Mangrove - Sumzero.com???
YMB had a series of posts that is a bombshell IF TRUE.
There was a post that was allegedly copied from a site called www.sumzero.com. This evidently is where asset managers can swap investment ideas. Supposedly Mangrove, which just filed (showing a 6, and then 8, percent holding of Celsion) posted this writeup.
I'm trying to find out if the above is correct.... Anyone here know of the site?
-Trond
-----
We filed last month with approximately 8% ownership of Celsion (NASDAQ: CLSN). The detailed commentary in our filing was prompted by a conference call we held at the request of the company’s management to discuss our statistical modeling of their ongoing clinical trial in Thermodox (the HEAT trial) and several recommendations we made as a result of our statistics work. So that you have no confusion after reading our filing, while we wrote in our filing that we “acknowledged” management’s view that the trial was unlikely to stop for efficacy at the upcoming interim look, our analysis strongly suggests that the trial will show overwhelmingly positive results. Indeed, if the existing Special Protocol Assessment between Celsion and the Food and Drug Administration is structured as we would expect, it is our strong belief that the trial will stop for efficacy.
While Celsion represents among the best opportunities in terms of expected value that I have seen in recent years, the value of the company hinges on a binary event: the outcome of the ongoing HEAT trial. To be clear, there is substantial downside should the trial be unsuccessful. Thermodox is a reformulation of the already approved cancer drug Doxorubicin such that it is encapsulated in a lipid barrier that breaks apart in the presence of heat a few degrees above body temperature. From a treatment standpoint, Thermodox is a very logical improvement of Doxorubicin, because it allows for a more targeted approach to treating patients with what is otherwise an extremely toxic drug. From an approvability standpoint, the prior approval of Doxorubicin should ease the process for Celsion. In short, we believe that the probability of a successful trial is high.
The HEAT trial is a test of Thermodox in conjunction with radiofrequency ablation in patients with hepatocellular carcinoma. Our confidence in the trial outcome is predicated on our biostatistics work, which suggests an overall median time of progression free survival across the entire trial of approximately 30 months, as shown in the 99.9% confidence intervals for median time of PFS across the entire trial in the table below:
< see attachment>
In deconstructing the trial’s performance between the treatment and control arms, we believe that it will be possible for the trial to achieve statistical significance at a p-value of = 0.001 at the interim look based upon a modeled median time to progression free survival in the control arm of up to 18-months. Commentary from Celsion suggests that the 190th event, which triggers the interim look, has already occurred and that the Data Monitoring Committee is now reviewing the trial results. In short, an announcement is likely in the next two months. To put an 18-month median of progression free survival in perspective, the Company has estimated that the control group should have a median time to progression free survival of 12 months and our literature reviews and checks with doctors have suggested similar, if slightly higher, times for progression free survival. As a result of what appears to be a meaningful separation between treatment and control arms that we have modeled, we see a high chance of the HEAT trial stopping for efficacy at the interim look and only a small chance that the trial will stop for futility.
As with several of our other statistically-focused investments, we modeled Celsion by reconstructing the enrollment data for the trial based on public information, sensitizing the date for the timing of the interim look and the estimated time for progression free survival for the control arm, and running monte carlo simulations based on exponential curves for each arm of the trial. Below is a table that shows the modeled number of months of progression free survival as well as the range of progression free survival for the treatment arm based upon 99.9% confidence intervals at a 70% recurrence rate:
< see attachment>
Having probabilities on your side when investing is only one part of investing with strong positive expected values – the second is your set of valuations under various scenarios. We believe that the most likely scenario, a halt of the trial for efficacy, results in an expected value of about $10 per share based on our view that Thermodox should be able to sell at least $150 million into the RFA market for HCC in the United States and Europe. We think there is a 60% probability of this. If the trial continues onto the final endpoint, we see the stock fairly valued based on the offsetting factors of a high likelihood of ultimate approval on the one hand and the time and expense of continuing the trial on the other hand. We place a 35% chance on this outcome. In the case of stopping the trial for futility, we see the stock trading to $1.00 or approximately the cash per share that is currently on hand. We place a 5% probability on this outcome. Our expected value is therefore roughly $7 per share or about 170% above the current market price.
Variant View
Our view differs from consensus in that we have generated highly sensitive confidence intervals for the upcoming interim look. Specifically, we have used a Metropolis Hastings Markov Chain Monte Carlo algorithm for modeling using derived distribution of lambda given data without differentiating between 2 different groups to generate confidence intervals for the upcoming interim look. While several other parties have tried to calculate the median time of PFS in the trial, they have been unable to determine whether or not this will result in a statistically significant result given the high threshold for efficacy that needs to be met for an interim stop. Our analysis / modeling overcomes this barrier.
AI
