SAGE-324 (GABA-A receptor positive allosteric modulator)

Sage Therapeutics/Biogen

SAGE-324 (BIIB324) is an oral neurosteroid that acts as as a positive allosteric modulator (PAM) of the gamma aminobutyric acid-A (GABA-A) receptor. Currently it’s in phase 2/3 trials for the treatment of essential tremor (ET; NCT05366751, NCT05173012), the most common movement disorder. It is being co-developed by Sage Therapeutics (SAGE 0.00%↑) and Biogen (BIIB 0.00%↑) in the U.S. and by Biogen exclusively outside the U.S.

Before diving in, some brief bio background: what’s the GABA-A receptor got to do with tremor?

The therapeutic relevance of the GABA-A receptor has to do with the GABAnergic hypothesis of ET. Mechanistically, it can be summarized in this diagram from this review

From Gironell Tremor Other Hyperkinet Mov (2014)

The relationship between decreased GABA-A receptor expression and manifestation of ET has also been described in post-mortem patients and through pharmacodynamic imaging in patients with ET.

So, the idea with SAGE-324 & other GABA-A PAMs is ↑ GABA-A receptor activity, ↓ tremor.

Gone fishing (or how I found this compound in the first place)

One pleasant Friday evening, I had gone fishing at my sea of choice: the EPO. Teach the girl how to fish and she will never go hungry—as the saying goes.

I had keyed in “solid forms” and “therapeutic” in the refined search and was absentmindedly scrolling through the thousands of (mostly Chinese) patents that had appeared.

Few things of interest to me. But it wasn’t all fruitless, for I had managed to catch this Sage patent.

Aha! A forms patent. AKA, a late-stage development patent. Low(er) hanging fruit, to mix metaphors. So I thought: I should solve this!

There are certain challenges to starting from a company’s pipeline vs. starting from the patent space. At least when starting with the patent space, I know there’s a very high probability that this compound is in (late stage) development. The first hurdle to overcome, though, is figuring out which compound this is.

So I familiarized myself with the Sage pipeline

Reelin it in

As of April 16, 2023, this is what we’re working with.

From the crystalline forms patent (‘584), I knew the compound in question was a GABA-A receptor PAM. Assuming that Compound 1 in the ‘584 patent is one of the pipeline compounds, my first thought was that this narrows down the choices to…

• SAGE-689

• SAGE-319

…because the listed indication is acute GABA hypofunction.

Both compounds are apparently in phase 1.

Problem. Could be that the phase 1 isn’t registered. In that case, no further insights here to help me figure out what Compound 1 in the ‘584 patent is.

So I turned back to the ‘584 patent to help me out. I find this bit in the summary.

So a variety of indications here. Turning back to the pipeline: now, I can’t rule out SAGE-324. I look to trials.gov and see trials listed for SAGE-324. More information = more clues.

At this point, I actually was leaning towards either SAGE-689 or SAGE-319 being the correct match. But since we had intel on SAGE-324, I start digging deeper.

Matching the correct structure does not actually require an in-depth understanding of the biology. Rather, it’s been my experience that a general biology fluency is sufficient.

I say this because, at this point, I actually did not know what the relationship between the GABA-A receptor and tremor, seizure, epilepsy, etc. was. But what I did know from the ‘584 patent was that such a relationship existed. That was all that mattered for reeling SAGE-324 back into the list of possible contenders.

Anyway: because SAGE-324 gave me something, I began my chase there. I jumped to the last page of the ‘584 patent, which contains the international search report (ISR).

This gave me the related composition of matter patent (WO 2015/180679) and the exact example (example 47, compound SI-4) that matched Compound 1 in the ‘584 patent. So I went there.

Good—it matches Compound 1 from ‘584. COM patents almost always give in vitro data and, if I’m lucky, some in vivo data too. Today I was lucky and got both. But what actually matters here was not the in vivo data itself, but the fact that SI-4 was dosed PO.

So the compound in question is an oral drug. Going back to our list of contenders…

• SAGE-689

• SAGE-319

• SAGE-324

…we can rule out SAGE-689. Why? Because it’s an IM drug.

Catch of the day

So now we’re down to either SAGE-319 or SAGE-324. I stayed the course with SAGE-324, since there was more info.

Moving on to SEC filings. I looked through Sage’s February 2023 SEC 10-K Annual Report. Not much in the way of showing data (sometimes you can get lucky). But I found these interesting bits of info:

Crystalline solid. Perhaps an allusion to the title of the ‘584 patent?

