Applied Therapeutics (APLT): When you call every failure a success
Executive Summary
The company is developing aldose reductase inhibitors, which have a history of lack of efficacy and high toxicity, including liver damage.
AT-007 and AT-001, the lead programs, have failed all large clinical studies, though the company doesn’t like to admit it.
Despite the above issues, the company recently filed AT-007 for approval with both the FDA and EMA. This is after a history of very frequently missing previous regulatory guidance.
AT-007 might be targeting the wrong pathway in galactosemia.
2023 was “The Year of the Four CFO’s”.
The company only has cash until the middle of 2024 and the fully diluted market cap is much higher than you think.
Applied Therapeutics is another one of those companies that just can’t seem to issue a straightforward press release when results are involved. On January 4, 2024, just one day after announcing that they filed for approval of AT-007, they announced Phase 3 data of AT-001 in diabetic cardiomyopathy. The trial failed the primary endpoint miserably, with a p-value of p=0.21 though the company still referred to this as a “strong trend”. They also data-mined their way to finding more attractive p-values by focusing on the subgroup of patients not receiving concomitant GLP-1 or SGLT2 treatment (p=0.04) and then on “clinically significant worsening in cardiac functional capacity of 6% or more” (p=0.035).
The kicker comes at the end of the press release though when they say that given these “encouraging results” they aren’t going to put more resources behind the program:
Given these encouraging results, the Company plans to focus on identifying an appropriate path forward through partnering in order to bring AT-001 to DbCM patients. Current resources are expected to be focused on the development, regulatory and commercial preparations for the govorestat rare disease program.
Anyway, let me back up. Applied Therapeutics is focusing on developing aldose reductase inhibitors, a class of drug with a very difficult past. These have been literally studied for decades and the results have not been great (emphasis added):
The aldose reductase inhibitors inhibit or reduce secondary complications induced by diabetes, specifically in tissues in which glucose uptake is not insulin-dependent (probably neural tissue, the lens, and glomeruli). Many of them (including alrestatin, imirestat, ponalrestat, and sorbinil) have been used in clinical trials, but have been withdrawn because of adverse effects or lack of effect [2]. Their main adverse effects include fever, nausea, diarrhea, increases in liver enzymes, skin rashes, including toxic epidermal necrolysis and Stevens–Johnson syndrome, marked thrombocytopenia, lymphadenopathy, splenomegaly, and adult respiratory distress syndrome.
Tolrestat was withdrawn because of deaths from fatal hepatic necrosis [4] and poor efficacy in clinical trials. Sorbinil was withdrawn because of hypersensitivity reactions in more than 10% of patients.
The company claims their alpha reductase inhibitors are “1000x more potent” and very specific. Yes, they are so potent that even after patients literally received grams of the stuff every day (1,500 mg BID), AT-001 still didn’t work.
And AT-007 didn’t work in their Phase 3 trial in pediatric galactosemia patients either, missing their primary endpoint with a p-value of p=0.1030. Though you really had to go through a lot to get that p-value as it is about 7 paragraphs (and about 500 words) in. And even then their wording is a bit funny (emphasis added):
While statistical significance defined as a p value of <0.05 was not met on the primary endpoint, systematic improvement over time was demonstrated for the overall primary endpoint (p=0.1030) and for a pre-specified sensitivity analyses including cognition (p=0.0698).
And note the second part about cognition. That is specifically something the FDA told them not to include in the primary endpoint, which focused on speech and behavior measures. See this from the conference call regarding the data:
Question – Brian Skorney: And then, could you just walk us through what happened with the primary endpoint protocol change to exclude cognition in the analysis. What was the FDA justification for requesting that protocol change? And what was the timing again on that?
Answer – Shoshana Shendelman: Yes, Brian, it was relatively recent. I cannot necessarily provide a rationale for that. We believe that cognition, especially, I mean, almost above many other endpoints, is very relevant and important in this population, as well as things like fine motor skills. So, I cannot necessarily provide a good reason for that. But we did communicate our belief in cognition and fine motor skills being an important element of the phenotype, which is why we compromised on the sensitivity analysis approach.
A nice non-answer. The analyst is asking why the FDA wanted the change and she is essentially not wanting to give it. This reminds me of a line from Father Guido Sarducci, the 80s SNL character “I can neither confirm nor deny, but I know the answer”.
Funny thing is that the sensitivity analysis that they ran which got them to p=0.0698 included both cognition and motor measures, neither of which were significant either. Cognition had a p-value of p=0.2625!
