Description

Thesis

CEL-SCI (CVM) is small cap biotech company with 90% downside when the results of the Phase 3 study are announced later this year. Styx1003 had a very timely write-up on the long side in July of last year when the stock was sub $1.00. Now with the stock at $4 and results imminent, we believe the risk reward is very different and now is the time to short the stock.

The previous write-up does a good job explaining a lot of the background and the history of the company so we will focus on the most relevant points of the short thesis here.

The Phase 3 drug being tested, Multikine, doesn’t work, and both the government and a panel of experts charged with monitor patient safety in the study have attempted to stop the trial early because the drug is not efficacious and may be harming patients.  In addition, CVM has a history of associating with paid stock promoters, a going concern qualification, and obscuring the data and the truth to keep the stock pump alive. The catalyst of Phase 3 results being released this year means that the stock will soon reflect the true value of Multikine - zero.  

Summary

The Phase 3 IT-MATTERS trial should have been stopped years ago, but CVM managed to delay the trial so that once patients were no longer on active treatment with the drug, the trial would be allowed to "continue" so that data could be collected. This trial has the dubious distinction of receiving the following accolades:

- A partial clinical hold from the Food and Drug Administration (FDA) in September 2016 due to:

unreasonable and significant risk of illness or injury to human subjects

a misleading, erroneous, and materially incomplete investigator brochure

a clinical trial design that is deficient in its design to meet its stated objectives

- A recommendation from the Independent Data Monitoring Committee (IDMC) that that the study be closed for safety and efficacy reasons

- A follow on letter from the IDMC stating that CVM was unresponsive to their concerns and that they were deeply concerned about patient safety in the trial based on its review of cumulative data

- An SEC subpoena that is investigating among other things disclosure around the Phase 3 study

- An attempt to increase the size of the trial due to slower than expected event rates (before the FDA placed the trial on partial clinical hold and prevented this from occurring)

- Switching Contract Research Organizations (CROs) in the middle of the Phase 3 trial due to slow enrollment

- Being unable to pay their current CRO for in cash for services rendered and requiring issuance of CVM stock to provisionally forbear collection of payables

Clinical Overview and History

In reality this Phase 3 trial never should have been started. The Phase 2 data that "supports" the progression into Phase 3 appears to be have been tested in a different population than the Phase 3, and offers almost no evidence of clinical benefit.

The Phase 2 study was done mostly in T2N0M0 as you can see from the table below.

In contrast the Phase 3 trial excludes patients with T2N0M0 classification. The Phase 2 appears to be mostly done in Stage II tumors, while the Phase 3 excludes Stage II tumors and focuses on Stage III and IV. These are very different patient populations and it very surprising that they went into Phase III at all.

The staging of cancers is very important in predicting the survival. Below you can see the survival curves segmented by the T stage. For T1 patients their 5 year survival is close to 80%, while for T4 patients the survival is only 40%.

Lydiatt 2018 - Major Changes in Head and Neck Staging for 2018

The first Phase 2 trial only look at pathology. The second Phase 2 used a historical control group and did look at response rates. But was very small population with 21 patients treated and only 19 patients assessable (2 of the patients actually did not have OSCC). This study was not designed to measure overall survival. But that hasn't stopped CVM from making statements about how Multikine saves lives. Below is a slide from their investor presentations.

So where did they come up with an improvement in OS of 33%? If you read the fine print in the slide you see that it comes from a Survey of 55 clinical trials; advanced primary H&N cancer (published 1987-2007). So they ran a small Phase 2 study with 19 patients, and then compared it to their analysis of 55 studies that were published over a 20 year period. This analysis holds no water. This was presented as an oral abstract at a medical conference in 2007, but never made it to be fully published by a peer reviewed journal.

Another thing to note is that if you actually look up the abstract of this presentation the data doesn't even match what they present in their investor presentation. The abstract lists only 39 trials were reviewed and their publication dates span 1987-2004.

