<Long post warning>. My gut feel a few months ago, which I have no qualification to be giving, was that the odds of passing endpoints in November was 50:50. Today, AI casually told me it was 1:7, which made me pause. So I decided to have a serious chat, in which AI did most of the talking and I pretended to understand all the big words. But I got what I wanted and I honestly don't think the AI knew what my earlier guess was although...wait...did it just tell me what I wanted to hear?....

A few of you follow pharma and biotech so you might find the output interesting. Anything missing?

Q. What approaches would sophisticated funds and big pharma take to assess the odds of success of the XanaMia trial, now that it is past interim futility and is half way through the OLE? The Actingen CEO alluded to asking for the harder futility test to be applied by the reviewers. Does that shift the needle in any way?

Logically go step by step from the broadest averages, and then adjust for all relavant additional information specific to xanamem and this trial. Show your working down from a broad average to specific odds for the upcoming readout in November.

Here's the summary table

And here's the very long answer.

A. Sophisticated funds and big pharma would treat this as a Bayesian / decision‑tree problem rather than “one magic number”, but it is possible to walk from broad industry priors down to a Xanamia‑specific range of odds. Below is one way they might logically do it, including how a “harder” futility test and OLE uptake might shift probabilities. This is inherently judgment‑heavy; the numbers are illustrative, not precise.

1. Start with broad Alzheimer’s trial base rates

From 2020–2023 reviews of AD programs, phase 2 and 3 disease‑modifying trials historically had low success rates, especially before the recent antibody wins. The share of phase 2 AD drugs reaching approval was in the single digits, with phase 3 success rates on the order of 10–20% in many analyses.pmc.ncbi.nlm.nih+1

For a “typical” phase 2b/3 disease‑modifying AD trial using CDR‑SB as primary endpoint (like lecanemab/donanemab programs), a neutral, pre‑information prior for success might be something like:

• 10–20% probability of hitting a clinically and statistically meaningful CDR‑SB effect.
• This is before adjusting for mechanism, enrichment, trial design quality, and any interim data.

So a sophisticated fund might start with ~15% as a crude base rate for “phase 2b/3 AD trial reaches its primary endpoint and is viewed as a true win”.

2. Adjust for modern design and pTau181 enrichment

XanaMIA is not an average legacy AD trial. Two big positives:

1. Biomarker enrichment (plasma pTau181):
2. The XanADu biomarker analysis showed that patients with elevated plasma pTau181 (the “high” group) progressed clearly faster than low‑pTau patients on CDR‑SB and other measures, and that in this enriched group there was a potentially meaningful Xanamem effect (LS mean CDR‑SB difference ~0.6 units, effect size ~0.41, p≈0.09) over 12 weeks.pubmed.ncbi.nlm.nih+1
3. This supports:

• Better “signal‑to‑noise” because progressors are selected.
• Prior evidence that this drug can move CDR‑SB in the right direction in the biomarker‑positive population.

1. Modern endpoints and trial design:

• CDR‑SB is now the dominant global function endpoint in early/mild‑moderate AD and has been accepted for recent mAb approvals.pmc.ncbi.nlm.nih
• XanaMIA follows a 36‑week, randomized, placebo‑controlled design in 247 patients with mild–moderate AD and elevated pTau181, across AU/US sites, described as one of two pivotal trials aligned with FDA feedback.ACW-HY26-results-XanaMIA-pivotal-Alzheimers-trial-progress.PDF+1actinogen

Relative to the “average, older‑style” AD trial, funds and pharma would typically up‑weight success odds for:

• Enriched, biomarker‑positive population that has already shown both strong progression and a trend to drug benefit.
• Alignment with contemporary regulatory thinking (CDR‑SB, disease‑modifying framing, etc.).

It would be reasonable to bump from ~15% to somewhere in the 20–25% range purely on trial design and biomarker enrichment, before looking at any drug‑specific data.

