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.
Did you catch the last lines of the AF article?
"One more thing: The DCVax efficacy interim analysis AWOL for months now may finally come out of hiding now that the financing is done. The company no longer has to worry about the stock tanking."
This statement suggests that AF believes that the efficacy data will be continue with enlarged sample.
The only missing piece is a vague announcement that the trial is being enlarged. Then we can all buy shares under 5. Nobody in their right mind is selling shares BEFORE huge news at ASCO in May. And expansion in Europe? I thought they already had that mostly covered. At least the setting up manufacturing side. I wonder what AF is doing right now? No need to wonder for long.
Hi Dok,
I'll try to address your first question as best I can.
Macrophages differentiate from different progenitor cells and I don't believe they can proliferate like some lymphocytes. Macrophages come from a different lineage, myeloid (vs. lymphoid).
Also, the macrophages are not exactly feeding on the cells as much as they are cleaning up cellular debris that might be released by cells undergoing apoptosis or pyroptosis. I could imagine that a large cancer cell might be a stretch for a macrophage to phagocytize. Either way, I doubt that there would be a net gain for the macrophage after breaking down a cell. I don't think infiltrating macrophages would necessarily have any energy disadvantage while mopping up. Like the other cells in the area they will be supplied with oxygen and nutrients (glucose, essential amino acids and the like) by diffusion from nearby blood vessels. Tumors are usually highly angiogenic.
I think the general order of how things are supposed to work for either -L or Direct is as follows:
DCs--> lymphatic tissue--> activate antigen specific naive T cells--> become effector T cells and proliferate (primarily CTLs)---> CTLs find the tumor and attack cancer cells---> cancer cells undergo apoptosis and die--> macrophages and maybe other phagocytic cells clean up debris. At some point antibodies can potentially come into play. This is a very generalized and over simplified view of course.
T cells do make chemokines that act on each other and other cells in the innate immune system so I suspect there would be a homing of T cells and maybe others to the tumor site.
One thing that bothers me about Direct is that cancer cells are generally characterized by not having much MHC. For the CTLs to target the cancer cells the T cell receptor needs to bind to its match on a MHC(I):antigen complex. I think there has been some discussion that chemo and radiation might help induce antigen presentation by the cancer cells. Without that then you are relying on what scant presentation might be present normally. Puzzling it is....
With regard to proliferation at the tumor site. I'm not sure. I doubt it, but don't know for certain. One thing I did read over the last few days that I didn't remember learning was that naive T cells can actually be programed by cells other than professional APCs. It can happen in the tumor, but rarely does because the cancer cells don't typically express the co-stimulatory ligands (CD80 and CD86). It's still my impression though that the dendritic cells injected with the Direct treatment still need to migrate, after taking up antigen, to the lymphatic tissues to present to naive T-cells. A CTL can kill more than one target.
So what is the deal with Bosch being in Seattle? And is he working with Tjoa there too? Do they have a lab at the University of Washington. Not understanding how this long distance relationship works. NWBO is in Bethesda; Cognate is in Hannover, MD and TN. Bosch and Tjoa are in the Seattle area. I noticed on the last paper with Bosch as author (2003) the address was listed as Bothell, Wa. Just curious to where the research is being conducted.
Flipper,
It's been a very busy day here and I didn't have internet at home last night so I am behind on getting back to you and your points. I will try to do that very quickly and maybe I can elaborate more later.
With regard to the site of activation of T cells. No, I wasn't suggesting that naive T cells are being activated in the tumor. That can't happen in my understanding of things. It's a whole maturation process that happens in the lymphatic tissue that is required for that. However, I think the real question is did the Triozzi et al. paper show any data that proves or even suggests that naive T-cells were activated to become effector T cells. I believe what Dr. Bosch is referring to in what you quoted him as writing is the data in figure 2 from the paper.
Let me first just say that the paper is a bit of a mess in how they refer to lidocaine injected control tumors as non-injected control tumors in parts of the paper and as control tumors in others and non-injected in still other parts. For a bit there I was beginning to think that the authors were comparing DC-injected to totally non-injected tumors. It's still a little unclear to me but i have to take it on faith that they would not be that deceptive and it also would not be able to get by a reviewer. That said, I'm really surprised a reviewer didn't ask them to standardize the language so that there is no confusion on what they are comparing to what. I think if anyone looks through all of the sections of the paper and reads how they waffle back and forth on the language for controls you will see my point...It's a flipp'n mess (no pun intended). Anyway, back to your question.
First off the IHC shown in figure 2 is not something I am familiar with reading. The staining is for Dendritic cells so I guess a pathologist is looking for lymphocyte cells based on morphology. It's worth pointing out that the DCs injected into the tumor are not pure. The authors describe them as being >60% DCs and the remainder of lymphocyte morphology. I don't know if this could contribute substantially to the difference or not. I kind of doubt it unless the section evaluated was cherry picked.