This analysis itself is already highly suspect, but without knowing exactly how they segmented out the different Stages of H&N cancer or if they examined only OSCC makes this analysis meaningless.

The Phase 3 trial began at the end of 2010.

In January 2007, CEL-SCI received a “no objection” letter from the FDA indicating that it could proceed with Phase III trials with Multikine in head & neck cancer patients. CEL-SCI had previously received a “no objection” letter from the Canadian Biologics and Genetic Therapies Directorate which enabled CEL-SCI to begin its Phase III clinical trial in Canada. Subsequently, CEL-SCI received go-ahead from the Polish, Hungarian, Russian, Israeli, Indian, Taiwanese and Ukrainian regulators.

CEL-SCI’s Phase III clinical trial began in December 2010 after the completion and validation of CEL-SCI’s dedicated manufacturing facility.

The original plan was to enroll 880 patients and get a 10% increase in overall survival in the treatment arm. The trial was scheduled to end when there were 298 death events.

The primary endpoint for the original protocol for this Phase 3 head and neck cancer study required that a 10% increase in overall survival be obtained in the Multikine group which also is administered CIZ (CIZ = low dose (non-chemotherapeutic) of cyclophosphamide, indomethacin and Zinc-multivitamins) all of which are thought to enhance Multikine activity), plus Standard of Care (Surgery + Radiotherapy or Chemoradiotherapy) arm of the study over the Control comparator (Standard of Care alone) arm. As the study originally was designed, the final determination of whether this endpoint had been successfully reached could only be determined when 298 events (deaths) had occurred in the combined comparator arms of the study. Under the original study design, the plan was to enroll 880 patients in order to be able to have 784 evaluable patients for the per-protocol analysis.

Then in April of 2013 they abruptly changed their CRO who was responsible for running the day to day operations of the trial. The previous CRO had been Inventiv Health, which CVM accused of enrolling patients too slowly. CVM ended up suing Inventiv and Inventiv countersued. CVM ended up winning on most counts in arbitration when it was all settled years down the road.

On April 23, 2013, CEL-SCI announced that it had replaced Inventiv Health Clinical, the clinical research organization (CRO) running its Phase III clinical trial. This was necessary since the patient enrollment in the study dropped off substantially following a takeover of Pharmanet by Inventiv which caused many of the members of the CRO’s study team to leave the CRO. CEL-SCI has hired two CRO’s who will manage the global Phase III study; Aptiv Solutions and Ergomed who are both international leaders in managing oncology trials. Both CRO’s will help CEL-SCI expand the trial by 60-80 clinical sites globally. As of April 2013, the last update given by CEL-SCI, the study had enrolled 117 patients. The 39 centers where the study was conducted include three centers in Israel where CEL-SCI’s partner, Teva Pharmaceuticals, has the marketing rights, and nine centers in Taiwan where the Company’s partner, Orient Europhama, has the marketing rights.

CVM ended up going with two other CROs - Aptiv Solutions and Ergomed. They also increased the number of clinical sites.

CEL-SCI has hired two CRO’s who will manage the global Phase III study; Aptiv Solutions and Ergomed who are both international leaders in managing oncology trials. The study is currently being expanded to about 100 clinical sites globally.

The agreement with Ergomed was an interesting one in that the CRO ended up being paid in a non-traditional way, effectively a risk sharing mechanism. Whether this is a sign that Ergomed had confidence in the trial success, or they were desperate for the business is up to interpretation.

In April 2013, CEL-SCI entered into a co-development agreement with Ergomed. Under the co-development agreement, Ergomed will contribute up to $10 million towards the study in the form of offering discounted clinical services in exchange for a single digit percentage of milestone and royalty payments, up to a specified maximum amount, only from sales of Multikine for head and neck cancer. Ergomed, a privately-held firm headquartered in Europe with global operations, has entered into five similar co-development agreements, including one with Genzyme (purchased by Sanofi in 2011 for over $20 billion). Ergomed will be responsible for the majority of the new patient enrollment since it has a novel model for clinical site management to accelerate patient recruitment and retention. For example, Ergomed has almost 25 physicians who can directly call on clinical sites to aid recruitment and retention. Some of the Ergomed physicians also have the experience of being clinical investigators themselves. CEL-SCI believes that this interaction on a physician to physician level is what is needed to help increase enrollment in the Multikine study.

