(Editorial Note: With the exception of this note, the article's subtitle, and the concluding section, the article was submitted to Seeking Alpha at 2 AM on May 18. Its previous subtitle was: "Despite Its 49% Jump in 8 Trading Days, Tonix is Still 2016's Extraordinary Bargain." Even though the material between this note and the final new closing section without knowing that the PTSD trial had turned out successfully for Tonix (NASDAQ:TNXP), current and potential investors will find it useful to reflect upon the original article's basic messages about (1) the similarities between PTSD and fibromyalgia (NYSEARCA:FM), (2) why if it works in one it should work in the other, (3) why crushed oral CBP (as opposed to low-dose sublingual CBP) doesn't achieve the same results, (4) what the company's prospects might be with the FDA and the U.S. Army, and (5) what its prospects are in the stock market). In the concluding section, I comment briefly on some of the key findings and their implications.)
Various Seeking Alpha authors (including Jason Napodano, Joe Springer, Kanak Kanti, and, most recently David Bautz) have published substantial details about Tonix (NASDAQ:TNXP), its management team, the conditions it seeks to address, its products, their FDA trials, and the performance of its stock (My March 2016 introduction to this article contains recent information about Tonix not mentioned in "The Third Time's a Charm"). Below, I succinctly highlight some key points to provide context for investors who have not followed Tonix closely.
Tonix has an attractive strategy: investigating potential new uses and new formulations for FDA-approved drugs that have long clinical track records. Part of this strategy includes developing ways to minimize known side effects. It used this strategy to develop new formulations for (1) cyclobenzaprine (NYSEMKT:CBP) for both fibromyalgia (NYSEARCA:FM) and Post traumatic Stress Disorder (PTSD), and (2) isometheptene mucate (NYSEMKT:IMH) for episodic tension headaches.
Results have been reported for two recent clinical trials. The Phase II trial of Tonix's CBP formulation for FM (TNX-102 SL, which is also called "Tonmya" when used for FM) demonstrated its promise but did not meet its primary endpoint. The Phase II trial for Tonix's IMH formulation for headaches (TNX-201) failed and Tonix has discontinued further work on it. The company's stock closed at $1.94 on May 6, primarily because of these missed primary endpoints, but also for other industry-related reasons (However, on May 17, it closed at $2.90; this 49% gain in just eight trading days illustrates both the extent of overreaction to the bad news and the anticipation of imminent positive results that is the focus of this article).
Two clinical trials are now underway, and a third one will be launched this quarter:
- A Phase II trial (called ATEASE) of TNX-102 SL for patients with military-related PTSD. I expect its results will be reported this month.
- A Phase III trial for Tonmya (for FM). I expect its results will be reported during the first two weeks of September.
- A second Phase III trial for Tonmya. This trial is expected to start by the end of June, 2016.
Oral CBP is swallowed, absorbed into the small intestine, and sent directly to the liver where it metabolizes into a problematic byproduct that (1) causes patients to feel tired, and (2) lingers in the body long enough to reduce the effectiveness of subsequent CBP doses, thereby causing diminishing returns after a few weeks.
TNX-102 SL improves on oral CBP in two important ways:
- According to Drugs.com, the standard dose is between 15 and 30 milligrams per day. Tonix's version is a low-dose formulation (2.8 milligrams). Patients take it just once a day, at bedtime.
- Oral CBP takes almost two hours to reach its maximum effectiveness. Patients take Tonix's patented eutectic formulation sublingually (under their tongues), so that the CBP is rapidly absorbed across the mucous membrane into their bloodstreams, and bypasses their livers on its first journey through their bodies (Placing crushed oral CBP tablets under the tongue does not produce a rapid effective dose or initially bypass the liver).
Although the CBP eventually travels to the liver and produces the problematic byproduct, these two improvements mean that its therapeutic effect begins within 15 minutes, and produces too little of the problematic byproduct to either cause drowsiness, or linger long enough to interfere with the therapeutic effect of subsequent doses.