What isn't shown in the figure is why are the lymphocytes infiltrating? Are they homing to the tumor? Are the lymphocytes effector T cells that are newly derived as a result of dendritic cells presenting antigen in a nearby lymph node? The authors themselves conclude that the DCs failed to migrate.
The experimental system I work in involves IC (intracranial) injection of an agent that is infectious but not particularly antigenic. In doing so there is a measurable proinflamatory response around the needle tract. Cytokines are elevated and macrophages, neutrophils and microglia move in to clean up the mess. I'm sure that some dendritic cells get in there too. The response is shut down pretty quickly by certain interleukins. You don't want a big inflamatory response in your brain.
Anyway, the point is that just the mechanical disruption of tissue is enough to get some response. Injecting cells and the mix of agents that would be in that vs. lidocaine might not be a fair comparison.
The more important data in my mind is figure 4. That figure sort of backs up that the infiltrating lymphocytes are more tumor specific than controls. However, it's important to keep in mind that only a tiny subset of the hsp's purified from tumor for that assay would actually be carrying tumor specific antigen. This means you don't know what is responsible for the proliferation. That's where the controls come in to try and show specificity to the tumor.
However, another thing of importance to note is that the authors state in the paper that none of the control tumors were characterized by lymphocyte infiltrates. However, they do isolate lymphocytes from control tumors for the proliferation study. So, I think we can conclude there are some, but there are less than in the tumors injected with DCs. That's important because the lymphocytes are clonal. Because there is an expansion phase in the isolation (read Materials and methods)process you are therefore amplifying fewer clones from the control tumors than from DC injected tumors. Therefore the antigen repertoire of the lymphocytes from DC injected tumors should theoretically be greater merely as a result of starting with more clones, irregardless of how they got there in the first place. Thus, when you expose the lymphocytes from DC injected tumors to hsp carrying antigens, having more clones means more antigenic matches and greater proliferation. The Lido control having fewer clones has a smaller repertoire and therefore fewer clones will find an antigenic match and a smaller proliferation. It's kind of not a fair comparison in my mind and that might explain the difference (black bars vs. open bars in figure 4). The allogenic control helps somewhat. I need to think about it some more but overall I think the study falls short of showing evidence of newly activated T-cells.
You might consider the following quote with regard to the requirement of pro-inflamatory signals for DCs to activate T cells to become effector t cells:
Triozzi et al.
Dok,
We are pushing the boundaries of my understanding of immunology....
As I understand it, the lymphocytes mainly arrive at the lymph nodes via blood vessels. At the node there are sticky molecules on the endothelial cells that specifically grab the lymphocytes. I used to show a time lapse micrograph in a biochem class I partially taught that showed a white blood cell rolling at the surface of a rat tail vein. The white blood cell eventually stops and extravagates into the surrounding tissues. This is all mediated by chemokines and cytokines. Something similar happens in the nodes. Once the T cell becomes an effector T cell it leaves, I believe via the blood vessel. Some may leave via the lymphatic vessels. There are interfaces for the two all over the body. The CTLs and antibodies and others use these like highways. Your organs and tissues are bathed in this stuff. If you have a good strong response then these agents of the immune system will canvas your body looking for the target. They move by flow of the blood or lymph, they crawl by diapedesis. The whole thing is remarkable and extremely complex.
I thought it was suggested by you and maybe others that in order to get everywhere the dendritic cells would need to spread out to other lymph nodes. My understanding is that doesn't happen and it isn't necessary to obtain a robust and long lasting response.
Flipper,
The response described in the Triozzi et al. study was not a systemic response. Why do you present it as such? The authors themselves state very clearly in the text that:
“Finally, although antitumor activity and antitumor-HSP lymphocytes were observed in situ, there was little clinical evidence to show that systemic antitumor immunity was elicited with IT DC therapy as it was applied; laboratory evidence in the form of PBMC proliferative responses also was not observed. The induction of systemic immunity has only rarely been observed with other IT immunotherapies.”
“So what are we really looking at? Why was the early systemic effect in the Triozzi study apparently only local?”
The reason is because it was not a systemic effect. It was a local effect:
“The persistence of the DCs on histologic examination and the absence of evidence of systemic antitumor activity would suggest that the injected DCs did not migrate.”