Then in the spring of 2014 the Data Monitoring committee for the Phase 3 study made a recommendation that the trial be stopped. The follow is the language from the 2016 10K.

On one occasion, in the spring of 2014, the IDMC made a recommendation that the study be closed for safety and efficacy reasons.

Why did we link to the 2016 10K instead of the 2014 10K that surely must have similar language? Surprisingly, this language does not exist in the 2014 10K. This is the language from the 2014 10K (filed in December 2014 - their fiscal year ends in September). It is very convenient of them to specifically omit the language around the recommendation to stop the trial the IDMC made in the spring of 2014. What is more impressive is that they commend themselves for not having a safety issue when in fact the IDMC had serious concerns.

In October 2012 and again in November 2013, in an interim review of the safety data from the Phase III study, an Independent Data Monitoring Committee (IDMC) raised no safety concerns. The IDMC also indicated that no safety signals were found that would call into question the benefit/risk of continuing the study. CEL-SCI considers the results of the IDMC review to be important since studies have shown that up to 30% of Phase III trials fail due to safety considerations and the IDMC’s safety findings from this interim review were similar to those reported by investigators during CEL-SCI’s Phase I-II trials. Ultimately, the decision as to whether a drug is safe is made by the FDA based on an assessment of all of the data from a trial.

So why did this disclosure show up in the 2017 10K (but not the 2018 10k)? What finally made the disclosure change?

However in August of 2016 they realized that the number of deaths was occurring at a lower rate than they had anticipated. Thus they decided to increase the number of patients enrolled in the study to 1,273 (1,146 evaluable). This is a 31% increase in the number of evaluable patients they were targeting. They also increased the number of events that needed to occur. The number of deaths went from 298 to 394 - a 32% increase as well. Meanwhile they changed the protocol to detect a smaller overall survival change (6.5%). This is a bit odd, as if it had just been an enrollment issue they could have just increased the number of patients enrolled and not needed to have changed the required increase in overall survival.

In August 2016, we announced that the currently available data from the Clinical Study reflected that the accumulation of deaths in the study was lower than that which was anticipated based on reported literature at the Phase 3 study’s inception. If the number of deaths continued to be accumulated at the current rate, it had been determined that it would take longer than originally planned to complete the study. To minimize this eventuality, we decided it would be necessary to enroll up to 1,273 patients to have 1,146 evaluable patients. There were also other changes in the protocol, such as the required number of deaths (394) and the required increase in overall survival (6.5%) in favor of the Multikine comparator arm. With this increased patient enrollment, we expected a corresponding increase in the number of deaths, and, if this plan were implemented, the study could be completed in a more timely manner. As required by law and in order to be able to implement the plan, we submitted an amendment to the existing Phase 3 protocol for our head and neck cancer study to multiple regulatory agencies in the countries abroad where the Phase 3 study is being conducted as well as to the FDA to allow for this expansion in patient enrolment.

It appears that the effort to increase the trial enrollment flagged additional scrutiny at the FDA because on September 26, 2016 CVM put out a press release announcing that the FDA had placed the Phase III IT-MATTERS trial on a partial clinical hold. In this case the FDA allowed existing patients to continue treatment but no new patients can be enrolled. So certainly a bad outcome, but not the nuclear option where the FDA could have requested that all patients stop treatment. Though given the way the drug is administered in the neoadjuvant setting only a small number of patients would have been receiving active Multikine treatment.