3. Adjust for Xanamem’s mechanism and prior clinical signal

Key Xanamem‑specific positives:

• Validated mechanism and CNS exposure:
• Xanamem is a selective 11β‑HSD1 inhibitor designed to lower cortisol production inside the brain. PET and pharmacology work showed high CNS target occupancy at 5–30 mg/day, and endocrine readouts (ACTH, cortisol dynamics) consistent with potent 11β‑HSD1 inhibition, with acceptable systemic cortisol effects at 10 mg.Jan-2026-ACW-positive-XanaMIA-Interim-Analysis-outcome.PDFactinogen+1ACW-HY26-results-XanaMIA-pivotal-Alzheimers-trial-progress.PDF
• Prior AD signal:
• In biomarker‑positive mild AD (elevated pTau181), Xanamem previously slowed CDR‑SB worsening vs placebo over 12 weeks, with an effect size in the small‑to‑moderate range.journals.sagepub+2
• Depression trial (XanaCIDD):
• In 2024, the 6‑week phase 2 depression trial showed clinically and statistically significant MADRS improvements and positive patient‑reported severity measures vs placebo, reinforcing the CNS cortisol‑control mechanism in another population.firstwordpharmaJan-2026-ACW-positive-XanaMIA-Interim-Analysis-outcome.PDF+1
• Safety:
• More than 500 individuals have now received Xanamem up to 36 weeks with no reported serious adverse events attributed to the drug, which supports the ability to dose long enough to see disease‑modifying effects.Actinogen-March2026-quarterly-4C.PDF+1

On the other hand, the magnitude of prior AD effect was modest (trend in a subgroup), and cognition in XanaCIDD improved similarly in both arms (placebo strong), so it is not an obviously “home‑run” profile.

A sophisticated buyer typically:

• Further steps up the probability because there is some prior disease‑relevant signal, in a biomarker‑enriched subgroup, not just target engagement.
• But does not jump to antibody‑like odds because effect size and duration are shorter and there is no prior large successful pivotal.

Reasonable adjustment:

From 20–25% (design/enrichment) to perhaps 25–35% based on mechanism + prior signal.

4. Incorporate the interim futility analysis outcome

4.1 What the DMC saw

The independent DMC reviewed unblinded safety and efficacy (for futility) on ~37% of the planned final dataset: 136 participants with at least one efficacy data point and 52 who had completed the full 36 weeks. The recommendation was to continue the trial “without amendment” after an earlier positive safety‑only review.Jan-2026-ACW-positive-XanaMIA-Interim-Analysis-outcome.PDF+2

In Bayesian terms, “passed futility” means:

• The data at ~37% completion were inconsistent with the trial being clearly futile under the pre‑specified futility boundary.
• That boundary, if more stringent (“harder futility test”), will itself shape how informative this decision is.

4.2 “Harder futility test” – what does that mean?

In practice, a “harder” futility test usually means a more aggressive stopping rule (e.g., higher conditional power threshold) such that:

• If the emerging effect size is modest and unlikely to reach significance at N=247, the DMC is encouraged to stop early.
• Only trials with plausible effect sizes consistent with success at full enrollment are continued.wikipedia+3

The Actinogen disclosures and your comment that the CEO alluded to asking for a harder futility test are consistent with:

• Sponsors pushing for a low tolerance for mediocre signals; i.e., they would rather the DMC tell them it is futile than spend money on a long‑shot.
• Given that the DMC, under that stricter rule, chose not to stop for futility, the posterior probability of ultimate success is meaningfully higher than the prior.

There is no published futility boundary, so one cannot quantify this exactly. But experience across oncology and neurology suggests that for a trial with aggressive futility rules:

• Passing futility at ~⅓–½ information can easily double the market’s fair probability, because the scenario where the true effect is “near zero” is heavily pruned.

From 25–35% pre‑interim, a disciplined fund might move to something like 35–50% post‑interim, with the center of gravity perhaps around 40%, acknowledging:

• The DMC did not say “overwhelming efficacy”, merely “not futile”.
• Some trials still fail after passing futility, especially if the emerging effect is borderline.

