Allergan, now that they’re not going to be Pfizer, has wasted no time signing a big deal with Heptares for Alzheimer’s. Heptares, as the receptor-centric people in the readership will know well, has been working on GPCR structure for some years now, and applying this knowledge to their own drug discovery efforts. They have what they believe are selective muscarinic agonists, which could be useful in Alzheimer’s and several other CNS diseases, and have been actively looking to partner with someone on the (no doubt lengthy) clinical development of these. Heptares gets $125 million now, and milestones that could reach into the low billions should something (a) show efficacy, (b) make it through the clinic, (c) get approved and (d) sell. That last one should follow should the first three come through, since there’s little or nothing for Alzheimer’s patients as it stands.
So how useful will M1 and M4 agonists be? That is the big question. I was working in the field when Eli Lilly was trying to get a selective M1 agonist (xanomeline) going, and selectivity in that receptor family is not easy to attain. The side effects will hammer you if you’re not, though, which in the end is what happened with Lilly. They did come up with some clinical data to suggest that the mechanism might be useful in Alzheimer’s before the whole program wiped out, though, which is what brought Heptares back around for another try.
And if anyone’s going to get selectivity, they’re probably the ones. They have some Phase I data that suggest that they’re engaging the target without the sorts of side effects that were seen in the Lilly effort, and these results are surely a big part of why Allergan signed up with them. In fact, I’m willing to stipulate that Heptares has a selective M1 agonist, for purposes of argument, although more data will be needed to make sure of that. But let’s assume that they do: how well will that work?
Here’s where the brow-furrowing starts. Even back in the 1990s, there was some room to wonder about this. M1 is the main postsynaptic receptor in the cholinergic signaling system, so one thing a selective M1 agonist would do is be a straight-up acetylcholine replacement. But there’s more to it than that – there pretty much has to be, because acetylcholine is released in discrete pulses on the presynaptic side, and just dumping in more (or a replacement) wouldn’t be expected to be a very useful signaling mode. I believe that one rationale for an M1 agonist is that it would increase the tone of the whole signaling network, priming the postsynaptic neurons to be able to respond to lower doses of acetylcholine from across the synapse.
Lower doses you will have, most likely, because the whole cholinergic system is known to be hit pretty hard by Alzheimer’s. That, in the end, is one of the worries about targeting it as a therapy. If you don’t have a disease-altering mechanism (and no one does), then you might be boarding a sinking ship by going after the cholinergic neurons. Nonetheless, that’s what Aricept does (donepezil, Pfizer and Eisai’s existing Alzheimer’s therapy, now generic), inhibiting acetylcholine breakdown in the synapse. It doesn’t work that well. Back in the 1990s, another approach was to try to make selective muscarinic M2 antagonists, because that was the presynaptic autoreceptor that sensed acetylcholine in the synapse and told the upstream neuron to stop releasing it. Blocking that, it was thought, would potentiate the signals, making them longer and more powerful, and overcome problems downstream. No one got that one to work, either, but to the best of my knowledge it never got a real trial in Alzheimer’s patients.
So here’s my question: how sure are we that potentiating cholinergic neuron signaling is going to help Alzheimer’s patients? There’s a bit of hope from the efficacy seen with donepezil, although that’s mildly palliative at best and goes away after a period of treatment, and some hope from Lilly’s agonist program, which went pretty far into the clinic in mild-to-moderate Alzheimer’s patients, insofar as they could be identified back then. No idea if that efficacy, if efficacy it was, holds up either, or for how long. What I’m saying is that this really is something of a leap of faith.
On the other hand, leaps of faith are about all we’ve got in Alzheimer’s, and it will be good to see this hypothesis finally put to a test again after more than twenty years. More clinical data are expected later this year, but it’s going to be a while before we see results from either of the Heptares agonist candidates in actual Alzheimer’s patients. Allergan’s up-front money is not huge here, which seems appropriate, given the history of the field. But their downstream milestone payments are huge, which seems appropriate, given the size the market. This will be very interesting to watch.