And even on motor skills they had a p-value of p=0.0937 and suspiciously didn’t include a chart on that despite the fact they graphed 5 other measures. They only included that as a footnote on slide 6.
Maybe that measure was actually worse than placebo for timepoints before 18 months? Another thing that might be suspicious (or might have some logical statistical reason), is that the original p-value was p=0.103, then they added one measure with a p=0.2625 and another with p=0.0937 and they somehow got to p=0.0698. That’s counter-intuitive to say the least.
And then besides wanting to include measures the FDA specifically didn’t want in the primary, they want to exclude one that they did want (speech), mainly because the p-values were abysmal (p>0.8):
And they blame the fact that speech therapy was available in both groups. I guess they don’t understand the concept of controlled trials and trying to prove that a drug is additive.
It also might be important to point out the safety. The company claims it is “safe and well tolerated” but it looks like a quarter of patients in the treatment arm had elevated liver enzymes. Which as I mentioned is a troubling feature of this class. Certainly something that bears watching.
As you can see the clinical data package they will be filing with the FDA for approval is far from what one would generally consider “clean”. Then there is the issue that the company has consistently under-delivered on any guidance related to the FDA for years.
In their S-1 from 2020 they mentioned they expected to file for regulatory approval based on their “pivotal” Phase 1/2 study which included only 14 adult galactosemia patients and repeated the same thing in March of that year. Then suddenly the language disappears.
In August of 2020, they suddenly announce (in an 8-K instead of in a press release) a partial clinical hold on their pediatric study, one that wasn’t lifted until February 2021. The company said the FDA “requested that the Company provide the FDA additional technical information relating to ensuring that every patient in the study has access to the prospect of direct benefit of the drug.” I have no idea what that means but if it was so innocuous, why did it take so long to lift the hold? Another thing, in that August 2020 8-K, the company had said they plan to submit an NDA in the first quarter of 2021, which they updated to the third quarter of 2021 upon the lifting of the hold.
In August of 2021, they reiterated the guidance of a filing in the third quarter of 2021, which would have meant they would have had to file in the following month. So obviously, they must have been pretty close in order to reiterate that guidance, right? Guess not. Nothing is heard about it until January of 2022, when they state (emphasis added):
Following discussions with the FDA at the end of the year, the Company has decided to hold on submitting an NDA for AT-007 for treatment of Galactosemia pending additional discussions with the agency. Although the Galactosemia program had previously been discussed in the context of an NDA for Accelerated Approval based on reduction in galactitol, the FDA has now indicated that clinical outcomes data will likely be required for approval.
If there is an FDA Advisory Committee meeting to discuss this NDA, I really can’t wait until the FDA presents the regulatory history of this compound to get more information on those previous discussions with the FDA.
This regulatory quagmire doesn’t stop there as in March 2022, the company says that it is possible that filing on 6-month data could be possible (as a reminder, they have filed on 18 month data). This was followed by a suggestion in May 2022, that maybe 12-month data will work. Also, in May, the company included this line (emphasis added) “the FDA confirmed that the pediatric study as it is currently designed would support an NDA submission if statistical significance is reached.” Was it reached in the final study data? Absolutely not.
As this is a global industry, in November 2022 Applied Therapeutics then started pushing an EMA filing, with the company exploring the possibility of filing on existing data. Neither filing was made until the end of 2023 until 18-month data was available.
I know I’ve spent a lot of time on this but it’s important to point out how the company has a history of over-promising and under-delivering on regulatory timelines. And I am not saying the company will definitely receive a refuse-to-file letter from the FDA in February, but given the history, I believe it is a possibility.
Now let’s move on to the mechanism, it’s possible that the company is targeting the wrong pathway in galactosemia. AT-007 reduces galactitol, but if you look at the literature, what they should be targeting is Gal-1-P:
Gal-1-P is considered a major key player in the pathogenic mechanism in classic galactosemia.
…
Since the exact role of galactitol in the pathophysiological mechanism of galactosemia remains unknown, other than for cataract, decreasing galactitol levels in other organs might show additional beneficial effects. However, caution needs to be taken since the effect of blocking the polyol pathway that converts galactose into the polyol (sugar alcohol) galactitol, is unknown.
And here is what they say in another publication:
Even if galactitol/galactonate metabolism plays a synergistic role with gal-1-P in the pathophysiology of GALT- and GALE-deficiency, it by no means diminishes the pathogenic role played by increased gal-1-P.
And as the company admits, AT-007 has no impact on gal-1-P levels:
Hence, it is possible the company has completely missed the boat on this indication. Though as we all know, knowledge about diseases change and experts can be wrong but I thought the fact that their mechanism of action might be completely wrong would be important to point out.
