I had a question from a reader about Substance P, a peptide that's been known since the 1930s as something that was involved in pain and neurotransmission. Its biological target is the neurokinin receptor subtype NK1, and there's been a huge amount of research on this system over the years, studying its role in the peripheral nerves, the spinal cord, and the brain.

And most of this work pointed to the idea that something that blocked this pathway would be an excellent analgesic. Stimulation of SP-responsive neurons produces sensations of burning pain, for one thing, and injection of the peptide is very unpleasant.) Weirdly, naked mole rats don't use the SP pain pathway, and are impervious to normally painful things like treatment with pure capsaicin. (Capsaicin, the hot pepper active ingredient, causes quick release of SP and the accompanying pain sensations, but ends up depleting it from presynaptic neurons, eventually raising the pain threshold.)

Pain relief is one of those things that some people think has been solved, but it really hasn't been. It's hard to knock down severe pain without knocking out the patient or using something with a high addictive potential. There are plenty of conditions - burn injuries, diabetic neuropathy and cancer come to mind - where a powerful analgesic with fewer side effects would be welcomed with rejoicing.

Several groups took a shot at making antagonists, but there were a lot of wrong turns along the way. For one thing, the NK1 receptors in mice are rather different from the ones in humans, something that was only worked out after many people had been led astray by mouse models of pain. (The good ol' guinea pig, which in spite of its reputation isn't really used much in drug research, turned out to have closer homology to the human receptor.)

And when good compounds were finally developed at Merck and other companies, and were taken into clinical trials for pain relief, an interesting thing happened: they didn't work. Not at all. The title of a review article from a group at Merck (Handbook of Experimental Pharmacology, p. 441, 2004) shows the frustration: "Substance P (NK1) receptor antagonists - analgesics or not?" They go on to say:

"Despite the identification of high affinity and selective substance P (NK1) receptor antagonists and a plethora of preclinical data supporting an analgesic profile of these agents, the outcome from clinical trials has been extremely disappointing with no clear analgesic efficacy being observed in a variety of pain states. This has led the pain community to seriously question the predictability and utility of preclinical pain assays, especially for novel targets."

Indeed it has, and the situation is far from being sorted out, from what I know of it. But Substance P had more life in it. As its distribution (and that of the NK1 receptor) in the brain began to be worked out, people noticed that it was often co-localized with the serotonin system, and these lines of evidence suggested a role in depression. Merck's MK-869 was the first compound to go into the clinic for this purpose.

And it died there, too. Initial results looked promising - check out this glowing report from 1998. But the next year, it was dropped, after failing to work better than placebo under controlled conditions. (That link is well worth reading for those interested in the topic of placebo controls, BTW.) After much searching around, NK1 antagonism was found to be imporatant enough in nausea and emesis for MK-869 to make it as an adjunct to cancer therapy. It's on the market as Emend (aprepitant), selling (as I put it in 2003, about one-fiftieth of what Merck had originally hoped. You have to wonder how long it'll take them to get their money back.

Many other companies have reported development of NK1 ligands, but I don't think that any have reached the market yet - and I don't know what they'll do if they get there. The whole area is an excellent lesson in the crazy complexity of drug target validation and drug discovery, and an interesting thing to consider when you wonder why drugs cost what they do. But we won't have to worry about Merck spending the time and money to learn such things if we sue them until they're crippled, now will we?