Below are five sections: "A Deeper Look at the Dynamics of PTSD and TNX-102 SL" "A Primer on Probability and the Meaning of This Article's Title," "Tonmya's Completed But Unreported Extension Trial," "Possible Options if the PTSD Trial is Sufficiently Successful," and "Tonix Valuation Models & the Odds That at Least One Trial Will Succeed."
A Deeper Look at the Dynamics of PTSD and TNX-102 SL
In an April 2015 article, I wrote about how the various symptoms assessed in the CAPS-5 (which is the interview protocol used to calculate the scores that will serve as AT-EASE's primary endpoint) create a vicious cycle, each symptom making the others worse. Below, I describe the role played by four other PTSD features (not included in the 2015 article) that make the pattern more insidious, show in greater detail how TNX-102 SL works to reduce the symptoms, and describe what each symptom cluster looks like when treatment is substantially successful via a diagram that shows how the intervention replaces the vicious cycle with virtuous feedback loops.
PTSD Features Previously Given Little Coverage. Before getting to the specific features, a note about diagnosis. As David Bautz pointed out here, the criteria for PTSD and many other disorders are variable. In many ways it is like the proverbial Chinese menu in that you have to have one or more from Cluster A, two or more from Cluster B, etc. Thus patients diagnosed with a disorders like PTSD are both similar to and different from each other.
It is interesting to compare the symptoms of FM with those of PTSD. There is substantial overlap, and the decision regarding which diagnosis to attach to any given patient's experience often has less to do with the symptoms themselves than the context under which the symptoms seem to have arisen (for example, the nature of the traumas, if any) and the patient's gender. This is why if TNX-102 SL works with one, it is likely to work with the other.
Below are four features associated with PTSD that I have not previously discussed or read about in Seeking Alpha articles.
- Pain. While pain is seen as a central part of FM, its frequent appearance in PTSD patients is viewed as "co-morbidity," which essentially means that the patients are suffering from pain at the same time as they are suffering from PTSD. In other words, pain is usually described as seen with rather than as part of the disorder. In a 2012 press release, Seth Lederman (Tonix's CEO) wrote: "PTSD patients suffer from poor quality sleep and many have widespread pain. . . . Similar to the pain in fibromyalgia, PTSD pain originates in the brain, so prescription pain killers, including addictive opiates, are not effective in reducing it. Unfortunately, many patients wind up taking prescription opiates out of desperation and some become addicted. We believe that an effective, non-addictive alternative to opiates for these patients will improve their health and well-being." In fact, the traumas often resulted in pain, and when PTSD sufferers experience similar pain in the present, the pain often triggers intrusive re-experiences.
- Maladaptive thought patterns. Ten "cognitive distortions" are listed here that are frequently associated with PTSD, but also appear in other disorders, especially depression. These maladaptive thought patterns foster and magnify the extent of the reduced engagement that is characteristic of PTSD patients.
- Fears. The psychological distinction between anxiety and fear is that anxiety is an emotional response to imprecise or unknown threats, whereas fear is an emotional response to identifiable threats. As stated in this New York Times article, "There are two criteria that must be present to qualify for a diagnosis of PTSD (1) the patient must have directly experienced, witnessed, or learned of a life-threatening or seriously injurious event, and (2) The patients' response is intense fear, helplessness, or horror." Most writers describe the fears characterizing people with PTSD as irrational, since what they fear is unlikely to actually occur. However, one of the key symptom clusters involves re-experiencing, and in recognizing what it involves, a PTSD sufferer wrote: "Some observers may think when PTSD sufferers react in certain ways to their triggers, they tend to think they are just being paranoid. Now this is most definitely not true. Paranoia is caused by irrational and nonexistent fears. With PTSD, the fears are real because the sufferer has endured them and many times over and over due to such triggers bringing back the trauma constantly back to memory. The PTSD sufferer remembers and relives the psychological and physical pain over and over again."
