Heisenberg Prin...'s  Instablog

Heisenberg Principle
Send Message
Heisenberg Principle is a full time private individual investor and trader.
  • Celsion: 5 Bear Arguments That Don't Hold Water 1 comment
    Jan 22, 2013 3:24 AM | about stocks: KERX, CASC, SRPT, CLSN

    Celsion (NASDAQ:CLSN) is literally any day away from its top-line release of its Phase III study, called the HEAT trial, designed to show that ThermoDox + RFA treatment improves the time to progression of HCC liver cancer by at least 33% for patients with large tumors versus RFA alone. As such bull and bear pressure and volatility are heating up more than ever as seen in last week's trading, nicely laid out in detail here. The bullish stance is simple: the science makes sense, the trial population is performed worlds better than RFA only would be expected, insiders keep buying, and the company has purposely stalled tapping a even a dime of its $75 million shelf much need shelf registration opting to wait release of the Phase III data, a note of confidence that doing so would be at much higher prices. Bears, of course, disagree and have been furiously presenting their case across the internet. The problem for their most popular arguments, however, is that I find they simply don't hold water. Below is the top 5 most common bear arguments I've run across:


    As in my previous two articles here and here have shown, I believe positive data conclusions about Phase III are already easily calculable to a highly probable degree. Simply put, Celsion management has guided for control arm to have a median progression-free-survival (NYSE:PFS) of 12 months yet the trial population is showing closer to 25+ months. This suggests the treatment arm must be making a significant difference possibly to the tune of hundreds of percent.

    Since those articles, I've gotten a lot of feedback and comments claiming models don't work along with a number of examples. The problem with this bear argument is that the counter examples had very weak models that forecasted mild improvements that left no room for chance or error. For instance, the most often thrown at me is the case of Oncothyreon (ONTY) found here. This is an example of a model failed to predict Phase III trial failure. The problem is that its model was poor from the beginning and set up to fail. The trial population was showing an overall median survival of 29 months yet a study mentioned in that same article called "The Kelly Trial" suggested the control arm could be as high as 33 months! This model took an optimistic unrealistic view of the control arm instead of the CLSN models where even the most conservative of estimates across the board yields slam dunk results. It was not exactly a surprise that the ONTY Phase III trial failed. I would have shorted ONTY myself; this "example" doesn't even compare to the CLSN models which absolutely blew away any and all control-arm studies, not missed a key one by 20% like ONTY has! With the trial population for ONTY having a median 6 months less than a similar trial, it's almost as if their drug not only wasn't any better than placebo, but it actually was worse and hurt patients. Indeed, there were 3 other studies at the time that suggested the trial population's median survival time matched closely to other trials and therefore the expected control arm.

    To me, for a model to be useful, it has to show some degree of confidence. The ONTY trial population was performing somewhat better than a few earlier control arm studies and worse at least 4 others and left so room for chance or error. Conversely, the CLSN trial population is performing astronomically well compared to any study by such a wide margin that it leaves plenty of room for both chance AND error. My Part 1 model estimates ThermoDox will show at least a 200% improvement over the control arm. Since the primary endpoint is only 33%, that leaves a very comfortable cushion. When you have a disparity this large which thus is the case with CLSN trial population vs. any reasonable expectation for the control arm, I'm going to go out on a limb and say: Models works.


    The problem with this theory is that people often mistaken take the word "distant" literally in the term "distant recurrence" when in reality all that distant means is anywhere outside of the original tumor, even if this recurrence is very close by and related to the original tumor, which is most often the case. As the Phase I study and its its CAT scans have shown, ThermoDox greatly increases the zone of ablation into the margins where RFA couldn't reach and well beyond outside the margins. It's this immediate area that are most likely to exist micro-metastases outside the ablation zone kill area. These are a potential site of recurrence if not treated RFA-only has no realistic chance of treating this is a specific area yet ThermoDox directly attacks with a high concentration of doxorubicin deposited by ThermoDox.

    This is where bears are often confused. As in one bear article mentioned a study and said: "With a median follow-up of more than two years....the majority of patients (56%) experienced liver tumor regrowth in parts of the liver that were not targeted by RFA ablation." It's this two-year window that should be of greatest concern to a bear. According to this study, "new lesions occurring within 2 years for treatment are related to occult dissemination of the original tumor, while lesions occurring in later periods are often true "de novo" tumors." In other words, they usually originate in the area of increased zone of ablation created and treated by ThermoDox, a matter of millimeters away from the original tumor, unreachable by RFA but bathed by doxorubicin, the active ingredient in ThermoDox. There's every reason to suspect that ThermoDox will lower the rate of distant recurrence.


