Short interest activity has been relatively quiet in the last months as the price per share was going higher.
A well known critic of Cel-SCI recently published a paying article which we do not recommend to purchase, not because we do not want to blind investors about the real risk of this company, but because it is a mixture of obsolete information _which has been repeated for years and years by Cel-SCI short interest and bashers_ and poor due diligence.
This piece of news has been published after hours and followed immediately by a short interest attack on after hours transactions pulling the pps down of 4-5%. We do not see this as a coincidence as we believe this relatively non-costly action of pulling down pps with a relative low number of shares is intended to scare the average investor to allow a cheaper exit window to the heavily invested short interest.
We will try to decode the information that has been given to us orally, not revealing the content of the original article which we did not purchase, and prove why it is erroneous.
A first answer has been addressed by North Shore Research
We will complement it by the following thoughts :
Regarding Opdivo and Keytruda
We think the key aspect of this analysis is that the blogger does not deny a longer survival than expected. It is probably the most important point to be noted from his article : the author recognizes an unusual survival pattern in this clinical trial !!!
The fact that he states that Opdivo and Keytruda are likely responsible for a longer survival in this clinical trial is however an incredible act of bad faith, as well explained by North Shore Research, the investors will have noted that very few patients from this clinical trial do reside in places where those drugs are used against Head and Neck cancer.
Provided that the US citizens of this clinical trial have all been treated with those drugs, most of the others have not, which is still huge, and with bad survival data as shown for instance in Bulgaria :
(source : Bulgaria Survival Rates)
So survival expectancy is pretty low in Bulgary and no approved drug whatsoever has been able to change this pattern... so far.
Indeed M. Blogger, all the possible conclusions aim to a better survival due to the effectiveness of a new drug !!!
Allegations that that Cel-SCI does not disclose IDMC information honestly while larger Pharmas do (like futility) because this is what happened in 2014 are erroneous. As we wrote in this article "CEL-SCI: On The Verge Of Approval" we know why IDMC expressed concern with a very limited number of patients treated and very few years of survival history and we know how Cel-SCI found corrective actions that seemed relevant to pursue the trial. Furthermore, since both FDA and the IDMC allowed the trial to continue in 2017 all IDMC decisions based on full set of unblinded data have been disclosed and those recommendations where that the trial should continue.
Allegations that Cel-SCI investors think that FDA had never a problem are a self-made statement. Every investor in Cel-SCI knows that FDA and the IDMC expressed concern in 2014 and 2016 and, most importantly, that those concerns were lifted.The allegations that IDMC and the FDA allowed trial to continue while they believed the trial had already failed are difficult to believe even for non-believers. So, American tax payers are paying for an administration that is following useless clinical trials ?
Ignorance on clinical trials methods to assess drug superiority
I was astonished by the lack of knowledge of a supposed biotech expert on the way endpoint success is measured,
First of all it would be unbelievable that FDA did not approve the protocol for assessing the superiority of the drug. This protocol is pretty standard and Cel SCI has not made up any kind of change or taken any distinct approach regarding it from the initial design of the trial.As indicate in the very official Clinical Trial site, the superiority of the Multikine Arm will be assessed using the logrank test on Kaplan Meier curves. The logrank test tests the “null hypothesis” and requires a large enough sample of data for both significance and power. Significance is required to avoid type I error and power is required to avoid type II error.
Type I error is the probability “p” to state based on a data sample that a drug is efficient when it is not in reality.
Type II error is the probability “β” of stating based on a data sample the drug is non efficient when in reality it is. In this study, p = 5% and β = 20%. The "power of the study" is 1- β = 80% in our case.
Finally, to assess the big enough size of the sample of patients that will require the very common p=5% and β = 20% values found in all clinical trials you need a pre-determined objective of effectiveness, what is called the “effect size” _ES .
Every clinical trial assesses an effect size. Cel-SCI had got approved by the FDA the p and the β value and the number of required events (298). The ES is the consequence of this and is the minimal detectable difference by the logrank test for given sample size, p and β value. The null hypothesis is the question “is my sample in control arm showing the same survival curve than my sample in the test arm”, it is the logrank test. The test will not detect any difference in the curves below the given ES for the study. The sample size is inversely proportional to the square of the difference of the ES.
Therefore for smaller ES to be measured with certainty, sample size should be larger. That’s why when you calibrate a clinical trial you will try a trade-off between a reasonable number of patients and ES. That is why Cel-SCI stated that more patients recruited (to accelerate clinical trial) would imply automatically a smaller ES would be detectable. That is also why Cel-SCI requires 298 events to happen for a 10% ES. It is not very usual for a Biotech to talk about the ES (I will come back to this below), but there is nothing bad about it !
May be the FDA is not really aware of the 10% figure, but no doubt that the FDA has agreed upon the logrank test based on 298 events p = 5% and β = 20%, which means the same ! The “null hypothesis” , if tested negative, will be stating exactly that the ES has been detected above 10% !
The 10% superiority, in fact, is a way Geert Kersten has chosen to communicate to the investors a way that is understandable, a more simple way to explain on Corporate Slides to investors than to state that “the null hypothesis will be checked upon the logrank test that will be ran on Kaplan Meier curve with p = 5% and β = 20%”.
Risks do remain !!
Don't tell me I told you there is guaranteed success
Of course risks do remain, this is Biotech !
- Stronger Overall Survival : I do not believe the OS has improved much in standard of care for the average patient. However it is possible that patients have been better treated in this clinical trial. Inclusion criteria is quite strict, but still it is a tendency for the investigators to chose patients that will not pass away while they are being treated, or before surgery. This might improve by a few percent the OS (not 20% !) especially in the very first years of the trial. Then, unfortunately for those cured by Standard of Care only the illness will progress and survival will be shown as for most of the population. That is why I do not think that the surprising long survival observed is due to a much stronger survival than real life in both arms (as per SEER Database)
- Large dropout numbers : If the number of dropouts stands well above the 30% mark, the long time to completion of the trial could be explained alone by the reduced size of evaluable patients. A small sample size implies a longer duration of a trial as less people in absolute numbers would die than from a larger sample size. I believe that, concealing such huge dropout numbers would be a real breach of the law by the trial sponsor _until very recently, Cel-SCI scientific presentation was stating that they required 784 evaluable patients as per the protocol (less than 15.5 % of dropouts)_but is still a theoretical possibility.
Disclosure: I am/we are long CVM.