Actinogen Medical — Differentiated treatment for Alzheimer’s disease

Actinogen, listed on ASX and headquartered in Sydney, Australia, is a biotechnology company with a focus on its lead product Xanamem, a novel cognitive enhancer. Cognition impairment in AD patients is the primary indication in the R&D programme at present, however, based on the known mechanism of action, from Actinogen and third-party data, Xanamem has the potential to be used in multiple conditions, where glucocorticoid metabolism is linked to impaired cognition. This includes depression, diabetes, schizophrenia, Parkinson’s disease, Down’s syndrome and others. The groundwork underpinning the cortisol hypothesis and Xanamem’s discovery was achieved at the University of Edinburgh, UK, in the 1990s and 2000s. In 2014, Actinogen acquired Corticrine, which had a licence to the technology IP that included Xanamem from the University of Edinburgh. Actinogen Medical was listed on ASX in October 2007 and its original plan was to develop new generation antibiotics. However, these activities were discontinued after the licensing of Xanamem in 2014, which is now their lead development drug. Of note, CEO Bill Ketelbey, who joined Actinogen in 2014, has highly relevant drug development experience and, as a former executive at Pfizer, led the local clinical development and was involved in the commercialisation of Aricept (donepezil), the best-selling AD drug.

Our valuation, which uses a discount rate of 12.5%, only includes Xanamem in AD. We currently do not value any additional opportunities, but will revisit once Actinogen initiates the development of other potential indications. We use a probability of success of 20% for Xanamem to successfully complete the R&D. A key element of our valuation is the base case scenario that Xanamem will show some benefit compared to existing symptomatic AD drugs. We developed assumptions that reflect this and calculate peak sales of US$2.0bn. If Xanamem demonstrates disease-modifying benefit, the peak potential would likely be a multiple of our current peak sales. If Xanamem’s efficacy proves comparable to current generics, the drug could still have substantial potential with significant sales, as was the case with all current symptomatic AD drugs. Our assumptions are discussed in detail below.

The fiscal H119 R&D expense of A$6.3m was the most significant cost item, up from US$3.0m a year ago, indicating that XanADu is in full swing, with a series of additional Xanamem studies in the second half of calendar year 2019 (Exhibit 3). We forecast an FY19 total loss of A$10.5m, up from A$6.2m in FY18. FY20 and FY21 results will substantially depend on the next steps after the XanADu trial results. Actinogen had cash of A$15.5m at the end of fiscal H119 (end of calendar 2018). This is sufficient to finish the XanADu study, to initiate and complete a series of additional studies, and to prepare for Phase III trials. Additional funding needs will depend on the timing and nature of a potential licensing deal and the scale of the next trials.

The main sensitivities are around the lead candidate, Xanamem. Although it has shown efficacy in animal models and proved safe in the Phase I/II trials, it remains to be seen whether this will translate into a clinical benefit. The data from the upcoming Phase II XanADu study will shape further development of the drug. Drug development is inherently risky and could fail at any stage. If the XanADu trial results warrant further development, this could include a variety of scenarios from second line to front line, in mild or even prodromal AD patients, as a symptomatic and potentially disease modifying drug, but these scenarios are all dependent on the data readout in Q219. Late-stage drug development is capital intensive, therefore fresh funds or a partner will be needed.

Xanamem is an 11β-HSD1 inhibitor designed to reduce excessive cortisol production in the brain. An abundance of accumulated literature and existing Xanamem data point to the conclusion that excessive cortisol is associated with cognitive impairment in patients with various chronic conditions, but most evidently in neurodegenerative diseases. Currently, the main indication under development for Xanamem is AD (XanADu trial enrolled mild AD patients). The Xanamem mechanism of action is unique within the AD drug development field. The field is predominantly focused on the development of antibodies targeting misfolded proteins (Abeta, tau), which are relevant to the so-called Abeta hypothesis of AD.

11β-HSD’s role in cortisol metabolism was discovered 1950s and in later years two distinct isoforms, 11β-HSD1 and 11β-HSD2, were characterised (Chapman et al, 2013). The origins of the Xanamem compound discovery date back to 1990s and early 2000s, with research undertaken by a team of scientists at the University of Edinburgh, who explored the effect of cortisol on cognition and possible pharmacological interventions. The initial promising results in preclinical and early clinical trials were demonstrated with carbenoxolone, a non-specific 11β-HSD inhibitor. Only 11β-HSD1 is expressed in the brain, so to improve the safety profile, researchers at the University of Edinburgh developed Xanamem, a specific 11β-HSD1 inhibitor that crosses the blood-brain barrier. This research and optimisation by the inventors was supported by the Wellcome Trust with a total investment of c $25m. Clinical development of Xanamem started in 2013 and was continued by Actinogen after it acquired the assets, including Xanamem, in 2014. The Phase II XanADu trial in mild AD commenced in March 2017 and Actinogen is expected to report results by end-Q219.