One more time…. To get a systemic effect the dendritic cells pick up antigen in the tumor and migrate into the lymphatic tissues. There, if the antigen:MHC complex is recognized by a naïve T-cell AND the proper co-stimulatory molecules are present then you get activation leading to a CTL. If you don’t have the proper co-stimulatory molecule then you get tolerance. THE DENDRITIC CELLS DO NOT (NOR DO THEY NEED TO) MIGRATE TO ANY OTHER LYMPH NODE TO GET A SYSTEMIC EFFECT. For the dendritic cells, the lymph in the periphery near the site of the antigen is a terminal migration point. Like a spawning salmon, they die after activation of naïve T cells.
“So Marnix Bosch and colleagues had to mature dendritic cells to the right stage so that when they were injected into tumorous cancer cells, the partially matured dendritic cells were predisposed to think whatever they encountered in the tumor was not normal.”
Dendritic cells have no way to determine if the antigen they are taking up and presenting is normal or not normal. Dendritic cells take up self and non-self (beneficial) antigens all the time. It’s just that in those circumstances there are no pro-inflammatory signals present that induce co-stimulatory molecules to be expressed and that are necessary to generate an immune response against the antigen that would include T and B cells.
“So, in doing so, along with a few other reasons for partially immature dendritic cell potency, the dendritic cells now went and expressed an unequivocal message to the T-cells and B-cells that the tumor is our enemy.”
Yeah, maybe. If they make it out of the tumor micro-environment with co-stimulatory molecules (think of them as their balls) intact. The dendritic cells that Triozzi et al. injected into tumors were partially matured cells that expressed the appropriate co-stimulatory molecules. But once in the tumor the cells didn’t migrate. In fact, they probably de-differentiated due to the fact that CD86 (one of those necessary co-stimulatory molecules) was down-regulated.
“In so doing, once the t-cells and antibodies overwhelmed the tumors in the body elsewhere, (and here is the additional important insight), the dendritic cells near those non-injected tumors near their lymph nodes mature to recognize the nearby tumors as abnormal.”
I don’t understand how you possibly conclude this. It’s true that the dendritic cells are constantly sampling the extracellular milieu by macropinocytosis among other more specific mechanisms, but without a proinflamatory response the dendritic cells will not be able to present the antigen in a way that will activate T cells. More importantly, why would you need too? If you already have tumor killing occurring that is mediated by CTLs and antibodies, what’s the point?
The absolute key to whether NWBO will be successful or utterly fail is whether or not their process for preparing the dendritic cells will overcome the countermeasures that have allowed the tumor to exist in the first place. The existence of tumors within the purview of our immune system only occurs through an incredibly complex process that involves our own immune system shaping the tumor cells into what they ultimately become. Overcoming that is momentous challenge and one that can only be overcome through thoughtful and ingenious devices. It would go a long way in my mind if NWBO would publish the preclinical data on Direct so that it might be examined and scrutinized by those with a better grasp of human and tumor immunology than myself.
The file case at the time point in the movie you pointed too belonged to the late Steve Warren, co-founder of take the fight to cancer. It is my understanding that Mr. Warren was diagnosed in April 2012 with brain cancer. He had SOC and it recurred. He received DCVaxL under compassionate use in Sept-Nov 2012. Mr. Warren passed away in in September 2013. Hence, the file case does not represent a patient on DCVax since 2011.
The real news is how much Steve Warren and his son David Warren have done in helping cancer patients nationwide, that probably felt helpless at first, feel empowered and once again in control of their lives.
Also, I think you are hitting on why a lot of people don't necessarily trust NWBOs approach. I was going to post some reviews of tumor immunology on the board but i've not gotten to it yet. There is a lot of information out their about the interaction of tumors and the immune system. I think a lot of it would surprise those that haven't reviewed the literature and understood what it is saying about cancer and the immune systems role. Maybe that is why the checkpoint inhibitors make some people feel more comfortable. There is no secret sauce. Pathways are known and the rationale for targeting at certain points is logical. People just don't understand why dendritic cells exposed to antigen in vitro or injected directly into a tumor are capable of circumventing the mechanisms that have allowed the cancer to exist in the first place. I have some ideas but I'm still looking through the primary literature.
Hi Dok,
This gets really complicated really fast. It's been a long time since immunology for me. I hear immunologist give talks weekly where I work. However, that is not the same as being an immunologist. What I can say quickly is that once the dendritic cells pick up antigen they become mobile and move into the lymphatic system where they "call" to lymphocyte cells. This calling is through chemotactic factors that pull these T-cells into the tissues through adhesion molecules on the endothelial cells of the vessels that the T-cells move through. Also, the T-cells are constantly moving through the lymphoid tissues sampling the dendritic cells for their match. When they find it they stop migrating and proliferate greatly and then disseminate. Right now, I can't recall what or even if there would be homing mechanisms for cytotoxic T lymphocytes to the tumor site. I know that macrophages like dendritic cells release chemokines that attract all kinds of cells including dendritic cells. Do dendritic cells in the tumor attract effector T cells? Not sure, maybe. Or maybe once a cytotoxic T (CTL) cell finds the antigen (cancer) it releases chemokines to attract more CTLs. I just don't remember. I would have to consult my 15 year old Janeway's immunobiology book at home to jog my memory.
