Tonix Pharmaceuticals — Update 1 April 2016

Tonix is an emerging specialty pharmaceutical company that was founded in 2007 as Krele Pharmaceuticals. The company changed its name to Tonix Pharmaceuticals in 2010 and went public via a reverse merger in 2011. It uplisted in August 2013 and is now listed on the NASDAQ.

Tonix has two programs in development both involving TNX-102 SL as a small, rapidly disintegrating tablet version of cyclobenzaprine (CBP) administered sublingually. It is currently being developed as a bedtime therapy for both fibromyalgia and PTSD using the 505(b)2 pathway. CBP is a muscle relaxant that was first approved in 1977, but was never approved for use in fibromyalgia or PTSD. Frost & Sullivan estimates that approximately 50 million doses are prescribed each year for patients off-label in fibromyalgia, although the use in PTSD is unknown. Data from Tonix’s Phase III AFFIRM study in fibromyalgia is expected in Q316 and data from its Phase II AtEase study in military-related PTSD is expected in Q216.

Using a risk-adjusted NPV model of its pipeline products, assuming a 12.5% WACC and no value past patent expiration, we value the company at $300m or $15.92 per basic share. On a fully-diluted basis, we value the company at $316m or $15.43 per diluted share. Key catalysts that would affect our valuation of the company would be the initial proof of concept Phase II data of TNX-102 SL in PTSD in Q216, although we view this trial as high risk, and Phase III data in fibromyalgia, which we view as having a high probability of success.

Tonix ended Q415 with $43m in cash and equivalents, which should be enough to get the company into 2017. We have reduced our forecasts for R&D spend for 2016 to reflect the termination of the TNX-201 clinical program in ETTH as well as the fact that the PTSD Phase II should be wrapped up in Q216, most likely in late May. However, we have also removed the associated expected upfront and milestone payments for the ETTH program, which leaves a shortfall. We have therefore increased the time to profitability to 2022 (from 2020) and the amount of financing required to reach that point to $140m (from $80m). We expect the company to raise additional capital in Q316 following some or all of the remaining clinical data on the fibromyalgia and PTSD programs. Although we currently model the company licensing TNX-102 SL after approval, licensing the compound earlier in its development could provide a potential alternative source of financing.

Tonix is subject to various sensitivities common to emerging pharmaceutical companies, including formulation, development, commercialisation, manufacturing and financing risks. While the clinical risk to TNX-102 SL in fibromyalgia has been lessened by the release of the Phase IIb BESTFIT data, the profile of TNX-102 SL in PTSD is largely unknown and has a high risk of failure. Military-related PTSD is considered by some to be a particularly difficult indication to see improvement through pharmacological therapy. Another major risk to the company is financing. We expect Tonix will need to raise an additional $140m prior to profitability in 2022 and its ability to raise additional funds will be dependent upon positive clinical data. If both the PTSD and fibromyalgia programs fail, it is unlikely that Tonix will be able to raise the capital necessary to continue in its present form, if at all.

Tonix has a pipeline of one drug for two indications (see Exhibit 1). Its lead program, TNX-102 SL (Tonmya) for fibromyalgia, appears to be a very safe and effective treatment for the disorder. Tonix's Phase IIb BESTFIT results, despite missing the primary endpoint, indicate a high probability of success for its Phase III program, which should read out in Q316. Data on TNX-102 SL in anxiety and sensitivity as well as its impact on sleep, points to applicability in PTSD. The company's Phase II trial in military-related PTSD is expected to read out in Q216.

The most advanced clinical program is for fibromyalgia, which is being investigated in a Phase III registration trial, AFFIRM, with results predicted in Q316 (a second Phase III trial is expected to commence in Q216 and will be of similar design). Fibromyalgia is a diffuse, chronic pain disorder where the areas of pain often fluctuate and there are a variety of comorbidities/symptoms, with sleep-related disorders (fatigue, stiffness, non-restorative sleep and difficulty falling asleep) being some of the most intense (see Exhibit 2). It is a relatively common problem from which approximately 2% of Americans suffer (3.4% of women and 0.5% of men)1 with prevalence highest among those between the ages of 55 and 64.