Finally in the 2016 10K we learn some data that had previously been withheld. We learn that the FDA is concerned about harm to patients. The there were 2 previously undisclosed recommendations made by the IDMC. The first was made in the Spring of 2014 to close the study entirely, and the second in spring of 2016 to stop enrolling new patients.

In its partial clinical hold letter, FDA identified the following specific deficiencies: a) FDA stated that there is an unreasonable and significant risk of illness or injury to human subjects and cited among other things the absence of prompt reports by us to the FDA of IDMC recommendations to close the study entirely (made in spring of 2014) or at least to close it to accrual of new patients (made in spring of 2016); b) FDA stated that the investigator brochure is misleading, erroneous, and materially incomplete; and c) FDA stated that the plan or protocol is deficient in design to meet its stated objectives. In its partial clinical hold letter, FDA also identified the information needed to resolve these deficiencies. In addition, FDA’s partial clinical hold letter included two requests relating to quality information regarding our investigational final drug product, which were noted by FDA as non-hold issues. CEL-SCI believes that its response submitted to FDA on November 18, 2016, addressed each of the deficiencies identified by FDA including detailing our belief that, under the applicable FDA guidance, there was no obligation to report the cited IDMC recommendations to the FDA at the time they were issued, and it also requested a face-to-face meeting with FDA, and outlined our commitment to diligently work with FDA in an effort to have the partial clinical hold for the study lifted. On December 8, 2016, the FDA advised CEL-SCI that the agency was denying CEL-SCI's request for a meeting at this time because FDA's review of CEL-SCI's November 17, 2016 response was ongoing. CEL-SCI was also advised that it will be receiving a letter addressing CEL-SCI's response by December 18, 2016.

Then in an attempt to lift the partial clinical hold, not only did they fail but by default the FDA upgraded it to a full clinical hold as patients had finished treatment with Multikine.

"Until you have submitted the required information and we notify you that you may initiate clinical studies, this IND remains on clinical hold and you may not legally initiate or resume clinical studies under it.” The heading on the FDA May 2017 letter is “Continue Full Clinical Hold”, whereas the heading on previous hold letters was “Partial Clinical Hold”. The FDA communicated to us that the IND is on Full Clinical Hold because the exception to the full hold (Patients enrolled in our Phase 3 head and neck cancer protocol prior to September 26, 2016, may continue to receive protocol-specified treatment at the discretion of the treating physician) no longer exists, since all patients had completed planned treatment with our investigational product, Multikine."

The final lifting of the clinical hold in August 2017 was anti-climatic as the company later stated in December 11th, 2017 that they would not be enrolling any additional patients into the study (even though in August 2016 they had determined that they needed to enroll more patients. Thus the lifting of the clinical hold is not meaningful if they are not enrolling more patients onto active treatment - as is the case. The trial is now "fully" enrolled with 928 patients. Active treatment of the patients with Multikine was completed in 2016, at this point allowing the trial to continue is simply a data collection exercise.

Phase 2 showed no evidence of efficacy and predicts Phase 3 failure

As we showed before almost all the patients in Phase 2 study were T2N0M0 patients, or Stage II patients. Stage II patient survival is quite good, and is the more relevant comparison (given 15 of the 19 patients were Stage II) than the advanced population comparison cited by the company in their literature review. It is important bear in mind that a more traditional or appropriate clinical development plan prior to entering Phase 3 would have been to run a randomized controlled study in Phase 2 instead of cutting corners and running an exploratory study from a single site in Hungary.