- Behavioral consequences. Two common consequences of PTSD patients' reduced engagement are increasing isolation, which results in fewer opportunities for reality-based feedback to correct the maladaptive distorted thinking. These are usually described as part of the avoidance pattern, but I treat them separately here because rather than being behaviors, they are the consequences of behavior. Separating them focuses attention on the need for a replacement set of more favorable consequences for when intervention proves successful (See the third figure below).
Figure 1 updates my 2015 article's diagram displaying PTSD's counterproductive feedback loops by incorporating these four features.
- Pain is in a reciprocal relationship with sleep disturbances: pain makes it more difficult to sleep and the lack of restorative sleep results in a higher level of experienced pain. Similarly, there is a reciprocal relationship between pain and distorted appraisals: the greater the pain, the harder it is to think straight. As Dr. Lederman pointed out in his letter to shareholders, "PTSD pain originates in the brain." Cognitive distortions can increase the intensity of experienced pain. In addition, as mentioned earlier, experiencing pain like that experienced during the initial trauma triggers intrusive re-experiences. Indirectly, the intrusions increase the intensity of pain by distorting how the brain processes the intrusion. And it is easy to see why increased pain leads to increased irritability.
- Increases in distorted reasoning and fear both increase the frequency of intrusions and hypervigilance, and they reduce the extent of restorative sleep and meaningful engagement with others. And the extent of distorted reasoning and fear is increased when PTSD patients experience more frequent intrusions and get less restorative sleep.
- Finally, the greater the isolation and the fewer the opportunities for interaction with others to provide corrective feedback about one's thoughts and feelings, the greater the frequency of distorted reasoning and fear.
Figure 2 updates my 2015 article's diagram displaying the symptom that BESTFIT (the Phase II trial for FM) already demonstrated to be improved by TNX-102 SL.
As before, TNX-102 SL increases the amount of sleep and the amount of restorative sleep (thereby reducing the frequency of nightmares), and also decreases the level of anxiety and the extent of the patient's sensitivity. What is new in this revision is that:
- BESTFIT found several statistically significant relationships between taking TNX-102 SL and experiencing less pain (including having a higher proportion of subjects report that their pain was reduced by at least 30%).
- While the previous version only showed the symptoms whose reduction was statistically significantly related to taking TNX-102 SL, this version shows that all the pathways through which this medicine works have as their first step the improvement of the amount and quality of sleep.
Figure 3 reformulates Figure 1 symptom cluster by symptom cluster. Collectively, it shows how the intervention replaces the vicious cycle with virtuous feedback loops.
Please keep three important points in mind.
- This article and Figure 3 both assume that TNX-102 SL will only work for some PTSD patients. The AT-EASE trial will give us a preliminary clue as to for what proportion.
- Figure 3 shows that for an important group of PTSD patients, the frequency or intensity of their symptoms will be reduced, but there is no claim here that TNX-102 SL will completely eliminate their symptoms.
- Seeking Alpha articles about PTSD have compared TNX-102 SL either to doing nothing or to taking pharmaceutical remedies that have serious side effects. However, there is a non-pharmaceutical remedy that has been effective with some PTSD patients in reducing the frequency or intensity of their symptoms: Cognitive Behavior Therapy (NYSE:CBT). Just like CBT works better for depressed patients who are taking antidepressants and for ADHD patients who are taking Ritalin, I suspect that CBT will work better for patients who are regularly taking TNX-102 SL. Only when they have had adequate restorative sleep and the intensity of their other symptoms is manageable can PTSD sufferers gather enough of their resources to benefit from talk therapy.
A Primer on Probability for Investors
and the Meaning of This Article's Title
Investors need to understand two concepts:
- The law for unrelated all-or-nothing events. If there is no relationship between two all-or-nothing events, then the fact that one of them occurs has absolutely no bearing on the chances that the other event will occur.
· The law for related all-or-nothing events. The greater the similarity between two patient populations, (1) the greater the chances that a medicine that works in one population will work in the other population, and (2) the greater the chances that a medicine that failed to work in one population will fail to work in the other population.