There are currently three approved drugs for fibromyalgia: Lyrica (also approved for neuropathic pain disorders); Cymbalta (an SNRI also approved for depression and anxiety); and Savella (also an SNRI, but only approved in the US for fibromyalgia). Importantly, Tonix has one of only three late-stage (Phase II/III or more advanced) developmental candidates in the pipeline (see Exhibit 3) and one of those is simply a longer-acting version of Lyrica, which is already approved.

In addition to potentially being only the second product that is once-a-day, the advantage of Tonmya vs the competition is that it addresses the issues of muscle pain and sleep disturbances associated with the disease. Tonmya would be taken at bedtime daily to improve sleep and lower pain. Its active ingredient, cyclobenzaprine, is a muscle relaxant that has been approved for short-term use in muscle spasms since 1977. Cyclobenzaprine has historically been one of the most widely prescribed treatments for fibromyalgia, albeit off label. In a poll conducted in 2005, 64% of fibromyalgia sufferers had taken the drug and 58% found it helpful.2

The problem with cyclobenzaprine, however, was that it had very variable oral bioavailability of
33-55% and that was highly dependent on what food people had ingested. The 18-hour half-life led to a hangover effect in some patients while other patients did not receive enough. By making cyclobenzaprine sublingual, Tonix has been able to improve the pharmacokinetics and allowed it to use a much lower dose for patients (2.8mg vs the most commonly used 10-15mg doses). Based on the data for Flexeril, one of the brand names for cyclobenzaprine, TNX-102 SL does have a superior safety profile (see Exhibit 4) and may be preferred by physicians and patients over generic cyclobenzaprine. It is important to note, however, that this comparison is imperfect as cyclobenzaprine for muscle spasms was taken during the day, instead of at bedtime, so there would be more opportunity to feel side effects.

As Tonmya addresses both pain relief and sleep quality, the 500-patient Phase III AFFIRM trial is designed to measure each of those. The structure of the trial is largely similar to the previous 205-patient BESTFIT trial, which, although it missed its primary endpoint, met the endpoint that has been approved by the FDA for AFFIRM (the proportion of patients with 30% or greater reduction in pain).

At the American College of Rheumatology (ACR) conference in November 2015, the company presented data supporting the correlation between sleep quality and improvement in fibromyalgia/pain symptoms. These data extracted from the BESTFIT trial presented a series of strong correlations between sleep quality and pain outcomes (Exhibit 5) and how Tonmya demonstrated efficacy across multiple endpoints and time points.

The improvement in sleep quality during the BESTFIT trial was unequivocal (2.5x more improvement than placebo, p<0.001), and we believe with the higher powering in the AFFIRM trial it will be able to replicate the statistically significant improvement in the primary endpoint of increase in 30% pain responders. The FDA approved this as the primary endpoint for the trial following an analysis of the BESTFIT data. The primary endpoint for BESTFIT was mean change in daily pain scores, which achieved a p value of 0.172. This type of analysis statistically favors a small number of strong responders over the broader but more modest improvement that was witnessed in the trial. The 30% responder analysis is much better at capturing the observed drug effect, and we do not see a reason to expect these results to change in the larger AFFIRM trial.

We currently model $343m in peak sales, the equivalent of c 30 million doses pa (equivalent to
1-2% of the total treated fibromyalgia patient population) with an initial cost of $6 per dose, approximately in line with the branded cost of other fibromyalgia medications. As CBP is able to achieve 50 million doses off-label, without detailing and effectively without any clinical data, we believe this sales level is achievable. This would also make it more successful than Savella but less successful than Cymbalta and Lyrica in terms of peak sales, which makes sense given the relative efficacy profiles of the drugs.

