Distinguished Research Scientist, Autodesk
To beat cancer we need more targeted drugs, made faster, made cheaper. And that means we need to fix the way cancer drugs are made. Andrew Hessel of Autodesk gives a presentation on how we can fix drug development.
Hessel: Good morning everyone—actually I guess it’s afternoon now.
So my name is Andrew Hessel and I’m somewhat of a scientist. I’m a microbiologist and a genetic scientist but I’ve kind of lapsed over the years; I’ve moved more into messaging and communications. But I’ve always been really interested in the intersection of genomics, of these technologies and business. And the area that has always captured my imagination is cancer.
Cancer for me is really dis-corruption. It is essentially a cell gone rogue in your body and instead of a microbe or a bacterium, it’s actually acting as an infectious agent. And if it’s bad enough it spreads through your body and can crash the network.
Cancer is really one of the diseases that we’ve made some progress with, but surprisingly little compared to other technologies. A hundred years ago, we weren’t so much worried about cancer; we were actually worried about bacterial infection. Bacterial infections, whether it was picked up by a food contamination, surgery, or other cut, could kill you in days. These are rogue cells that have taken over your body and these were life threatening—until this molecule was discovered.
This is penicillin. Penicillin changed the game in microbiology—penicillin and its chemical cousins, because really there’s a family of antibiotics that are so selective for microbes, they’re killed off but they don’t really hurt our normal cells. And so from what was really a scary disease, an infection that would kill us, almost overnight it became something that we just didn’t worry about. And we still don’t worry about it much today.
In contrast, the therapies that we use at the front line of cancer are essentially carpet-bombing approaches, very broad-spectrum. You know, you’re trying to differentiate between a normal cell and a cancer cell, but there is no antimicrobial, no antibiotic that is so specific for that, so we just kill everything that’s fast-moving. And it’s very effective but its toxic, and you may not get it.
Now some other newer drugs that we’ve been making, drugs like Herceptin and Gleevec, are very targeted. The molecular targets are known and they’re usually used as cleanup. But when these drugs are available, when the target is in the cancer cell and combined with chemotherapy, the change in outcomes is as radical as what we saw with antibiotics. It can be potentially a cure.
So everyone wants more these magic bullets that can selectively target cancer cells, everyone—every doctor, every patient. And the thing is, medicine—if we can make these, I believe the face of cancer will be changed.
Unfortunately I don’t think we’re going to get them anytime soon, and that’s because drug development is a process that takes ten to fifteen years and costs over a billion dollars. We all know the numbers, we don’t know exactly the numbers because the accounting isn’t very good, but I spent seven years of my life working with a large bio-pharma and the numbers are pretty close. It’s very expensive. It’s very slow. It’s difficult work.
This is actually the most important slide in my presentation. It is the trend line of drugs developed per billion dollars invested over the last sixty years—data that was compiled by Boston Consulting Group. Now I love exponential technologies. This whole museum is a testament to exponential: faster, better, cheaper, Moore’s Law. This is actually the reverse Moore’s Law. It is an exponential trend but not a positive one; it’s a negative one. It’s a negative slope. This is not one drug company; this is all drug companies in general. You know, this is their business model. It is slowing down. This is the number of drugs that were actually approved by the FDA last year: twenty-seven. This is not cancer drugs; this is all drugs. And this is one of the reasons why, really, drug making is so difficult and why our ability to cure disease is becoming compromised—not just manage disease.
Now the business model of drug making isn’t that complicated. It’s kind of like a Hollywood movie studio. They go around the world, they find interesting projects, they bring them in-house, they clean them up, they polish them, they get them past censors, and then they try and get them out to the largest market possible. Cancer drugs are the summer blockbusters of the drug industry, really—lineups of people waiting for the drugs, very low expectations, easy money. About $13 billion dollars worth of cancer drugs were sold last year. The targeted medicines that we need to really start fighting cancer, because no two cancers are exactly the same—remember it’s an infection if your cells in your body—they’re kind of like little indie arts films. They don’t really work in the Hollywood model; they’ll never go mainstream.
