Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
To those of you concerned about the publication - the fact that it has not appeared yet, or may not appear even in the next couple of weeks, has no bearing on its' quality or its' eventual publication status. I publish for a living, and can tell you it is often a long and painful process (but worth the wait). Reading between the lines of what was said at the ASM, I think they are at least at the stage where they have submitted revisions and are awaiting a response from the journal reviewers as to whether the initial reviewer concerns have been adequately addressed. They might get it accepted as is, or might have to go through another round of revisions. Then it has to be sent to the publisher, galley proofs edited, etc. It could come out tomorrow or it could be April or May, but it's definitely coming. If they did not get a positive response from the journal initially, I see no reason for the delay of the ASM, especially since there were no votes on any key issues that took place. Clearly, the company is as irritated as the rest of us at how long it is taking, but that is how the sausage gets made
Raf - That is what I was assuming. They are way more expensive than print ads.
Cool - That would be a reasonable person to do it. Print ad I assume? I know he was listed as a person taking V on their website.
FWIW - I was thinking about Bosch citing Bohsie and SOS in his presentation a while back, which many laughed at. I am wondering if the company's blinded analyses have shown exactly what those two had concluded, and because their posts were already public information, citing them was a way to say "we have found the same thing in our internal analyses" without having to spoil the data to be presented in the upcoming publication. Just a thought.
Actually - They have talked about a trial of direct to consumer advertising even before the readout. I think HDG indicated he agreed with my opinion that it made no sense for the company to do this stock deal now if they simply wanted to raise cash for use after R-It fails. They could have raised much more by waiting til closer to the readout after the run-up, which will assuredly happen even if R-It ends up a failure.
In event of failed primary outcome, what happens depends on whether V is found to reduce MACE in specific subgroups (e.g., diabetics). If so, given that it is already an approved drug and the first amendment win, they could legitimately promote it for use in a larger population than covered by the current indication. For example, diabetics with moderately elevated TGs. This would increase revenue and PPS (eventually) even in context of a formally failed trial. I am sure the company has planned for this contingency.
Zip - Maybe the increase to mid-4's was driven by insiders who knew what was coming and hoped to profit off short sales once news came out? I still think no real run-up will occur until early June based on the 60% and 80% IA behavior.
My guess is no real movement until next Fall (when results of re-do of phase 3 with booster should be announced), or if there is a crisis somewhere involving Zika, Ebola, etc that gets prominent news airplay. Always the possibility of a surprise buyout announcement unexpectedly though.
AVI - Interesting idea. So you think the paper could report even a failure of PFS as primary, but then slip in positive info on the long tail OS results? The company's statements seem to suggest that they are excited about the results presented in the paper, and certainly reporting failed PFS would not be exciting to them. Is there a precedent for this type of mixed blinded/unblinded publication? If you are correct, this could certainly help them get this manuscript describing blinded long-term survival findings (their top focus now) into a top quality journal.
Regarding the number of authors, who cares. Whether 5 or 65 is not going to really matter. I don't think anyone judges the value of science by the number of authors associated with it. The only thing I take from the company's "number of authors" comment is that they had a lot of administrative hassles in getting this manuscript done.
It is quite common in many journals to provide a review due date that is two weeks after the reviewer accepts the review (although reviews sometimes are late, prompting automatic e-mail reminders to do the review). If you factor in editorial processing time, a 4-6 week review process is very do-able for an initial submission. When you then have to factor in getting the initial reviews sent out to the authors, having time for authors to do revisions, getting all authors' input on the revisions, and then starting the review process again for the revision, that is when it starts to drag out more. Even after it is accepted, they will have to have the publisher prepare galley proofs and have the first author review the proofs, followed by a delay until the journal issue with space to publish it is available. Some journals publish electronic-only versions of the publication (e-pubs) once the proofs are corrected, and this could speed up release a little bit.
Mr. Main - Think about it. Everyone knows there will be a run-up to a much higher PPS even if R-It ultimately fails. If all the company wanted to do was screw us and make sure they had plenty of cash when R-It fails, it would make more sense for them to wait until July to does this offering, after the run-up. Instead, they do it now at a lower price, with the reason given that they want extra money for direct to consumer advertising. Doesn't that make sense from a strategic perspective - get money now to start right away promoting V more widely so that if (when) R-It succeeds, they are off to the races right away? Seems like a reasonable and pretty transparent transaction.
No - The 5.2% noted for placebo rate had to do with the assumptions in their statistical power analysis.
