A recent correspondence on the topic of “Why aren’t there more drugs for the big CNS disorders” got me thinking about the topic. My take, having worked in the field, is that there is still so much unmet need in that area because we just don’t understand what's going on. It’s hard to come up with disease-altering therapies when you don’t really understand a single disease in the whole field.
Does amyloid cause Alzheimer’s, or does Alzheimer’s give you amyloid, or is amyloid just a sideshow? What sets off the chain of events that ends up killing off cells in the substantia nigra in Parkinson’s? What are the detailed molecular mechanisms of depression, or schizophrenia? Why don’t neurons remyelinate in multiple sclerosis? We don’t know. We know a lot more than we used to; we know more every year. But we don't know enough to cure anyone yet. Even in the areas where we know more than average, we still don’t know enough to step in with therapies that can do what people really want them to do.
By that, I mean do for these diseases what insulin does to Type I diabetes, or what antibiotics do to infections. To any working CNS researcher, such results in their field would be hard to distinguish from magic. We can’t even touch the surrogate endpoints, and do what statins do for LDL levels, or the various antihypertensives do for blood pressure. We understand those areas a lot better than we understand the brain. Even so, we still get surprised, as witness the controversy over Vytorin, and the various ongoing attempts to find something that will raise HDL – you push a bit beyond the mechanisms that you’ve worked out, and all sorts of things start to happen.
The best way I can illustrate how difficult it is to find a disease-stopping therapy for something like Alzheimer’s is to point out the incentives for one. Any drug company that came out with such a therapy would immediately have one of the most profitable drugs on the market, and they would go on to reap more and more money every year. Think of the sensation that a treatment that stopped – just plain stopped – schizophrenia. As I said, indistinguishable from magic. But the success that such a thing would have would be immense. The incentives are there; it’s just that the barriers are very, very high.
Of course, it may not be possible to do some of these things. I’d be very careful to rule anything out, at our current stage of ignorance, but schizophrenia may well be one of these things where a dozen (or a hundred) different pathways lead to the same roughly similar disease state. (Cancer, as I’ve said here before, is the best example of something like this). And even if it’s not quite that bad, it may be that the tangle of the disease just doesn’t lend itself to a single agent – that, I’d say, is quite likely. I strongly doubt if just stepping in and adjusting the D-whatever dopamine receptor a bit will turn out to do the trick. This doesn’t mean that it’ll be impossible to treat, it just means that it’ll be very complex.
And so it is, and so are most of the other big CNS conditions. I find it hard to explain to people outside the field just how complex these things are, and why progress has been so painfully slow for the patients who need these things now. It’s not that there’s no explanation. It’s that actually finding a drug that works for anything is ridiculously hard and expensive, a very difficult task by anyone’s standards. And CNS drugs are fiendishly difficult even by the standards of drug discovery.
1. RandDChemist on October 15, 2008 9:15 AM writes...
My understanding is the in terms of Alzheimer's, the amyloid plaques are gravestones.
I think that as imaging methods for the brain continues to grow, so does the chance of understanding CNS pathways. Or I hope so...
Permalink to Comment2. Jose on October 15, 2008 10:28 AM writes...
When "best methods" in CNS research are still essentially "drop-a-mouse-on-a-hotplate" you realize we still don't have even the slightest grasp of the systems involved.
Permalink to Comment3. petros on October 15, 2008 11:10 AM writes...
And there are all those drugs for stroke, which work well in the standard animal MCA occlusion models, but all fall and burn in phase II or phase III studies. That's another area littered with expesnive failures
Permalink to Comment4. Sili on October 15, 2008 11:16 AM writes...
I vaguely recall that it's possible to find distinct differences between 'Summer kids' and 'Winter kids' that grow up to be schizophrenic. Seems very likely that there are indeed more than one illness at the bottom of it, and that will indeed screw up just about any trials for efficacy until we find some sorta (epi)genetic marker.
Is it customary to slice and balance human guineapigs according to their birthdates to look for effects like this? Or is that considered too close to astrology for comfort?
Some years ago I was lent the memoirs of the retired chief pathologist here in Denmark. He'd been looking into cot death (informally*) and he had the hunch that it was related to schizo.
* He'd checked SID victims with unusual surnames against the phonebook on the hypothesis that if there was a genetic cause there was bound to be 'near misses' of SID around. Of the medical histories he could collect that way, the only condition he could correlate to SID was schizophrenia.
Permalink to Comment5. c on October 15, 2008 3:39 PM writes...
Nature has some useful reviews on neurobiology this week (455,16).
Permalink to Comment6. clinicalpharmacologist on October 15, 2008 4:36 PM writes...
Derek. I have spent most of my 20 years in pharma working in neurological and psychiatric disorders and I could not agree with your comments more.
Rational drug development requires a solid rationale and we don't have (m)any in this area. So, lucky drug development is our best hope and the current development process has had all the opportunities to pursue luck designed out of it. Which is a shame because the approach used to work rather well. And productively.
Permalink to Comment7. milkshake on October 15, 2008 4:57 PM writes...
One CNS area that can be approached on rational basis: There are kinase cascades that start with insult stimuli and eventually trigger neuronal death. If you can interrupt this maybe you can also slow down the progression of Alzheimer, Parkinson and glaucoma. Or you could reduce the stroke damage.
Permalink to Comment8. srp on October 15, 2008 6:49 PM writes...
#6: Welcome to the "maybe trying to target specific receptors with specific molecules isn't the best way forward at this point" club. If you have any specific points about what needs to be done to restart "lucky drug development" I'd be interested.
Permalink to Comment9. Ty on October 16, 2008 9:15 AM writes...
Agreed with #6 and #8. Reductionist approach (targeted drug) is especially futile in the CNS area for the reasons I don't have to go over. Let the (molecular cell biology) brain research go on, but drug discovery had better be (or should have been) driven by phenotypic assays (in vivo or ex vivo). At this point, it's too early for drug discovery to reap any benefit of CNS biology. We should make (lucky) drugs first and let the biologists to connect the dots.
Permalink to Comment10. G on October 16, 2008 2:27 PM writes...
Agreed with #9. Just look at modafinil. It was discovered 20 years ago and we still don't really know how it works, aside from a laundry-list of neurotransmitters involved (gaba, histamine, dopamine, hypocretin, etc). It definitely is an example of "lucky drug development" that likely originated out of the observation that it had stimulant properties in animals.
Permalink to Comment11. Anonymous on October 17, 2008 4:51 AM writes...
Where are the drugs: China
[at least that's what GSK thinks...]
Permalink to Comment12. JBK on October 20, 2008 3:56 PM writes...
A neuroscience professor of mine in college put the state of CNS treatment as follows: "Everybody likes the analogy of the brain as your body's personal computer. Now imagine that your computer is broken, and the only thing you're allowed to use to fix it are a car battery and a set of jumper cables."
Permalink to Comment13. Glen on October 27, 2008 12:46 PM writes...
I think part of the problem is that there is to much focus on finding the smoking SNP. Auto mechanics can teach us something here. If our car is stumbling, he/she knows that that the same symptoms can be the result of myriad of different problems.
Permalink to Comment14. K spigos on March 6, 2009 6:06 PM writes...
I infer that the real problem could lie in our current nosological system...so let's hope for the revolutionary advance in neuroscience or even in medicine, that will lead us to a new one...
Permalink to Comment