The number of people living with dementia worldwide is currently estimated at around 50 million, with approximately 60% of whom have AD. Given the lack of a preventative or curative treatments, this number is set to almost double by 2030 and more than triple by 2050 (World Alzheimer Report 2018). In 2018, the direct costs to American society of caring for those with AD and other dementias totalled an estimated US$277bn, which is expected to increase to with approximately US$1.1trn by 2050 (2018 Alzheimer’s Disease Facts and Figures). Despite this unmet need, no new drugs have been approved by the FDA for AD since 2003 and the existing four drugs (cholinesterase inhibitors: donepezil, rivastigmine, galantamine and a NMDA receptor blocker: memantine) provide only symptomatic benefit, which is not sustained over the long term and does not influence patient survival.

The lack of disease-modifying therapies leaves a vast unmet clinical need. Drug development for AD has become a major political, academic and industrial effort, as evidenced by initiatives such as the Global Dementia Discovery Fund and the big pharma collaborations between Biogen and Eisai, AstraZeneca and Eli Lilly. This illustrates the scope and willingness of the industry to support the development of novel treatments in AD. However, the length and expense of clinical trials, with numerous failures to date, underline the risks associated with the field. While the risks cannot be eliminated, we believe that greater understanding of the pathophysiological mechanisms of the disease process spurred on by the substantial unmet medical need, will maintain the interest in the area.

There is a notion reiterated by executives in the industry that current late-stage clinical trials in AD, especially those involving antibodies against amyloid, are ‘based on 1990s science’. We believe that focusing on the Abeta theory, which was perceived by the industry as the best bet to develop a disease modifying therapy, meant that other technologies targeting AD were marginalised and did not break into large pharma’s mainstream interest. Some of these ‘low profile’ technologies slowly progressed into more advanced clinical testing, but we believe that these programmes still suffer from a general lack of enthusiasm from investors who recall recent high-profile failures for drugs targeting the Abeta hypothesis of AD. While this likely translates into lower company valuations relative to other disease areas where there are more success stories, it also means that if any of these ‘undiscovered’ molecules delivered positive results, the unlocked share price upside would likely be amplified.

Cortisol, also known the ‘stress hormone’, is a steroid hormone (glucocorticoid), tightly regulated via feedback loops connecting the hypothalamus, pituitary and adrenal glands. This is also known as hypothalamic pituitary adrenal (HPA) axis. Most cells in the body have cortisol receptors, so this hormone has profound effects on the body’s metabolism. For example, as a part of the response to stress, cortisol is a powerful inflammation suppressor, making glucocorticoid anti-inflammatory drugs became popular in many autoimmune diseases.

Classical understanding about chronically elevated cortisol is that it is associated with a variety of conditions characterised by various metabolic disturbances. One of the best-recognised syndromes resulting from chronically elevated cortisol is Cushing’s syndrome, the main symptoms being central obesity, dyslipidaemia, hypertension and glucose dysregulation.1 While not a hallmark symptom, neuropsychiatric problems also occur in patients with Cushing’s syndrome, particularly in memory impairment. This sparked interest in exploring the effects on cognition of excessive cortisol exposure in key areas of the brain.

Various published research studies now agree that increased glucocorticoid exposure in the brain is linked to age-related cognitive impairment and a person’s cortisol level correlates with reduction in hippocampal volumes and memory impairment1. Corticosteroid receptors are highly expressed in the hippocampus, as well as in neocortex, including frontal lobes associated with cognition. It is known that elevated levels of glucocorticoid concentrations in vivo and in vitro promote biochemical, electrophysiological and structural changes in neurons in these regions2. One of the suggestions made was that that excess glucocorticoids increase levels of amyloid precursor protein (APP) and the APP-cleaving enzyme (beta-secretase), leading to increased amyloid beta formation, reduced degradation and increased tau expression.2

More targeted exploration of the role of elevated plasma cortisol and its effect on brain in neurodegenerative diseases was also conducted. Specifically in AD, elevated plasma cortisol was associated with accelerated disease progression, while high cortisol level in cerebrospinal fluid was associated with more rapid worsening of patients with mild AD (Cernansky et al, 2006; Popp et al, 2015). Animal models showed that if mice are treated with glucocorticoids, this leads to increased amyloid beta formation, reduced degradation and increased tau expression in the brain (Green et al, 2006).