And of course if you’re trying to make these small, targeted drugs, they’re only going to be for a small number of people, and whether you’re making a blockbuster or a targeted drug, it really costs about the same. So the result is, the smaller the market, the higher the price of the drug, the harder it is to get anyone to pay for it, health insurance companies or individuals. And so the reality is: the best drugs today for cancer are actually used by the fewest people. This is just the reality. And any new drug coming out of the pipeline today is going to suffer this type problem.
So I started asking myself a decade ago, could this trend be reversed? I saw it in the drug company I worked at. I saw it in all drug companies. And I’m kind of a weird guy, you know, if everyone’s running one way, I’ll go the other. And the very simplest thing to look at was, well, if everyone, if the entire system is built on this business model, what does the opposite of this business model look like?
And really it looks something like this. Instead of making mass-market—you know not mass-market, but instead of making a broad approach to cancer, make a targeted approach, exclusively. Instead of being closed and proprietary and patenting everything, make it open. Instead of worrying about making a dollar, just make money irrelevant in the process. And instead of making drugs for more than one person, a mass-market, just make it for one person at a time, just one, because every single cancer is different.
Now that sounds a little weird, but when you think about it, it’s achievable. So I actually made a little provocative company, a co-op biotech company, and our mission was this: to make the best cancer medicines possible, as fast as possible, and give them away for free. Now it’s mainly a thought exercise, it’s meant to be provocative, but it’s not meant to be silly.
When you think about it today, companies like Google—information giants, profitable as hell—give most up their tools and services away to people for free. There’s plenty of things you can give away for free that actually can have an underlying economy that’s more powerful. So I actually thought it was doable for pharma.
Plus, I’ve been tracking the digitization of pharma, of genetics, for fifteen years now. This is the field of synthetic biology, and you’ll probably hear more of it, you may work in it now. But synthetic biology is really just genetic engineering done with digital tools. This makes it faster, cheaper, and easier to do.
This is how we did molecular biology and genetics for as long as I’ve been around. It’s a complicated kitchen. There’s complicated mixing devices, there’s complicated liquid handling devices, there’s complicated ingredients on the shelves. And it works, it works fine if you’re in an individual kitchen. But it’s slow and expensive and it’s hard to do.
This is the future genetic engineering. This is a printer that prints DNA. You can see it’s not really a consumer device but it really will just print out DNA. It’s like a 3D printer for the DNA molecule, and it’s one of the most powerful tools we’ve ever made. It’s like a laser printer for DNA.
So I work for a software company. They make software tools for designing just about everything—cars, buildings, trains—which is kind of strange. So why is a genetic engineer working for a software company? Our business is changing.
This is our new facility down at Pier 9. This is the first time we’ve stopped writing bits and started to move into atoms. It’s brand new, we just opened it up, and it’s one big giant maker space. It’s like TechShop on steroids. And we’ve got artists in residence, but we’ve also got all this manufacturing capacity, from 3D printers to metal shop, woodshops, electronics, sewing. We bought a company called Instructables that’s housed there and really teaches people how to go make stuff, and the whole maker movement is so cool. It’s so cool—your kids are probably involved in today if you’re not. It is how to actually use some of these new tools that are available everywhere—and some of them quite sophisticated—to go make stuff. And one of my favorite applications that we have today for the new makers is a company called Creature. And with this you can design virtually any creature using an iPad-based app and you can send it to get it 3D printed. It’s really neat. So you can start imagining that.
