What to listen for on CVM Investor Call
CVM reported its long-awaited Phase 3 results from its head & neck cancer trial for its immuno-therapy agent, Multikine, on June 29, 2021
The company press release had a "positive" tone, but the market reacted very negatively
Investors reacted negatively because the trial missed its primary endpoint
The company has taken a positive public view of the data based on analysis of certain subgroups
There is not enough data on the subgroup analysis for the market to share CVM's positive view of that data
On tomorrow's announced investor call, investors will need to see much more detail about that subgroup analysis, especially whether there was balancing of key characteristics for the experimental Multikine vs. control arms for that subgroup analysis.
See post-call comments below added on 7/1/21 at 10:20 AM.
CVM's long-awaited Phase 3 head and neck cancer readout for its immuno-oncology agent, Multikine, read out on June 29, 2021: "CEL-SCI’s Multikine® Immunotherapy Produces Significant 14.1% 5-Year Survival Benefit (62.7% Vs 48.6%) in the Group Receiving Surgery Plus Radiotherapy in a Landmark Head and Neck Cancer Phase 3 Study". Although the press release had a positive tone, the market reaction and primary readout were negative. Given the history of CVM, this trial, and the relationship between CVM and the biotech investment community and media, it should come as no surprise that this readout created a huge uproar (See e.g., FIG. 1). And CVM's CEO has engaged in some of the back and forth (See e.g., FIG. 2).
FIG.1 Adam Feuerstein tweet from 6/28/21
FIG. 2 Geert Kersten (CVM CEO) tweet from 6/28/21
As we pointed out in our subscriber post from May 17, 2021, with detailed facts, there are a lot yellow/red flags swirling around this CVM trial. And there are a lot of prior negative articles on the trial from STAT news/Adam Feuerstein (See this 5/17/21 article here and this 5/5/21 STAT article here) and others (e.g., see this previous Seeking Alpha article). Ahead of tomorrow's announced investor call to share the data that was announced last week, we provide further information to try to help investors to make decisions given this complex situation
Now that we have the actual June 29, 2021 data readout in hand, we are challenged with the task of trying to understand it and use as the basis for investing decisions. In our subscriber post from May 17, 2021, we concluded the following with respect to this catalyst readout: "From an investment standpoint, we find way too many yellow/red flags on CVM and this trial to buy CVM stock or make any other bullish investment in this company." The press release and the resulting media backlash highlights the challenges faced by retail biotech investors, as we discussed in the above forum post and our prior free blog post on the CVM readout: "As investors we are often caught in the middle of CEOs, who by their job description are excellent salespeople for their company. And journalists, who make a living on hyping up issues." Thus, although CVM may be an extreme example, it is common to read a "positive" clinical trial readout press release, only to find that the key predefined endpoints were missed. Furthermore, the CVM readout highlights another huge challenge for biotech investors: We need to understand complex diseases and data readouts, or have consultants/friends/advisors we can trust to understand and explain these to us, which unfortunately, in many situations is incomplete.
Why did investors react negatively to the CVM data readout?
Despite the positive headline: "CEL-SCI’s Multikine® Immunotherapy Produces Significant 14.1% 5-Year Survival Benefit (62.7% Vs 48.6%) in the Group Receiving Surgery Plus Radiotherapy in a Landmark Head and Neck Cancer Phase 3 Study, the study did NOT meet its predefined primary endpoint.
This is evidenced by a a very important sentence that is pretty deep in the press release (probably should have been a headline bullet point in the PR):
"When the complete study population to which the Multikine treatment regimen was administered (i.e., the combined lower risk (no chemotherapy) and higher risk (with chemotherapy added)) was compared to control, the study did not achieve its primary endpoint of a 10% improvement in overall survival."*
How could CVM put a positive spin on negative data?
The press release indicates that the trial is positive because of the following:
"Patients treated with the Multikine treatment regimen followed by surgery and radiotherapy (no chemotherapy) demonstrated statistically significant OS (ITT, p=0.0236, HR= 0.68) advantage vs. Standard of Care (SOC) alone; the 3-year survival advantage was 4.9% (72.4% vs 67.5%) and the 5-year survival advantage was 14.1% (62.7% vs 48.6%) for the pre-defined population receiving no chemotherapy."
Doesn't that positive subpopulation result warrant a positive reaction from investors?
Probably not. There isn't enough information in the press release, and given the limited information and the warning signs around CVM and this trial, investors are not giving CVM the benefit of any doubts.
An immediate issue that comes to mind with this sort of result (overall trial misses, but a subgroup hits) is were the baseline characteristics of the subjects in the experimental vs control arms balanced. For example, was their balance with respect to the numbers of patients in each arm as well as their disease state and any other important baseline/entry criteria. With respect to the eligibility criteria for this trial (see the clinicaltrials.gov entry for this trial (NCT01265849)), standout criteria include Karnofsky Performance Status (KPS) and tumor category. KPS categorizes patients based on their functional impairment (See this document) and although patients needed to have a KPS of greater than 70 to enter the trial, the CRO running the trial likely tried to balance the KPS scores between the Multikine and control arms for the entire patient population. What about for the low-risk patients?