TNX-102 SL is currently fully enrolled in a Phase II study for the treatment of PTSD. AtEase is a 240-person, placebo-controlled study examining two dosage strengths of TNX-102 SL in veterans with PTSD over the course of 12 weeks.

According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), to be diagnosed with PTSD sufferers have to exhibit symptoms across four categories: intrusions, avoidance, mood and cognition and arousal (see Exhibit 6). Importantly for the development of TNX-102 SL in PTSD, according to the guidelines from the Department of Veterans Affairs the first signs come in the form of complaints of sleep and chronic pain (two areas where TNX-102 SL has shown efficacy, per the BESTFIT trial).

There are a number of shortcomings associated with available treatments for PTSD. There are currently only two products approved for the disorder, paroxetine and sertraline, both SSRIs with the associated side effects and only modest effect size. Based on the Cohen’s d statistic, where 0.2 is considered a small effect size, 0.5 is considered moderate and 0.8 is considered large,3 both approved therapies have only small to moderate impacts (see Exhibit 7). Among other pharmacological therapies, some of which are used off label after SSRI failure, meaningful effect sizes were generally seen only in small trials (the effect size for topiramate is heavily skewed by a 67-patient trial in Iran where the drug was tested as adjunct therapy rather than monotherapy for PTSD). Psychotherapies of various forms appear to be the only type of therapy today with consistent results; however, there is the issue of compliance as psychotherapies require frequent office visits.

Furthermore, patients with military-related PTSD are considered more difficult to treat. Only 20% of military-related PTSD patients were effectively treated in previous SSRI studies.4 In one study of sertraline, which was able to get statistically significant results in the general PTSD populace, in 169 veterans in an outpatient VA setting, sertraline missed every primary and secondary efficacy measure.5

It is unclear exactly why military-related PTSD appears to be more difficult to treat. Tonix theorizes that it might be due to the fact that those in combat have repeated traumatic events during their tours of duty, which makes their form of PTSD more ingrained. There are some data that support this theory, including that the more deployments a soldier has, the greater the risk of experiencing psychological problems and being on mental health medication (see Exhibit 8).

Another potential reason could be that 48% of veterans from Iraq and Afghanistan who have PTSD also have traumatic brain injury (TBI) according to the Rand Corporation. As TBI can involve physical damage to the brain, the underlying pathology may be different to non-military-related PTSD. In addition, TBI shares many of the same symptoms as PTSD, such as irritability, cognitive deficits, insomnia, depression, fatigue and anxiety – if those symptoms are caused by physical damage rather than PTSD, they may be more intractable.

Patients with “significant” TBI are excluded from Tonix’s AtEase trial. The company defines “significant” as those diagnosed with moderate or severe TBI, which means anyone with brain abnormalities on a scan and/or someone who was unconscious after a traumatic event for less than 30 minutes and had less than 24 hours of post-traumatic amnesia.

There are a few problems with these criteria, mostly due to underreporting and misclassification of TBI. In terms of underreporting, it is important to note that access to medical care for soldiers involved in blasts was not uniform. According to the Joint Mental Health Advisory Team 7 report from 2011, which investigated the mental health of soldiers in Afghanistan, only 57.4% who reported a head injury and only 55.6% who lost consciousness following a blast were seen by a medical professional.

Also, according to the Veteran’s Administration/Department of Defense Clinical Practice Guidelines, computed tomography (CT) is the method of choice for brain scans to diagnose TBI but it has a high rate of error. According to a study of 135 TBI patients, 27% of those with normal head CT had abnormal brain MRI scans.6 In turn, MRI can miss brain abnormalities that would be detected if diffusion tensor imaging had been used.7

We currently assign a 30% chance of success for TNX-102 SL. Safety is not an issue and based on data from the BESTFIT trial in fibromyalgia, the drug has demonstrated benefits in anxiety, sensitivity and sleep quality, which are all issues for PTSD patients.