5. Open‑label extension (OLE) uptake as a soft signal

The OLE began in March 2026, with the first participant treated 31 March, and early data showed 15 of 17 (88%) eligible participants rolling over into the OLE within the first month.Actinogen-March2026-quarterly-4C.PDF

Funds and pharma will see OLE uptake as:

• A strong safety signal: Participants and investigators are comfortable continuing on active drug for up to 25 months.Jan-2026-ACW-positive-XanaMIA-Interim-Analysis-outcome.PDF+1
• A maybe‑weak efficacy signal: If the drug were clearly hurting, investigators might discourage OLE participation, but with AD and no other curative options, high OLE uptake is expected even with neutral efficacy.

Given AD context, 80–90% OLE uptake is not enough to materially change efficacy odds by itself. It likely confirms:

• No obvious catastrophic safety issue emerging.
• Engagement of sites and patients, which reduces operational risk and missing data but does not say much about CDR‑SB separation.

So sophisticated assessors might tweak probabilities marginally up for “no obvious negative signals surfacing”, but not more than a couple of percentage points.

6. Overlay current AD competitive landscape

Compared to anti‑amyloid antibodies:

• Lecanemab’s CDR‑SB separation at 18 months was ~0.45 units (27% slowing) with baseline CDR‑SB ~3.2.eisai+3
• Donanemab and others showed similar 20–30% slowing in early AD.pmc.ncbi.nlm.nih+3

These set the bar for “clinically meaningful” effect size: regulators and payers now expect at least modest slowing on CDR‑SB.pmc.ncbi.nlm.nih

For a first‑in‑class oral cortisol‑modulator:

• A smaller effect than antibodies might still be commercially compelling (especially if safe, cheap, easy to administer, combinable).
• Statistically, however, XanaMIA still needs a detectable and robust CDR‑SB effect at 36 weeks.

Given the earlier 12‑week effect around 0.6 CDR‑SB in the enriched subgroup, one can reasonably hypothesize:pmc.ncbi.nlm.nih+2

• If that effect sustains or grows modestly over 36 weeks, the trial could hit.
• If the effect attenuates with time or the pTau181 enrichment plus trial noise reduces signal, it may miss.

This context probably keeps a sophisticated analyst from moving probabilities above ~50%, because:

• There is no prior large, clearly positive AD trial; everything rests on one biomarker‑subgroup signal + mechanism.

7. Putting it together – worked‑through “house view” range

Pharma BD and sophisticated funds might end up with ranges rather than a single number. A reasonably defensible step‑down could look like this:

<summary table I moved to the top>

Where a pragmatic central estimate could easily cluster in the 35–45% band for “meeting the primary endpoint in November,” with a bull‑case investor perhaps leaning toward 50%+ and a bear staying closer to 30–35%.

This aligns with the Space instructions: the bull case focuses on:

• Enriched biomarker‑positive population and prior Xanamem signal in that exact setting.pubmed.ncbi.nlm.nih+2
• Validated mechanism and multi‑indication evidence (AD and MDD) with a good safety profile.pmc.ncbi.nlm.nih+1ACW-HY26-results-XanaMIA-pivotal-Alzheimers-trial-progress.PDFpmc.ncbi.nlm.nihActinogen-March2026-quarterly-4C.PDF+1
• A positive, stricter futility read, which strongly de‑weights the “true effect ≈ 0” scenario.pubmed.ncbi.nlm.nih+3Actinogen-March2026-quarterly-4C.PDF+1

The bear case highlights:

• Only modest prior effect sizes and short‑duration data, not a prior pivotal win.journals.sagepub+1
• History of failure of most non‑amyloid AD mechanisms, even with reasonable biology.explorationpub+3
• The fact that passing futility is a necessary, but far from sufficient, condition for success; many such trials still fail.fda+2

Crucially: the CEO’s comment about asking for a harder futility test does reasonably “shift the needle” relative to a standard or soft futility rule. If the rule was genuinely stringent, a continuation decision at ~37% information justifies moving from, say, ~25–30% into the high‑30s or low‑40s as a central probability, but not into the 70–80% territory that you might reserve for a drug with a prior clear pivotal win.