"However, what if the DMC recommended an early termination for efficacy, and the company agreed? Next a PR is released as to the situation as a report goes to the FDA. Then we await a finding. The FDA may agree, or may order the trial to continue. If a continue was ordered, the trial would have been un-blinded, to some extent, in this second order way, if the DMC recommendation was given more credence by the patients than the FDA finding of continue."
Firstly, it's NWBO's trial. The FDA can't order a continue. They can only recommend or provide their thinking (unless safety is in question, then they can order you around). Secondly, stop for efficacy threshold is set at the onset of the trial. The scenario you proposed is exactly why those thresholds are so important and why the FDA warns against actions that are not in accordance with the trial guidelines. If you you violate the upfront agreement then you greatly risk the data validity.
I don't think it's the patients being unblinded that is the main problem. It's the practitioners that might change how they are managing the patients in the trial or the sponsor that might change how they conduct the trial.
I think if the decision by the German's hinged on the PIII it would suggest to everyone a trend is being observed either way. Maybe more importantly, what if the Germans looked and then denied the exemptions? What would that tell the sponsor and practitioner about how the trial is trending? You don't know if the cat is in the box until you look. Of course the quantum mechanics folk might say that the cat wasn't there until you looked!
After taking up antigen in the tissues (tumor), dendritic cells move through lymphatic vessels to local or regional lymph nodes. There they present antigen and produce other molecules that activate naive T cells and B cells. The T-cells become the cytotoxic, helper or regulatory cells you mentioned. The B cells become plasma cells with a minor population becoming memory cells (the T-cells can do this too). The extent of the activation is highly dependent on a number of "things".
Did you cover? I don't really understand the short side of things. You short at price X and cover at price Y and the profit is in the difference?
If one were to assume that the German's saw PIII data from this first interim look and then gave the approval, wouldn't that be a little like unblinding the trial(declaring the trend)? I don't think they saw anything from the PIII trial(IMO).
Wow, I would say deception for sure. But now the question is how does the company respond. Shareholders have become accustom to a rebuttal. Simply pointing out that this isn't new information will just cast an even brighter light on it. Perhaps the company will use this as an opportunity to elaborate on the PFS vs OS primary endpoint thinking.
I think there are even better examples than avastin and primary literature to support PFS. I posted this previously. I will try to respond after I get to the lab.
So if the FDA didn't agree to the endpoint, how did NWBO set the boundaries for stopping due to efficacy? These boundaries would need to be outlined at the beginning and it was my impression that they would need to be something that the FDA had agreed too.
The amateur show continues. This was a CYA move by NWBO. I posted this previously but I will post again. PFS is acceptable as a primary endpoint with the requirement that OS is protected. It's a little late in the game for NWBO to be letting shareholders know that they actually don't have a prearrangement with the FDA to accept their endpoint.
see post 7815
He comments on seekingalpha articles often, Larry smith has published his comments and he has now commented on the Immunotherapy note that came out of the motley fool propaganda machine.
The sp is pulling back a little now. Maybe day traders or maybe it's lunch. Don't know, but I don't like sitting on the sidelines (mostly) so i will probably put an order in for 1000 shares today. And yes I will pay more than 7.25 probably. This is why I don't day trade. Still hoping not to get blindsided by train of expanding trial size...good or bad in the end, I think it will be picked apart by AF and the like and sp will decrease on that kind of news.
I don't need to reread it. I think LTT and AF may be of the same mind about NWBO, but not the same person. I work in literal meanings. Sorry. If you say you think something is another thing, I will interpret that literally. The scientist in me I guess. Actually, I think LTT is more positive than AF.
i honestly think the stock is going up because of DocLogics comments. A lot of people read fool letters and some might see the comments. That and all of the news is slowly sinking in....
Maybe not nuts but definitely paranoid. AF is doing a live blog on Afrezza FDA panel right now. He is busy following discussion in real-time and writing about it. LTT does not equal AF.
I'm probably mixed up on how the trial arms are being run. Don't know. Flipper said he would get back to me with something to help me understand. I'm still kind of trying to understand the structure with the pseudoprogression group. I think I have something wrong in my head (no pun intended).
