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.
"If one Lithium mine has anything happen that affects production the Gigafactory will continue to operate."
I agree Nomad. In fact Musk seems to be very much about contingencies. I believe he is a man of average intelligence and fore site on the scale of the Divinci era in a world of foggy brained moderns including myself. He sees clearly the odds of upcoming potential disasters, and prepares accordingly. Ie manning Mars.
In that vein, in the near future, the US wants to ween itself from dramatic artificially low interest rates. The potential for negative scenarios that end with foreign goods costing a fortune in US $ is real. Maybe not overwhelmingly likely... I hope, but the chance is significant enough to warrant contingencies. Lithium is currently a foreign commodity. Just sayin.
I am not all in yet, so I will say no more positive things about this stock for now. My big $ are in, but my small Roth is still awaiting a more solid bottom.
Prudent for Tesla to buy WL for $500M now. That would be less than 10% of the cost of their factory. Oh... I'm long.
I leaped with a brief first look... upon seeing the surge and then hearing the news, so I am in at a higher basis than most here. So of course I want sugar, and lots of it, now.
But having looked further... I think I did the right thing. The two local competitors (Developed and undeveloped brines) do not seem to have much of a resource compared to WL. And WL could have their eye on other properties, awaiting the right moment.
Li may not quite qualify as a rare earth metal, but given the large volume projected use, some of the same rationale exists to ensure US control of whatever local assets exist. The same reasoning should compel Tesla to control their own destiny in that regard. It seems to me that their best option would be to acquire WL while it is cheap at around $4/sh, develop the current operations to the point of readiness at a large throughput... then put it on hold and use Li from foreign sources for now. That would give them a large, local, ready reserve... kind of like that multi-source power grid available to them at the plant.
With a little coinage available, WL (as a subsidiary of Tesla) could also acquire any local resources that they have their eye on. Further... by delaying large scale production, Tesla could continue improvements in mill / extraction efficiency before using up the ore/clay.
If Tesla waits to purchase WL, then a bidding war could start, which could cost them, and could lead to a foreign country owning the mine. I say do it now Tesla, for your sake and for the sake of the USA. God Bless America!
Pyrrhonian; You weren't talking to me. So pardon that... but the issue of early progression and cd4 counts is interesting.
True early progressors would correlate with exceptionally low cd4 counts.
Pseudo early progressors would correlate with exceptionally high cd4 counts.
Not saying there is a direct correspondence. But correlating does not require that. This damage to the immune system is just one factor. But when you separate it out, you would certainly see the above relationship.
Another primary determinate for progression would be the methylation issue. I speak indirectly about it because the methylation is a tad hard to remember for me. There is the methylation of the DNA repair enzyme which prevents that enzyme from functioning properly. Then there is the damage to the DNA that Temador does as it's normal function. It damages the DNA by methylating it. So their are two aspects of methylation.
At any rate, the patients with the damaged / methylated repair enzyme would correlate with pseudo early progression. I think it is correct to say that patients with the healthy repair enzyme would correlate with true early progression.
Although there are many other variables, as you say, I think these two parameters probably do a pretty good job of explaining most of the observed response behavior. And these two variables are the ones Bosch talked about as significant.
RRR: Are those Asian Pharmas all Japanese? Any Chinese?
If Direct pans out; The simplicity of it for the physician selecting treatment, and for mfg, both in terms of not needing to determine subgroups, would make it something that a government should consider buying rights to. The Chinese government being the most logical.
But it would not make sense to sell exclusive rights to a country like China. NW would sell China the right to duplicate the process for use only in China, and would continue to market the therapy to the rest of the world. They would sell that right to China for a loooooot of money. Not having to do surgery is an enormous cost reduction. If you get much improved efficacy and immune memory on top of that, then the value is hard to estimate.
Just how blinded is the trial to OS? Are the patient's names available? Home towns? SSN's?
That's not PFS, but they know the enrollment dates. If they know the dates that the corresponding patients passed away that is a lot of information. They could project PFS pretty well from OS.
That's for both legs combined, but that is still a lot of info if it is way off what they expect.
Evalute, "Some current/past examples:
Pending? Outstanding? where do things stand?"
Reflexively defending in I believe a fair way... this change to the terminology, in retrospect, likely reflected a change from expecting the results of a DMC review of efficacy to not knowing if the DMC was going to do a review of efficacy.
