Cel-Sci ($CVM): Why expecting positive Phase 3 data in June 2021?

My opinion, no investment advice. Twitter: @dinhvesting

Why is it unlikely that Cel-Sci will report data in May?

On May 18, 2021, Cel Sci issued a letter to shareholders which states:

Ok, so what does it mean?

• On May 11, 2021, the date of the letter, the statistical analysis is still being conducted by the experts.

• Then the results will be double-checked by another statistical team, a typical peer-review process to assure quality of the analysis. This double-check could take easily a couple of weeks given the quantity of the information, and the primary analysis itself has already taken 6 months.

• Then ICON will communicate the results with Cel-Sci.

• Then Cel-Sci will have up to 4 business days to file an 8-K to communicate the results with its shareholders.

Given there are only 14 business days between May 11, 2021 and end of May, it is unlikely that all these events can happen in May.

Why is it likely that Cel-Sci will report data in June?

Several reasons here:

Number 1: Geert (and likely ICON) initially expected data much sooner than June.

On December 7, 2020, Cel Sci announced in a press release that data lock had been completed and that “the statistical analysis of our Phase 3 study data is being performed according to a statistical analysis plan that was approved in advance of data lock.”

That means by December 7, 2020, ICON should have had given to Geert a preliminary ETA.

One week later, on December 15, 2020, Geert presented at the 13th Annual LD Micro Main Event Conference. At 15:55 of the presentation, Bill Langein, the founder of Sana Currents threw Geert a joke:

“We’ll see the data in the middle of June?”

To which Geert answered:

“Ha, that would really really surprise me if it took that long, really really really surprise me. Thank you. A good one though.”

So Geert, with the preliminary ETA indication from ICON, did not expect data analysis to take that long. My guess is that he would expect something early April just ahead of the usual shareholder meeting time around mid-April. The reason of the delay is unclear, perhaps ICON underestimated the workload or the delay was related to Covid-19.

Now the Geert’s big surprise is becoming reality.

Number 2: Shareholder meeting has been scheduled on July 1, 2021

If June was already a big big surprise for Geert and ICON, data in July and August or later would have been a huge surprise for him and ICON.

On May 18, 2021, Cel Sci announced that that shareholder meeting would be held on July 1, 2021. Last year shareholder meeting was held on April 17. My guess is Geert has been dragging on fixing the date since 2 months in order to have certainties on the ETA from ICON. It would be hard for Geert to face CVM shareholders without the results.

Number 3: We are not far off

On April 26, 2021, David Moadel posted a tweet about Geert rejecting an interview invitation. The reason was “We may well be really close to getting our final data readout.” If the data were still 3 months away, I would not expect such an answer.

David Moadel @davidmoadel

$CVM Interview rejected for now, but it's understandable $SPY $QQQ $DJIA $DIA #stockmarket #investing #finance #stocks $AAPL $TSLA $AMZN $INTC $NFLX $AMD $NIO $GME #economy

So all in all, I put a high probability for a data readout in June.

Why Phase 3 data is likely to be positive?

In the latest shareholder letter, Geert reminds again about the results of the Phase 2:

Multikine increased overall survival by 33% in a final Phase 2 study when compared to published data from 55 clinical trials in the same patient population. In addition, some Multikine treated patients experienced complete tumor elimination, confirmed by pathology, in just 3 weeks before surgery in these Phase 2 studies

Now that was Phase 2. But remember the phase 3 is a event-driven study (study ends when 298 deaths are reached) and it took 2 years longer than expected. Based on available data and Fosco’s initial analysis, I’ve done the following statistical model. Click here to download.

This model works in 4 steps:

• Step 1: Enter the drop-out assumption. Typical drop-out is around 15% for this type of study.

• Step 2: Choose the source of survival data for the placebo arm.

• Step 3: The model will then find the best parameters of log rank function to fit the observed number of deaths on various dates. As you can see below, the model works pretty well with the fit measure, or average error between the model and reality of 2.8 deaths per observation.

• See the results highlighted in orange. Overall, the 3 years survival improve is between 24% to 57% depending on the survival rate source. 5 years survival improvement is between 34% to 67%. These give a healthy buffer vs. the 10% improvement target.

Now when we play with the assumption, we will see that drop-out is a key factor. Above 25% drop-out, the results will be at risk. However, Cel Sci has confirmed several times that drop-out is not an issue like in this press release:

We do not think the increased survival of the patients in our study is related to the dropout rate of patients either. While we are blinded to the study results, we have never heard from the CROs who run the study that the dropout rate is a problem. In addition, the actions by the Independent Data Monitoring Committee (IDMC) speak against that as well. The IDMC, which meets periodically to review the study and data in an unblinded manner (i.e., they see everything) is tasked with focusing on the following areas:

• Efficacy: to assess the primary efficacy measure as well as the conditional power and sample size

• Safety: to assess the magnitude of adverse events and monitor for safety concerns

In reviewing the conditional power and sample size the IDMC would be considering the dropout rate. If there was a problem with the dropout rate, then per the Charter the IDMC is required to tell us to enroll more patients. They have not done so. As recently as October 2019, the IDMC recommended “…to continue the trial until the appropriate number of events has occurred”. In their letter to us they said that they reviewed “…progression free and overall survival and limited demographic and safety data available for the aforementioned protocol.” This language tells us that they are following the guidelines outlined in their charter.

My analysis is based on public information and my humble statistical knowledge. I expect Geert, although blinded to full data, has hired professionals to do the same analysis with more insiders’ information. That’s may be why he is so confident as illustrated by his recent tweet:

Geert Kersten @GeertKersten1

$CVM

Conclusion

I am expecting positive data release in June. Given the potential of Multikine to be the new standards of care for head & neck caner and the ability to use it off-label for solid tumor cancer, Multikine will likely be a blockbuster if approved.

Based on this huge potential, my valuation for Cel-Sci is conservatively between $10B to $20B or roughly $200 to $400 per share or 10-20x vs. current share price.

Based on public information and analysis, I give 80+% success probability.

I’m long Cel-Sci. Please do your work before investing.