I didn't remember seeing this in the july '13 ppt from the company. But in one slide there was this:
• 312 patient, randomized (2:1), double blind, placebo controlled Phase III trial: the “gold standard” in clinical trial design
– Primary endpoint: PFS (progression free survival)
– Secondary endpoints include OS (overall survival)
– 2 interim analyses for efficacy & 1 interim analysis for sample size
– 3 DCVax-L treatments upfront (Day 0, 10, 20), then 3 boosters (months 2, 4, 8) then 4 treatments twice/year for maintenance phase (months 12, 18, 24, 30)
Maybe this isn't informative, maybe it is. It's not clear if this interim look would be announced. maybe it's happened and they are thinking about it? Maybe they are thinking about looking?
The conclusions you've drawn are accurate. I think it would be perceived as negative by the investment community as a whole. I'm not really concerned about what you or I or Afford or anyone else thinks. My concern is with the institutions and funds that are now holding a lot of shares. If one or two or more decide they don't want to wait for a longer trial to complete then they could dump their position and seriously hurt the share price for a good long time. I don't really understand the arguments that enlarging the trial makes things go more quickly. I think if you enlarge the trial you move the checkpoints from 88 and 110 to later. I don't see how you don't do that with a larger sample. I know that we don't know enrollment numbers, but couldn't it be possible/likely that we hit 88 or even 110 events before they even get done enrolling 312 patients? If that's possible then why enlarge. You'll be done before any of those patients enter the trial.
As for the other point, does this reflect negatively on the trial outcome. Well, maybe yes or maybe no. I still hold that if you don't hit your endpoint, miss by a month or two, but the extended PFS is very statistically significant (p=0.02 or better) then you can still go for registration. I'm not sure that if that happened everyone would take that same view.
As to your point of whether the company knows hitting the endpoint is in doubt. I really never said they did. It's clear now from looking at the guidance from the fda that one of the jobs of the DSMB (if so written into the protocol by the sponsor(i.e. envisaged)) is to at the first interim look check that the original assumptions that went into calculating the trial size are still valid. If they are not valid then you recalculate and resize. This is not biasing for success. It is simply correcting an error in the power calculations. It could be viewed as properly powering the trial to show that there is no benefit in PFS just as easily as it would be powering the trial to show that there is benefit in PFS. Either way, to get a valid answer you have to conduct the right experiment (correct sample size).
Afford,
I didn't say it had to be negative. I just would like to understand the potential reasons. NWBO can only enlarge if they can justify. You can't enlarge just as an empowering act. Look back at the paper Pyrrho just quoted. overly large trial are costly and unethical. If something has changed with the powering of the trial then you can adjust. These are people for god sake. You can't just add more souls getting placebo because you want too.
I like that you are quoting passages from guidelines that I first posted and you discredited as being old and out of date...
You do realize that when there is a DSMB involved there are numerous sample re-sizing strategies that use uncoded data and protect the overall type I error? It doesn't have to be the sponsor re-calculating.. that's been my point all along
I'm sorry to say that I think it's entirely pointless to discuss anything that doesn't involve label expansion or the impending halt of the trial. I've been censored in my last post. Basically we can't have an open discussion in which a person post their opinion and provides primary literature or existing FDA guidelines to substantiate their thought process. It is now standard practice on this board for a moderator to post excerpts from primary literature articles that have been modified to remove only information that might go against the moderator's argument. Doing that in science would immediately and irreversibly destroy credibility. And again with the insinuation that I have ever said Linda is in possession of uncoded data and will change the trial based on this information:
"I think you and Mol and (apparently) iclight and (maybe) fox should just go ahead and believe that somehow Linda and management (not firewalled) got their hands on either unblinded data, or were told by the DMC via the steering committee that there were variances noticed in the data that (probably) called into question the trial's ability to achieve its primary endpoint, and that they should consider re-sizing the trial to lower that expectation (because I mean, the independent DMC is all about that, you know), and that Linda Powers et al are currently considering just that under those specific pretenses."
I've re-read my posts and I never said any such thing. And then more about not understanding the variances, event rates or survival experiences:
"Really, I think you all should. Please. I mean, you have the exact location in an FDA document entitling you to, regardless of whether or not you understand what overall response variances, event rates or survival experiences are. And, what the hell, let's just assume any variance is a bad variance while we're at it. Cool. Good job, guys. Great posts."
This whole discussion has morphed from the moderator arguing that the addition of patients is because "they can and why not" to now some argument that the change is because of a good variance.