Misinterpreting the law for unrelated all-or-nothing events can lead to two contradictory fallacies:
- The gambler's fallacy holds that if you have had a string of bad luck, you are due for some good luck, meaning that if the probability of a coin landing on heads was normally 50%, after a string of one or more tails, the probability of the coin landing heads would be higher than 50%.
- The slump fallacy (which is the opposite of the fallacy of the "hot hand") holds that if a long-time .300 hitter hits around .250 for several games, the chances are that in the next few games, his or her batting average would be below .300.
Here's how these principles and fallacies relate to this article's title and to thinking about the chances for success of Tonix's AT-EASE and AFFIRM trials.
- Taken literally, the maxim that the third time is a charm is an example of the gambler's fallacy. All other things being equal, if the roulette ball landed on black twice, it is not now more likely to land on red the third time. Thus if we assumed that the chances for success in the AT-EASE (PTSD) trial is not related to the outcomes of either of the recent trials (the FM Tonmya trial and the headache TNX-201 trial), then the fact that the first two trials did not meet their primary endpoints does not improve the chances for AT-EASE succeeding; the third time would not necessarily be a charm.
- On the other hand, under the same assumption that the chances for success in AT-EASE is not related to the outcomes for either of the recent trials, then investors who believe that the probability of an AT-EASE failure is larger than it would otherwise have been are swallowing the slump fallacy. Thus under the assumption that the outcomes of the three trials are unrelated, the fact that the first two trials did not meet their primary endpoints does not increase the chances that AT-EASE will fail.
- Both Paragraphs 1 and 2 assume that the chance for success for (or outcome of) any of the three trials is unrelated to the chances for success for (or outcomes of) the other two trials. This assumption is partly right and partly wrong. The part that is right is that there is absolutely no relationship between the outcome of the headache remedy trial and the chances for success for either of TNX-102 SL's indications. Not only is IMH a different substance then CBP, but the hopes for success for the IMH trial were based on experiments with rats and mice, whereas the hopes for success for the TNX-102 SL trials are based on human trials. In short, the failure of the IMH headache trial has no predictive power regarding AT-EASE.
- The part of the assumption that is wrong is that there is a clear relationship between the chances for success of the various TNX-102 SL trials. For one thing, the trials are based on the same medicine. For another, they both use the same pathways (with the first step being improving the quantity and quality of sleep). Finally, despite the fact that BESTFIT missed its primary endpoint, it established some important points: First, as Figure 2 shows, compared to a placebo, TNX-102 SL does a better job of reducing the following symptoms that are common to both FM and PTSD: amount of sleep, quality of sleep, pain, anxiety, and sensitivity. Second, not only is CBP a well-studied drug, but the only reasonably common side effect for Tonix's low-dose sublingual formulation was a mild or temporary numbing of the tongue. Third, there were no serious adverse effects. Thus, according to the abovementioned law for related all-or-nothing events, the fact that BESTFIT demonstrated it produced the desired results (and even met what will be AFFIRM's primary endpoint) means that there is a good chance that AT-EASE will be sufficiently successful (I'll explain what I mean by "sufficiently successful" in another section). Fourth, in my April 2015 article, based on the positive results in BESTFIT, I demonstrated why the chances that Tonmya's AFFIRM will be successful were around 78%. I used an unorthodox way of predicting these chances, so it was gratifying to discover that Tonix management set the sample size based on an 80% probability that AFFIRM will be successful. For the rest of this article, I will use their 80% estimate.