    I am immediately reminded of Sarepta Therapeutics (NASDAQ:SRPT) whenever this concern comes up as a counter-example. SRPT has a similar institutional percentage ownership as CLSN (15% and 14% respectively). On October 3, 2012 SRPT was already up 477% from its low. Yet none of that stopped it from releasing Phase II data that blew the street away and sent shares soaring as much as 200% in a single day from 14.99 to 45.00 per share. Comparatively, CLSN is up 329% from its low, less than SRPT was.

    Dare I say, the lack of institutional ownership and already high price appreciation won't stop CLSN from reporting positive data and rising. Interesting to note that many of the same bears will argue that if the data was going to be good, the stock price would be higher while at the same time they'll say that the price is has risen too much so good results must already be priced in. I'm not sure how both can happen at the same time. I'll note that the rise began in August 2012 along with knowledge that 380 PFS hadn't been hit, the first true indication mathematically that the 701 patient population was doing phenomenally well.


    Essentially this rule states that 120 days before the release of Phase III data for cancer drugs that if a company has a market cap under $300 million, the trial will be a failure. The theory being smart money would bid up the market price of the stock if it had a realistic good chance of showing positive Phase III data and since CLSN's market cap is under $300 million it must be failure ahead.

    The problem with that theory: most companies. as we saw with SRPT only after the crown jewel: results of Phase II data. In CLSN's specific case, however, 11 regulatory bodies around the world including the FDA were so confident, so excited, and so encouraged by CLSN's Phase I data, they were allowed to skip the Phase II trials entirely and move onto Phase III. The Feuerstein-Ratain Rule shouldn't apply here because CLSN is the equivalent to a company presenting Phase II data rather than Phase III data and therefore the Rule wouldn't apply. Not surprising, the example cited in the Feuerstein-Ratain Rule article was Keryx Biopharmaceuticals (NASDAQ:KERX), a company which article itself cites has having botched the Phase II trials. Not exactly an apples-to-apples comparison with CLSN who the FDA specifically asked to skip the Phase II trials.


    I will let these quotes from CEO Michael Tardugno speak for themselves:

    Quote #1:
    "interactions with other companies...they are happening and they are frequent...our expectation is that post positive data will be entertaining multiple term sheets."

    Quote #2:
    "I do not have any questions that post positive data that we have partnership relationships that we can come to the conclusion we'll have terms, economics that benefit the company."

    Quote #3:
    "Typically, what is required before a discussion of partnerships is demonstration, a robust demonstration of feasibility, and we are prepared to do that, assuming, of course, positive data from the HEAT trial which we have no reason to believe won't be positive."

    Quote #4:
    "It's not a matter of if we'll complete a license it's a matter of when...if you open your medicine cabinet you will likely see a number of companies performing diligence on us"

    Quote #5:
    "The medical world, I daresay, is watching."

    As a reminder, even though I disagree with what I find to be the 5 most common bear arguments in CLSN, I have my own list of 5 risks for CLSN stock:

    (1) We all could be missing something big in our models. As an example, however unlikely, if all of the recruiting for the trials was done on the internet, that would inevitably lead to a younger and therefore much more healthy trial population. Younger people tend to have better outlooks when it comes to disease. This is just one example of potential recruitment bias.
    (2) There could be some unknown, unexpected side effects that are unacceptable and only came to be realized with final analysis of the data.
    (3) I could be correct in my analysis, but the stock could have already priced in a positive data reaction in the short term and the stock could sell off on the news (buy the rumor, sell the fact).
    (4) New more superior competition could emerge.
    (5) The company itself could stumble, make an error, something could happen to a key man in the company, there could some sort of internal power struggle drama within the company that develops, etc. Things such as this happen sometimes, especially with tiny companies;

    Disclosure: I am long CLSN. 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.

    Stocks: KERX, CASC, SRPT, CLSN
Back To Heisenberg Principle's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (1)
Track new comments
  • Robert Schwartz
    , contributor
    Comments (184) | Send Message
    Relying on historical data to extrapolate the results of contempory trials can be hazardous to your portfolio.
    31 Jan 2013, 09:19 AM Reply Like
Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.