Mr Main - One way to look at it is that the trial is on track to be powered to detect treatment effects with the placebo rate exactly where they thought it would be (after adding 1,000 patients). I assume that the 5.2% they used in their updated power analysis is actually lower than they thought it would be in reality (if they were trying to be conservative).
Not correct - I have much more than that bought through Fidelity.
Thanks for your efforts Kjones - Your modelling showing a 4.72% composite rate is consistent with what my simplistic estimates based on total patient years and 90% of events recently reported produces. A placebo rate of around 5.2% or more is going to be required to meet the 15% RRR target.
Flip - The other thing to mention is that anyone who is a scientist or academic MD is not going to be swayed by some internet idiot's warnings to "stay away from that publication." The authors actually know how the study was done and what the data show. They would have confidence that it is real. Since the results would be groundbreaking to some degree even though still blinded, I cannot imagine anyone initially wanting to be an author would change their mind, and lose their opportunity to be associated with this. They are not trying to publish this because the data are bad. The only way they get these blinded data published in a decent peer reviewed journal is if the results are outstanding and the arguments presented are very compelling. Everything indicates that this is the case, so I am very much looking forward to seeing it. Should be a fairly convincing summary of what the final OS readout will look like (even though PFS will be harder to interpret in blinded fashion and I suspect will not be highlighted in the article)
If you are serious, no, not very likely. Much easier to drop authors than add them.
Flip - As in "LP: damn, the publication is not going to be ready by the conference date"? She previously cancelled that conference in London due to pending blinded data release on Sept 22 (day of the 8-K I think). This might indicate they submitted the manuscript before that date rather than November as Ike seems to think was being implied at the ASM. Obviously both would be "in the Fall" though.
Ex - I agree that consent issues would make sense as an alternative reason for the hold. The fact that the hold was lifted indicates that the FDA did not think it was a safety issue for those who had already initiated treatment.
Good find Long. Sums up the issue nicely.
Agree that there is definitely a pub waiting in the wings. The company has said so verbally. Their behavior also says so - cancelling two previously scheduled conference presentations at the last minute "pending release of blinded interim data," then agreeing to the January conference presentation late presumably because they thought the paper would be coming out by then (then cancelling at the last minute when it did not appear). They would not have done the latter unless they had received fairly encouraging initial reviews of the manuscript. Given my publishing experience, I suspect they sent in a revised manuscript in early December or so that they thought would have been accepted by the conference date last week, and that either: a) processing of the revisions by reviewers was delayed due to the holidays (many reviewers fail to respond quickly over the holidays), or b) the reviewers unexpectedly asked for more changes before the journal accepts the article. This could mean publication any day now if (a), or might be another 6 weeks or so if (b). The company's statements and behavior indicate that they feel they are past the point at which articles are outright rejected (usually on the first review, within 6-8 weeks of initial submission).
Just spitballing here - Maybe the interim analysis was done at the originally planned time. I think it is certainly plausible that PFS appeared to have "failed," maybe related to the difficulty of accurately identifying pseudoprogression using their original study protocol. Maybe it even looked like DCVAX was making progressions occur faster than in SOC alone. This could make the FDA say, "you need to stop enrolling because the patients might be harmed by the treatment." The company could come back and say "it may look that way but here are some data indicating that this is not real progression. In fact, our patients appear to be living much longer. We think the drug is having significant clinical benefits." After some back and forth, eventually the company convinces the FDA that their arguments are valid, and the FDA lifted the hold because no new patients would be exposed to the treatment, so there was no added patient risk. The company then decided to continue the study as long as possible to maximize the OS data (going beyond the minimum target), knowing that PFS may look bad (but for explainable reasons). They would have even greater incentive to do so given the apparent FDA changes planned for these types of immunotherapies, particularly given that this is an orphan indication. The publication describing blinded OS data indicating that patients are living much longer than expected compared to a variety of SOC data sources then comes out. Everyone begins to realize that the treatment actually works in terms of increased survival. The rest is (future) history.