- Immediately above I have argued that there is a substantial correlation between the chances that both AT-EASE and AFFIRM will be successful. However consider the following question: Because FM and PTSD have highly overlapping symptoms, both trials are using TNX-102 SL, and the pathway via increasing the amount of restorative sleep is the same, do the laws of probability suggest that a failed AT-EASE trial would lower the odds that AFFIRM (the second FM trial) will succeed? Well somewhat, but not very much. After all, the primary endpoints of the two trials are totally different. Specifically, AFFIRM's primary endpoint focuses solely upon pain (it's whether the difference between the proportion of Tonmya-taking subjects who report that their pain has decreased by at least 30% and the comparable proportion of placebo-taking subjects is statistically significant). But AT-EASE's primary endpoint focuses solely upon the CAPS-5 score (it's whether the difference between the average reduction in the CAPS-5 score for medicine-taking subjects is greater than the comparable average reduction for placebo-taking subjects), and the CAPS-5 score does not include an assessment of pain. In brief, despite three similarities that suggest that if the medicine works well for FM patients it should work well for PTSD patients and vice-versa, the primary endpoints are mutually exclusive. The bottom line is that whether or not AT-EASE is insufficiently successful to justify further testing, the probability that AFFIRM will successfully meet its endpoint will remain between 78% and 80%.
Tonmya's Completed But Unreported Extension Trial
As recounted above, Tonix argues that because it is using a low dose of CBP and a sublingual formulation that allows it to be absorbed into the bloodstream rapidly without first going through the liver, the amount of the problematic byproduct never gets high enough to make patients feel tired and undermine the effectiveness of subsequent doses. I find the logic persuasive, but know that a lot of ideas that had solid theoretical formulations have not worked out in practice. Investors want to know what the evidence that subjects taking Tonmya for more than 12 weeks don't get turned off by the side effects or by the medicine's becoming less and less effective over time. Tonix needs to demonstrate that its reasoning works in the real world.
Fortunately, the BESTFIT study was extended. Any subjects who had taken either Tonmya or the placebo who wanted to continue beyond the initial 12 weeks could participate in the extension study, and those who had taken the placebo were switched to Tonmya. The results of the study have been tabulated, but have not been released. Why not? There are at least two reasons.
- The gold standard for testing medicines is randomly assigning patients to either a medicine-taking or placebo-taking group and "blinding" both the patients and any medical staff either handling the medicine or making judgments about the outcomes as to who is in which group. The first part of the 12-week study adhered to the gold standard but the extension portion had no control or placebo group; as already implied, everyone participating in the extension study knew that they were taking the placebo. Therefore, any claims made about the results have less credibility than if the extension study had remained a double-blinded random assignment study. The company may not make any claims about what the data means, and many scientists would view such data skeptically. Therefore, why open the company up to criticism by presenting data that is deemed to have relatively low scientific value?
- My position has been that the company should release the data, for example within an SEC form with lots of qualifications about how viewers of the data should be skeptical of reading anything into the results since there was no control group and neither subjects or medical staff were blinded. After all, even the fact that a substantial group of subjects were motivated to keep taking Tonmya for many months means something for investors, even if it doesn't pass muster with the FDA. The company agreed to take my suggestion under advisement and ultimately agreed to make the key data available, but to cut to the chase, not until November. The reason for the delay is that the scientists who did the work want to get publication credit. The only way to do that is to present it at a scientific conference (or to publish it in a respected, refereed journal, which takes longer). To present at such a conference, you can't have already publicly released most of the information about your results. The next relevant scientific conference was deemed to be in November. We may get a hint in October when the abstract(s) are published, but we won't know before that.
Possible Options if the PTSD Trial is Sufficiently Successful
This section takes up three issues: (NYSE:A) What does it mean for a Phase II clinical trial to be sufficiently successful?, (NYSE:B) If AT-EASE is sufficiently successful, might the use of TNX-102 SL for PTSD achieve a breakthrough designation, and if so, what does that mean?, and (NYSE:C) If AT-EASE is sufficiently successful, might the U.S. government pay some or all of its future PTSD-related R&D costs?
What does it mean for a Phase II clinical trial to be sufficiently successful? Normally, a study is considered successful if the endpoint is achieved well enough so that it is statistically significant at or below the .05 level. Given that I have previously discussed the meaning of statistical significance in the April 2015 article cited earlier, I won't explain it here. However, the company argues, and I consider the argument to be persuasive, that the company, the FDA, and the U.S. Federal Government would all conclude that it makes sense to proceed to Phase III, if AT-EASE comes close to being statistically significant at the .05 level, but doesn't actually make it. Of course, any subsequent Phase III would be designed with a sample large enough so that the chances of achieving success at the .05 level was 80%. Tonix's in-house designation for this more relaxed standard is "the .06 scenario."
