Throwing it over the wall

People who’ve worked at several different drug companies will have noticed several problems that everyone knows about, that everyone tries to avoid, but that still keep coming up. One of the biggest is the ’throwing it over the wall’ style of drug development.

Which wall? There are several. A drug candidate has to be handed along to several different groups on its way to the market - there’s no way that the same team can stay with it from the earliest discovery days until the time comes to pick out the colour of the packaging. That’s because the expertise needed at different stages covers a wider range than any one group can supply. For example, at most large companies, one group will develop the idea for the target, get it through screening, and do some preliminary work on the most interesting leads that come out. Then another team of chemists and biologists takes over, driving the compound series along until it’s been narrowed down to a single clinical candidate. At that point, another set of natural handovers takes place, as the scale-up chemists find ways to make enough drug and the clinicians take the compound on through advanced animal screening and then through human trials.

The problem here is that this system can become a bit too compartmentalised. The people working on Phase II human data don’t really know that much about high-throughput screening protocols, and the hit-to-lead chemists would not be the people you’d want running your extended mammalian toxicity trials. (I’ve been one of those chemists; you really don’t want me anywhere near the rats). So in the course of sticking to what they know, each group may decide, naturally, that their work is done when the compound moves on. It’s a short step from that to thinking that their job is to wash their hands of the compound as quickly as possible. And that leads to the belief that any weak points in the development package are someone else’s problem.

Sowing dragon’s teeth 

So the synthesis of the lead molecule involves a starting material that’s hard to source, and a potentially explosive exothermic reaction along the way. Well, that’s the process group’s problem, right? Isn’t that what they’re paid to do, find ways around that sort of thing? So the compound shows good blood levels, but only when dosed in solution, and said solution is fit to strip the paint from the walls. But why else do we have a formulations group, if not to clean up these details? And so on. Each group can feel as if it’s doing its job, while sowing dragon’s teeth for the teams that come after them.  

There are several ways to deal with this tendency. One is to make more explicit the boxes that have to be checked off along the way. That’s not always a bad idea, but it runs the risk of being counterproductive. Once you start adding criteria and cutoffs, it can be hard to stop. Some personalities are peculiarly suited to putting up roadblocks and toll booths, and they may be disastrously attracted to the whole idea of coming up with the perfect checklist. You can find yourself with a list of criteria that no real compound can ever quite meet. That can leave everyone feeling as if there’s something wrong with every compound that advances (which is probably true, but it’s not good for morale to make a policy of pointing it out). At worst, it encourages bending and polishing the data in order to meet the targets, and that way lies sorrow. 

The other fix that I know of for this problem isn’t so much managerial as personal. Simply put, it’s harder to pass problems along to people you know well. If the next group along works at another site completely (or even on another continent), that makes it easier to just mail them a package and be done with it. But if you run into these people in the hallway regularly, you may think twice about that approach, for both altruistic and non-altruistic reasons. (That’s the beauty of this solution; it works with several different personality types to achieve the same end). But even widely separated groups can stay in enough contact for this to work, if the company’s structure allows. 

The downstream groups will also have a chance to spot potential trouble - that is, if the presentations they’re seeing are forthright enough to show them. Now, there’s another behaviour that you’ll see across companies, for sure. But a cure for ’sanitised slide disease’ has yet to be found. 

Derek Lowe is a medicinal chemist working on preclinical drug discovery in the US.