months, or three years, whatever it takes to do their work with us collaboratively. We want to break down the silos between different institutions to make this happen.


This is part of the team, it’s grown a lot even since this picture was taken, but these are the folks that do the work and I want to acknowledge them. Finally I want to acknowledge the community. When we started this project, in addition to the challenges that the pharmaceutical companies faced, our decision was to not spin this out into its independent biotech company. If we did we’d be responsible for investors for getting a maximum return on it, that would almost surely mean selling that company to a pharmaceutical company, where the decisions about which drugs were made and which ones weren’t made would be in the hands of one small committee. We didn’t know if we could trust them with something that we think can really change the world.


We started Project Violet in honor of this little girl Violet, who is a patient of mine and who asked us to take her brain at autopsy after she died, and make research tools to share with the world. In her honor we started Project Violet. It became the front porch of this whole new drug discovery program that we’re talking about, and through Project Violet we have raised over $8 million dollars in the last two years that allows us to hire these 33 scientists, to buy robots, and to just go full-speed ahead while keeping these libraries available to do our own research, to partner with pharmaceutical companies, to partner with biotech companies, and to partner with academics who are working on rare diseases that can’t be treated in other ways.

I appreciate your time, encourage you to follow us on Facebook for Project Violet. … Thank you so much.

Subscribe to GeekWire's Space & Science weekly newsletter


Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.