• CEL-SCI: An Imminent Phase 3 Failure

Available for free download (on right) is our comprehensive analysis on Cel-Sci Corp. (select "Add to Cart" and checkout--no bank or credit card info is required).


Table of Contents:


Short Thesis Brief………………………………………………………………………….........….. 4

Corporate History…………………………………………………………………………............. 8

Pipeline…………………………………………………………………………………………........... 10

Capital Structure………………………………………………………………………………....... 16

Clinical Data Portends Harm……………………………………………………………........ 19

DMC Recommendations and FDA Clinical Hold………………………………...…... 43

Modeling the Survival Curve………………………………………………………………….. 75

Allovectin-7 Trial Parallels……………………………………………………………………... 97

L.E.A.P.S. and COVID-19……………………………………………………………………..... 103

Where are the Data?....................................................................................... 104

Grand Conclusion………………………………………………………………………….....…. 108

Appendix………………………………………………………………………………………..........112


Short Thesis Brief:

Cel-Sci (NYSEA: CVM) has been developing Multikine (“MK”) since the early 1980s, or for nearly 40 years. Initially, the drug was based on the novelty (and thus, patentability) of a “natural” derivation of the cytokine IL-2 for the treatment of cancer. As time went on, the Company’s language on MK shifted from IL-2 primarily to the entire “cocktail” of MK and the potential synergy of IL-2 with other cytokines in the mixture. Cel-Sci has to-date conducted only one randomized controlled (let alone phase 3) trial with any sort of clinical endpoints: the currently ongoing IT-MATTERS trial, a 928-subject trial testing MK (their only therapy in clinical development) in squamous cell carcinoma of the head and neck (HNSCC), or more specifically, oral SCC (OSCC). The trial hit 298 primary events (deaths) in late April of 2020, the target number to trigger the process towards data lock and final analysis. It has been ongoing since early 2011—a little over 9 years.

Paramount to the short thesis, the DMC has on two separate occasions (spring 2014 w/ ~150 subjects enrolled and spring 2016 w/ ~800 enrolled) issued formal recommendations to halt the IT-MATTERS phase 3 trial for “safety [harm] and efficacy [futility] reasons” (brackets ours). CVM responded to the first rec with unknown data that assuaged the DMC and the trial continued, and it responded to the second rec by seeking to amend the protocol to allow 1,273 subjects to be enrolled vs the originally designed 880—and, importantly, to increase total events from 298 to 394. However, when the protocol amendment was submitted to FDA and FDA became aware of the two futility recs (CVM had not notified FDA of either), they placed a clinical hold on the IND. FDA even stated outright that “patients enrolled in the study are exposed to unreasonable and significant risk of illness or injury” and “the study is unlikely to demonstrate superiority and thus should be terminated for futility.” This hold was not lifted until CVM withdrew the protocol amendment and closed the trial to further recruitment. 

Cel-Sci has conveniently removed any mention of the clinical hold and statements from the DMC and FDA from its most recent 10-Ks (2018 & 2019). The only 10-K it can be found in is the SEC-filed 2016 10-K (the 2017 10-K mentions the hold was imposed and then lifted, but gives no additional color). The absence of any mention of the hold or its cause from the most recent 10-Ks appears to be in violation of securities law, in particular the “duty” of “ongoing” disclosure of material information (e.g., https://www.sec.gov/about/offices/oia/oia_corpfin/princdisclos.pdf). 

Although the P1/2 studies testing MK in HNSCC produced some interesting data, including a handful of alleged responses, these “responses” could have been the result of a variety of factors besides a targeted immune response (e.g., inflammation causing “accidental” or “unprogrammed” necrosis, rather than programmed apoptosis/autophagy). Also, in the papers on these studies, the authors describe responses as portions of the resected lesions noted as fibrotic, while not meaningfully shrinking in size. These are clearly not valid responses by RECIST. 

Furthermore, the formulation of MK has differed across studies, and few of the studies conducted by Cel-Sci involve the same formulation used in the IT-MATTERS trial (e.g., Timar et al., 2003/2005—who at the time was the recipient of monetary compensation from CVM). But importantly, even if MK in its current form can sometimes cause shrinkage of OSCC lesions, the totality of the mixture would likely confer tolerogenicity/tumorigenicity rather than immunogenicity, due to the overwhelming presence of IL-8 and significant presence of IL-6 and TXB2 (a metabolite of TXA2 and representing COX-1 activity, implicated in the formation of metastatic intravascular niches). Corroborative evidence strongly suggests Multikine treatment led to an exacerbation of progressive disease/time-to-death, and is almost certainly why the DMC overseeing IT-MATTERS stated in earnest that they were “deeply concerned about patient safety based on [our] review of cumulative data.” 

