Oryzon Genomics — More data support vafidemstat for aggression

Oryzon has developed a broad R&D programme for its lead CNS-optimised LSD1 inhibitor vafidemstat and is targeting a range of neurological disorders (Exhibit 1). The newer Phase IIa REIMAGINE and REIMAGINE-AD projects are exploring vafidemstat’s potential in managing aggression in four separate patient cohorts: borderline personality disorder (BPD), attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD) and moderate-to-severe Alzheimer’s disease (AD). Two earlier Phase IIa trials are testing vafidemstat in mild-to-moderate AD (ETHERAL) and multiple sclerosis (SATEEN).

Data from the three REIMAGINE cohorts have been reported throughout 2019 and were promising enough for Oryzon to start planning Phase IIb trials in these indications (Exhibit 1; more details in our recently published outlook report). On 3 April 2020, Oryzon reported interim data for the first time from the REIMAGINE-AD and ETHERAL trials. The Advances in Alzheimer’s and Parkinson’s Therapies (AAT-AD/PD) 2020 meeting was held in an entirely virtual format due to the COVID-19 pandemic with presentations available online.

As with other patient cohorts in the original REIMAGINE trial (ASD, ADHD and BPD) the key endpoints in the REIMAGINE-AD study are focused on vafidemstat’s potential to control agitation and aggression seen in moderate to severe AD patients. This open-label, single-arm trial recruited 12 patients. Initially the study was planned to run for 13 weeks, a similar period to the original REIMAGINE trial, but later it was extended to 24 weeks to better assess the effects on memory (not the primary endpoint, however). One patient withdrew consent during the first two months of treatment. Nine patients completed four months of treatment and seven patients completed six months. As in the other REIMAGINE cohorts, Oryzon used a variety of different neuropsychiatric scales:

Results showed that after six months of treatment with vafidemstat a statistically significant improvement (p<0.05) was seen on all these scales in patients who completed the treatment. In addition, the 1.2mg dose was confirmed to be safe and well tolerated. These results are similar to those seen in other indications from the original REIMAGINE trial.

Memory loss is a key symptom in AD. Although vafidemstat’s effect on memory was not the primary goal in the REIMAGINE-AD trial, Oryzon did evaluate 11 patients’ MMSE scores (Mini-Mental State Exam, the most common method to evaluate memory). After the completion of two months of treatment, seven patients had improved their MMSE scores, three patients had stable scores and one worsened. The improvement, however, was not maintained after four and six months of treatment (Exhibit 2A). Two of the four moderate AD patients (Exhibit 2B red boxes) demonstrated consistently better scores during the treatment. Oryzon decided to extend the treatment of these two patients to 12 months.

Given the small numbers it is difficult to come to any firm conclusion on vafidemstat’s effect on memory, but in AD research the general notion is that once the disease progresses to severe stage, memory impairment is advanced and likely irreversible. Since this trial focused on aggression, the severe AD patients were the appropriate target patient population in this trial. The prevalence of agitated and aggressive behaviour increases as AD severity increases. The MMSE assessment was exploratory and it is interesting to see that two out of four patients in the moderate AD group seemed to benefit in terms of memory improvement. These two patients continue the treatment, so some follow up data could still be reported from this trial.

Vafidemstat’s effects on agitation and aggression in REIMAGINE-AD were in line with those reported from the other cohorts in the original REIMAGINE trial. The improvement seems to be very consistent across different scales. We note that by now Oryzon has accumulated data sets from a very diverse group of patients with neuropsychiatric or neurodegenerative diseases. Even within the same indication, for example BPD, the clinical picture of a patient can vary drastically from one subject to another. In our view, the main achievement by Oryzon in its CNS programme is the identification that agitation and aggression seems to be the common denominator in all these different conditions, which can be improved with vafidemstat. Vafidemstat is the most advanced epigenetic drug in development for CNS diseases. As opposed to targeted therapies, like monoclonal antibodies for cancer, an intervention with an epigenetic drug can have profound effects on biology. The emerging evidence that vafidemstat can help control aggression could turn out to be the key breakthrough in its development.

As this was a Phase IIa stage study, it recruited a relatively small number of patients, like the other cohorts from the REIMAGINE trial. The next step is to start a larger Phase IIb stage study. Since the results were consistently positive across all indications, Oryzon can prioritise the indications that are most lucrative from a commercial point of view (borderline personality disorder is likely and we include it in our valuation). Oryzon planned to initiate a Phase IIb trial with vafidemstat in aggression in BPD as the first indication, but this will likely be somewhat delayed because of the ongoing COVID-19 pandemic (more details in Financials section below). We expect updates from Oryzon about the next steps in due course.

The Phase IIa ETHERAL is a randomised, double-blind, three-arm trial (placebo, low dose 0.6mg and high dose 1.2mg) that enrolled mild to moderate AD patients in Europe and the US. After the randomisation the patients are treated for 24 weeks. This is then followed by a 24-week extension when the placebo patients are randomised to receive vafidemstat. The primary goals of the study are safety/tolerability, while secondary endpoints include initial efficacy (cognition, agitation, apathy, depression, quality of life, volumetric MRI) and biomarker results. In total eight biomarkers are assessed including traditional CSF biomarkers, but also novel ones related to inflammation and neuronal damage.