I started to create a compilation of the posts that went back and forth because I thought there was some good information in there, but I really just don't feel like putting any more time into this. However, I do want to say one more thing that was going to be a lead in to the compilation of the posts on this point:
I think that everybody knows at this point that during Linda's presentation I interpreted something she said to mean that NWBO is considering increasing the patient size of the trial. Others also picked up on this as well including long time defender and NWBO guru Larry Smith. Smith wrote in his most recent article that he thought trial enlargement is likely. Frankly, I don’t know what the odds are of the trial going bigger nor do I know what rationale would be used for this purpose. That is in fact what spurred this discussion; too identify why the trial might be enlarged. Some on the board have painted me as a heretic and basher. As a person that is spreading misinformation and is trying to persuade shareholders through fear to sale your shares. I am not trying to persuade anyone to sale, buy or hold anything. I don’t want that responsibility. Next week just about any news could be released that pushes the share price one way or the other. I don’t want to feel responsible for someone dumping shares if positive news comes. Instead I have tried to have a fair and open discussion among the collective that could take in all possibilities. Moving on…
One possibility is of course that there is no discussion going on to enlarge the trial and that Linda just spoke poorly when addressing this issue in her presentation. I concede that this is a possibility. In fact, that is the very best possibility in my mind beyond all of the other arguments that a trial size increase will further de-risk the trial and that is a good thing. I don’t like that argument for many reasons that if you read on will become clear.
Another possibility is to take Linda’s words literally, which would lead one to believe that NWBO is in fact now actively discussing enlarging the trial. This is the interpretation that I choose to believe. This of course begs the question too why would the sample size need to be adjusted. Hence the discussion that has taken place.
BB, you have no credibility. In your world this message board is only for pumpers and not for people with an open mind seeking to hear from others that might have a different perspective on any given news.
I think I've been pretty clear from the start that I believe L will ultimately be successful. That doesn't mean I am going to just stick my head in a hole and hope nothing comes along and bites me in the butt in the meantime.
I think the thing that bugs me the most is that when I make an argument for debate I try and substantiate that argument with factual guidance that would have been used in the design of the trial and also in the continued conduct of the trial. I've not heard too many counterpoints that are based on anything other than "what the company says". I'm a scientist. Scientist are critical thinkers. I question everything. I rely on fact and data, not conjecture or heresay.
Okay, so I've been accused of disseminating misinformation now. Pyrrho, what is your background in dealing with the FDA and clinical trial data? I think you know mine.
What exactly did I misrepresent. This is not just googling for information. When working on a bla for a very real vaccine that very much did get approval, FDA guidance was like a bible when writing sections for review. What I pasted was the factual guidance from the FDA for industry. Not specific to any institute or entity. What context do you think that I've missed.
I'm very much open to being convinced otherwise. I want to be wrong. But I'm not going to pander to "what I want" by discarding logic and blindly believing that everything I'm told is the gospel when evidence exist otherwise.
Pyrrho,
I don't know if you just don't read my posts or if maybe you ignore their content. I believe your logic that the company can just add to the sample size without any reason is based on hearing Linda say that they wrote "the option to enlarge the trial" into the protocol. That statement from Linda doesn't mean that there is not some underlying reason for enlarging the trial. And no, because we can is not a valid reason in the eyes of the regulators. You must provide justifications for administering an investigational drug or PLACEBO to more human beings than is necessary (based on calculations). Look, if you write a study protocol that details the assumptions you've made to generate a sample size of 312, then you also have to justify why you might increase the sample size later on. Look at the FDA guidelines on determining sample size. No where does it say that you can calculate sample size for a trial and then add the option to increase that later because you can afford it or just want even more cushion.
"The method by which the sample size is calculated should be given in the protocol, together with the estimates of any quantities used in the calculations (such as variances, mean values, response rates, event rates, difference to be detected). The basis of these estimates should also be given. It is important to investigate the sensitivity of the sample size estimate to a variety of deviations from these assumptions and this may be facilitated by providing a range of sample sizes appropriate for a reasonable range of deviations from assumptions."
Note the statement about range of sample sizes and range of deviations. "Deviations" implies that one or more of your assumptions in calculating the sample size deviated from the expected. Doing these kinds of calculations with deviated assumptions allows you to model how it changes your sample size and expected outcome. This I believe is how you would justify increasing the sample size. What if your assumption about the mean and median of the placebo group for PFS were underestimates? Correct me if I'm wrong, but didn't IMUC discover that their control group lived longer or had a longer PFS? Maybe I misheard that or maybe that was based on extrapolations from the disclosed information. I honestly don't know.
If you are on the DSMB and you are reviewing the data and you see that the placebo group is surviving longer progression free than you modeled in your powering of the trial; then wouldn't it make sense to change the assumptions and see how that changes the sample size needed to be statistically significant? Please read the excerpt below about amending the sample size. Note that despite the fact that you keep harping on double blind this and double blind that, and that nobody can make an educated decision about the trial because it's double blinded, there is mention in the passage of remaining blinded but using the data to change the sample size.