The history of MK itself is sordid. The initial MK therapy (we’ll call it “MK1”), an unsophisticated and antiquated product developed in the 1970s, featured the Company’s licensed, “natural” human-cell derived IL-2 as the advertised active ingredient (with all other matter unidentified). MK1 was later shelved in favor of what we’ll call “MK2” sometime in the late 1980s/early 1990s. MK2, as it turns out, is indistinguishable from ImmunoRx’s “IRX-2” (since acquired by Brooklyn ImmunoTherapeutics) and developed by Dr. John Hadden, whom Cel-Sci sued, along with ImmunoRx, accusing them of “copying Multikine,” and were engaged in IP dispute from 1996 to 1997. And perhaps because they realized the case was futile (it was dropped in Oct 1997 without settlement), moved on to testing what we’ll call “MK3” in clinical trials. 

The MK1 patents (ca. 1980) did not disclose how much IL-2 (the only validated cytokine shown to confer a clinical benefit; e.g., “Proleukin”) the mixture yielded per 1 mL dose, but in Dr. Hadden’s patents (ca. early-mid 1990s), we find the impetus behind developing MK2 being the need to yield a higher percentage of IL-2 in the overall formulation than MK1, and that because of the very low levels of IL-2 in MK1, the process to make MK1, and MK1 itself, were deemed inferior. Given that the later (ca. 2003-2005) Cel-Sci patents for Multikine are exclusively method-of-use, and not only adopt the same name as was used in the 1980s, but break down the cytokine profile—including IL-2, which is quite low—and given that both the original 1980 patents and later patents report “serum-free and mitogen-free IL-2 containing supernatant” via “buffy coat cells,” we can safely conclude that MK1 and MK3 are one in the same. The later patents reveal the full cytokine profile of MK3 (and thus, MK1), and the IL-2 content per 1 mL dose is noted as only ~0.6 IU, or 0.18% (“preferably”). The language in the 10-Ks also alters from the point it became clear Cel-Sci could no longer use MK2/IRX-2, in that their touting of “IL-2” in the mixture is largely removed (ca. 2003 on).

MK3/MK1—which we’ll now refer to simply as “Multikine” or ”MK”—consists of a high percentage of IL-8 (est. ~72.6%), followed by IL-6 (~8.2%), TXB2 (~4.3%), MlP-1β (~4.0%), MIP-1α (~4.0%), and lesser amounts of 14 other cytokines, including, as mentioned, trace levels of IL-2 (~0.18%). Some components of MK may induce tumor cell lysis, such as TNF-α (although TNF-α can also have a tolerogenic effect), but the vast majority of what MK is—especially IL-8—is clearly tumorigenic (in fact, it might be more appropriate to think of Multikine as IL-8). These effects are apparent in the deleterious alteration of the CD4+/CD8+ ratio, EMT induction, and increased Ki-67 expression after MK dosing, inter alia. 

Modeling a blended event rate based on the clear enrollment curve provided by the Company and event milestones, together with a lower-risk subject demographic that likely make up the IT-MATTERS trial—especially when considering the ever-increasing prevalence of HPV16+ OSCC—fits well, and if anything shows the blended event rate should have been slower than that observed in IT-MATTERS. Thus, there could be no separation between groups, or even a better-performing control group, and the blended rate observed thus far would not be out of line. 

In sum, the DMC recommendations to halt the trial for futility, especially the 2nd rec in 2016 with ~800 subjects enrolled over a 5-year period; the DMC’s strong allusion to harm in the MK groups; shutting down of recruitment by FDA, noting futility directly, while also noting harm in the MK arms; a well-fitting or even underperforming blended event rate via lower-risk OSCC prognostic factors; clear rationale for why MK is very unlikely to benefit patients (tolerogenic effects of prominent cytokines in mix); elevated market cap (~$500mm @ $13/pps) and near-term catalyst of ~80-90% immediate loss of valuation upon reporting the IT-MATTERS trial failed to meet its primary endpoint, due to no other candidates (or even indications for MK) in development and little cash—make CVM a highly compelling short candidate.

Near-term price target: $1.25/share




CEL-SCI: An Imminent Phase 3 Failure

  • Product Code: CVM Report
  • Availability: In Stock
  • $0.00


Tags: Cel-Sci, CVM