From Monday's presentation, this slide shows a possible trend over 12 weeks for biomarker-positive patients in the phase 2a Alzheimers trial (n=34). If this trend is proven out for the current larger phase 2b/3 trial (n=245), and extends out to 36 weeks, then we have a winner.

Because this bio-marker group (n=34) was a subset of the overall trial (n=72) the above result is not statistically significant. It is nonetheless a clue that the current trial is designed and powered to tease out, because all 245 patients were screened to have this bio-marker, and AI tells me that that should be enough for a significant result. Also, the CEO has 30 years of experience including at FDA, the team is high powered, and they appear to know what they are doing.

Actinogen just passed halfway and the interim futility analysis for their phase 2b/3 trial. It’s a novel Alzheimers drug that reduces brain cortisol and may be disease modifying.

They then revealed a pre-arranged $12m raise and offered a further $5m in shares to existing shareholders that altogether will carry them past the final data readout in Nov 2026. This will be a binary event with big pharma watching.

The CEO tipped in another $500k, taking his total initial investment to $2.5m. No options were required. The current MC is $175m.

At yesterday’s webinar, the CEO mentioned that they had asked the independent reviewers to apply a relatively hard statistical futility test to the data. While I am not an expert in statistics, he implied that this means there would be a bit higher probability that the drug candidate curve is pulling ahead of the placebo curve at this stage (as expected), than if a softer futility test had been applied. The test is on the sum of a standard range of psychological and functional indicators for Alzheimers. They are confident that their drug has already shown effectiveness in earlier data, and it seems they made a calculated bet here to show confidence to the many pharma companies watching.

The open-label extension phase will also start soon as patients who have completed their 36 weeks on the double-blind trial can elect to roll onto it.

An important thing to note about odds of success is that in this trial, they used their previous trial data and advances in p-tau biomarker testing to select a trial group that is more likely to really have Alzheimers, they have an extra biomarker beyond just a clinical diagnosis. Analysis of their previous trial data showed that this sub-group showed a clearer response to their drug candidate verses placebo, than the broader trial group did.

November will be a binary outcome for Actinogen. If the final data shows effectiveness, a deal will be made within a few months with a big pharma to take the drug to a larger but otherwise identical phase 3 trial. Given that the Alzheimers market does not currently have a satisfactory drug, this novel once-a-day drug with excellent safety profile would be a no-brainer (no pun intended).

It also has potential as a depression drug (pending further trials), and the Alzheimers patients might show indications of this as time goes on. So a pass in November leads to the deep pockets who can opt to explore that.

I am betting the odds of success are more than than 50:50 here, and while the downside is close to zero the upside is large, which is why I will be adding a bit in the SPP available to existing holders at 42c. (Disclaimer, trials can go wrong for unknown reasons, I’m a novice).

Actinogen are now ramping up their phase 2/3 XanaMIA Alzheimer's trial, with interim data expected in Q3 CY2025, following the interesting results from the depression trial. This year has a lot of potential catalysts, that may be enough to bring in a partner to take it to the next level.

Here's an IR presentation where the CEO lays out the state of play, and why it might be a step up from the current Alzheimer's drugs, which are a bit controversial. The more I hear from him, the more I feel it's a good bet (I do have a small stake).

They are going for the holy grail - a safe, once a day pill for the modification of Alzheimer's and then also depression. It's a pill that might be able to be safely paired with other medication. It's a novel mechanism of action - reducing cortisol in the brain (while not affecting the normal hormonal function). It hasn't shown the kind of side-effects associated with SSRIs and other established anti-depressants.

They have evidence that it improves cognition, even in healthy adults. If the trial results prove consistent, it's the kind of thing I think anyone who's interested in maintaining good mental health might take a look at in the long run.

