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.
Flipper,I went over my calculations and I now believe that my math was somewhat wrong and the public PR that suggests that 233 will hit in June and July is much more In the ballpark. However, I believe that 233 in June or July gives one the opportunity to take a rough guess as to whether the trial will succeed. Using RK's midpoint for enrollment as December 2013, 42 months later we are in June 2017 with sadly a 70% loss of life. I think that I am very conservative to assume that at least 80% of the blended so called placebos (late cross over + no cross overs) had experienced the second event). This leaves the early L recipients with a 65% fatality rate at 42 months. If correct I am guessing that the early treated group would have an mOS of around 30 months. Hopefully there is a decent gap between the mOS of the early L treated group and the late L recipients. Another confounding factor would be the dropout rate
but as long as it is not too high (i.e. less than 20) it should not affect the results.
GLTA and Best Wishes
csf, many thanks for your comments and attached link to the publication. After reading your comments I feel that while proper statistical analysis is very important, the possibility of a skewed trial if present would reside most likely in the design of the trial rather than in the statistical analysis which have yet to be carried out. However the article thoroughly explains how many things could go wrong when carrying out this and any other trial. I asked myself whether randomization with respect to age,
sex, disease severity etc. was handled inadequately. I don't know whether the DMC revealed to the regulators one or more trial design flaws and whether that alerted the FDA to order the partial hold. I guess anything is possible.
I also wonder whether other factors could weigh on long trials. I am almost certain that in a long trial with 331 middle aged participants at least 10 would die from causes unrelated to brain cancer. This would be a loss of some information. The failure of 46 placebo participants to cross over is also not a plus. If I were a participant who became sick and progressed three months after initiation of treatment I would strongly suspect that I was part of the placebo group and would be more likely to move to another treatment. However, with. all this I am still hopeful. GLTA
What kind of PFS data can be revealed after May 22? Since the final results are still blinded, the maximum I can think of is something about blended mPFS results although this too may not be ready to be released. Any ideas?
flipper, I had a chance to do some thinking about your though that we may hit 233 during June or later. Remembering the recent Korean study (that I have been unable to retrieve) that was a DC vaccine study of primary brain cancer and somewhat similar to Phase III. They found a mPFS of about 15 months and an mOS of over 28 months. I will speculate that our 220 early DCVax-L treated group has a similar 13 months time gap between progression and the final event. Now even if the two remaining groups of late DCVax-L (63 patients) and true placebos (46 patients) had shown a similar time gap of 13 months it would be difficult to conclude that the 233 events would hit as late as June. Even if all the 248 progression events occurred in January 2017, we would expect about 248 final events during February 2018. My estimate is that we may have at this stage 4 final events per months. With all the progression events occurring during January 2017 we would hit 233 final events around October 2017. If the 233 were to appear during June, the gap between PFS and OS would still be long, around 9 months. Since in real life most of the participants have sadly already experienced two events, the average remaining gap between progression and survival would be at most 50% of the time gap (MAX about 6 months). If my speculations are true, we may have hit 233 by now.
csf, I don't know what you mean by your opinion that the confidence level of the regulators with respect to the Phase III trial was skewed. If the regulators had proof that the trial was skewed (i.e. a cherry picked trial) they could have submitted the evidence after completion of the trial and allowed the last 17 patients to enter the trial. Alternatively perhaps the regulators had no doubt about the safety of the patients already enrolled but some new information they received may have pointed to something that may endanger the well being of the new patients. Manufacturing probably was not the issue because that most likely would have been resolved. It is also hard for me to believe the often repeated theory that only so late in the trial the regulators had evidence that DCVax-L was so effective that to start new patients on the placebo treatment was not ethical. Had they been aware of such a degree of efficacy much earlier would they have stopped new enrollments and thereby put this very valuable trial in jeopardy?
In summary when it comes to the partial hold I am still in the dark like almost everybody.
csf, if RK matters' enrollment schedule for Phase III are realistic (I think they are), the midpoint for enrollment occurred around 12/13. Even if the 248 PFS events hit as early as 11/16, the 248 PFS would have been recorded 35 months after midpoint enrollment. If DCVax-L affords no benefit, the blended PFS would be the historical 8 months and let's even go overboard to an incredible 11 months. My estimation is that at 35 months past midpoint and with a PFS of 11 months, at least 85% (instead of 75%) would have progressed. Not all DC vaccines are necessarily the same. I am hoping that L will make the cut perhaps because of its diverse antigen load and imho the estimated numbers point in a positive direction.