There is also the question of how much the drug actually does penetrate the brain considering the primary endpoints in the pediatric trial were speech and behavior oriented. The answer though is not really very much if you compare the concentration in plasma vs. what got into the brain.
As an excellent (though granted a bit dated) report from Biotech Research Partners stated in 2020:
There should be a clear correlation between the concentration of the drug in the relevant organ and inhibition of the enzyme If a drug concentration of 40,000 ng/ml inhibits the enzyme to reduce galactitol by ~30-35% (our generous estimates when placebo correcting APLT’s data), with 2,350X less concentration in brain ~17ng/ml) the inhibition and reduction should be negligible.
I highly recommend a look at that report which goes through the early data the company presented in quite a lot of detail. That report was so good that the company issued an ill-advised press release two days after it was posted (emphasis added):
On July 1, 2020, a fraudulent “short report,” was posted on an online platform from an untraceable and anonymous source. Immediately following, individuals believed to have taken short positions in Applied Therapeutics’ stock shared the report on social media. The report includes fabricated graphs and fraudulent data which it falsely attributes to Applied Therapeutics for the apparent purpose of manipulating the company’s stock price.
Applied Therapeutics will work with law enforcement and regulators to ensure that this criminal activity is prosecuted. The company has hired Reed Brodsky of Gibson, Dunn & Crutcher, a former federal prosecutor at the United States Attorney’s Office for the Southern District of New York, to assist in this matter.
And then just a little more than a month later they got that partial clinical hold from the FDA!
Now I’d like to just briefly talk about their Phase 3 in 56-patients with sorbitol hydrogenase (SORD) deficiency. It’s a 24-month trial with a 12-month update expected to be released soon. This one is the hardest to call as there hasn’t been a lot published on the disease and there just isn’t a lot of clinical data to analyze. One reason AT-007 might have a greater chance of working here is because it doesn't necessarily need to cross the blood brain barrier as the primary endpoint in the trial is the 10-meter walk-run test (10MWRT). It’s also reducing sorbitol which seems more directly linked with the condition (though again, research here is a bit slim) than galactitol is in galactosemia. One issue though with the next update as 12-months just might not be long enough to show a statistical difference in the 10MWRT endpoint in such a small trial. Plus, it’s unknown what kind of data they will divulge and what spin they will put on it.
Something else to keep an eye on with management is the CFO position. 2023 was “the year of the 4 CFO’s” for this company which is not something I think I have ever seen (though granted most had titles of “Principal Financial Officer” but the functions are the same. In February of that year, Chids Mahadevan resigned as Vice President of Finance and interim Principal Financial Officer. He had been in that position since January 2022 when Charles Silberstein, Chief Financial Officer and Principal Financial Officer, resigned. Chids Mahadevan was replaced by Steven Ortega, the Chief Accounting Officer who also became the interim Principal Financial Officer. Just 4 months later, in June 2023, Steven Ortega left, to be replaced by a consultant, Catherine Thorpe. Then in November, the company names Les Funtleyder, a portfolio manager and a Director of the company since 2016, as CFO. Why is it so difficult for them to externally recruit (and keep) a bona fide CFO? Could just be that the company culture is toxic of course rather than something actually nefarious but still it’s strange.
And it’s not like they don’t need a good CFO. The company had $37.5 million in cash as of September 30, 2023, and had an operating cash burn of $38.2 million through the first 9 months of 2023. The company is guiding for that cash, plus the $12.7 million in net proceeds from an ATM on October 13, 2023 and expected milestones from Advanz Pharma, will get them through the middle of 2024. So really, they need to raise additional money very soon. And unless this is through an ATM, the raise probably won’t be too clean if you look at the capital structure of the company. As of November 8, 2023, they had 77,229,207 shares outstanding and a whopping 45,875,618 warrants. That means that if you fully dilute the market cap (adding in those warrants plus the 4.7 million stock options and 3.3 million RSU’s), it goes from $225 million to $381 million. So the company is not necessarily as cheap as you think.
So, if you’re an investor there just is a lot to consider here.
Please like, share and subscribe!
Disclaimer: I do not hold a financial position in Applied Therapeutics (APLT) nor do I have any financial incentive for the stock to move in any direction or for it to become more volatile. This post is purely for informational purposes only and is not meant as financial advice nor as an encouragement to trade. I definitely suggest you do your own research and/or have a discussion with a financial advisor.
Stocks mentioned: