New pathways for schizophrenia treatment

20 February 2023

Stephen Brannan MD is the chief medical officer at Karuna Therapeutics in Boston, Massachusetts. At the 35th ECNP Congress in Vienna he presented the phase 3 results from the EMERGENT-2 trial, ‘KarXT (xanomeline–trospium) in patients with schizophrenia: results from a phase 3, randomised, double-blind, placebo-controlled trial’. The presentation generated significant interest.

KarXT is an oral, investigational M1/M4-preferring muscarinic agonist in development for the treatment of psychiatric and neurological conditions, including schizophrenia and psychosis in Alzheimer’s disease. The drug combines the muscarinic agonist xanomeline and muscarinic antagonist trospium.

Here Stephen Brannan talks to ECNP press officer, Tom Parkhill.

TP: What’s your background, how did you end up working with Karuna?
SB: My background is as an academic psychiatrist. When I was a senior resident, a new-fangled thing called “Prozac” came out, so my early career coincided with one SSRI after another coming out, and my focus was on mood and anxiety disorders. I also did some neuroimaging, mainly with PET scanning. Then our chair left, and the new chair asked me to take over the research ward. This got me involved in studies, and I saw more of the industry side of stuff. I was already being asked to talk about depression, and I had several pretty imaging pictures, which they also loved. Then one day Eli Lilly made me an offer I couldn’t refuse (as they say in The Godfather), and so I went into industry; that was almost 25 years ago. I’ve been in both small and large companies, and I’ve been fortunate enough to have four compounds come to market – and now potentially a fifth (fingers crossed). When I was with Takeda, a new CEO was appointed, and he changed the emphasis to oncology and GI (and away from neuroscience). I then went to Forum, a small biotech in Boston. I was familiar with the company, and I knew some of their people. They were working on cognitive impairment in schizophrenia as well as Alzheimer’s. I went there, but their phase 3 trial didn’t work. A couple of months later a couple of people I knew at Karuna approached me. They showed me some of their data, and I thought, “Wow, if this works it won’t be just another SSRI or atypical, it might really change things”. I wanted to work on something meaningful. And now, with KarXT, it looks like it might come to fruition, which is pretty exciting.

The response to the phase 3 results has been very positive.
It’s been such a long time in schizophrenia since there was anything terribly new. Atypicals were around when I was young, and everything new since has pretty well been like an atypical. But this has changed things. It’s not just us: the TAAR1 programme is novel and other people have revived their muscarinic programmes after seeing our phase 2 data.

Schizophrenia is largely treated through dopamine pathways at the moment.
Yes, dopamine and serotonin. And the D2 receptor is probably the primary key, which I was taught as a resident, and it hasn’t changed. In schizophrenia there are positive, negative, and cognitive symptoms, which have been something of an interest of mine.

You anticipate that this will allow treatment of people who don’t respond to other drugs. Where do you see the drug sitting?
A couple of years ago at SIRS, Oliver Howes was asking why we can’t do adjunctive treatment in schizophrenia. You add one drug to another, but it doesn’t really add that much ‘oomph’. He was saying that there is the new TAAR1, and muscarinic compounds, that might get some synergy with D2 compounds.
When I was a young doctor, I’d always want more than one arrow in my quiver. I would have wanted something which combined different mechanisms. It seemed to be “rational combining of mechanisms”. When you introduce a new mechanism, it should have different properties, different side effects. Schizophrenia is very heterogeneous, it’s very likely more than one thing that we lump together. People talk about it having hyperdopaminergic and hyperglutamatergic subtypes, and so on. And now, in the last few years, we are thinking about cholinergic systems. Again, my background is as a neuroimager, and the brain does interconnect. You can’t affect one neurotransmitter without affecting others – we’re not a big bucket of chemicals. But as a shorthand, it’s useful to think about these things, and when you combine, you might pick up more efficacy and the ability to help more people. As our programme has advanced, we have been pleased both by the efficacy and the tolerability of this drug in the schizophrenia population. The big news is that there are lots of long-term side effects which we haven’t seen in clinical trials, such as EPS, tardive dyskinesia, and the whole metabolic syndrome (including weight gain), as well as anti-cognitive effects. So it would be useful for subjects who are having difficulties with these sort of things. But there’s also a pretty strong ‘oomph’ in terms of the efficacy. Over the last couple of years we have seen drugs which are quite efficacious, but have a lot of side effects, or drugs which are less efficacious but which have fewer side effects. Now we have something which seems to be a game-changer with both. We may not completely understand it, but we are pleased to see the results.

