Worldwide investors commit millions of dollars to drug companies, hoping that a new drug will become a blockbuster. However, before committing, investors should consider a thorough scientific due diligence to reduce risk. On May 7th, Actinogen, a small-cap Australian company, announced that their phase-2 trial of Xanamem showed no cognitive improvement in Alzheimer’s disease (AD)patients. Below is an analysis of the scientific rationale for Xanamem’s trial, which identifies issues that could explain the recent and future failures of Actinogen’s Xanamem.
Xanamem is a small-molecule inhibitor of the 11β-hydroxysteroid dehydrogenase type 1 ((11β-HSD1)). Xanamem blocks the regeneration of inactive cortisone into active cortisol within cells. Previous trials ran by Abbott ((aka AbbVie)) and Corcept with drugs that inhibit 11β-HSD1 or that block cortisol signaling have been unsuccessful, however, Actinogen’s phase-2 Xanamem trial differed in keyways.
Abbott’s 11β-HSD1 inhibitor ABT-384 failed for futility, but the patients treated had moderate to severe AD, (MMSE scores ranging from 10-24), whereas Xanamem patients had mild cognitive impairment (NYSE:MCI) (MMSE scores 20-26). Abbott reported full brain inhibition of 11β-HSD1 by ABT-384, but Abbott’s calculation was based on the assumption that brain-regenerated cortisol could be measured in cerebrospinal fluid (NASDAQ:CSF). This assumption was incorrect because 11β-HSD1 functions in the brain as an intracellular amplifier of cortisol and the regenerated cortisol is not exported from the brain parenchyma.
Moreover, the CSF concentration of ABT-384 was 0.3-1% of total plasma level, whereas Xanamem CSF was 7.5-11%. The mean CSF concentration with 50-mg doses of ABT-384 was 2.3 ng/ml ((after 5 daily dosages)), whereas the mean CSF concentration with 35-mg doses of Xanamem was 69.8 ng/ml ((after 4 daily dosages)). The inhibitory constant ((Ki)) of ABT-384 was 5-10 nM, and Xanamem’s was 22 nM. Although ABT-384 had a lower Ki, it appeared that brain penetration, along with including patients with moderate to severe dementia could have impacted the efficacy of ABT-384. Differences in brain penetration may be attributable to chemical differences shown below:
Cortisol regulation can be inhibited with glucocorticoid receptor antagonists. Corcept Therapeutics terminated a trial of glucocorticoid antagonist mifepristone ((RU486)) after patients ((MMSE scores 18-27)) showed no treatment effect after 4 months. It has been speculated that low enrollment contributed to futility. However, chronic RU486 treatment results in significantly increased cortisol levels due to feedback inhibition. Also, considering that an effective dosage for Cushing’s disease can reach 1200/mg day, the 300mg trial dosage may not have blocked enough glucocorticoid receptors. Furthermore, RU486 may have been ineffective because it is a potent progesterone receptor antagonist, and progesterone receptor signaling has been shown to be neuroprotective in animal models of stroke.
Given the explanations for previous trial failures, Actinogen considered it worthwhile to explore this approach again, especially since cursory research seems to support the notion that chronically elevated cortisol causes cognitive damage and increased cortisol predicts progression from mild cognitive impairment to AD. Although cortisol is well known for its immunosuppressive effect, depending on dose, length of exposure, and tissue/cell type, cortisol can also be pro-inflammatory and even neurotoxic ((as in Cushing’s disease)). Indeed, high cortisol effects are associated with AD, such as altered glucose uptake and utilization in the hippocampus, modified glutamate transmission, increased β-amyloid production, increased GSK3β-mediated tau phosphorylation, and destabilized microtubules.
Blocking cortisol regeneration with acute treatment of 11β -HSD1 inhibitor significantly improved cognitive performance in mice and rats, and long-term inhibition of 11β -HSD1 in transgenic mice overexpressing a mutant form of amyloid precursor protein significantly improved fear-conditioned memory. The inhibition of 11β -HSD1 increased verbal fluency in elderly men and increased verbal memory in patients with type-2 diabetes. With this scientific evidence indicating that cortisol reduction through 11β-HSD1 inhibition might be effective in improving cognitive function, why didn’t Xanamem work for AD?
