ChEMBL

I'm writing a foreword for a book on drug side effects, and thought of something that doesn't really fit in with the theme of the book, and so thought I would post it here (I'd better get a free copy of the book when it comes out, or the foreword will be really bad ;) ). Combination drugs are quite common, and typically contain agents that target different processes in the same pathway (there is a database DCDB that records a wide range of therapeutically used drug combinations if you're interested in seeing some real examples), there a have also been a few biotechs that have searched for synergistic (sometimes referred to as potentiating) combinations of drugs, foremost has to be CombintoRx (again have a click around their website for an idea of their approach).

However, what about thinking about co-dosing me-too drugs as a rational way to minimise side effects. By this I mean taking two different statins, or two alpha-antagonists (there's nothing special about these targets or drugs, they just sprang to mind while writing; and of course it may not need to be two, it could be three or four....) and co-dosing them. So assuming in the first instance that they have similar PK properties, and imagine that they have a 100 mg once daily dosing scheme, and furthermore, suspend reality and pretend they have identical potency against their therapeutic target - remember this is a thought experiment after all. The idea would be to combine them in a new pill with 50 mg of each component. This would minimize side effects through a really simple mechanism....

Drugs are not selective, they interact with many, usually tens, but maybe sometimes hundreds, of proteins in the body at clinical concentrations. The higher the Cmax of a drug, the more targets will be affected (so, think about drugs having an activity spectrum, analogous to a classical energy spectrum - the more energy available (equivalent to the concentration of the drug) the more ernergy states will be populated (targets will be affected)). So the higher the concentration of a drug, the more likely side effects are likely to occur (in fact it is well established that most adverse events are associated with peaks in free drug concentration). So two drugs targeting the same target can be combined to achieve an equivalent coverage of the receptor/enzyme. However the Cmax for each drug will be lower, and since the binding to off-targets will likely be different for the two drugs, halving the dose of each component would halve the Cmax for each drug and should greatly minimize the likelihood of adverse events. A key assumption here is that there are different binding profiles for the drugs, this seems a reasonable assumption, and it would be possible to explore this idea, if one had access to the data. However, even if they are identical in binding profile, it cannot be more dangerous with respect to off target side effects, than dosing with a single agent. The optimal combinations would have uncorrelated activity profiles, other than sharing similar affinity for their efficacy target.

Some negatives for this scheme include an increased chance of reaction to a drug containing a mixture of two (or more) active components, secondly, co-dosing drugs that interfered with, or saturated, the others metabolism is bound to lead to trouble, and thirdly, the business benefits of implementing this aren't obvious. Anyway, does anyone know of any attempts to try this idea in patients? One that occurs to me is the parallel dosing with COX inhibitors (paracetamol, ibuprofen and aspirin), but are there any other examples.