However, in addition to the "p" value, the decision to continue moving forward will involve several considerations. For example, does TNX-102 SL's previously disclosed side effect profile remain favorable?, and at least one more that I describe in the next paragraph.
Remember that AT-EASE contains three different groups: Two that took TNX-102 SL and one that took a placebo. One of the medicine-taking groups took 2.8 mg., and the other took 5.6 mg. The study's primary endpoint is whether or not the difference between the average reduction in the CAPS-5 score for the 2.8 mg. group and the comparable reduction for the placebo group is statistically significant. Well what if that comparison does not achieve statistical significance at the .05 level, but when the two medicine-taking groups are combined to produce a sample that is twice as large, it turns out that the difference between the average reduction for the medicine group and the comparable reduction for the placebo group achieves statistical significance at the .05 level? Because the decision to proceed will take into account the answers to questions such as these, I am refraining from using Tonix's in-house designation (which might prompt investors to fixate solely upon the p value), and am instead using the phrase "sufficiently successful."
If AT-EASE is sufficiently successful, might the use of TNX-102 SL for PTSD achieve a breakthrough designation? Tonic management is optimistic that it might. However, receiving such a designation is not the same as receiving a "Yes" to the R&D subsidy question. Also, it does not mean that you can skip some of the clinical trials that are normally required. So what do you get if you receive a breakthrough designation? The equivalent of special, priority, or VIP handling. You can find details here (and also learn about other beneficial options), but from a cursory review it seems to mean that the turnaround time on responding to your inquiries or to set up meetings is reduced, and the processing of your submitted applications is speeded up to the extent possible without sacrificing standards.
The cited Brookings Center for Health Policy discussion guide states: "In order to qualify for a breakthrough therapy designation, a candidate therapy must be intended to treat a serious or life-threatening illness, and preliminary clinical evidence must indicate that the therapy may demonstrate a substantial improvement over existing therapies on at least one clinically significant endpoint." The issue of such a designation cannot even be broached to the FDA until there is an evidence-supported application, in other words, until after the AT-EASE findings have been analyzed, written up, assembled in an application package, and submitted. Tonix management does not doubt that military-related PTSD counts as a serious condition, and that it is sometimes life-threatening (given the suicide rate of military-related PTSD sufferers. If the results are good enough (which includes the side effects being as minimal as expected), they think it is possible that the second criterion may be met. However, there are arguments on both sides. There are antidepressant SSRI's that have meaningfully reduced previous versions of the CAPS score, but their side effect profile is considerably worse than TNX-102 SL's. That should be taken into account, but no one knows whether it will be.
If AT-EASE is sufficiently successful, might the U.S. government pay some or all of its future PTSD-related R&D costs? Yes. There are two ways that the government may contribute. The most important and likely is related to Tonix's Cooperative Research and Development Agreement (CRADA) with the U.S. Army Medical Materiel Development Activity (hereafter referred to as "the Army"). While CRADA will not result in Tonix receiving a grant or contract, the Army may design, conduct, and pay the total costs of its own Phase III studies. This Army unit is fully aware of FDA protocols and standards, so we can expect that any studies that it conducts will pass muster.
Even if the Army does conduct its own Phase III trials, Tonix can also seek financial assistance from other military units to conduct related studies. And if the Army does not conduct its own Phase III trials, these other units may still subsidize some or all of Tonix's future PTSD-related R&D costs.
Under either or both of these subsidy scenarios, Tonix's need to raise additional money will be reduced, thereby minimizing the dilution of TNXP shares.
Tonix Valuation Models and the Odds
That at Least One Trial Will Succeed
I developed my first valuation models for Tonix in my March, 2015 article, and revised them in July 2015 here. This section contains provides two updated valuation models. Following the presentation of the models, I provide a basis for investors to examine the impact (on the probability that at least one of trials AT-EASE and AFFIRM will succeed) of varying both the probability that AT-EASE will succeed and the correlation between the two trial's chances for success.