First, inefficacy could stem from increased cortisol levels during the nighttime low ((nadir)). Actinogen reported that Xanamem did not significantly elevate plasma cortisol in subjects. However, they did not sample nighttime cortisol. Glucocorticoid signaling follows a circadian rhythm, which is altered in an aged rat hippocampus, and significantly correlates with up-regulated 11β -HSD1 expression during the inactive phase ((in old rats compared to young)). In AD, plasma cortisol levels are abnormally high during the nadir ((2am -9am)) and 11β -HSD1 inhibition might reduce the nighttime cortisol spike. Instead, nadir cortisol is elevated in 11β-HSD1 knock-out mice, although interestingly, this apparently depends on the genetic background, since Seckl’s group reported no cortisol elevation at nadir in C57BL/6J 11β-HSD1 knock-out mice. An increase in nighttime cortisol with 11β-HSD1 inhibition could negate any benefit in AD.
Second, young 11β-HSD1 knock-out mice show increased inflammation and impaired resolution in inflammatory models. Macrophage phagocytosis is reduced from 11β-HSD1 knock-out mice or in the presence of 11β-HSD1 inhibitor. While 11β-HSD1 deficiency increases peripheral inflammation, there are several reports of an opposite effect in the brain. Mice deficient in 11β-HSD1 had reduced inflammatory response in the hippocampus after exposure to a potent peripheral inflammatory stimulus. Moreover, blocking glucocorticoid signaling in an aged brain reduced hippocampus microglia activation to inflammatory stimulus. The activation of microglia and their release of inflammatory molecules such as reactive oxygen species are thought to promote degeneration. Although dampening microglia activation by 11β-HSD1 inhibition might slow cognitive decline, it’s also possible that some inflammatory molecules are beneficial, nonetheless, reduced phagocytosis is probably not desirable in AD.
Third, cortisol binds the glucocorticoid receptor at the membrane or in the cytosol to produce intracellular, genomic, or mitochondrial effects. Cortisol’s effect depends on the expression of multiple glucocorticoid receptor isoforms, which changes with age. A study using dietary restriction cast doubt on the potential of 11β-HSD1 inhibition. Dietary restriction is known to increase neurogenesis, plasticity, and neuroprotection, but paradoxically increases cortisol and 11β-HSD1 expression in the aged mouse hippocampus. However, long-term dietary restriction in aged mice changed the glucocorticoid receptor isoform expression pattern to that resembling young mice. This study indicates that glucocorticoid receptor isoforms are important to maintaining cognitive health, not low cortisol levels.
Finally, there is a rare allele for 11β-HSD1 that increases AD risk 6-fold, and yet the sequence polymorphism decreases 11β-HSD1 transcription in an in vitro assay, (although the effect could be cell-specific or even opposite in vivo). However, because of technical difficulty it is unknown whether 11β-HSD1 activity is truly upregulated in an AD brain. NADPH is a cellular reducing agent and is required for 11β-HSD1 reduction. NADPH could be in short supply in AD brain regions. For instance, NADPH also provides the reducing power for the antioxidant’s glutathione peroxidases and thioredoxin reductases, and NADPH would be in high demand under oxidative stress. Although they are reduced in different locations, both enzymes depend on redox power of the pentose phosphate pathway (NYSE:PPP) substrate, glucose-6-phosphate. Accordingly, the PPP is upregulated in AD brain regions.
In summary, Xanamem recent trial futility may have resulted from elevated nighttime plasma cortisol, altered microglia activation, reduced phagocytosis, or inhibiting an enzyme that is already downregulated. Glucocorticoid action has complex systemic and local regulation, multiple cellular targets with fast and slow response, changes with circadian rhythm and aging, and regulates unique sets of genes in different cell types. Sporadic AD is also complex, with unknown etiology and mixed pathology. In view of issues raised above, and the complexity, the unknown glucocorticoid physiology and pathogenesis of AD, the futility of reducing intracellular cortisol regeneration in MCI and AD patients was somewhat predictable. Further, it is unlikely that increased Xanamem concentrations or a longer treatment duration would be effective.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.