The recent presentation at the AAT-AD/PD 2020 described the preliminary six-month treatment data from the European part of the trial. In total, 117 patients were enrolled, of which 96 completed the six months of treatment. Drop outs were randomly distributed among the study arms, while seven patients were reported to have severe side effects. Of those seven, four patients, however, were in the placebo group, two in the low-dose vafidemstat group and one in the high-dose vafidemstat group. Such distribution shows that vafidemstat is not causing more side effects than placebo and again demonstrated good safety/tolerability profile. Other key results include:

The primary endpoint of safety and tolerability was achieved. This is important, as epigenetic drugs are known to suppress blood production process, but such an effect is typically seen in cancer treatments where the doses are much higher. Vafidemstat’s use in any CNS diseases will likely be chronic, so a good safety profile is crucial.

Some interesting biomarker data provided evidence that vafidemstat has an effect of the pathophysiological processes of the disease. Ultimately, Oryzon will have to prove that this translates into a clinical effect, ie some lasting symptomatic effect or disease modifying effect in AD patients. The released data did not yet prove that, but Oryzon pointed out that this is preliminary data from six months of treatment and more mature data will be reported. The patients are receiving the treatment in the extension phase for another six months. In addition, there is a US study running in parallel. So, Oryzon should obtain much more mature data, which will help to inform how to design the efficacy trials later on. Final 12-month treatment data is expected in Q221.

The ETHERAL is a Phase IIa trial, so it is fairly early stage and not designed to provide a definitive answer to whether the drug is clinically effective (this is addressed in Phase IIb/III trials). The ETHERAL trial is meant to explore the effects of vafidemstat in patients and to help inform the design of the subsequent trials. As a result, the biomarker results are the key data to be collected in this trial.

Overall biomarker results suggest that vafidemstat has a pharmacological effect in the brain at both doses in mild to moderate AD patients. YKL40 deserves special attention, as the interest in this biomarker in the AD context has grown significantly over the past several years (Muszyński et al, 2017). In animal studies, an increased expression of the mouse homologue member of the same YKL40 family has been observed in the brains of mice models of AD when compared to aged-matched controls (Colton et al, 2006). Increased expression of YKL40 was also found in brain samples obtained during autopsy of patients who had confirmed AD (Colton et al, 2006). YKL40 levels in CSF were also found to be significantly elevated in AD patients (Rosén et al, 2014 found a 77% increase compared to normal controls).

Oryzon also has its own preclinical data that demonstrates that vafidemstat has a specific effect on YKL40. In 2017 the company presented findings from one of its in vivo studies designed to test vafidemstat in multiple sclerosis (MS), one of the indications of interest for vafidemstat. The accepted animal model for MS is autoimmune encephalomyelitis (EAE). Vafidemstat was compared against fingolimod (Gilenya, Novartis) and the results showed that its anti-inflammatory effects were faster and stronger compared to fingolimod. We described these data in detail in our December 2017 report.

While the aim of that study was to collect more evidence and support the clinical trial in multiple sclerosis (Phase IIa STEEN study is currently ongoing), it also assessed gene expression changes. One of the genes assessed was the mouse homologue member of the YKL40 family. The expression of it was strongly induced in the EAE model when the animals developed the disease. When the animals were treated with vafidemstat, the expression of this biomarker was significantly reduced, especially in the spinal cord (among other improvements in encephalitis symptoms) (Exhibit 5). The new biomarker data from ETHERAL now show that vafidemstat can also lower the human version of YKL40.

Oryzon’s total spending in 2019 was €14.1m, up from €9.7m in 2018 as the company has accelerated its clinical R&D. Cash was €35.1m at end-2019 (net cash €21.9m). We expect similar operating spending in 2020 and 2021, so its cash runway is to 2022. This is a relatively comfortable position given the ongoing volatility in the markets due to the COVID-19 pandemic, and will also allow Oryzon to deliver multiple catalysts from the ongoing trials.

With regards to the impact of the ongoing COVID-19 pandemic, Oryzon indicated that some delays with trials may be expected, as patient visits to clinical trial centres are currently restricted. However, to protect Oryzon from data loss, different measures are undertaken, such as remote patient assessment or reduction of site visits. The supply of Oryzon’s lead drug candidates, vafidemstat and iadademstat, to patients is uninterrupted. The opening of new sites is restricted at the moment, so we do not expect Oryzon will initiate any studies in the coming weeks or months (the Phase IIb trial in BPD was close to starting). However, given the circumstances, a manageable trial delay is much more preferable to data loss, in our view. If some of the clinical trial work is delayed, Oryzon may be able to save on CRO costs, which will act as a mitigating factor.

Our valuation is marginally higher at €470m or €10.3m per share due to rolling our model forward. We make no changes to our rNPV for the time being. To reflect vafidemstat’s potential in the treatment of agitation and aggression in CNS patients we have already included BPD, as Oryzon indicated that this could be the first indication from the REIMAGINE cohorts to be included in Phase IIb stage trials. If more indications are confirmed to progress to Phase IIb, we will include them in our model as well. We have also included vafidemstat for mild to moderate AD in our valuation. Based on the preliminary findings in the ETHERAL trial we do not yet make any changes. More mature data is needed to understand vafidemstat’s potential in this indication.

A detailed review of all of the company’s ongoing programmes including upcoming catalysts can be found in our recent outlook report.