Derek Lowe considers the problems of addressing drug development out of sequence

Medicinal chemists should always be aware of where they’re positioned in a drug development project. Ideally, you want to be near the front of the process: where the compounds are synthesised, before you wish them luck on their way to the assays. You learn what you can from the resulting data, repeat the process as necessary, and deal with problems as they crop up.

That’s how it’s supposed to work on paper, which is fine, but where you can run into trouble is when the chemistry is pencilled in as a key step too late in the project. A typical example might be found in a central nervous system (CNS) effort. What you want to avoid is the ’and then the chemists take the best compound and find a way to get it into the brain’ step. This is bad, because we still only have rather vague ideas of how to get compounds into the brain - and these may or may not be compatible with the chemical you have at that point. If you’ve spent too much time optimising other things (potency, selectivity, etc), it may already be too late to turn this into a CNS drug.

Avoiding traps 

There are many other similar pitfalls. ’Now just make this compound orally bioavailable’ is a classic, and the anti-infectives people often get to hear requests to ’make a version of this compound that will get into (or stay inside!) wild-type bacteria’. The problem with all of these problems is that they put you at a real disadvantage - you shouldn’t choose to make any of them a key battlefield if you can possibly help it. Your colleagues who are making the requests may honestly not realise how difficult these steps are. If they’ve been around longer, or are more devious in general, they may just want to make sure that these problems end up with someone else’s name on them. 

Guidelines to success 

It’s worth thinking about these situations in general. The medicinal chemists reading this will know these difficulties well, but people outside the field are often taken aback to find out how little we understand about these issues. For example, one might think that drug absorption after an oral dose would be a pretty well worked-out phenomenon by now - but it most assuredly is not. There are some broad guidelines, mostly having to do with molecular weight, polarity, and the like. Other things being equal, these are best adhered to, but exceptions abound. Oral dosing remains stubbornly and expensively empirical, as do most other membrane penetration problems. And medicinal chemistry, once it gets out of the very first stages, is often a saga of membrane penetration, which makes it all the more unnerving that we don’t have better ways of dealing with it. 

For oral dosing problems, there are all sorts of formulation steps (particle size, different vehicles, salt forms, etc). It’s not a good sign if you have to resort to exotica in this department. And just as it’s not good form for people to pass on such problems to the chemists, it’s not a good idea just to assume that the formulations people will be able to solve your problems for you in turn. 

So what can be done about this? My first advice, particularly to med-chemists who are still learning their trade, is to be aware of these problems, respect their complexity, and don’t promise to solve them at some convenient date. Rather, get to work on them as soon as possible, because you may need all the time you can possibly arrange. The first step is to make good-looking compounds, if you possibly can. A naphthalene ring with five fluorines on it may improve your potency in the primary assay wonderfully, but you’ll have an awful time turning that beast into a drug. Better not to even look under such rocks - you could find things. 

And addressing these problems also means working as close to the real system as you can possibly get, at all times. Model systems and in vitro assays have their places, but will produce no drugs on their own. Drugs succeed or fail in whole animals, whether they are mice or humans. The sooner you can find out how your best compounds behave in them, the better off you’ll likely be. There’s a tendency to delay such tests until better compounds come along, but if you wait too long, you’ll find that you’ve waited until the only compound that can possibly work is all you have left.  

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