In January 2017, the Australian Imaging, Biomarker & Lifestyle (AIBL) study (R H Pietrzak et al, 2016) was published. The 6-year multicentre prospective cohort study evaluated the relationship between Abeta, plasma cortisol and cognitive impairment in 416 normal older adults. One of the findings was that those subjects with a higher plasma cortisol had a much greater chance of developing AD. Another major study (J B Echouffo-Tcheugui et al, 2018) was published in October 2018 in Neurology, the leading peer-reviewed neurology journal. The data for this analysis came from the famous Framingham Heart Study, which is a large study following the health of residents of Framingham, Massachusetts, and their families since 1948. In this particular publication, the authors evaluated a group of more than 2000 men and women with an average age of 48 with no sign of dementia. MRI brain scans, psychological evaluation of cognition and blood chemistry, including cortisol levels, were performed at the beginning of the study and after eight years. The conclusion was that participants with the highest levels of serum cortisol had the highest levels of impaired memory and that that people with higher cortisol levels had lower brain volumes.

Historically specific 11β-HSD1 inhibitors were first researched in various conditions closely related to diabetes and the metabolic syndrome. But growing understanding about cortisol-cognition relationship prompted researchers to look for a feasible pharmacological intervention aimed at reducing the excess cortisol in conditions where impaired cognition is a prevailing symptom, in particular AD. Direct manipulation of the HPA axis in AD is unlikely to be suitable for AD patients, as it would likely compromise the stress response in the long term. Glucocorticoid action on its intracellular receptors is determined not only by circulating steroid levels, but also by target tissue concentrations. The latter is regulated by 11β-hydroxysteroid dehydrogenase, which in effect ‘gates’ access of these steroids to their nuclear receptors2,3. 11β-HSD1 is present predominantly in the liver, adipose tissue and brain and converts inactive cortisone to cortisol, while 11β-HSD2 is expressed mainly in the kidneys and catalyses the reverse reaction3. Therefore, tissue-specific modulation of intracellular cortisol in relevant brain areas (including the hippocampus and frontal lobes) can be achieved by inhibiting the 11β-HSD1 enzyme without adversely affecting the normal regulation of circulating cortisol.

The researchers at the University of Edinburgh showed that in aged rodents the genetic knockout of 11β-HSD1 or its pharmacological inhibition improved memory1. In the first clinical trials this concept was tested using carbenoxolone, which was known to be a non-specific 11β-HSD inhibitor potently inhibiting both 11β-HSD types. Sandeep et al (2004) showed that in healthy volunteer studies (elderly men and patients with type 2 diabetes) treatment with carbenoxolone led to improved memory. More specifically, in two randomized, double-blind studies, administration of carbenoxolone improved verbal fluency (p=0.006) after 4 weeks in 10 healthy elderly men and improved verbal memory (P=0.005) after 6 weeks in 12 patients with type 2 diabetes. We note that such an effect was achieved after a relatively short treatment period and that patients in the XanADu trial are treated for only 12 weeks.

It is interesting to note that the non-specific 11β-HSD inhibitor carbenoxolone is derived from the liquorice plant, which is native to the Middle East, southern Europe and parts of Asia, such as India. In traditional medicine, liquorice has been used in senile dementia and to treat inflammation, among other conditions. According to one published literature survey covering traditional Chinese medicine formulae published over the past 10 centuries, liquorice was in the top 5 herbs used for senile dementia.