NW says they petitioned for the new trial terms 9 months ago. In the interim, the efficacy review was "Outstanding". They didn't know if there was going to be an efficacy review. But prior to the DMC or statistician deciding or recommending the petition for trial changes, NW thought they were going to be receiving a recommendation from that review. So at that earlier time the results of the efficacy review were believed to be "Pending".
Pyrrhonian, "The categorization of low CD4 enrollees as their own subgroup seems to have been the only accepted protocol change regarding them, in other words the inclusion of their analysis. But it doesn't seem possible that the direct result of their identification was sample size enlargement. Thoughts?"
It appears to me, as betweenthelines speculated, that the new small group to need statistical boosting is the complementary "not low CD4 enrollees" group. This would be the fallback group if the larger group fails.
With Linda Lau working on ways to monitor the damage to the immune system caused by radiation and or chemo via CD4 counts, there might be a decent method, down the road, of detecting how much radiation or chemo is enough. Where the precipice of too much of a good thing starts. When the toast is golden brown.
So, if the larger group fails, but that "not low CD4" subgroup passes, then the FDA would be compelled to approve DCVax-L with the provision of having to have sufficient CD4 levels. If that became SOC, then there would be many patients initially rejected, but there would be great pressure to test CD4 levels to see when too much radiation and chemo begins, and terminate it at that point. This could allow all patients to benefit from DCVax-L, at an unknown cost from less SOC. Thus that follow-up would probably have to be a label extension.
Not a word from AF (in his TheStreet article on yahoo finance news just now) about the IMVACS conference last night, even though it was in his home town. And he dissapeared off twitter for a couple of hours during that conference, suggesting he may have attended.
Bazaar. No word from anybody on this side of the fence, and now word from anybody on the other side of the fence. And the stock price is flat.
I don't believe that DCVax-L can restore the immune system in a lasting way once otherwise permanently damaged by radiation and or chemo. I have asked Flipper for his source for this information many times over the last 8 months, but he has never produced it.
I'm not sure that DCVax-L can even restore the immune system per se on a temporary basis. It can certainly restore dendritic cells, but that doesn't restore the rest of the immune system.
In fact, if the immune system is completely wiped out, then they would not be able to properly process / create the DC's in the lab to begin with. But assuming that there are a small number of the needed stem cells in the blood, then I wonder if there is a way for NW to generate more stem cells from them, rather than guide the maturation of the stem cells toward DC's. Rather, for some of the stem cells, in this case, it might be best to grow more stem cells rather than DC's.
If you inject such stem cells back into the blood stream, then you would restore all the types of immune cells that those stem cells can become, at least temporarily. That might not be all the immune system, but it would be a wide swath. Whether those stem cells would also repopulate the bone marrow or other locations that normally produce them would be of interest.
But I doubt that matured DC's could go backwards in development and become stem cells. Maybe. That is what would be required for even a temporary re-establishment of this wide swath of immune function. And again, for permanent restoration, it would also require these stem cells becoming the population in the bone marrow. I don't know if the stem cells in the bone marrow are the same types of cells that they produce and release to the blood.
There are similar questions about T-Cell production in the Thymus. I don't remember if that production stops at some age... but another dimension to the problem.
Not a word on IMVAC?
110 * 2 * (348/312) = 245 Meaningless. My apologies.
If there had been 110 events prior to this change in trial plans, then this math might make some sense. But there weren't. That was the end goal, but they were not there yet, of course. So...
Never mind.
Working in the medical instrumentation sector there was an FDA policy that I saw as horrific. It was the requirement that all obstacles in development be identified up front so that performance is accurately predicted in the marketing submittals to the FDA, long before the core development takes place. If the device performance is not as good as stated in those early marketing docs, then the device does not get FDA approval. Or this was my impression of how it works.
Such a policy ignores important truths in development. Engineering is about doing very quickly things you have done before. It is about using your training and experience. R&D is not just engineering. It usually has a scientific element which is more about speculation on how things that have not been done before are likely to work. The best of scientists do not do a perfect job predicting such things. To have to toss a project and start over because it did not meet the original marketing goals just doesn't make any sense. This begs for scientists to lie about data. And I am sure it goes on all the time.