The first model differs from those I have published previously in several respects:
- It no longer includes the headache medication whose trial failed.
- It assumes that the company's value will be divided by 26 million shares, even though its May 9, 2016 investor presentation said that there were only 18.9 million shares. Yet SA author and commenter Jonanne has calculated that if you include warrants, etc., the total of possible potentially dilutive existing shares is 22,774,320. In their SEC filings, Tonix has said that they have enough money to last through March, 2017. However, Jonanne estimates that Tonix will need to issue about 900,000 options and RSU's and anywhere between 1.35 and 2.7 million new shares to support their on-going programs. Adding these together, he estimates that the resulting total of diluted shares would range between 25.0 and 26.4 million shares. My 26 million figure is a nice round number at the high end of this range. Nevertheless, until a combination of more than 3,335,680 new options, RSU's, and new shares are issued, assumptions based on 26 million outstanding shares underestimates the stock's actual NPV/share.
- It encourages you to examine the impact on the Net Present Valuation (NYSE:NPV) per share of varying the market penetration rates for each indication.
- It does not multiply the NPV per share by each indication's estimated probability for success (As Seeking Alpha commentators have pointed out, with or without factoring the chances of success, the values calculated are merely heuristic fantasies, given that they are based upon "What if's" that may never materialize. The second model provides a partial remedy by allowing readers to examine the impact on the valuation of varying the chances that AT-EASE will be successful).
Table 1 displays the first model. Its major value is to recognize the stock's potential Net Present Valuation per share if both AT-EASE and AFFIRM are successful.
Previously, I have been criticized for asserting that there are 4.25 million PTSD sufferers seeking treatment. The figure comes from two statements that Tonix has made in its investor presentations, including on slide 17 of the presentation mentioned above. One statement (supported by a 2013 citation) says that there are about 8.5 million adults in the U.S. affected by PTSD. The other says that half of those who are diagnosed use professional care to treat their condition.
The criticism is that the number of people with military-related PTSD is much smaller. While it is true that AT-EASE only enrolled current or former military personnel and that the U.S. government's primary interest is in treating current and former military personnel, there is nothing about the label that the FDA would confer that would restrict Tonix from marketing TNX-102 SL to non-military sufferers. The medicine should work the same way, and it might be even more effective for much of the non-military populations because for many civilians, the traumas they have face have been less frequent and occurred over a narrower time span. I have discussed this directly with Tonix management, and if TNX-102SL receives FDA approval, they are interested in reaching the largest potential market.
Table 2 displays the second model. Based on the BESTFIT results, it assumes that the probability that AFFIRM will meet its primary endpoint is 80%. Table 2 encourages you to explore the impact of varying the probability (from 40% to 60%) that AT-EASE will be sufficiently successful to continue R&D. Like Table 1, it encourages you to examine the impact on the NPV/share of varying the market penetration rates for each indication.
I recognize that many readers will consider estimate from 40% to 60% to be wildly optimistic. That is why I carefully laid out how PTSD symptoms form a vicious cycle, pointed out that TNX-102 SL has already demonstrated that it reduces key symptoms in a very similar disorder, shown how reducing those symptoms reduces related symptoms, and emphasized the point that the probability being estimated is not for AT-EASE to meet its primary endpoint (which it will eventually need to do in subsequent properly-powered Phase III tests), but only to provide evidence that further R&D is justified.
Note that all but one of Table 2's 48 NPV/share estimates are above $30. The only one that is below $30 assumes that both Fibromyalgia and PTSD will only capture 3% of the U.S. adults who are seeking or receiving treatment. Given the side effects of the available FDA-approved alternatives, I believe 3% is a very low ball estimate for both indications.
Finally, it is possible to estimate the probability that both trials will fail, and then subtract the result from 100% to discover the probability that at least one of the trials will succeed. As I explained in Points 4 and 5 of the Primer on Probability for Investors, the probability that two all-or-nothing events will succeed (or fail) depends not only upon the probability that each trial taken individually will succeed, but also upon the extent to which their individual probabilities for success are correlated.