Following these early insights the University of Edinburgh team’s efforts were focused on the discovery and optimisations of a specific, brain penetrant 11β-HSD1 inhibitor. This work was supported by the Wellcome Trust investments. An animal study using Tg2576 mice (AD model) showed that inhibition of 11β-HSD1 with a Xanamem analog led to improved memory and a reduction in amyloid beta plaques after a relatively short 28-day treatment period. The cognitive improvement effect was sustained out to 13 months2. Authors of that study highlighted the disease modifying effect potential of the compound in this model. After the optimisation stage Xanamem (formerly UE2343), was tested in a Single Ascending Dose Phase I study (n=48). The original team of Edinburgh researchers, Professors Jonathan Seckl and Brian Walker, and Scott Webster co-founded a company called Corticrine, which in-licensed the rights to Xanamem from the University of Edinburgh. In December 2014, Actinogen acquired Corticrine and continued the development with a number of studies, including a Multiple Ascending Dose Phase I study. No safety issues were detected in the Phase I. As part of this programme Actinogen also assessed Xanamem’s bioavailability while administering with or without food (no effect). Actinogen also confirmed that Xanamem penetrates the blood-brain barrier and achieves adequate concentration levels in a PK sub-study1. This is a critical element of the investment case, in our view. AbbVie has tried to develop an 11β-HSD1 inhibitor for AD in the past, but this drug failed to demonstrate efficacy in a Phase II trial. Subsequent analysis clearly showed that it crossed blood-brain barrier very poorly (more details on page 13). We interpret this as large pharma recognizing the potential of the approach, but not having the right compound.

The Phase II XanADu trial in mild AD commenced in March 2017. It is a double blind, placebo-controlled Phase II trial, which is now fully recruited (n=186) with mild AD patients, who received 10mg of Xanamem daily or placebo (1:1) for 12 weeks. The study is being conducted in 25 sites in the US, the UK and Australia, which is particularly beneficial for Phase II proof-of-concept AD studies, in our view, as this eliminates unwanted confounding factors, such as language and cultural differences. In addition, high-quality data can be expected to be generated in these locations. In late stage trials, a more diverse patient population will be needed to more closely replicate the real-world setting. Two independent Data Safety Monitoring Board (DSMB) reviews of unblinded safety data have been conducted on XanADu with no safety issues raised. The XanADu results are expected in Q219 and will be critical, along with data from a series of additional ongoing studies (Exhibit 3), for Actinogen and they will define the way forward for further development and commercialisation of Xanamem.

The endpoints used in the study are designed to measure changes in cognition or a slowing in cognitive impairment, since Actinogen is investigating Xanamem as a cognitive enhancer.

The XanADu trial does not have biomarker endpoints centred on biomarkers, which was Actinogen’s decision based on several factors. There are no known 11β-HSD1 biomarkers. We believe that the absence of these in the XanADu trial is not an issue as Xanamem, as an 11β-HSD1 inhibitor, was designed to prevent the production of excess cortisol and not to directly target Abeta plaques (even though preclinical models suggest effect on plaques as well). In addition, biomarkers can substantially inflate trial costs, therefore for all these reasons biomarker testing was not included in the trial design and for a proof-of-concept Phase II study we see this decision as commercially sensible.

When it comes to what would be considered a positive efficacy outcome, we believe there is no single answer. The CDR-SOB test constitutes about half the weighting for the ADCOMS score, while ADAS-Cog and MMSE share the remaining half. Therefore, the most ‘valuable’ endpoints are ADAS-Cog, CDR-SOB, MMSE and ADCOMS, while the others add support and value in particular domains. Failure to meet all endpoints would clearly be negative, while all endpoints met would be a best-case scenario. However, we believe there are many other possible combinations of primary and secondary endpoints that could provide a viable path forward. While elements of the seven endpoints in the XanADu trial overlap, each endpoint measures various unique features of cognition. For example, if one of the co-primary and some of the secondary endpoints are positive, further development would likely continue.

The XanADu dataset will be instrumental in understanding what precise target population would be best suited for late-stage development. In addition to that, in the second half of 2018, Actinogen initiated several additional studies with Xanamem. These include target occupancy studies (results expected in Q219), a number of long-term toxicology studies (around end-2019) and a higher dose safety study (XanaHES, Q219). These are complementary to the XanADu trial in several different ways. For example, target occupancy data will improve the Xanamem data package, while results from the higher dose and long-term toxicology studies will be useful for chronic use human studies and will provide flexibility for adjusting the dosing in AD if needed or its potential use in other indications.

Actinogen is currently developing Xanamem as a cognitive enhancer and the primary focus is to establish its effects on cognition. The general consensus in AD is that so-called disease modifying therapy (DMT) would represent a breakthrough. However, since there are no such drugs and development of MDT's has been spectacularly unsuccessful, a clear definition of what constitutes DMT and how it differs from symptomatic therapy is also lacking (J Cummings and N Fox, 2017). According to the EMEA guidance document 2018, ‘symptomatic improvement is defined as a treatment effect that does not change the overall course of the disease’, whereas ‘a disease modifying therapy is one which delays the underlying pathological or pathophysiological disease processes’, which would correlate with ‘a significant effect on adequately validated biomarkers’. The choice of biomarker as well as the type of analysis is left open. Disease modification can be understood as different outcomes that range from cure to slowing of the natural course of the disease and increasing patient survival. This concept is shown in Exhibit 4.

