"We know there's a lot of interest both from doctors and patients in a much more convenient option for Alzheimer's. If we are indeed, as the pilot data suggests, five times better than these antibodies, I don't think anybody's going to take the antibody."
Steven Gourlay trained as an internal medicine physician, earned a PhD in clinical trials and biostatistics, did a postdoc at UCSF, spent six years at Genentech, moved to the investor side, and then became CMO of Principia Bioscience — a startup that took an autoimmune drug from pre-clinical through Phase 3 before Sanofi acquired it for several billion dollars. He moved back to Australia for family reasons, imagined semi-retirement, and ended up as CEO of Actinogen because the molecule was too compelling to ignore.
That molecule is Xanamem. It's a once-daily pill. Its mechanism targets an enzyme called 11-beta HSD1 that manufactures cortisol from inert cortisone in tissue — specifically in the brain, liver, and fat. By inhibiting this enzyme at low doses, the drug reduces cortisol in brain tissue without touching adrenal cortisol production, which is essential for normal stress response. It's the first brain-penetrant molecule ever developed to do this.
This conversation covers the cortisol hypothesis for Alzheimer's and depression, what the clinical data actually shows, why Actinogen is prioritizing Alzheimer's over depression, and why the field spent $42 billion going in what Gourlay believes is the wrong direction.
The Cortisol Hypothesis
The scientific foundation here is decades old. Elevated cortisol in the brain has long been associated with cognitive decline, Alzheimer's risk and progression, and the severity and incidence of depression. There are tens of thousands of published papers on the connection. What has been missing is a drug capable of selectively inhibiting brain cortisol while leaving the body's normal cortisol function intact.
Gourlay walked through three lines of evidence that convinced him the hypothesis holds. First, implanting cortisol directly into mammalian brain tissue causes neuronal dieback around the implant site. Second, long-term therapeutic use of prednisone — a cortisol analogue used to treat inflammation — produces well-documented side effects: depression, mood swings, cognitive impairment, bone thinning, elevated glucose. These are not placebo effects. Third, a significant proportion of patients with severe depression show a blunted dexamethasone suppression test, indicating dysregulation of the cortisol system.
Previous attempts to block related pathways had either failed on safety grounds or been deliberately designed to stay outside the brain — because the clinical target was type 2 diabetes, not neurodegeneration. Actinogen's molecule crosses the blood-brain barrier. No one had done this before. Gourlay joined because the data on binding and safety were the best he'd seen in 35 years of drug development, and because the only way to validate a pathway is to block it.
The Clinical Data
The sequence of trials Actinogen has run is methodical. PET imaging confirmed high-level binding to the target in brain regions of interest at doses as low as 5–10 milligrams — 60 to 80% inhibition across virtually all relevant brain areas. Animal models show cognitive protection at 30–60% inhibition, well below what the drug achieves. A Phase 1 trial in healthy older volunteers showed improvement in attention and working memory over 12 weeks, an unusual result in a cognitively normal population.
The most significant result came from a reanalysis of stored plasma samples from an earlier Phase 2 Alzheimer's study — one that had originally read as negative because the trial was too short to show disease modification. When Gourlay's team went back with modern blood biomarkers not available at the time of the original trial and looked specifically at patients with elevated p-tau181 — a marker indicating both Alzheimer's pathology and progressive disease course — the treatment effect was large. In that biomarker-positive subgroup over just 12 weeks, the drug showed roughly five times the effect size of the approved amyloid antibody therapies over 18 months.
A subsequent prospective trial in patients with both cognitive impairment and major depressive disorder showed clinically significant improvement in depression symptoms over six weeks, providing the first direct evidence of effect on depression itself at a 10 milligram dose.
Alzheimer's or Depression: Where to Focus
Actinogen is prioritizing Alzheimer's. The reasoning is straightforward: the unmet need in Alzheimer's is greater, the approved standard of care is weaker, and the regulatory bar — set by the amyloid antibodies — is measurably low.
The existing approved therapies fall into two categories. The older drugs — donepezil, memantine — modestly rev up neurotransmitters without changing disease course. The newer amyloid antibodies, including lecanemab, have been approved on a treatment effect of roughly half a point on the 18-point CDR-SB scale over 18 months. One point is considered the minimum clinically significant difference. They require IV infusion every two to four weeks, carry a real risk of brain swelling, and their cost and access profile limits uptake. Actinogen's pilot data, if it replicates in the confirmatory study, would represent a treatment effect several times larger, delivered in a once-daily pill with a clean safety profile across more than 400 patients treated to date.
Depression remains a compelling indication — the prospective trial data is real — but it will wait for partnership funding or grant support. The Alzheimer's trial comes first.
Patient Selection and Trial Design
One of the central challenges in Alzheimer's trials is noise. The primary endpoint — CDR sum of boxes — is an 18-point scale assessed through structured clinical interviews. Different raters, patient variability across visits, and the subjectivity inherent in the instrument all add noise that has contributed to many failed trials. Gourlay's approach involves rigorous rater training, standardized assessment environments, and in some cases centralized rating to keep assessment consistent for the same patient across visits.
Patient selection is equally important. The goal is patients with early symptomatic Alzheimer's who will progress meaningfully over 36 weeks — not so mild that progression is too slow to detect, not so advanced that a treatment effect becomes impossible to show. Elevated p-tau181 in blood has turned out to be the right filter: it confirms amyloid pathology and flags progressive disease. It's the same marker that drove the dramatic subgroup result in the earlier trial.
The current pivotal Phase 2b/3 study is ongoing in the US and Australia. Interim analysis is expected roughly six months after the hundredth patient is enrolled. Final results for 220 patients are targeted for late 2026.
The Amyloid Question
Gourlay is direct about the field's trajectory. About 80% of new NIH grants are now directed at non-amyloid targets. The scientists at large pharma companies, he says, largely agree that the pathway has been exhausted — even at companies whose marketing departments are still promoting the approved antibodies. The approved therapies removed amyloid efficiently and produced modest clinical benefit. If removing the protein almost entirely doesn't change the course of disease meaningfully, reducing its formation is unlikely to do better.
He sees neuroinflammation, metabolic disruption, and cortisol as the more promising areas. The APOE4 gene — which codes for a lipid carrier protein in neurons and is the major measurable genetic risk factor for Alzheimer's — points toward metabolic and inflammatory mechanisms rather than amyloid directly. The GLP-1 class of drugs showed a lower incidence of Alzheimer's in diabetic populations and has randomized trials underway, but their GI side effects and weight-loss mechanisms are poorly suited for the typical Alzheimer's patient, who is already underweight and often struggles to maintain adequate nutrition.
Actinogen's view is that if the confirmatory study replicates the pilot data, Xanamem and the GLP-1 class will be the two most viable approved therapies in Alzheimer's in the near term — and that no one is going to choose an infusion over a pill when the pill has a larger effect size and a cleaner safety profile.