AVI - I see your point from the graph, and agree that the event rate should slow over time as people event and "drop out" from the future first event pool. I just have a hard time believing that a company with this much money at stake would make such a basic error in making their projections. Maybe they generated this graphic for presentation purposes only without using real numbers for the curve, just for instructive purposes. I still assume (maybe wrongly) that their projections were more sophisticated in their assumptions (and I assume Deepak Bhatt at Harvard was involved), and that the "surprise" drop in event rate (i.e., the trial went longer than they expected) reflects at least some portion of a real reduction in event rate. Also, in followup to your comments about my loose projections (FWIW), I went back and assumed 1470 events (90% plus an extra 21 cases to correct for the small % of unadjudicated cases) and only 31,000 PYs (assuming about 33,500 currently but dropping 2500 PYs to roughly adjust for eventers still in the study). It still results in a placebo rate of slightly less than 5.2% resulting in reaching the 15% RRR target.
AVI - Good point re: 1st vs. 2nd+ events and continuing in the trial. Agree it is rough but it is the only way to even attempt to put a number on it. Obviously, people who die do not continue to accrue patient years. But, given the unavoidable error in these estimates, I agree it might be prudent to go with the worst case scenario I just posted in response to Raf, which requires a placebo rate of just over 5.2% to reach the 15% RRR target. This placebo rate is the most recent one assumed by the company in their power calculations, so in this sense the trial appears to be tracking their targets. Obviously, none of my calculations capture the actual survival curve over time. Mainly a way to have fun while waiting on actual results.
FWIW - I have no idea what is most likely, but under the scenario that you state is your best guess (1,550 events and 33,300 PYs), a placebo rate of about 5.05 or greater would be required to meet the 15% RRR target.
In your best case scenario, we hit 15% RRR at about a placebo rate of 4.72 or higher, whereas under your worst case scenario, the placebo rate would have to be about 5.25% or greater.
Given that the RRR projections I posted indicate the primary MACE outcome will meet 15% RRR under the most likely placebo rate scenarios, seems that the continuation was driven by a desire to get more events (and significance) on key secondaries (as JL and others have stated). Whether the DMC would go against their instructions and recommend stopping the trial for futility if that were the case would be total speculation. My guess would be they would though if warranted.
Based on prior similar trials, I think a placebo rate of < 4.8% is very unlikely, especially given the high proportion of diabetics that JL noted.
How do you consider their stated expansion plans anticipating success of the trial (including some changes costing them money even prior to R-It readout) as "jumping into lifeboats"? Ridiculous.
Kiwi - He acquired just as many shares as he disposed of...
Mr. Main - Agreed that overall event rate is lower than before. However, this fits with the final event being delayed from the end of 2017 (as stated last year) until March 2018 more recently.
Excellent news - Thanks for checking!
JL - Hope you are correct. I agree the high % of T2DMs in R-It makes a good RRR more likely.
Also have to wonder whether there might be pending news on the middle east approval? It is already overdue.
Raf - Agree that the calcs might not be entirely accurate, as there is wiggle room ("more than...") for both numbers in the calculations. They are estimates only, but I find it comforting to know that V is a drug showing a strong enough risk reduction effect that results should be significant even if the placebo rate is 4.8-4.9% (I believe the PSK9 trial last year had a 4.9% rate).
Good point - So, 33,000 PYs currently might be correct (average of 500 PYs month).
33,000 now (my read of it) vs. at final makes a difference - RRR will be lower if PYs are now below 33,000. They accumulate about 500 PYs a month. If now it is at 32,000 PYs, this means that the minimum placebo rate to reach the RRR target of 15% would be back to 4.9% instead of 4.8%. This is about where it was at the 80% IA.
Based on today's PR, we know that at least 90% of events have occurred (so minimum = 1,450 events) and that there are at least 33,000 patients years to date. Based on these numbers as a minimum, estimated RRR for the primary MACE outcome given various placebo rate assumptions are below. No way to capture the actual KM curve, but this provides a rough estimate:
Placebo Rate/RRR
4.60 /8.83
4.70 /12.89
4.80 /16.79
4.90 /20.53
5.00 /24.12
5.10 /27.57
5.20 /30.89
5.30 /34.08
5.40 /37.15
5.50 /40.11
5.60 /42.97
5.70 /45.72
5.80 /48.38
5.90 /50.95
6.00 /53.43
Zip - Sounds right to me.
The problem is that because the company is blinded (and don't know exactly how big an effect DCVAX has had), they are guessing as to when the right time is to stop the trial. If you think the patients are living longer, then the linger you wait, the stronger the results get. I think they are delaying in order to be conservative and make absolutely sure that there is the biggest home run possible. Stopping now might be fine, but there is no way for them to know. Waiting a bit longer increases their odds of success with the one shot they have.
NEJM or Lancet would be great, if it happens, but historically, they do not take blinded articles like this. I understand your point though.