When I was a PhD student in the early 2000s, my boss, who was under 50 at the time as he's a bit of a scientific prodigy, used to claim confidently that we'd have an effective cure for cancer by the time he retired. He must be around retirement age by now, though scientists as famous as he is don't really retire. Anyway, even if he's active in research for another 10 years I don't think we're close to fulfilling his prediction. Which sounds negative, but it is also true that the rate of progress in real treatments for cancer has massively accelerated over the past 20 years. There are several cancer types, many childhood cancers, most breast cancer and some forms of leukaemia that are definitively curable now, though at the turn of the millennium they were a certain death sentence. There are many more cancer types which still kill people in the end but most people get at least several years of decent health after diagnosis, which wasn't true when I first started being interested in cancer.
I concur with Magnanti's view that the vein of discovering 'magic bullet' treatments by means of combinatorial chemistry is basically mined out. And she's right that there are things that need to be fixed on a social level as well as scientifically. Yes, the myth of the Great Man working entirely in isolation impedes progress. Yes, the way that pharmacological research is funded is horribly messed up. There have been some magic bullet treatments discovered, but the process of discovering them, testing them for safety and then finding out whether they actually work is prohibitively expensive, and the current international intellectual property structures are inadequate both for giving businesses a return on their investment, and for making affordable drugs available for patients.
Before you despair, though, I want to tell you that there are scientific approaches that are getting round the problems inherent in magic bullet treatments. Not just hypothetical ones that people might use, but actual approaches that are being studied in the lab, in animal trials, and in clinical trials. I've just been to a cancer research conference, mostly academics (rather than people working for pharmaceutical companies or medical oncologists, who are the other two main groups of people I meet at conferences), and a lot of what's going on is serious debate about what non magic bullet approaches are most likely to work. People were talking to fellow scientists, not to investors, they were not trying to "sell" their amazing discoveries of The Cure, but at the same time they were talking seriously about how best to apply their research in ways that will directly help patients.
Magnanti is skeptical about "personalized treatment", and I think she's right in the sense of the idea that was popular in the early 2000s after the sequencing of the human genome, and still has some traction, that we would go around sequencing patient genomes and predicting essentially from first principles what treatment would work for each patient. That's not feasible, not even with modern sequencing technology and unlimited computing power, because there are tens of thousands of genes and we don't know what most of them do and we certainly can't make an accurate prediction of how they will all interact. But I think we can and likely will in the next few years see more and more of something a lot less sophisticated than that: people were talking about simple, single gene or small patterns of gene tests which predict accurately whether patients will respond to a particular drug, or whether they will have unusually horrible side-effects, or whether the drug will just be ineffective and therefore not worth the costs of either paying for the treatment or making the person endure the side-effects. Some of these ideas have worked, for admittedly small numbers of real people, but real people, not just a theoretical idea or something that works in mice in ideal experimental conditions.
There's also people who are moving away from the rational treatment paradigm, and trying to make so-called conventional chemotherapy better. Chemo (as distinct from rational treatment, where you use a drug that has a known target and molecular mechanism) is mostly based on giving people nasty poisons and hoping that the cancer cells die faster than the healthy tissue. And mostly that's kind of true, but only mostly. But what I was hearing about last week was people who are studying in detail what the difference is, why cancer cells die faster. The usual simplification is that they die because they're growing faster, but it seems like there's more to it than that. They have disturbances in their metabolism and their programmed cell death systems. And people are trying to come up with ways to widen the gap between killing cancer cells and saving healthy cells. Could be predicting which particular poisons the cancer cells are most sensitive to, could be finding ways to protect the normal cells, thereby improving the general approach of giving people additional drugs and treatments to minimize side-effects.
Another one is to use several drugs in combination; that's not a new concept but it's something that people are getting better at studying. The reckoning is that if you have three or four effective drugs then it's statistically near-impossible for cancer cells to develop resistance to all of them. This is the approach that's been used in mostly successfully treating AIDS. Or perhaps there will be combinations of chemotherapy with rational secondary drugs that make cancer cells extra-sensitive to chemo, meaning lower doses.
And there's research going on into just better drug screening and testing from the start. It's not very glamorous or headline-grabbing, but if you can get rid of drugs that look promising but won't actually work at an early stage, before you start doing expensive, economically and to the patients acting as guinea pigs, clinical trials, that makes everything a lot cheaper and makes it far more likely that people will be able to discover new drugs that are cheap enough to be offered to everyone who needs them. So more realistic animal models, better replacements for animals such as artificial tumours grown in petri dishes, more back and forth between genetic testing and empirical research to improve predictions of what will work before trying it for real. There are economic pressures against this, because investors want to hear about drugs that definitely will work, not better ways of ruling out ones that won't, and even academics get more prestige for publishing positive results than negative ones. But in spite of this, this kind of work is definitely going on; indeed the prize for the best poster at the conference was awarded to a guy who definitively showed that a particular drug doesn't work as claimed. It's a p53 drug I worked on briefly during my PhD, decided, this isn't really working, and moved on from it as I had higher priorities than trying to debunk it. So I feel a bit vindicated that my gut feeling this wasn't a good drug proved right.
Names to watch, IMO:
- Thorsten Stiewe who's doing really good rigorous work on the protein chemistry of p53, which probably isn't interesting to lay people in itself, but he's really exploring the implications of his work for actually developing better treatments.
- Thomas Helleday, who is kind of a self-publicist but he's got some pretty cool ideas about improvements on chemotherapy and deliberately poisoning cancer cells by messing up their DNA and protection from oxidants. And also social hacking; he's doing a thing called "open innovation" where he shares reagents and protocols with anyone who asks rather than trying to keep everything secret until he can get publications and patents.
- Tony Letai, who just impressed me. He's the one who's got really good stuff on predicting whether treatments will work or not, and he's thinking outside the box and doing some fantastic experiments to test his theories.
So yeah, I feel positive about the idea that research is leading to better cancer treatments, even though I don't believe that The Cure is just round the corner.
I prefer comments at Dreamwidth. There are currently comments there. You can use your LJ address as an OpenID, or just write your name.