Currently available treatments and most of the competitor trials appear to be focused on reducing brain plaque buildup. I asked Perplexity why these treatments were considered controversial, and why Actinogen could improve the situation.

Discussion of the amyloid-beta treatment controversies

Aducanumab (Aduhelm) is a monoclonal antibody targeting amyloid-beta (Aβ) plaques in Alzheimer’s disease (AD), controversially approved by the FDA in 2021 under accelerated approval pathways. Here’s an analysis of its trajectory and Actinogen Medical’s potential role:

Aducanumab: Current Status

What Works

13. Amyloid Plaque Reduction

• Demonstrated 59%-71% amyloid plaque clearance on PET scans in phase 3 trials (EMERGE/ENGAGE), with dose-dependent effects.

• Biomarker data showed reductions in plasma p-tau181, a downstream tau pathology marker.

14. Regulatory Precedent

• First amyloid-targeting drug approved since 2003, paving the way for newer anti-amyloid therapies like lecanemab and donanemab.

15. Subgroup Benefits

• In EMERGE, high-dose aducanumab showed a 22% relative reduction in cognitive decline (CDR-SB score) over 18 months vs. placebo.

Challenges

Issue Impact

Divergent Trial Results EMERGE (positive) vs. ENGAGE (neutral) raised validity concerns; post-hoc analyses deemed statistically fragile.

Safety Risks 35%-41% incidence of amyloid-related imaging abnormalities (ARIA-E/ARIA-H).

Cost & Access Priced at $56,000/year, with Medicare restricting coverage to clinical trials due to uncertain benefit.

Clinical Relevance Amyloid reduction ≠ cognitive improvement: CDR-SB difference of 0.39 points (vs. ~1.5-2.0 MCID).

Actinogen’s Position

Actinogen’s emestedastat (Xanamem) represents a paradigm shift from amyloid-centric approaches:

16. Mechanistic Advantage

• Targets brain cortisol via 11β-HSD1 inhibition, addressing neuroinflammation and synaptic dysfunction linked to Aβ/tau pathology.

• Phase 2 trials showed cognitive improvements in AD patients without amyloid-related safety signals.

17. Strategic Differentiation

• Oral administration vs. aducanumab’s IV infusions.

• Targets broader neurological conditions (e.g., depression, Fragile X) beyond AD.

18. Clinical Progress

• XanaMIA phase 2b/3 trial (n=220) uses pTau181 biomarker stratification, with interim results expected Q4 2025.

Key Debates

• Amyloid Hypothesis Validity: Aducanumab’s approval reignited skepticism about Aβ’s causal role in AD.

• Regulatory Trust: FDA’s accelerated approval faced criticism for prioritizing biomarkers over clinical outcomes.

• Market Dynamics: Lecanemab/donanemab now dominate amyloid-targeting pipelines, while Xanamem offers a cortisol-focused alternative.

Conclusion

Aducanumab’s legacy lies in accelerating AD drug development despite its commercial and scientific setbacks. Actinogen aims to supersede it by addressing cortisol dysregulation—a mechanism orthogonal to amyloid—with a safer oral therapy. While amyloid remains a therapeutic pillar, the field increasingly recognizes multifactorial approaches as essential for meaningful disease modification.

In case anyone wants to dive in to the pharmacology. I also asked Perplexity to summarise it!

https://pmc.ncbi.nlm.nih.gov/articles/PMC11788964/

Perplexity summary:

This publication discusses the clinical development of emestedastat (also known as Xanamem), a drug being developed by Actinogen Medical for potential treatment of Alzheimer's disease and other neurological conditions. Here's a summary of the key points in plain English:

Background

• Emestedastat inhibits an enzyme called 11β-HSD1, which is involved in producing cortisol in the brain.

• High cortisol levels in the brain are thought to be harmful and may contribute to cognitive decline and Alzheimer's disease.