Now we’ve been developing some more sophisticated tools for the bio-nanospace, a kind a platform for it. I won’t go into that, but basically it’s so we can start to actually work at the molecular level with serious scientists on some of these tools, and I believe that some these tools actually allow us to make programmable drugs. Anything that can be encoded in DNA actually is tractable through synthetic biology. But what drug could you possibly make that would work in cancer? And the answer for that was actually dropped my desk by a colleague nine years ago. He put a review paper of oncolytic viruses on my desk and said, “Read this, you’ll love it.” Now I’d never heard of an oncolytic virus before, but it’s actually pretty simple. An oncolytic viruses is a really weak human virus. It’s so weak that if it infects a normal cell, it can’t get a toehold. The cell has all sorts of virus protection mechanisms in place, it shuts the virus down and it basically stops the propagation. But they found that cancer cells actually, because they’re broken, have some broken viral defenses. And some viruses have been matched to different types of human cancers and they can get a toehold in the cancer cell, start to replicate, break it open, and actually release more viral particles, which can go back and infect cancer cells again. So it’s actually hacking cancers to become, really, little cancer drug companies.
Effectively cancer cells are catching a cold. These are not really toxic viruses and they’re naturally oncolytic without any genetic engineering. But if you do genetically engineer the virus, you make it safer, you make it more targeted, and in fact you enhance the various cancer-cell-killing abilities. So I fell in love with these viruses.
And it was this paper that I saw from 2003 where Craig Venter and his partner Ham Smith, a Nobel Prize winner, had demonstrated that they could make a synthetic virus. Now this is kind of complicated but it reduces down to three steps: Design the genome of the virus using computer tools; synthesize the DNA of the virus using those DNA printers I showed you; and then ultimately boot and test the virus. And this is not a cancer-fighting virus, I want to be clear, this is just a virus that infects E. coli cells. But what’s neat is it took them two weeks ten years ago to do this. This is one of the first synthetic genomes. So I’m fascinated by this technology of synthetic virology.
And in fact I got Autodesk to start exploring this. This is a 3D model of the ϕX genome, the same one Craig made. We ended up working with the number of different suppliers over that time. We also 3D printed the model of the actual virus. That’s kind of cool when you can hold it in your hand. That’s my colleague, Jackie Quinn. We sent it off to a DNA synthesis company. In fact, we sent it off to a number of them. And they printed the entire genome of that virus, which is 5,386 bits, and a number of companies sent us the DNA. And in a very simple few-minute experiment, we booted up that virus. What you’re seeing here is a growth plate of E. coli, and wherever you see a spot, that’s where a synthetic genome has booted up and started replicating in killing E. coli cells. This is the first organism that Autodesk ever made. It started its life as a file on my laptop. And it really got people interested in the company and this work, because really it’s a 3D-printed virus. It took fourteen days, a thousand dollars, and you didn’t need a lab.
So I’m seeing the start of a whole new drug development pipeline, one that really starts with the cells you’re trying to target and everything you can actually quantify, digitizing that—the sequencing, all the stuff you hear around genomics—getting that into a design tool. I don’t care who makes it, but it’s really powerful. This is going to grow fast. Autodrug—take those designs, turn them into genomes, and boot them up and get them into treatment centers. I think we’re ready to do this in cell cultures really quickly. I think will move into animals fast. And I think we’ll get into humans very soon. And this is the part I like: the cost of making this DNA is falling faster than Moore’s Law. It’s going to plummet in 2015, and already I see that most viral genomes are going to be so cheap to make after 2015, they’ll be a few dollars. This is the possibility of really changing drug development. I’m not saying the first drugs will be that great, but we can build a community around doing it because now we have a programmable therapeutic.
So really digital biology is allowing makers to become drug makers too. It also changes the idea of the business model. All of a sudden instead of buying one drug, maybe you can get a subscription to an endless number of personalized cancer drugs—the Netflix model. No one knows how this is going to work, but beyond that, because these tools are already democratized, digital, and available, you could even see the emergence of YouTube you know, make your own drugs for any reason. I don’t know where this will go, all I know is we’re at the start of this journey, it’s extremely powerful, and we should think about it and explore with an open mind.
Thank you very much for your attention.