It seemed to me that under-performance relative to marketing goals, in some aspect of a device, is likely common. And my experience is that it is very common.
After getting scrwd at that MRI company (that I talk about all the time) I was cast out into the temporary workforce. A common job available to temporaries with Validation and Verification background is the testing of devices that don't quite make the marketing requirements in some way. It is the job of the temporary to make the device pass. If you don't, they can get another temporary. And if you make waves, they can ruin you. Blackballing is very real.
How do they just get a new temporary. You can do V&V testing on a device at one of the companies and there will be no record of you working there. If things go badly, you worked for the temporary agency (who typically knows nothing about the work you perform for the client). If things go well you worked for the client.
The point is that the rigidity of the FDA in this regard creates horrible situations in industry. It should be fairly standard to simply modify the original marketing requirement. The FDA review of those changes should be reasonable and take a reasonable amount of time.
Here, it appears the FDA has finally taken that attitude. While LP's connections may have helped, there has been a great deal of pressure for the FDA to be a cooperative body rather than a competitive body in the development and approval of new drugs, such as cures for cancer. In this case it is not a change to the marketing requirements, but it is a similar change in attitude by the FDA that is required.
On the other hand it might not make sense to call it starting over again. The now have a very large pool of patients. Waiting for more events is not nearly as difficult or expensive as starting patients in the trial. If they were previously dealing with a huge percentage event rate, then they would be saturated for that early group. It would be very difficult for them to gather so many more, but they were not at a huge percentage event rate. Though the giant tail becomes less useful.
Not to say the numbers listed are not a serious squeeze for events.
110 * 2 * (348/312) = 245 which is very close to the new 248 events for the full trial.
This suggests that they do not have the CD4 details to correct the existing data in the post analysis, and had to start over. Sort of. Maybe.
I think this is consistent with what you were saying. For one thing, this puts the second 110 at 110 because it is expected to be grouped closer in time than the first 110.
But this is a little confusing... so I'm not sure that I am not full of crp here.
Thanks for throwing in that "Shoenfeld Formula" term Pyrrho. Now I just have to get off my butt and look that up.
I think that is definitely one of the possibilities. I wish it was the only possibility... but it is definitely a decent possibility.
What showing? There is no webcast... so far.
Could be the SA article dragged NW down. But I would think the PR by NW would be of more weight than a repeat of AF's article yesterday. Particularly with the head of the DMC talking so strongly against AF, and so clear on the blinding of the data to NW itself. Further, that DMC head being so credible and forceful alludes to the kind of meetings that went on between NW and MDA about the Dr. Butzar incident and future releases of Direct data... Given that a DMC, if not the same DMC, reviews that Direct patient data and presents it to NW prior to those PRs.
On the other hand, there was no morning PR announcing a webcast. I think that shifted the balance negative.
It would be strange to PR mid-day... but can they legally do so? That is, without a halt, which might be a little too dramatic even for the best of data.
I realize the view my last post describes would mean a change to what most think was in the works in the US; not just a change to the European side. Including all the early progressives means including the pseudo-progressives... which were believed to be separate from the main trial. But putting all the progressives, including the psedu-progressives, into the main trial in the US and Europe makes a lot of sense from the perspective of inclusion that AV kept pounding on. And as smart as some of you are, he had the most experience in dealing with the FDA.
I can't quite follow many of these arguments. Almost... but not quite. That gives me the advantage of being able to step outside of the box a little, while gaining from the gist of what I read.
Doesn't a move to include the early progressives into the trial in both Germany and the US fit? All that is missing is for the EU trial description which still says that they do not include the early progressives. Maybe that is the only typo here.
This would fit with the stuff AV or whatever his name was on IV kept saying about the FDA wanting to include all patients.
And for the trials in Europe and the US to have such different protocols doesn't make sense. So perhaps the PEI has approved some aspect of the change and is reviewing the rest. Maybe they did a partial update of the protocol.
A massive extension of events would be required if a subgroup had only a small improvement and you wanted that to show within the noise by the end of the trial. Pyrrhonian just did a quick guestimation of a hypothetical split population in the group of 55 between true and pseduo early progressors. The guess part is because we don't know the population split, if any. Assuming there was one or more pseudos in that group, and the true progressors showed only 5 months improvement in PFS, then it would take a lot of events to make sure that 5 months showed up in the end as greater than 4 months with statistical significance. True even if that improvement was 7 months... not sure how high it could be and still need this large increase in events to stand out clearly enough over the goal of 4 months...