D. Sample Size Adjustment (4.4)
In long-term trials there will usually be an opportunity to check the assumptions which underlie the original design and sample size calculations. This may be particularly important if the trial specifications have been made on preliminary and/or uncertain information. An interim check conducted on the blinded data may reveal that overall response variances, event rates or survival experience are not as anticipated. A revised sample size may then be calculated using suitably modified assumptions, and should be justified and documented in a protocol amendment and in the clinical study report. The steps taken to preserve blindness and the consequences, if any, for the Type I error and the width of confidence intervals should be explained. The potential need for re-estimation of the sample size should be envisaged in the protocol whenever possible (see section III.E).
Does that sound to you like an amendment to the sample size cannot be accomplished without unblinding the trial? Another point, wasn't the trial sample size already enlarged? I don't know this so I am asking. I thought it went from a phase ii with x sample size and then to a phase iii with y sample size. But I also thought I remember seeing a respected poster say that another sample size change was made to fine tune things. I don't have time to look on clnicaltrials.gov right now but I thought I remembered that happening. Maybe not.
Finally, I'm not pulling this stuff out of my butt. I'm honing in on the fact that teh placebo group could be surviving longer than the assumption for the power calculation because Linda herself posed this in a roundabout covert way when discussing being able to show statistical significance even with a pfs difference that is smaller than 6mos. I know you would argue that was not the context, but I would argue it hint into the logic of going bigger. She may not have been told the pfs of the placebo group is longer than expected, but if the DSMB recommends, through the built in option, to go bigger Linda might be guessing why that is.
Otherwise, you are right and the FDA doesn't care how many people you put in your trial. If you can afford them then add them and you don't need to show any justification for the change (counter to the guidelines I've posted a couple of times).
One last point. If DCVax misses the 6 mos pfs endpoint but shows a 5, 4 or even 3 mos pfs benefit that has a p=0.02 that will not be a failure in my view. And I do not believe that they wouldn't be able to go for registration. But, to show that p=0.02 or better they will almost certainly need more sample size.... That's it.... I'm done, stick a fork in me.
This information should help answer a lot of questions. It will take a good while to digest. Note, there is no icing on the cake section when estimating sample size. Sorry...
You can find the full document by google'n
Guidance for Industry
E9 Statistical Principles for Clinical Trials
III
E. Sample Size (3.5)
The number of subjects in a clinical trial should always be large enough to provide a reliable answer to the questions addressed. This number is usually determined by the primary objective of the trial. If the sample size is determined on some other basis, then this should be made clear and justified. For example, a trial sized on the basis of safety questions or requirements or important secondary objectives may need larger numbers of subjects than a trial sized on the basis of the primary efficacy question (see ICH E1A).
Using the usual method for determining the appropriate sample size, the following items should be specified: A primary variable; the test statistic; the null hypothesis; the alternative (working) hypothesis at the chosen dose(s) (embodying consideration of the treatment difference to be detected or rejected at the dose and in the subject population selected); the probability of erroneously rejecting the null hypothesis (the Type I error) and the probability of erroneously failing to reject the null hypothesis (the Type II error); as well as the approach to dealing with treatment withdrawals and protocol violations. In some instances, the event rate is of primary interest for evaluating power, and assumptions should be made to extrapolate from the required number of events to the eventual sample size for the trial.
The method by which the sample size is calculated should be given in the protocol, together with the estimates of any quantities used in the calculations (such as variances, mean values, response rates, event rates, difference to be detected). The basis of these
21
estimates should also be given. It is important to investigate the sensitivity of the sample size estimate to a variety of deviations from these assumptions and this may be facilitated by providing a range of sample sizes appropriate for a reasonable range of deviations from assumptions. In confirmatory trials, assumptions should normally be based on published data or on the results of earlier trials. The treatment difference to be detected may be based on a judgement concerning the minimal effect which has clinical relevance in the management of patients or on a judgement concerning the anticipated effect of the new treatment, where this is larger. Conventionally, the probability of Type I error is set at 5 percent or less or as dictated by any adjustments made necessary for multiplicity considerations; the precise choice may be influenced by the prior plausibility of the hypothesis under test and the desired impact of the results. The probability of Type II error is conventionally set at 10 percent to 20 percent. It is in the sponsor's interest to keep this figure as low as feasible, especially in the case of trials that are difficult or impossible to repeat. Alternative values to the conventional levels of Type I and Type II error may be acceptable or even preferable in some cases.
Sample size calculations should refer to the number of subjects required for the primary analysis. If this is the full analysis set, estimates of the effect size may need to be reduced compared to the per protocol set (see Glossary). This is to allow for the dilution of the treatment effect arising from the inclusion of data from patients who have withdrawn from treatment or whose compliance is poor. The assumptions about variability may also need to be revised.