However, there are several reasons to be cautious:

In summary, the AtEase trial is testing a product with an unknown effect size using a relatively unknown scale in a difficult to treat population. That said, given all of these unknowns and the fact that this is essentially a proof-of-concept trial, a p-value that comes close to statistical significance may be a win as a subsequent trial may be able to hit significance through some trial design enhancements and greater power. If AtEase does hit significance, Tonix is likely to apply for accelerated approval and we believe it would need to run a second confirmatory trial prior to FDA approval.

Commercially, if TNX-102 SL is shown to be effective in PTSD, it should be able to have meaningful sales given the size of the population, even when you adjust for those who are successfully being treated by SSRI therapy and those unwilling to seek treatment. We are currently assuming $113m in sales to the military-related PTSD market and $690m to the rest ($803m in total). These estimates assume that TNX-102 SL is able to capture 10% of the PTSD market (c 2% of the total patient population) that is refractory to SSRIs and with a price of $6 per pill at launch. We are currently assuming that Tonix will find a strategic partner to commercialize TNX-102 SL in PTSD with a focus on those that currently fail SSRI therapy.

Until recently, Tonix had been investigating TNX-201 for the treatment of ETTH in a Phase II clinical trial. Tonix announced in February 2016 that the clinical trial failed to meet its endpoints and that it would discontinue development of the compound. The 147-person trial had three different primary endpoints with progressively more permissive requirements: resolution of headache pain, 70% reduction in pain on the Visual Analog Scale (VAS), or improvement from baseline on the VAS. None of these endpoints reached statistical significance and the company did not provide any more detail on the outcomes. It stated that TNX-201 was very well tolerated, but the failure of the trial and the company’s decision to discontinue the program strongly suggest that the compound has little clinical utility. The result raises some questions considering that TNX-201 is the R-isomer of isometheptene mucate, which has been marketed in combination with acetaminophen and dichloralphenazone (under the brand Midrin) for the treatment of tension and vascular headaches. However, the compound never underwent FDA approval as it was grandfathered in 1962. It is unclear at this point where this molecule failed, given the lack of information regarding its activity. Regardless, we have removed all forecasts related to TNX-201 from our model.

Using a risk-adjusted NPV model of its pipeline products, assuming a 12.5% WACC and no value past patent expiration, we value the company at $300m or $15.92 per basic share. On a fully-diluted basis, we value the company at $317m or $15.49 per diluted share. Key catalysts that would affect our valuation of the company would be the initial proof of concept Phase II data of TNX-102 SL in PTSD in Q216, though we view this trial as high risk, and Phase III data in fibromyalgia, which we view as having a high probability of success.

The company reported a loss of $13.4m for Q415 and $48.1m for FY15. The loss for the quarter was in line with the previous quarter ($13.3m) as a slight reduction in R&D was offset by an increase in SG&A of approximately $1m. On a yearly basis, R&D was up 91% ($16.9m), reflecting the increase in spending with the advancement of the company’s clinical trials. With the discontinuation of the TNX-201 trial and the conclusion of both the AFFIRM and AtEase trials, we expect total R&D spend to be lower in the near term. G&A expenses for the quarter were up approximately $1m on a sequential basis to $3.9m. G&A for the year was up 40% to $12.7m, with the increase approximately equally split between compensation and professional services. Cash burn for the year was better than our expectations and Tonix ended 2015 with $43m in cash.

While the termination of the ETTH program has lowered R&D expenses, we have also had to remove the upfront and milestone payments associated with the program, resulting in a shortfall. We have therefore increased the time to profitability to 2022 (from 2020) and the amount of financing required to reach that point to $140m (from $80m). We expect the company to raise additional capital in Q316 following some or all of the remaining clinical data on the fibromyalgia and PTSD programs. Although we currently model the company licensing TNX-102 SL after approval, licensing the compound earlier in development could provide a potential alternative source of financing.