This assumes that a subgroup by subgroup pass is required, with statistical significance of some kind. And the added events also gives some cushion for the aforementioned need to show better performance than competitors for all subgroups. Less noise gives a higher minimum performance that can be attributed to a given observed mean or median. As you all know.
Maybe not exactly right here... but I think this might be the gist of the situation. Probably already said by you guys a dozen times... but that is my view from out of the box.
I used to have a boss that would repeat things that I said in meetings, maybe 2 minutes later, with the preface, "I just had an epiphany!". But this does follow many things I too have said in the past.
But what constitutes a response?
---------------------------------------------
The 80% response rate comes from NW presentations. It has appeared on the corp highlights slide on multiple occasions. Here is a cut and paste from a Jan 2014 pres:
Quote:
Consistent & Striking Results In Diverse Trials
• Clinical trials in brain, ovarian & prostate cancers; >80% of patients respond.
• Median PFS & OS extended by 1-1/2 years or more beyond results with SOC.
------------------------------------------------
I noticed that too Austin, and have the same concern.
----------------------------
The 55 patients were not eligible for the Phase III trial because they were either definitely or potentially “rapid progressors”
----------------------------
As some have argued, post analysis by the Germans, if nobody else, probably eliminated the possibility that any of these were pseudo-progressors... but if that is the case, that needs to be said. As stated, I agree that the PR leaves open that possibility. In fact I said so after reading the PR and got shot down from about 4 different directions.
Any pseudo-progressors would certainly be in the methylated MGMT group. Maybe there are other known genetic traits at this point, for pseudo-progressives that could be looked for. Maybe more telling would be if the OS numbers for the 55 fell into two distinct groups. One large group at say... 12 months, and one small group at say 60 months, for example. Probably not the case because everyone would have noticed. But this would not be counter to their PR. If they gave us the OS data, we would be able to eliminate this minor possibility.
That makes sense to me Pyrrhonian. To allow statistical significance for these smaller subgroups based on CD4 count, you would have to have a lot more events. And this would be the right way to handle this situation.
And I think you are right that the placebo patients would not be grouped altogether here. A 1 for 1 comparison in corresponding subgroups makes sense. One minor alteration to Anti's model in post 17074.
As LP said, it is generous of the FDA to allow this change. NW was responsible for having large enough trials before Phase 3 to ensure, statistically, proper representation of such subgroups. And apparently they did not.
What makes it reasonable in my opinion for the FDA to allow this exception is that this damaged/dead immune system issue is total FUBAR for immunotherapies that should have been better anticipated by the rest of the medical community including regulatory. It was quick sand for all the immunotherapies that include radiation as part of SOC. I do not understand why this issue has not been extended to chemo. I read an article that showed that many patients do lose their immune systems, at some point, from Temador and that this is often the reason they die. But if the tell on that damage is also a drop in CD4 counts, then this method of sorting out the issue will work for the effects of chemo as well. Down the road, or sooner, they need to learn how much is too much for both radiation and chemo when used with immunotherapies, and these CD4 counts might prove to be the proper feedback.
One concern I have about these hypothetical plans for subgroup comparison by CD4 count is whether they have that data for all the patients. Maybe for living patients, they can just add that test at this point going forward, but what of patients that have passed away? Do they keep blood samples for post analysis? Did the white counts that were likely done along the way include CD4 measures?
-----------------------------------
My apologies to the MD's on this message board for lashing out at the medical community. You represent some of the most valuable posters on this message board. But I hope you understand my frustration. I think the downside of free enterprise and competition in medicine shows here. I am not saying it should be more socialized. I am just saying that the FDA needs to take on more responsibility in examining this type of issue if the industry is not going to do it. These individual immunotherapy sponsors should not have been on their own here.
But Temador does help the performance of DCVax-L. I have seen the graphs from the study which demonstrated this.
It is possible that radiation is also synergistic with DCVax-L. I haven't seen the study on that.
But as I have been saying about Temador for a long time now, maybe it makes sense to use a little less of those treatments now that you have a third treatment. In particular, now that you have an immunotherapy as a third treatment.