I gave arguments why the correlation between these probabilities might be reasonably high, as well as why it might not be as high as it might at first have seemed. Table 3 is set up to encourage you to study the impact (on the probability that at least one trial will be successful) of varying both the probability that AT-EASE will succeed (from 20% to 60%) and the two probabilities' correlation (from 0.00, which would mean that there is no relationship between them, to 0.50, which would mean that 25% of the variation in one probability is explained by the variation in the other probability).
Note that even though the table includes rows in which the probability that AT-EASE will succeed is only 20% or 30%, all of the probabilities that at least one trial will succeed range from .76 to .92. I think the six most reality-based cells are the ones in the middle two column in the bottom three rows. These probabilities range from 82% to 88%.
It is highly likely that at least one of the trials (AT-EASE or AFFIRM) will succeed, and if it does, then investors will not only cover their recent losses but also celebrate their decision to learn about this impressive company.
Addendum: Some Finding Highlights & Their Implications
The press release and conference call discussed both the findings and future plans. Some important points:
- The arm of the study that used the 2.8 mg. dose was in the right direction but did not achieve statistical significance, but the arm of the study that used the 5.6 mg. dose achieved statistical significance, with p = .038 or .031, depending upon the data analysis method used. Note that the lower the p value, the better.
- On a global basis, clinicians judged that patients taking 5.6 mg. of TNX-102 SL improved more than those taking the placebo, with p = .041.
- On a global basis, the difference between the average improvement for patients taking 5.6 mg. of TNX-102 SL and that for those taking the placebo was statistically significant, with p = .035.
- Patients taking TNX-102 SL had a greater reduction in startle response intensity than those taking the placebo, and this is important because SSRI's usually do not produce the same result. This finding considerably increases the chances that the FDA will confer a Therapeutic Breakthrough Designation.
- There seems to be a substantive difference between FM and PTSD not discussed in the body of the article: FM patients tend to have problems with non-REM sleep (when they are not dreaming), whereas PTSD patients tend to have problems with both non-REM and REM sleep. The 2.8 mg. formulation effectively improves non-REM sleep, but is not very effective in improving REM sleep. In contrast, the 5.6 mg. formulation of TNX-102 SL improves both kinds of sleep.
- Tonix is currently contemplating having two Phase III trials using the 5.6 mg. dose: one for military patients and one for civilian patients. This justifies going after the 4.25 million people diagnosed with PTSD who are currently seeking or receiving some kind treatment. Furthermore, Men are often viewed as not willing to seek treatment for PTSD. Perhaps the biggest reason is that the only currently FDA-approved treatments (antidepressant SSRI's) tend to interfere with sexual functioning. TNX-102 SL does not have this unwelcome effect. When the word gets out that there is a treatment that makes a difference without impairing sexual functioning or causing other serious side effects, many of those affected by PTSD who have not sought treatment may indeed explore what TNX-102 SL can do for them. Note that while the Army may conduct and pay for the military Phase III trial, Tonix will probably have to sell new shares to pay for some or all of the civilian Phase III trial. However, despite the expected dilution, proceeding with a civilian Phase III trial will reap handsome dividends for investors.
Table 2 assumed that the probability that AFFIRM will meet its primary endpoint is 80%. It varied the probability of AT-EASE being sufficiently successful from 40% to 60%. I used the procedures that led to Table 2 to calculate the NPV/share under two assumptions: (1) that Tonix or the Army will power the Phase III military PTSD trial so that its probability for success is 80%, and (2) that there would be a total of 26 million diluted shares. If the market penetration for both indications is 3%, 4%, 5%, or 6%, then the NPV/share is, respectively, $40.92, $54.56, $68.19, and $81.83.
Fellow investors: Celebrate the fact that you remained invested.
Potential investors: Welcome aboard.
Disclosure: I am/we are long TNXP.
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.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.