As far as tumor category/grade, patients could enter the trial if their tumors were "T1N1-2M0,T2N1-2M0,T3N0-2M0,T4N0-2M0 (T4 allowed only if invasion of mandible is negligible)". Those classifications relate to the size of the tumor and lymph node metastasis (see e..g., CANCER OF THE ORAL CAVITY, Surg Oncol Clin N Am. 2015 Jul; 24(3): 491–508). Again, the CRO likely balanced those characteristics for experimental vs. control arms for the overall trial population. What about the low risk subjects? We see in the clinicaltrials.gov entry that "higher risk subjects" were "subjects determined at surgery to have positive surgical margins, 2 or more clinically positive nodes or extracapsular nodal spread". Thus, this may have been difficult to balance because the high vs. low risk status was not known until surgery was performed, but Multikine was given before surgery. On the other hand, maybe it was observed and could be assumed that if the stages at entry were balanced between the Multikine and control arms, that balance would hold true for low risk Multikine vs. control patients if lower stage at entry was much more likely to mean lower risk after surgery. It gives some hope to CVM bulls that these high risk vs. low risk groups were identified in the trial setup, as was the fact that standard of care would be different for these groups (See clinicaltrials.gov entry for this trial (NCT01265849)). Other criteria between these subgroup Multikine and control arms that would be good to see balancing include age, sex, and race.
In addition to the Multikine vs. control arm balancing issue, it is also unknown to me, whether CVM had the FDA's agreement in advance, that a positive readout for subgroup analysis of just the low-risk or high-risk, would be sufficient for approval. It seems unlikely. We don't see that specified in the clinical trials.gov entry. In the CVM CEO's tweet he indicates that the analysis was in the Statistical Analysis Plan agreed to by the FDA (FIG. 2). Again, investors aren't going to give CVM a benefit of the doubt here. That statement without more detail is ambigous.
So what do investors need to see at tomorrow's press conference to turn their reaction positive?
Much more data about the subgroup that yielded the positive results. Were the baseline KPS scores in the low risk Multikine vs. control patients balanced? Were their baseline tumor grades balanced? Did the FDA agree that a positive primary endpoint readout for the low-risk subgroup would be enough for approval?
If Mr. Kersten's tweets since the June 29 readout are any indication, we won't get that detail tomorrow. To his defense, it is common for companies to hold details of their data for a later publication at a scientific meeting or in a scientific journal. However, in this case, with so much negative press and with a primary endpoint miss, the investor community is unlikely to read much positive sentiment into the CVM press release, unless they release data such as that which I indicate above. It will be interesting to see what is presented to investors tomorrow. Although, I often like to play catalyst events, usually with a bullish bet, I continue to stay on the sidelines for this one. Stay tuned for tomorrow's fireworks.
If you are interested in CVM and this phase 3 head & neck cancer trial, there is another very long-term cancer immunotherapy trial that is pending data readout from another small cap biotech stock: The NWBO Phase 3 trial for its cell therapeutic product - DCVax-L in glioblastoma (For BPIQ subscribers see IQ card HERE). In our prior full forum post on BPIQ.com, our BPIQ team provides a detailed table comparing the CVM trial to NWBO's Phase 3 trial, as well as to a 3rd long-term cancer trial looking at overall survival. Check it out. The data lock for that trial was Oct 5, 2021, month's before CVM's and that trial has not read out yet.
7/1/21 10AM - Post call comments
I heard very little new data on the call today. CVM's CEO and CSO appeared to repeat the data presented in their prior headline. As far as I heard, none of the data above was presented. There was only a disclose that 308/318/380 (I couldn't distinguish from the CSO's words) patients were in the low risk group. We still don't know how many of these patients were randomized to the Multikine arm vs. the control arm. And we don't know whether there was good baseline matching of these low-risk patients between the Multikine arm vs. the control arm.
An interesting point was made on one of the questions that the CEO read: If Multikine was used in treatment, how would a physician know whether this was a patient who would be high risk upon analysis during surgery, when this assessment is performed later. What I thought I heard CVM's CEO say is that they have a test that provides this information before surgery. This seems to be another warning with respect to approvability of this treatment. I do not recall seeing anything about a companion diagnostic or other test that tells which patients are low or high risk before surgery.
All in all, I am disappointed that CVM did not share more data on the call today. The style of the company is very unorthodox in the way that they focus on share price and shorting on their investor calls. For me, I am still not investing in CVM and I am resisting making a negative options/short bet because I would prefer to share my views on bpiq.com and don't want to be accused of a bias because of my trading position. Plus, I am generally pro-biotech and like to make bullish bets on companies. CVM had every opportunity today to present data details, especially around baseline balancing for the low risk patients as I've mentioned in the article above. The question around determining which patients in advance would be low risk (benefit from Multikine) vs. high risk (benefit from chemo) raised on the call, is yet another question/uncertainty in the approvability of this treatment.
Manny Vacchiano and the Amp Bioresearch team at BPIQ.com
Footnote
*The clinicaltrials.gov entry for this trial lists the primary endpoint as follows:
"Overall Survival (OS) in LI + CIZ + SOC vs. SOC [ Time Frame: 3 year]
OS will be assessed using Kaplan-Meier life-table and compared using a logrank test and confirmed further with tumor stage location and geographic stratified log rank tests. The unstratified logrank test constitutes the primary analysis. A two-sided p-value of 0.05 or less will be considered statistically significant for comparing the two groups."
(See the clinicaltrials.gov entry for this trial).
And even going back to the original entry for this trial in clinicaltrials.gov (https://clinicaltrials.gov/ct2/history/NCT01265849?V_1=View#StudyPageTop), the final primary readout appears substantively identical.
It is not clear to us whether the 10% difference mentioned in CVM's press release was agreed to with the FDA and whether that agreement related only to the full dataset, or the low-risk or high-risk subsets on their own too.
This article is not investment, tax, or legal advice. Please do you own diligence and seek advice from professional advisors representing your interests.
Article history:
6/30 initial published at ~ 10PM Central
6/30 updated to provide info re: tomorrow's call ~11:35PM Central
7/1 Added post-call comments ~10:15 AM.
#CVM #mulitkine #NWBO #DCVAXL #DCVAX