Best would be a test to give you some kind of feedback as to when the patient's immune system is starting to show effect from the radiation and chemo. That would allow each patient's radiation and chemo to be optimized individually with one protocal. Maybe that is what Dr. Lau has been working on.
Difficult with chemo and radiation both. If excess radiation is the primary culprit, the maybe you focus on that somehow in the way you do these tests. But I know Temador can also cause immunosuppression at some point for some patients. I saw that study.
But take the unmethylated MGMT group as a separate case. Temador doesn't help them much anyway, if at all, so one would think that the optimum dose to start with would be little or none. That would put it well on this side of doing damage to the patient's immune system. And that is a very large subgroup.
They should have found an alternative to Temador for that subgroup (such as that SA article talked about... though it would not cover all this subgroup) a long time ago. The FDA should have forced that. Why didn't they? How long have we known that Temador does not help the unmethylated MGMT population but can damage the immune system at some dose level for some people? Why is NW having to deal with this on their own, now?
This issue of immune system compromise by SOC in conflict with emerging immunotherapies is an issue of enormous scale and weight. It is like too galaxies colliding.
But it need not be a war. It needs to be a coordination of the two technologies as much as possible.
One little galaxy and one very big galaxy.
The PR today said in so many words that the results of the L trial are heavily effected by this issue of GBM patients so often having compromised immune systems. It was so strongly implied that if it is not the case it was misleading.
Nothing specific in the PR about DCVax-L, just about SOC for GBM in general, but the trial size is large enough, with 40% stated effected, that to not have been effected by this issue would be a miracle. If you assume, as I do, that the PR was not misleading.
Reefrad; does the staggering of patients effect the power only because some patients will not have had time to progress, or is there a separate downgrading of the value of staggered data over grouped data, as has been stated today? If so, what do they call that downgrading due to staggering?
PR Replay
-----------------
Important recent research has found that as many as 40% of GBM patients have such severely depressed white blood cell counts following radiation that their level is comparable to the level at which AIDs patients are put on continuous antibiotic treatments, prophylactically. Further, this research has found that these GBM patients’ white blood cell counts do not recover with the passage of time. See, for example: Grossman et al., 2011: Clin Cancer Res 17(16):5473-80; Ellsworth et al. 2014: Oncoimmunology 3(1):e27357. Epub 2014 Jan 3.
------------------
What does this mean to the DCVax-Direct trial? Since the efforts to make changes to the trial started 9 months ago, this was probably a well recognized issue before patient selection for Direct. That said, this means Hodge's wife probably has a healthy immune system, which is very good news.
The serious issue of compromised immune systems from SOC radiation and chemo was apparent before immunotherapies ever started. And so, it was an obvious issue and an obvious obstacle for immunotherapies.
For that reason, it is very disturbing that the extent of immune system damage for this significant fraction of GBM patients is news. I'm not blaming northwest. I am just saying this is irresponsible of the medical community in general. Irresponsible for this to be news at this late date. Unbelievable. But I do believe it.
I am not certain what to make of the entire picture with the massive increase in total events for full trial, but I believe the PR about the high number of patients with compromised immune systems, and this is a huge issue. These numbers, below are enormous.
By finding some way around this, NW is pushing through a wall here. They are not letting this obstacle stop them.
Again, I have not followed all the complex arguments well enough to be totally confident that all is well here. I am very concerned still, and I have trimmed my oversized sails accordingly for the possible short term fallout. But I will continue to review what has been written here, and learn. And I am certain that Northwest got a bad deal in having to test their immunotherapy on a patient population with such a large percentage of damaged immune systems.
PR Replay
-----------------
Important recent research has found that as many as 40% of GBM patients have such severely depressed white blood cell counts following radiation that their level is comparable to the level at which AIDs patients are put on continuous antibiotic treatments, prophylactically. Further, this research has found that these GBM patients’ white blood cell counts do not recover with the passage of time. See, for example: Grossman et al., 2011: Clin Cancer Res 17(16):5473-80; Ellsworth et al. 2014: Oncoimmunology 3(1):e27357. Epub 2014 Jan 3.
------------------
Did I miss the point here? Is the point of the expansion to have the option of removing the patients with the radiation compromised immune systems from the post analysis?