You can see below that survival in Stage II Head and Neck Cancer is almost 70% at 3 years and approximately 60-65% at 4 years. Therefore the 63% OS cited at 3.33 year median follow-up by CVM in their Phase 2 follow up seems fairly pedestrian. It is important to note that this follow up of 3.33 years for CVM may be post-surgery while the charts below are from diagnosis, so one may need to add back 1-2 months for time from diagnosis to surgery.

https://www.seattlecca.org/diseases/head-neck-cancers/head-neck-cancers-overview/survival-rates

Other things to note are that: the JCO 2005 paper cites 21 patients treated while the 2007 paper notes that there were 22 patients treated. Consent for follow-up was only obtained it appears for 19 patients, but for something like OS you should be able to obtain death certificates. There were also 2 of the 19 patients that were included in this calculation even though they were not found to have OSCC. CVM did not present sensitivity analysis explaining if the statistics would have changed if these patients had been excluded. Though we will note that one of these patients had adenocarcinoma of the floor of the mouth which at Stage 1 or 2 appears to have higher survival rates than OSCC.

OS was calculated using the entire treatment population that consented to the follow-up portion of the study (19 subjects), including two subjects who, as later determined by three pathologists blinded to the study, did not have oral squamous cell carcinoma, or OSCC. These two subjects were thus not evaluable per the protocol and were not included in the pathology portion of the study for purposes of calculating the complete response rate, as described below, but were included in the OS calculation.

In summary, at Phase 2 trial that used historical controls, different population, and unpublished literature analysis to move into Phase 3 has virtually no chance of showing success. The FDA's actions and the IDMC (which has access to unblinded data if they wish) both tried to get stop this trial. CVM was able to stonewall taking action to stop active treatment with Multikine and thus the trial is allowed to "complete" to collect the final OS data.

Phase 3 Readout Timing

In September of this year, all patients in the trial will reach their 3 year anniversary. Given the Stage III or IV patients enrolled in this study we should expect results this year though exact timing is uncertain. Clinicaltrials.gov does note that the end of the study is September 2019 though that is likely a guess and just chosen to be 3 years from last patient enrollment.

Valuation

The company has 33.7M shares outstanding according to their latest filing. They also have a number of warrants outstanding. There are 10.7M warrants outstanding of which approximately 7M have strike prices in the money (<$4.00).

Warrants outstanding from 10Q filed for the quarter ending March 31st, 2019

CVM does not have anything else in the pipeline other than a pre-clinical program. While they do have $16M of property plant and equipment listed on the balance sheet it is not clear what they would be able to get for it to a 3rd party.

They ended the March quarter with $5.5M in cash and burned $3.6M on operations. As you would expect they have a going concern qualification in their filings. The cash burn this last quarter was partially offset by $2.6M in cash from exercise of warrants. On April 15th, the company put out a 8K announcing that they had received $8.9M since the beginning of 2019, which means an incremental ~$5.3M in cash had been received from warrants in the first 15 days of April.

With cash remaining able to fund operations until approximately the end of the 3Q19 (before proceeds from additional warrant exercises), we believe the stock is worth close to $0 per share when the Phase 3 trial fails.

Other observations

CVM received a subpoena from the SEC as per their 2016 10K, but no mention of it in their subsequent filings.

CEL-SCI has received subpoenas from the Securities and Exchange Commission, which is conducting a non-public, private investigation relating to certain of CEL-SCI’s private and public financings as well as reports, articles and other publications prepared by third parties concerning CEL-SCI, the pending arbitration between CEL-SCI and CEL-SCI’s former CRO, inVentiv Health, and CEL-SCI’s Phase 3 clinical trial. This is the first SEC investigation involving CEL-SCI. While CEL-SCI is confident that it has the appropriate policies and procedures in place to ensure compliance with all SEC rules and regulations, CEL-SCI cannot predict when the SEC will conclude its investigation or the outcome of the investigation. CEL-SCI is cooperating fully with the SEC in this matter.

CVM has issued stock to pay for "investor relations" services.

During the six months ended March 31, 2019 the Company issued 140,233 restricted shares of common stock to consultants for investor relations services.

Perhaps related to the above, certain Seeking Alpha pieces written by a consultant have been turned into press releases. Adam Feuerstein interviewed Sharon di Stefano who published a number of pieces for companies such as Neurotrope and CVM that were then turned into press releases.

Catalyst

Phase 3 trial failure later this year.