Flipper, does your "June or later 233 release" refer to a statement that the 233 events have occurred or does it refer to release of most of the Phase III statistical results of PFS and OS? To me it seems most likely the former since occurrence of the 248 PFS events was only announced in early February. Would it not take much longer than 4 months to reach 233 final events and complete data lock, scrubbing and stat analysis?
What is the limit for share dilution at this time, is it 500 million? If that limit is reached what does the voting procedure to expand this limit entail?
Doc logic, if Dr, Liau's calculation that the mOS for SOC is nowadays almost two years, what is in your opinion the chance of success of this trial? My calculation for the entire blended trial resulted in about 16 months for mPFS and about 27 months for mOS. If according to Dr. Liau the mOS for SOC is about 23 months (almost two years) and the mOS for the blended trial is indeed 27 months, the mOS for the early L group would be about 29 months. This would be a 6 months difference between the two trial groups and may be significant but I was hoping for a larger difference. My figures are mere estimations and may be completely wrong. I Would very much like to look at alternative estimates. Any ideas?
Sorry Virgilio, I read it wrong since you were referring to OS not PFS. (haste makes waste).
Virgilio, when I read the 10K yesterday I must have missed the statement that PFS for SOC was 15 months. I am surprised that they would reveal this since the trial is still blinded. I was unable to retrieve the 10K and read it again because except for part of the first page everything was blocked. My calculations suggest that the trial would have a blended PFS of about 16 months. Hence if the placebos alone have a PFS double that of the historical results, (about 6-8 months PFS)we are in trouble if my estimate is correct because the L treatment arm would have a very similar PFS>
But that's the whole point abc. If NWBO succeeds AF will be regarded as an "abject failure". He is therefore desperately trying to destroy NWBO and prevent the release of Phase III results. AF is simply fighting to save his reputation and to some extent his professional career.
Thanks abeta for the info.
While it is reasonable to assume that radio chemo would suppress the immune system it is also reasonable to assume that this suppressive effect is dissipated before the tumor regrows; with this immune suppression greatly weakened DCVax-L would then be able to inhibit subsequent tumor growth.
My intuition is that although DCVax-L may be effective in patients experiencing regrowth of their tumor (progression), it will be even more effective when administered to patients right after resection when the tumor load is at its lowest. All we can hope is that this guess is correct and results in a statistically significant difference. We can also hope that the small group of placebo patients who did not cross over can yield results that will prove the efficacy o L.
If I recall it correctly, the Company informed us in February that on reaching the desired 233 it would take multi months for data lock and verification of results and another few months for statistical analysis. I can understand that data lock and final verification of results of 331 cases can take several months but why should stat analysis take that long? If we hit 233 during May, when do you think we will get the yearned for results?
abeta's estimate that the median enrollment period occurred during December 2013 (44 months ago) seems quite reasonable. However I do agree with you that as the trial progresses, the number of survivors enrolled per quarter should gradually increase. For example the possibility that 4 of the 11 patients enrolled during 2008 (more than 8 years ago) are alive is questionable. In fact although interesting, I don't see an absolute necessity to estimate for each q the number of patients still alive. The fact that more than 30% of the participants are still alive at 44 months, suggests to me that at a minimum 50% are alive at 24 months. If we are lucky, there will be a significant time difference for both PFS and OS between Program, Controls who crossed over and perhaps a group of Controls who failed to cross.
Barunuuk and exwannabe, If we accept the calculation of RKmatters, Flipper and others that the midpoint of randomization for the 331 participants occurred sometimes around December 2013, we would be already 40 months past that time and not there yet. I expect the final 233 will be reached June-August or 42-44 months past midpoint. With an OS of 70% at 40 months (if it happened this month), a 50% OS at 20 months is still IMHO extremely conservative. Now if the midpoint of randomization occurred as late as JUly 2014 (a highly unlikely scenario) we would still be 33 months past the midpoint now and 35 -36 months past that point in JUne/July. Even if July 2014 were the correct midpoint I would expect more than 70% of the patients to have sadly experienced the second event by April-July 2017. In summary I expect an OS of no less than 24 months for the entire trial.
antihama, G-d forbid, just not a reverse split.
marzan, I forgot to comment on the possibility you raised that the FDA could intervene to end the trial Is there a frequent precedent for that? Does the FDA usually or sometimes intervene to end a successful trial before it is completed? This would be the right thing to do if the company running the trial has financial difficulties but I am not aware that this occurs in real life. Do you know of cases where the FDA intervened in this manner?
marzan, I know that the later the 233rd is recorded the better the results but how long can NWBO survive financially. If the SP is much lower by June/July the company may be in real financial trouble by the end of the year. What is your take on the financial defenses that NWBO will have if the final results are obtained much later than June/July?