I thought they were saying that the bar needed to be lowered due to these patients. That would imply that these patients would be included in the analysis.
This sure puts pressure on radiation as a treatment, regardless. I had read that radiation can weaken the immune system, like some chemo can, but it is hard to visualize damaging all of your bone marrow, or all of your lymph nodes with radiation. It is generally very targeted, right? I guess the free radicals that it generates float around and do a great deal of damage all over the body.
Of course, as Flipper once said, not all radiation was created equal, as is true with chemo. And there is too much of a good thing. There is also too much of a bad thing.
The PFS threshold for approval dropped from 6 months to 4 months, a factor of 1.5. (if the PEI approves).
The number of data points was increased 2.25 fold from 110 to 248. The RMS noise in that data dropped by the square root of 2.25 or 1.5.
If the way the FDA works is to want a mean efficacy and confidence such that some common threshold of likelyhood for the "true" mean is above / at 0 efficacy, then the 2.25 fold increase in data points corresponds directly with the 1.5 fold decrease in the required PFS.
Does anyone know if that is how the policies work. In the end, is the net measure of sufficient efficacy such that there is simply net positive efficacy with X certainty, where that X is always the same. Ie, for all trials?
I don't want to sound too positive, because I am not sure what to think of the extreme changes to the event numbers required.
However, I want to point out that the statement about the great response in the early progressed patients foreshadows an application for AA (as others have suggested should happen) and an AA would make the increase in the event numbers to end the trial nearly irrelevant.
If the application for AA was to await the new first interim, or second interim, then those delays would matter a great deal. But Pyrrhonian keeps saying that AA is usually applied for after a phase 2. So... there is no need for them to wait until those new interim thresholds. (In fact Pyrrhonian has been a fan of the idea that AA may have already been applied for. Not sure if that changed yesterday...)
In that case, what those increases in required events would do is increase the amount of paperwork that has to be done in the final years of the trial, but delay the ultimate judgement for many years at the same time lowering the (already small?) risk of failure. This would ensure time to develop and market direct and other products with absolutely no risk of interference once $ from AA approval come in as well as HE $. De-risking the company allows large investors to take a more serious look at the upside.
But as I previously questioned, I wonder if lowering the bar for the final measure on DCVax-L efficacy also lowers the bar for AA approval? You would think the bar for AA would be higher due to the lower sample size. So the need to lower that bar is understandable. Does anybody know if lowering the final bar lowers the bar for AA? I would think it might.
Pyrrhonian: Maybe you are right about the data being staggerd changing the statistics. Thanks for the relevant buzzwords for me to search in the CRC literature you previously linked. I wrote my last response before reading your last response. So.. ignore it. Probably too late. You probably blew your top. If so, no worries. I am owed a few of those all around.
I think you were right Pyrrhonian that the trial delays effected the needed trial design, but not quite for the reasons you quickly came up with. Not that I see this clearly... but going from where you where correctly leading us, before you may have taken a wrong turn...
The patients being so staggered means that more would not have progressed yet. That means less data samples. That would make the data less statistically significant. So, maybe you were right after all though for a little different reason than you stated. Not sure if technically that would increase the variance, but I think so. If not, it would certainly make the data less statistically significant in some other technical way.
Pyrronian: I know you see this regulatory stuff much more clearly than I do. But I think you error-ed in your view of the data being spread out.
If a group of patients all start at the same time, but the PFS numbers are distributed widely over some mean, then the certainty of that mean is poor... the variance is high.
If a group of patients each start at very different times, so that the absolute times to progression are all very different, but the individual times to progression are all the same, then that group has a small variance. I don't think when they started has any bearing on the variance.
You might come to the same conclusions about the trial with this other view of the effects of staggering (ie no effect), but I think you should re calibrate and see.
For those with more statistics background than myself: A clue is that the increase in PFS events is 2.25 fold, the square of 1.5. The noise goes as the square root of the number of samples. A 2.25 fold increase in samples lower the noise to 2/3 of what it was.
So you have to increase the samples a lot to lower the noise a little. One reason to start the trial with a cautious low number of patients. You can easily waste a lot of money if you try to reduce the noise by just a little more than what you need.
I say for those with more background than myself because that is as far as I can go with this.
If you are sure these 55 are true progressives, then I would agree that those results are great. Good to know. Thank you Pyrrhonian. Also thanks for that link to the CRC on trial statistics yesterday. Excellent resource.