Until the results of the XanADu trial are published, we believe that all potential scenarios are possible. If in the best-case scenario both co-primary endpoints and all/most secondary endpoints are met, then this would open the path to explore Xanamem as DMT in late stage trials. If some combination of efficacy signals is obtained, then Xanamem’s development as a DMT or a symptomatic drug will depend on the totality of results, which would include results from the additional Xanamem studies that will read out at the same time. This would also determine Xanamem’s market opportunity. The revenue potential of new entrants in the AD market would depend on the benefit relative to the currently approved symptomatic drugs. This could range from direct competition with existing drugs if efficacy is similar, to a market disruptor with a multibillion dollar potential if the drug modifies the progression of the disease. There could also be a variety of scenarios in between these extremes, in our view. For example:

The existing understanding of excess cortisol theory and Xanamem data does not exclude any of these scenarios, therefore the upcoming results from the XanADu trial (by end-Q219) will substantially shape the future development of the drug.

Current treatment practice and AD market

Four pharmacologic treatments have been approved and are established in clinical practice that can moderately improve symptoms of AD, however, these do not slow its progression nor improve patient survival. These are acetylcholinesterase inhibitors (AChEIs): donepezil, rivastigmine and galantamine; and the NMDA receptor antagonist memantine (Exhibit 5). AChEIs are the standard first-line therapy and the only approved drugs for mild AD. Memantine is approved for moderate to severe AD, but is used off-label in mild disease as well, usually added to ongoing therapy with AChEIs. Preferences between specific treatment options vary widely across geographies.

Memantine was the last branded therapy approved in 2004. Some of the off-patent drugs have been reformulated into extended release (eg Namenda XR) and patch formulations in efforts to protect their market share, but ultimately revenues are being eroded due to generic entry. The current AD market consists of symptomatic therapies and GlobalData estimates the value of the seven major markets (7MM)4 was US$2.9bn in 2018. According to GlobalData, the US and Japan are the largest markets currently (US$1.3bn and US$1.2bn respectively in 2018), whereas EU5 is relatively smaller (US$445m in 2018). In our view, high sales in Japan reflect a relatively high proportion of Japanese that are older than 60 years.

With regards to market size, currently the trend of ageing populations and the increasing prevalence of AD is mostly offset by the increasing use of lower-priced generics, as branded drug protection expires. However, due to a high unmet need and the vast patient population, any new, effective drug could disrupt the market. In our view, any new disease-modifying drug would have a realistic potential to reach annual peak sales that are a multiple of the current combined sales of all generic AD drugs. Even an effective symptomatic new branded drug, which adds benefit to the current standard of care, could reach blockbuster status or meaningful sales, as was the case with all four existing AD drugs.

The average cost of a generic standard treatment per patient per year in the US is approximately US$1,200 vs US$4,500 (GlobalData) for a branded version, which represents a 72% average discount even though patent protection for the drug has expired. The average for EU5 is US$300 generic vs US$530 branded, and for Japan it is US$170 generic vs US$560 branded (GlobalData). The potential pricing of Xanamem will substantially depend on what benefit the drug will add to the standard of care.

Large pharma is still active in the AD drug development field despite many recent clinical trial failures and they are still carrying out AD programmes and executing deals in the space. Some recent deals include:

Xanamem’s target population in mild AD

Actinogen is currently developing Xanamem as a cognitive enhancer for mild AD. Mild AD represents the largest AD subgroup to target. If MCI is taken into account, then the distribution of diagnosed and treated patients is shown in Exhibit 7 (GlobalData). Approximately 31% of all treated MCI and AD patients in the 7MM have mild AD (GlobalData). This represents roughly 1.2m patients, of which the largest number come from the US and Japan. The addressable population of mild AD patients is likely to grow over time due to underlying factors such as ageing populations and improvements in diagnostics. Depending on R&D strategy, expansion into other populations is also possible: moving into earlier stages such as minimal cognitive impairment, or potentially even moderate and severe AD. Actinogen’s patents have been approved in all major markets.