MD1225, I should have stated in my POST # 112024 that I was questioning your opinion concerning the revealing of the final L results during 6/17 or 7/17. I know that you are basing you opinion on talks you had with the company. I would very much like you to be right but the results obtained from the Korean study (a very similar study to L) make me wonder whether June/July is overoptimistic. So as to regain my optimism I would like to be wrong. Any ideas?
A Korean study (POST 107505) discussed on this board seemed to be quite similar to our L study in that their 13 primary brain cancer patients were subjected to intradermal injections of a similar vaccine. With their small sample of patients their median PFS was 15.6 months and the OS 28.4 months. This 13 months interval in a small cohort may not be entirely reliable but it still points to a long period between PFS and OS results. If our results were similar to the Korean results than even if 248 PFS events were already obtained during November 2016 we may have to wait longer than July 2017 for final results Based on the pace of patients progressing in this trial I am making the assumption that roughly each month 4 patients that have already progressed will sadly suffer the second final event. With a 13 month average interval between progression and death I estimate that during November 2016 we had about 196 patients who passed away [248-(4x13)]. To reach 233 final events we would have needed in November 37 more events or around 9 months. So even if the 248 PFS number occurred already during November 2016, the 233 OS number would occur only during August 2017. We would then be waiting several additional months for the data lock procedure and stat analysis.
Of course this is all based on my assumptions which I hope are wrong because I am tired of the waiting and dilution. Any thoughts?
You make absolute sense with respect to the urgency to release positive results ASOP by early June. However I just don't see how they can end up with 233 secondary events, scrubbing, data lock and stat analysis by early June. If we had arrived with 248 PFS during July-September 2016 they could have had the full PFS and OS report maybe by early June, 2017. However, it is hard for me to see how 248 progressing patients occurring between November 2016 and late January 2017 can generate 233 final events and a complete preparation of final reports by early June 2017. Nevertheless I really hope that you are right and that my doubts are unfounded.
I was under the impression that the complete results concerning PFS and OS will be unveiled at the same time and until then most of the final results will be blinded. I just wrote a post (111855) in which I postulated that we may get the complete PFS and OS results during Q4, 2017. I was aware that some data concerned with the PFS results can be exposed after May 23, 2017 but it is not clear to me how much information is going to be released. What is your take on the information to be release after May 23 and perhaps during ASCO.
Several weeks ago a Korean trial which also employed a DC based vaccine designed to treat primary brain cancer patients was discussed on the MB. I was unable to relocate the study but I do recall the 13 patients involved had a PFS of about 15 months and an OS of approx 28 months. Given that only 13 patients were treated, the 13 months difference between PFS and OS represent a rough approximation of reality. However, assuming a range of 11-13 months between PFS and OS, that is still a long time interval.
The results of the NWBO study may end up being somewhat similar to that obtained by the Koreans although I suspect that the patients receiving L from the get go in the NWBO study will have a somewhat longer PFS and OS.
Still it is quite possible that the interval between PFS and OS is similar to the Korean result and a 11-13 months interval is realistic. I am making an assumption that 67% of the NWBO patients (who received L from the get go) have a PFS to OS interval of about 13 months. Now even if the 19% late crossovers and 14% non crossovers have an interval between PFS and OS as low as 8 months, the average interval would still be a little above 11 months.
Even if the 248 PFS number had already been reached during November 2016, I don't see a PR by June or JUly 2016. Assuming about 4 secondary events per months and an 11 month interval, we would have during November 2016, 248-(4X11)=204 secondary evented patients. We would need 233-204=29 additional secondary events and given 4 events per month, 8 additional months to reach 233. That would take us from November 2016 to July 2017 and at least 3 months for data lock, stat analysis etc. will land us in October 2017 (at the earliest). If 248 PFS occurred later (December or January) we may have to wait till November, December 2017.