But I remain concerned, as some others about the gist of the change to the PFS end goal from 6 months to 4 months. I am not familiar with the dependence of data quality on how well it is grouped in time as you have stated. I am not familiar with a statistical consideration of grouping in that sense. For the start times to be staggered broadens when (in absolute time) progression occurs, but it does not broaden the statistics of progression. That is not the same thing. Without that, the data having been been staggered beyond expectations just elongates the trial, which can be compensated for by adding patients, enlarging the trial, but not by adding to the number of events. From my view, adding to the number of events would only be done to reduce the noise or variance, allowing a lower efficacy to be statistically meaningful. They would do this for one of two reasons. Either they think the efficacy is lower than they expected, or they are wanting to lower the target / goal to make AA easier. Not sure the latter makes sense, but within my clouded understanding, it might.
I am a little confused BSB. It is not clear to me what to make of these 55. There can be more than one reason for early progression.
If the stats they state are accurate, then these results sound very good. But early pseudo-progressives would have great responses that could swing the mean positive much farther than poor response from the true early progressives swings it negative and I am not sure if the simple mean is the normal measure for PFS efficacy. I thought there was some downplaying of extreme good numbers. I don't remember the method used by the FDA. (is the median used? That would downplay the extreme positives if they are not at least half of the population. That sounds familiar.)
As for the belated concern about radiation damaging the immune system... this bodes well for immunotherapies down the road as it makes it more clear that the change from existing radiation and chemo to immunotherapies is going to be turbulent. Those old technologies, however good or bad, were not optimized or selected for use along with chemotherapies.
But what does this mean at this point? Why would this have not been part of early statistics, whether they knew the underlying mechanism or not? Was there no radiation in the early data? Or just not enough data points... and the 40% that the radiation damages did not happen to be properly represented in that small group for the early studies? Without that knowledge, this post revelation does not make sense to me.
I remain very hopeful for DCVax-Direct. And I believe that the ability to target tumors directly and to not have to do surgery is such a wonderful possibility that I do not want to be negative about the company here. DCVax-Direct is much more valuable than DCVax-L. And it is more beautiful if I may use that word. So while early data remains promising, I remain very positive on NWBO. But I am not sure that the positive news today about L is not somewhat of a spin. If so, just hope the SP hangs in there until news on Direct becomes a focus. But this might be a short term problem.
Flipper is the progression vs pseudoprogression expert on this message board... and he is very smart. He catches things that I miss, and he sees things very clearly. And he is quiet this morning. Perhaps he is on a hike!
There were two PR's in a row from NW this morning. The part about the 55 compassionate use patients was in the one you may not have read yet. I think it was the second one.
Clarification to Long's post (2 posts back).
I'm sure I understood what you where saying, but there was a point of ambiguity. Adding new below removes the ambiguity.
------------------------------------------------
Quote:
Now, tell me again why you think full enrollment has any bearing on AA being granted?
The AA full enrollment theory is this:
- AA is granted subject to the agreement of the sponsor to continue the PhIII (double-blind, placebo controlled) as a confirmatory trial.
- To serve this, the PhIII must have the placebo cohort
- But AA means every new patient from AA date onward gets DCVax, not placebo
- Therefore, the placebo cohort must be fully in place before AA can be granted; and to get placebo cohort fully enrolled, the trial must be fully enrolled.
The only way AA could happen without the above is if the confirmatory trial is not the same trial, i.e., the current trial is effectively ended due to not being able to fill the placebo cohort, and a new non-placebo-controlled trial is started.
Perhaps it could happen, but it seems IMO the FDA would want the placebo-controlled results.
-----------------------------------------
Flipper, you lost me very early in your first post. I want to understand it, so please spell it out further.
"348 x .66 x .2 = 46 treated but non-responders."
Are these .2X non-responders in the treatment group non-responders to DCVax-L, or to Chemo and Radiation? If to Chemo and Radiation, are these the progressed + pseudo-progressed combined or only one of those two groups?
Maybe you are right that the trial size change is just the small 36 add. If so, the time to full enrollment should not slide much, if at all. So even if HE enrollment has to wait for full enrollment in the trial, then it might not be so far off.
That is scenario B that Flipper described.