The traditional amyloid hypothesis postulates that insoluble Abeta plaques are the key neurotoxic culprit in AD and that drugs targeting these plaques could have a disease modifying effect. Vast resources in the industry have been aimed at intervening in the amyloid cascade to prevent or slow these toxic effects, however, no drugs have been successfully developed. Several hopeful Abeta antibodies still remain in late stage studies, such as aducanumab (Biogen), BAN2401 (Biogen/Eisai/BioArctic), gantenerumab and crenezumab (Roche). We note that Xanamem is the only 11β-HSD1 inhibitor in mid-stage R&D for AD. Although there are multiple other drugs with different mechanisms of action in clinical trials in various stages, these technologies are not necessarily direct competitors. Theoretically, Xanamem could be used in combination with many of those therapies, for example with amyloid beta drugs or cognitive enhancers that are in development.

Biogen tested Aducanumab in a Phase I trial and the results showed a dose-dependent reduction of the amyloid load (measured by PET), which was associated with a significant cognitive benefit in the MMSE and CDR-SOB test. Although this was an early study, the results were sufficiently impressive for Biogen to move directly into a Phase III trial in 2015 (results expected in 2020).

One of the more notable recent developments in the Abeta antibody development arena was Biogen and Eisai’s announcement of the Phase II results in July 2018 with their amyloid antibody BAN2401 (in-licensed from the original developer, BioArctic) in patients with MCI due to AD and mild AD dementia (n=856). Although there were some ‘skeletons in the closet’, overall the data hinted at the potential for disease modifying effect (slowing cognitive impairment accompanied with dose dependent Abeta reduction in the brain). The main issue was a regulator intervention requesting a stop to recruiting APOE4 carriers into the highest dose arm due to side effect fears. This substantially skewed the distribution of APOE4 carriers among the treatment and, even though the headline data seemed positive, the bias raised many questions and it is unlikely that BAN2401 would receive early conditional approval (Eisai has indicated that it is still working on it). The latest announcement from the developing partners is that the drug is moving into Phase III development in Q119.

Beta-secretase or BACE has become a popular, more upstream target in the amyloid cascade. BACE is one of the two enzymes that catalyse the production of Abeta (Aβ1-40, Aβ1-42 and Aβ1-38) from amyloid precursor protein, the other being gamma-secretase. BACE therefore acts as an upstream target in the amyloid cascade and inhibition of BACE should prevent the formation of Abeta, as well as its downstream products including Abeta oligomers, aggregates and plaques. According to amyloid hypothesis, this should prevent further AD pathology and the worsening of symptoms. BACE inhibitors were viewed as having great potential and many large pharma players were heavily invested in this strategy. However, 2018 alone saw three BACE inhibitors failing Phase III studies: verubecestat (Merck; large reduction in CSF Abeta, but no cognitive benefit), atabecestat (Janssen; stopped because of liver toxicity) and lanabecestat (Eli Lilly, AstraZeneca; did not pass the futility test).

There have been many attempts in the past to develop ‘cognitive enhancers’, mostly neurotransmitter strategies aiming to replicate the results seen with donepezil etc but no notable results have been achieved so far. There are many ongoing clinical trials that use cognitive endpoints, but it is difficult to determine whether these companies are pursuing a similar strategy to Actinogen or aiming to develop as DMT. According to the Cummings et al (2017) review of the 2018 AD development pipeline, one Phase III trial and 21 Phase II trials are investigating symptomatic cognitive enhancers. Exhibit 8 summarizes some of the more interesting cognitive enhancers currently in Phase II stage of development.

Most companies in the space are small biotechs. However, Boehringer Ingelheim is developing BI 425809, a glycine transporter inhibitor expected to improve cognition through NMDA receptor antagonism, in a large Phase II study (n=585) with sites in the US, Europe and Japan. The estimated primary completion date is February 2020. Suven Life sciences is developing a serotonin antagonist (5-HT6 receptor antagonist) in moderate AD. This was a popular strategy for cognitive enhancement in AD, with several past attempts. GlobalData found that KOLs were quite optimistic about this strategy (in its 2017 Alzheimer’s report). However, two notable late stage failures in this class were: Axovant’s intepirdine in Phase III MINDSET trial (mild to moderate AD as an add-on to donepezil, n=1,315) which did not meet co-primary endpoints ADAS-Cog and ADCS-ADL, and Lundbeck’s idalopirdine in Phase III.