Study Approach

The researchers used several methods to determine the best dose of emestedastat for future clinical trials:

1. Traditional pharmacokinetic studies to see how the drug moves through the body

2. Brain imaging (PET scans) to see how much of the drug reaches its target in the brain

3. Cognitive tests to measure any improvements in thinking and memory

4. Hormone measurements to check for effects on the body's stress response system

Key Findings

• Earlier studies suggested 20 mg daily might be needed, but newer data showed lower doses of 5-10 mg may be sufficient.

• Brain scans showed the drug reaches its target in the brain at these lower doses.

• Cognitive tests in healthy older adults showed improvements in attention and working memory with emestedastat treatment.

• The drug was well-tolerated with no major safety concerns identified so far.

Conclusions

• Using multiple types of evidence helped refine the dosing strategy for emestedastat.

• A dose of 10 mg daily appears promising for future clinical trials in Alzheimer's disease.

• While no single study was definitive, the combined results support further development of emestedastat.

The researchers believe this approach of using various biomarkers and tests can help improve drug development for complex brain disorders like Alzheimer's disease, where many potential treatments have failed in the past.

A couple of introduction pieces:

• 10 min Spotify podcast on the upcoming Alzheimer's trial and science overview. October 2024.
• Stockhead article from September (before the depression trial result which was not a home run). Some good colour on the twists and turns the company has already taken, and the depression and Alzheimer's landscape.

The very basics: Chronic cortisol is toxic to nerve cells. This drug inhibits an enzyme that something something reduces cortisol in the brain.

Why was the depression trial result not a home-run?

The drug has a good safety and side-effect profile so far in hundreds of patients.

They did see clinically and statistically significant improvements in 'depression symptoms'.

But, because the placebo group also improved over the 10 weeks of the trial in the cognition tests, they cannot make any claims about cognitive improvements in depression. In case this is confusing, 'cognitive impairment' is apparently a different aspect of depression than 'depression symptoms' and they are tested independently.

I think they want to circle back for a longer depression trial with a funding partner after the Alzheimer's trial, but they would have immediately been in a better bargaining position had they hit a home-run. They want to prove out the cognition side because that would broaden the market...a lot.

Regardless, they are focused on Alzheimer's now, and if they stick to the schedule, the trial is fully funded as an independent pivotal trial (one of 2 that would be required for FDA approval).

• 220 patients, 9 months, Aust and US sites
• Interim results expected in mid-2025
• Final results mid-2026

' The patients are assessed on both cognition and Alzheimer’s progression.

“We believe we have already validated the target by showing improved cognition in healthy older volunteers and potentially a big clinical benefit in biomarker-positive patients with Alzheimer’s,” Dr Gourlay says. '

CEO Dr Steven Gourlay put $1m into an $11m capital raising. From what I can gather, he now owns at least $2m in shares of a 75m market cap.

They state this should fund trials until late CY2026.

From what I can gather Dr Gourlay has worked in biomedical research for 30 years, including at the FDA and the European equivalent. He may have already done rather well if he had skin in the game at prior ventures, but he seems to have good experience anyway.

Actinogen is entering a Stage 2b/3 trial to test their compound Xaneman over 36 weeks targeting an inhibition of brain cortisol for Alzheimer's (as well as in depression).

I got interested here after hearing him chat with Alan Kohler and seeing the volatile share price after a recent depression trial. The market decided it didn't like the trial results. Results were somewhat promising in early weeks against 'depression' and the signal strength improved between 4 and 10 weeks. But they lucked out in a test of 'cognition' because the placebo control group also saw a positive result there. Not an expert but apparently placebo groups doing well in depression trials can be a thing, because taking action of any kind tends to help (including entering a trial).

They wheeled out the chief science officer who did a series of talks and then the price recovered (only to tank again at the announcement of the raise).

I am not a Doctor but: I am tempted to believe them, the theory makes sense that reducing brain cortisol would help a lot of people (and it doesn't interfere with non-brain cortisol). I imagine (guessing) that they can't go back to test depression for a longer period since the money is spent. So they are moving forward with Alzeimer's in a longer test (36 weeks).