The timing of this report from the FDA is surely no accident, but it’s always a good time to think about this: the great majority of all drugs that enter clinical trials fail. They fail because they don’t do anyone any good, or because what good they might do is outweighed by some serious and unexpected harm. Around 90% of all compounds that start in the clinic never make it out. Even by the time you get to Phase III – and these are drugs that have apparently already worked in sick patients by that point – the failure rate is still nearly 40%. Drug projects fail constantly.
It’s hard, sometimes, for people who’ve worked in other industries to appreciate this. Drug development is a unique combination of very high regulatory burden and very high failure rates, so it’s temping to say that the regulations cause the failures. But that isn’t true. Biology causes the high failure rates – specifically, our lack of understanding of biology. And that’s the essential part that it’s hard for people outside the field to grasp – just how little we know. We know a lot more about how some new code will run, how a new car will corner, how a new recipe will taste, how a new building will look, or how a new movie will do on its first weekend. A lot more. Those all have their uncertainties, of course – the code has bugs, the recipe needs less salt, the new movie actually sort of tanked in the end. But those uncertainties are, believe me, tiny compared to the uncertainty of taking a drug into the clinic and giving it to people. I don’t think that there’s anything else quite like it in the modern industrial world. It’s terrifying. Once you grasp the situation, you won’t believe that we do it this way, but we do, because we don’t have any other way to do it.
That how you end up with the examples in that FDA report. You have things like bitopertin, which was a glycine reuptake inhibitor developed for schizophrenia. No one had ever tried that mechanism before – this was new science. It made it through Phase I, and in Phase II, where real schizophrenic patients first took the drug, an eight-week controlled double-blind study in 320 patients showed a real effect. On to Phase III, 1800 patients for a year, and guess what? Nothing. This is also how you end up with darapladib, which was developed for atherosclerosis as an inhibitor of liproprotein-associated phospholipase A2. That was another promising new mechanism that made a lot of biological sense, and the compound’s Phase II trials moved blood chemistry in just the right ways, as predicted. People forecast that it was going to be a blockbuster. But for atherosclerosis, you need a lot of patients in your Phase III trial, and you need to follow them for a long time. 15,000 patients over a median of almost four years of treatment later, and guess what? Nothing. No benefit. Another Phase III trial looked at 13,000 patients who had just had heart attacks, and followed them over the next two or three years looking for a benefit. Nothing. The results weren’t statistically significant, but if the trend had been just a bit stronger, then the drug-treatment group would actually have been worse than the placebo.
These aren’t musty cases from the files; both these failures took place in 2013. The FDA report linked to above has plenty more, and if you want to go further back, I (and anyone else with experience) can send you to a long and expensive list of drugs, drug classes, and whole research approaches that have been complete wipeouts, despite looking really exciting and promising at the time. These things really do happen constantly, and it’s not because we’re a little slow on the uptake around here. We in the industry would, in fact, rather not blow hundreds of millions of dollars at a time on trials of drug candidates that don’t do anything. We are (and have been) frantically trying to find ways to keep from doing that sort of thing, but what you see is the actual state of the art in 2017.
So, to sum up, most investigational drugs don’t work. “Freeing up” the system in order to give them to more people will merely allow more people (and more insurance companies) to pay for drug candidates that don’t do anything, or actually make people worse. That’s what we’ve got for you at the moment. That’s why we run those big, expensive clinical trials, the ones that just look like stupid delays imposed by bureaucrats – turns out that those trials are not only a good idea, scientifically, but it’s also ethically unacceptable to do otherwise. If you think the current system is expensive, wait until you try tearing it down.
Update: see this by Matthew Herper for more – an excellent history lesson.