A couple of the 1β-HSD1 inhibitors that did progress into Phase II trials in neurology were AbbVie’s ABT-384 (originally developed at Abbott) and Astellas’ ASP3662. Astellas was developing its compound ASP3662 for AD (Phase I) in addition to neuropathic pain (Phase I), and painful diabetic peripheral neuropathy (Phase II). The Phase II study for painful diabetic peripheral neuropathy was terminated at futility analysis for efficacy and there appears to be no recent update on the other two indications. AbbVie’s ABT-384 did progress into a Phase II study for AD after reporting Phase I safety study results. The Phase II study was conducted in 267 patients with mild to moderate AD and results were published in October 2014. The study was stopped after futility analysis for efficacy found no benefit from baseline to 12 weeks measured by ADAS-Cog. Webster et al (2017) discussed these results and concluded that ABT-384’s main issue was poor blood-brain barrier penetration. It has been shown in preclinical development that Xanamem crosses the blood-brain barrier and achieves adequate CSF concentration levels1.

Currently, according to EvaluatePharma, there are several other 11β-HSD1 inhibitors in development, but all are in preclinical stages and target non-CNS indications. Previously, 11β-HSD1 inhibitors have been studied in multiple early-stage clinical trials by large pharma including Amgen, Astellas Pharma, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Novo Nordisk, Pfizer, Roche. These studies have mainly targeted diabetes, obesity and various metabolic syndrome indications. However, none of the programmes appear to have advanced further than Phase II and there are no approved 11β-HSD1 inhibitors on the market. We believe that historically such direction for 11β-HSD1 was chosen mainly because of a very well-known connection between glucocorticoid overproduction (Cushing’s syndrome) and wide-ranging symptoms that are hallmarks of the metabolic syndrome: obesity, dyslipidaemia, hypertension, insulin resistance leading to type 2 diabetes and others. Additionally, unlike Xanamem, none of the 11B-HSD1 inhibitors were designed to cross the blood-brain-barrier and penetrate the brain. For these reasons and the fact that the understanding of the effect of excess cortisol on cognition is still only now entering mainstream knowledge, we believe that HSD1 inhibition in neurodegenerative indications has been overlooked.

The main sensitivities are around the only drug candidate, Xanamem. Although it has shown efficacy in animal models and proved safe in the Phase I and Phase II trials, it remains to be seen whether this will translate into a clinical benefit. The data from the current Phase II XanADu study will shape further development of the drug. Drug development is inherently risky and could fail at any stage. If the XanADu trial results warrant further development, this could include a variety of scenarios from second line to front line, in mild or even prodromal patients, as a symptomatic or disease modifying drug, all dependent of the data obtained from XanADu and the additional Xanamem studies in Q219. Late-stage drug development is capital intensive, therefore fresh funds or a partner will be needed. We have assumed a partnership deal in our valuation based on terms of comparable industry deals. However, we have limited visibility on the timing and terms of any future deals, although we believe that positive Phase II data should help attract and secure a partner. Future pricing and market dynamics are hard to predict, hence assessing the commercial opportunity for Xanamem is not straightforward. Our current scenario assumes that Xanamem has a symptomatic effect that is a beneficial addition to standard of care. We therefore assume a premium to the price paid for existing generic AD drugs. This scenario is more conservative to one where Xanamem could be shown to have a disease modifying effect. In such a case, the commercial potential would be much higher in terms of market penetration and achieved pricing. On the other hand, if Xanamem has an effect comparable to existing generic AD drugs, then it would compete directly with these.

We value Actinogen at A$195m or A$0.17/share, based on a risk-adjusted NPV analysis, which includes an estimated A$9.9m in cash (no debt) at the end of FY19. The breakdown of our rNPV valuation, which uses a discount rate of 12.5%, is shown in Exhibit 9. Our valuation only includes Xanamem in AD. We currently do not value any other potential indications, but will reconsider it if Actinogen develops other indications. We use a probability of success of 20% for Xanamem to successfully finish the R&D development.

A key element of our valuation is the assumption that Xanamem will show meaningful benefit compared with existing AD drugs. We use a set of assumptions (see below) that would reflect that scenario and calculate peak sales of US$2.0bn. As discussed above, if Xanamem shows disease modifying results, the peak potential would likely be a multiple of our current calculation. If Xanamem’s efficacy was comparable to current generic drugs, sales might still be substantial as AD is a large market. The underlying assumptions in our model are discussed below.