The sample size of an equivalence trial or a noninferiority trial (see section III.C.2) should normally be based on the objective of obtaining a confidence interval for the treatment difference that shows that the treatments differ at most by a clinically acceptable difference. When the power of an equivalence trial is assessed at a true difference of zero, then the sample size necessary to achieve this power is underestimated if the true difference is not zero. When the power of a noninferiority trial is assessed at a zero difference, then the sample size needed to achieve that power will be underestimated if the effect of the investigational product is less than that of the active control. The choice of a clinically acceptable difference needs justification with respect to its meaning for future patients, and may be smaller than the "clinically relevant" difference referred to above in the context of superiority trials designed to establish that a difference exists.
The exact sample size in a group sequential trial cannot be fixed in advance because it depends upon the play of chance in combination with the chosen stopping guideline and the true treatment difference. The design of the stopping guideline should take into account the consequent distribution of the sample size, usually embodied in the expected and maximum sample sizes.
When event rates are lower than anticipated or variability is larger than expected, methods for sample size reestimation are available without unblinding data or making treatment comparisons (see section IV.D).
Section IV
D. Sample Size Adjustment (4.4)
In long-term trials there will usually be an opportunity to check the assumptions which underlie the original design and sample size calculations. This may be particularly important if the trial specifications have been made on preliminary and/or uncertain information. An interim check conducted on the blinded data may reveal that overall response variances, event rates or survival experience are not as anticipated. A revised sample size may then be calculated using suitably modified assumptions, and should be justified and documented in a protocol amendment and in the clinical study report. The steps taken to preserve blindness and the consequences, if any, for the Type I error and the width of confidence intervals should be explained. The potential need for re-estimation of the sample size should be envisaged in the protocol whenever possible (see section III.E).
I only have one post left so I'm adding this on in response to another post...
Nothing about my doubt or concern with the trial size has to do with AF. I posted a concern about what I heard just a few minutes after I heard from Linda..just wanted to be clear on that.
An option to do something is not a reason. They are administering an experimental drug to a patient population. Some patients are receiving placebo. To add additional patients to the trial you would have to show that you need those patients in your power calculations.
That brings up the question of what it is that has changed in their power calculations if it isn't the data. If I want to do an experiment in monkeys, it's not how many monkeys I can afford. Rather how many monkeys do I need to get reliable data.
There is no backhanded slipping of data. I'm talking about a preestablished mechanism for which the trial size can be enlarged without unblinding the data to the sponsor or the clinicians.
Sorry, but you have to provide justification for the number to enroll. The "Why Not" is because the FDA is not going to go along with modifying the study protocol to add more patients for a "Why not reason". And i didn't say that the DMC told Linda or the steering committee and then the steering committee told Linda. I think I have been clear that the mechanism for deciding to enlarge is not clear. Who makes the decision to recommend enlarging?
Ultimately it's up to the NWBO since they have to foot the bill. But their is a big difference between a recommendation to enlarge the study coming from the DSMB or steering committee directly to Linda without any additional information (BUT based on data)and someone actually telling Linda what the data are and asking her to bless adding more patients. I think the former is the way it would go down not the latter and certainly not because of the "Why not" hypothesis.
If you look at the post I made to Astavarka I think you will see that there is some allowance for passing information to a responsible party in order to do their job..... (e.g. ascertain if the sample size is appropriate). I don't know who else that could be if it isn't one or more members of the steering committee that make the recommendation based on looking at the data. I'm sorry but I simply cannot accept that to enlarge or not enlarge will be made on zero information about the ongoing trial. You cannot really believe that they will add months if not more to the endpoint just to be extra...extra...extra....quadrupley doopuley extra certain that should they see a separation in PFS that it will be statistically significant. Linda clearly stated they are thinking about it. That implies discussions. What would that kind of discussion sound like?
Linda: Joe, what are your thoughts?
Joe: Well I know nothing of the data so I flipped a coin and say we add 40 more patients!
Linda: Okay, that's one vote for more patients. Tom what do you think?
Tom: Well I also don't have any facts to go on but heck I say that if you got option to add more why not! Let's add some more cost and overhead to the equation.
Linda: I see, very interesting. So that just leaves me here on the management team to voice my opinion and I say......welllll helllll yeah! Let's do this THING!(cue the music; And everyone gets up and starts dancing to MC Hammer "U Can't Touch This"]
it you should use it so I sure wouldn't want to find out that if we
It just means they have to have the info otherwise they can't do their job.
For instance, if it is your job to determine cohort size based the study protocol guidelines, you might need to see some of the interim data. Maybe the data is coded or something so as to not be totally unblinded.