Read very carefully. They have issued COM patents on SAGE-324 that expires in May 2035. They also have US and foreign patent applications covering SAGE-324, SAGE-319 etc.

This distinction is important because the line states the COM patent covering SAGE-324 is OFFICIAL (aka, not just an application). It also expires MAY 2035. 2035 - 20 years (patent lifespan) = 2015. Remember this?

Looks like we might have something! The COM patent describing SI-4, aka Compound 1 from ‘584. Published December 3, 2015 but FILED May 29, 2015. Dates match up. SAGE-324 is now my front-runner.

What’s cookin good lookin

After some more meandering, I decide to explicitly search “sage therapeutics” in the EPO database. Just to see if I’ve missed anything that could be related to the compound in question. And indeed, I come across another patent—this time, a methods of use one (WO 2022/221195).

Oh hey, the same Compound 1 as Compound 1 in the ‘584 patent! Methods of use patents are super useful pickups. Often, they have animal data. But if you’re extra lucky, you’ll get actual clinical data. Once again, I am lucky because we have clinical data!

The very first example describes “A Phase 2, Double-Blind, Placebo-Controlled, Randomized Study Evaluating the Efficacy, Safety, and Tolerability of Compound 1 in the Treatment of Individuals with Essential Tremor.”

I scanned the SAGE-324 trials listed on trials.gov for anything similar. Lo and behold…

NCT04305275

…just sub “Compound "1” for “Sage-324”!

Cool find. But this alone does not prove anything. Let’s see if the rest of the trial details described in the (‘195) patent line up.

Dose checks out ✅

Primary endpoint checks out ✅

Participant number checks out ✅. This one was a strong push in the right direction just because the number was very specific.

Okay nice. There’s a high likelihood that this trial (NCT04305275) is the same one that’s being referred to in the patent.

One more check: are the patent filing dates plausibly consistent with the trial dates?

Trial was completed March 3, 2021. The treatment patent application, which contains the the (likely) matching clinical data was filed on April 12, 2021. So, entirely plausible. Good.

Since this trial was completed a while back now, I tried to see if any results had been posted. I mostly just found press releases with the trial highlights. No data. But, I did find the NCT04305275 trial had a name—KINETIC.

From Practical Neurology

This was helpful because in the course of my rummaging, I had browsed several Sage decks referring to the KINETIC trial data. They hadn’t tied an NCT number to this KINETIC trial in the slides, so I couldn’t do much with that information. But now that I knew that ‘195 treatment patent clinical data = NCT04305275 = KINETIC, I could make some useful connections.

From Sage’s 3Q2022 Investor Presentation (November 2022)

Here, they present clinical data from the phase 2 KINETIC study. There aren’t any graphs in the ‘195 treatment patent, so I don’t zero in on the graphs first. Instead, I look at the AE profile and compare the %s to what I see in Table 12 of the ‘195 treatment. For the seven most common AEs, the %s are the same!

At this point, I’m ready to make my final judgement. But just for fun (lol), why not just plot the ‘195 patent data and see if lines up with their investor slides?

So the ‘195 patent has a lot of data (13 tables, to be exact). The graphs investor slides describe:

1. Change from baseline for TETRAS performance subscale upper limb tremor total score in SAGE-324 and placebo treatment groups

2. Change rom baseline for TETRAS ADL subscale total score in SAGE-324 and placebo treatment groups

So I gotta grab these values from the patent. Slight problem: a lot of these tables have similar headings as these graphs and I was too tired to parse the nuances.

To make it easier for myself, I just noticed that for graph 1 (TETRAS performance), there were asterisks (p < 0.05) only on days 8 and 29. For graph 2 (TETRAS ADL), there were asterisks (p < 0.05) only on days 8, 15, and 22. So I kept an eye out for these distinguishing features in ‘195 patent tables.

Tables 3 & 9 match Graphs 1 & 2, respectively. Time to graph.

My graphs of the patent data are on the left. On the right are the company slides. X-axes are a bit off, but you can see that the shape of the graphs are the same.

Ding ding ding. Looks like we got a winner!

Compound 1 = SAGE-324.

Phew. Put in more elbow grease in this than I expected. But a satisfying conclusion, I think.

For those following this field, you will notice that SAGE-324 is quite similar to zuranolone (SAGE-217/BIIB215). Hm, wonder what that means.

This one was fun—hope you enjoyed! See you in the next one. Peace out.

Have a compound that you want me look into? Suggest a compound here.