We used the seven major markets globally: the US, the top five EU countries and Japan. While the US and Western European markets tend to assess novel drugs in a very similar way, there are regulatory specifics in Japan that often prompt companies to consider a separate R&D step to seek approval in this market. However, Japan is a very important market for AD because it is a wealthy country with one of the oldest populations in the world. In addition, AD is much more prevalent among elderly. These factors combined mean that Japan is a large market for AD drugs and could be the reason why many other Phase III trials in AD include centres in Japan as well as other major markets.

As we discussed, Xanamem could potentially be used in different stages of the AD. Because XanADu trial includes mild AD patients, we use this population in our model. GlobalData is using a combination of primary research and secondary analysis of available epidemiological data assessed by an in-house epidemiologist team and estimates that in 2018:

We note that our approach is likely conservative. Published epidemiological studies can vary substantially in different regions. GlobalData acknowledged this fact and indicated that, for example, in the US market it is using the Aging, Demographics and Memory Study (ADAMS) study, whereas other industry representatives are using the 2010 U.S. Census and the Chicago Health and Aging Project (CHAP) study (for example Alz.org), which has much higher prevalence rates (total AD prevalence of 5.7m in the US alone). Brookmeyer et al (2010) discussed the differences between these data sources and came to a conclusion that diagnostic criteria were the main cause for the difference (GlobalData chose ADAMS study so that the diagnostic criteria and prevalence numbers would be comparable globally). Diagnostic rates and treatment rates are even more difficult to estimate, hence we rely on GlobalData’s primary research and use its data. Overall, we view current market size estimates and even existing treatment paradigms in different countries as not limiting Xanamem’s potential. A novel effective AD drug would likely disrupt the status quo.

We assume US$6,000 per patient per year in the US (a premium to branded generics currently) and a 50% discount in RoW. We assume Market penetration of 20% with peak sales of US$2.0bn achieved in seven years after, we assume, the launch in 2025.

We assume that after the XanADu trial Actinogen proceeds without delay with preparations for the Phase III studies, which will last four to five years (launch in 2025). At the same time the company will be conducting licensing discussions with potential partners. Several scenarios could follow, for example if Actinogen found a partner willing to step in immediately, then a full out-licensing deal could be concluded before the Phase III trial even begins. This is a possibility if the XanADu trial results show a clear potential. To err on the conservative side, we assume that Actinogen will initiate the Phase III without delay, but the licensing deal will be done in 2021. Until then Actinogen will invest US$20m in the drug’s development (implies a need for further capital). Other development scenarios are possible Phase IIb and then a Phase III. Licensing deals that we view as comparable are listed in Exhibit 10. We use the average values: an upfront payment of US$55m and total milestone payments of US$550m spread across the R&D (one third of the total amount) and commercial phases (two thirds of the total amount). We assume Actinogen will receive tiered royalty rate of 18-20%.

Actinogen has a portfolio of patents that includes Xanamem (composition of matter) and its use in AD and other neurological and metabolic diseases associated with chronically raised cortisol. The portfolio covers all major markets including the US, the UK, EU, Japan, China, Canada and Australia and is expected to provide protection to at least 2031. A 5% royalty back to the original inventors would be payable on commercialisation.

Actinogen follows local GAAP and the fiscal year starts on 1 July. In its H119 results (end of CY18) the company reported A$80k in revenue and other income, which corresponds to financial income (interest) and a business growth grant (A$20k). The H119 R&D expense of A$6.3m was the most significant cost item and was up from A$3.0m a year ago. This indicates that XanADu is in full swing and approaching completion with final study payments, as well as the cost of initiating nine additional Xanamem studies as indicated in Exhibit 3. Total H119 expenses were A$7.2m, up from A$3.8m, almost exclusively due to the already mentioned higher R&D costs. Actinogen booked A$3.2m in tax rebates in FY18, and Edison expects it to book another A$4m and A$3m in FY19 and FY20. As a result, we forecast an FY19 total loss of A$10.5m, up from an A$6.2m loss in FY18. FY20 and FY21 results will depend substantially on the next steps after the XanADu trial. Actinogen had cash of A$15.5m at the end of fiscal H119 (end of CY18). This is more than sufficient to finish XanADu and all the additional studies initiated in the second half of CY18 and to prepare for the Phase III trial. Additional funding needs will depend on the nature of any potential licensing deal and the scale of the next trials. Our current scenario is described above.