A recent Medscape article on your ECNP presentation mentions that Eli Lilly had looked at M1/M4 receptor agonists as a treatment for patients with Alzheimer’s disease, but that the side effects were a problem.
The cholinergic system is hard to harness. I’ve been in companies over the last 20 to 30 years that have been trying to work at this, and we could never figure out how to do it. We would try to make compounds more specific for one receptor or another, sort of thinking that, “These are the good receptors, these are the bad” – I’m not talking here about dopamine versus serotonin, I’m talking about muscarinic 1 versus muscarinic 2, etc. When I came to Karuna they said, “Yes, we’ve thought about that, but we decided to do something different”. So in simple terms, the efficacious part of the drug would be central and the “problems” would be peripheral, so Karuna combined xanomeline with a peripheral anticholinergic to mitigate the peripheral problems, but leave the central benefits alone. It’s a simple idea, but the moment I heard of it I thought, “Why didn’t I think of that?”.

You presented the phase 3 results at ECNP. Why did you come to ECNP?
There were a couple of reasons. ECNP is a very large meeting of psychiatrists, so that makes it a good place to talk about a promising medicine for schizophrenia. The timing was right. We got our results in the summer (our phase 2 results were already out), and we had been speaking to some people at ECNP who said that if the phase 3 results came out in time, it would really interest delegates. When I went to the SIRS (Schizophrenia International Research Society) meeting in Florence in the Spring, I got really good feedback. So ECNP became a natural choice.

Can you tell me a little bit about the trial itself?
It was a very basic design which has been used for almost all antipsychotics. It’s looking at acute psychosis. We did five weeks – one week titration and four weeks of treatment. One thing which was a little different was that we used a flexible dose, so it was just KarXT versus placebo, which is a way to control placebo effects, which have been highly problematic in schizophrenia trials in the last ten years. The other thing is that because of the tolerability issues that we were worried about when we first started, we had a gradual build-up dose in that first week. 85 to 90% of patients went up to the top dose. We may have undershot, but we had very strong efficacy, so there was no need to push it further. When we started this at phase 2 we didn’t really know if we were at the right dose. It turns out from the results that we were! One of the things I was taught is that phase 3 is confirmatory, you don’t change what you did in phase 2. We changed very little from our phase 2 design.

We used the standard PANSS (Positive and Negative Syndrome Scale) as the primary outcome measure. It was an inpatient study, five weeks in a unit. There were two reasons for that. First, many schizophrenia patients are known not to take their medication, so we needed to make sure that people were actually taking the drug. It sounds like a straightforward, silly thing, but it’s an important detail. The second thing is that we really don’t want people taking other things which are psychoactive. Here I’m not just talking about illicit street drugs: people will have other antipsychotics at home; let’s say they are on placebo and they don’t feel well, the habit is to take one of their old pills. That adds to the placebo effect.

The trial met its primary endpoint, with KarXT demonstrating 9.6-point reduction in Positive and Negative Syndrome Scale (PANSS) total score compared to placebo at week 5. KarXT also met key secondary endpoints, demonstrating a statistically significant reduction in both positive symptoms and negative symptoms of schizophrenia. It was generally well tolerated and was not associated with common problematic side effects which are seen in current therapies.

You’re also looking at psychosis in Alzheimer’s?
The original studies on this drug, xanomeline, were for use as a pro-cognitive drug in Alzheimer’s. They discovered, really by accident, that it was good for the behavioural problems associated with Alzheimer’s disease – this was in the 1990s. The only drug we had (for Alzheimer’s) was tacrine, so it was competing with tacrine. Of course, it had the typical muscarinic agonist effects – diarrhoea, nausea, vomiting – things you just don’t want in an Alzheimer’s patient. It also produced fainting – syncope. However, since we have added trospium, in addition to the xanomeline, we have not seen a single case of syncope. So it looks like this combination does significantly de-risk the chance of fainting. This combination also decreases the GI effects. In the original trials they also started participants on the full dose, straight away. We don’t do that anymore. We start on a low dose and increase it.

You are scientific programme chair at the International Society for CNS Clinical Trials and Methodology (ISCTM). The ISCTM has collaborated with ECNP over the last several years. What’s the background to this, and how do you see it going forward?
We’ve been working together for about a decade or so. Rich Keefe and some other people were initially involved in pulling things together. The ISCTM is very involved in the sort of things I have been talking about. In fact, I was talking to our president, Gary Sachs, straight after we announced our results and he said, “So, methodology really does matter, doesn’t it”. And he’s right: it can really help you in these trials. There are obvious parallels between what ISCTM does and what ECNP does. The ISCTM meets twice a year, once in Washington DC in the winter, the other (fall) meeting has tended to move around. So the idea was that we meet with ECNP once every two years in the fall. This gives us a European point of reference, so we have a co-participant day involving both ECNP and the ISCTM. The day is co-chaired by someone from the ISCTM and someone from ECNP. We are just now going through a process of planning for the next joint day, in Barcelona. We’re looking forward to it.