Here’s another one for the file of huge, long, costly drug discovery efforts that came to nothing (and that no one outside the business ever hears about): cathepsin K inhibitors. I remember “Cat K” from my own (relatively brief) days in osteoporosis drug work some years ago. It’s a target that’s been around since the mid-1990s, and Merck alone has been trying to develop their drug, odanacatib, for over twelve years now. But not any more.
The enzyme is able to break down collagen, elastin, gelatin and other proteins, and it’s a key player in the activity of osteoclast cells as they break down and resorb bone tissue. Recall that your bones, at a microscopic level, are constantly being remodeled – osteoclasts are breaking things down, and osteoblasts are building them back up. A longstanding theory of osteoporosis is that the balance between these two activities has gone off, with bone destruction running ahead, and you’d certainly think that anything that slows down osteoclast activity would help with bone loss. It’s very hard to do that, though, without slowing down osteoblast activity as well, so some osteoporosis therapies sort of settle for bringing things to a standstill rather than allowing bone regrowth to take place at the levels you’d probably want.
Thus cathepsin K inhibitors. Osteoclasts secrete acid to break down the mineral component of bone tissue, and they set cat-K loose on all the protein components (the enzyme works best at acidic pH). It’s found almost entirely in this cell type, giving you a real hope of targeting osteoclasts selectively. Protease inhibitors themselves are a type of small molecule that, in general, we have pretty good success with in drug discovery, so what else do you want? Well, there is one little thing. A lot of people have looked at this target over the years, but osteoporosis (although a big market) is not one that you get into lightly, because the patients are elderly and the disease moves so slowly. The clinical development costs will eat you alive unless you have significant piles of cash to risk.
A number of companies took a crack at it (here’s a table from 2012), but Merck was widely thought to be in the lead. The Novartis compound on that list, balicatrib, wiped out in Phase II because of dermatologic side effects, and the GSK compound, relacatib, went into Phase I and then just sort of hazed off into the limbo that awaits problematic clinical development candidates in that organization. Both companies published patents showing that they were following up with other chemotypes, but I’ve no idea what happened to all of those. Sanofi had a dual cathepsin K/S that also dropped out of the clinic for reasons unknown. A number of smaller companies have also been developing compounds, no doubt hoping to sign a deal with a larger outfit to get through Phase II and III (I think that Medivir is the only one left in the game).
Merck’s odanacatrib phase II trials went on for five years, but they showed that the compound did, in fact, slow down bone resorption and increase bone mineral density. The Phase III trials ended up with over 16,000 female patients, 65 and older, at nearly 400 clinical research centers, and if you think that one wasn’t a beast to run then you ought to come on down and try it sometime. The good news for everyone was that the trial was actually stopped early for efficacy, whereupon the control group got a chance to switch over to the drug. Analysts cheered and raised their earnings forecasts.
But no story is simple in this business. Cathepsin K, it gradually became clear, was involved in a lot of other things besides bone resorption. (Here’s a good review). Thyroid function, learning and memory, processing of vasoactive peptides, adipose tissue – it may not be expressed at high levels in some of these tissues, but it might well be a player nonetheless. People kept waiting for news on odanacatib after the success in Phase III, but as time went on, it became likely that there was something complicating the picture.
And earlier this month, Merck announced that they were discontinuing development of the compound. Analysis of the clinical data showed that while it was indeed efficacious, it also increased the risk for stroke, which is something that you really have to be on guard for in this patient population. As far as I know, no one quite knows the mechanism behind this problem or saw it coming. It’s quite possibly a mechanistic problem, which would probably doom any other cathepsin K inhibitors, but who knows? You’ll need to get a really good compound together and dose a few thousand patients to be able to answer that one, and if that sounds like a good time to you, again, come on down. Note as well that by the time you make it through the clinic, your patent will have the majority of its lifetime eaten away, leaving you with a narrow window in which to charge for the fruits of your efforts before the compound goes generic. Merck, for its part, spent at least 12 years on odanacatib, and who knows how much money. We now know much more about cathepsin K than we ever did, but this knowledge did not come cheaply. . .