Of course these are just my assumptions and I would love to be wrong. I hope that someone will prove to me that June, July is a more likely scenario because like everybody I am tired of waiting.
Thanks Rk. It is great to receive your clarification that the dropout rate is likely to be negligible. With 331 middle aged and older patients one would expect a very small number to die from causes unrelated to GBM before and after the onset of PFS, However this number would probably be very small and should not greatly affect a positive outcome (which we are all hoping for) except of course the shorts.
Thanks RK. That clarifies for me the phase III enrollment schedule. Another factor affecting the results may be the final drop out rate. Logically this would include participants that have died from non-GBM causes as well as those (especially placebos?) who sought alternative solutions. At the time that 233 participants are no longer alive does that mean that all 98 are still participating in the trial? This assumption has been stated in several posts. I hope that the final number of patients that cannot be evaluated, censored is low (i.e. less than 10%) but I may be wrong about that. Any input by you and others would be appreciated.
Also very sorry for the initial message redundancy but I am new at this and could not correct the errors quickly because I had to run out quickly to fulfill other responsibilities.
For the phase III enrollment we have seen some very interesting models by abeta, RK and others. However, the March 7, 2013 guidance for enrollment presented by NWBO still gives me a headache, This guidance states that by the end of Q1 2014 or the beginning of Q2 2014, 312 patients could be enrolled. So, if about 90-100 patients were enrolled by March 7, 2013, NWBO would have had to enroll an additional 212-222 patients by April 30 or let's say by May 31, 2014. Therefore roughly within 4-5 quarters NWBO was planning to enroll about 42-55 patients per quarter (212/5; 222/4). This would seem to me to be a very difficult task. I would very much appreciate any explanation or comments.
Correction AVII77, I should have my eyes examined. From your graph midpoint of enrollment looks more like December 2013 (at best).
Thanks AVII77. I just posted that I was making a rough guess that the midpoint of enrollment was in mid 2013. According to your graph it looks more like early to mid fall of 2013. This is still a little more than three years earlier than the occurrence of the 248th PFS. Consequently I am still optimistic (though not euphoric).
Thanks RKmatters. That clarifies a lot for me. I can venture a rough guess that according to your information the median patient was enrolled sometimes in mid 2013. This is still more than 3 years on trial and hopefully long enough for a successful outcome.
RKmatters,Due to the tremendous number of posts on this MB I don't know if anyone has commented that the above link concerned with the phase III enrollment timing strongly suggests that most of the patients were already enrolled before the end of 2013. This March 7 2013 release states that at an enrollment rate of 20.8-22.3 patients per quarter, 312 patients would be enrolled by Q1 or early Q2 of 2014. This fast rate of enrollment suggests that during early March 2013 approximately 220-230 patients were already enrolled and perhaps 165 patients (50% of 331) were enrolled sometimes during the fall of 2012. By November-December 2013 when NWBO started its struggle to get permission to enroll a total of 348, we could have had at least 270 patients in the trial. If this is true, it took NWBO about 20 months to enroll the last 60 or so patients to reach the total of 331. The bottom line is that the midpoint of enrollment could go back to the fall of 2012 or more than 4 years ago with about 80% of the 331 patients starting more than 3 years ago. This bodes very well for both PFS and OS>
http://www.nwbio.com/wp-content/uploads/2013/06/press2013.03.07.pdf
http://www.nwbio.com/wp-content/uploads/2013/06/press2013.03.07.pdf
Due to the tremendous number of posts on this MB I don't know if anyone has commented that the above link concerned with the phase III enrollment timing strongly suggests that most of the patients were already enrolled before the end of 2013. This March 7 2013 release states that at an enrollment rate of 20.8-22.3 patients per quarter, 312 patients would be enrolled by Q1 or early Q2 of 2014. This fast rate of enrollment suggests that during early March 2013 approximately 220-230 patients were already enrolled and perhaps 165 patients (50% of 331) were enrolled sometimes during the fall of 2012. By November-December 2013 when NWBO started its struggle to get permission to enroll a total of 348, we could have had at least 270 patients in the trial. If this is true, it took NWBO about 20 months to enroll the last 60 or so patients to reach the total of 331. The bottom line is that the midpoint of enrollment could go back to the fall of 2012 or more than 4 years ago with about 80% of the 331 patients starting more than 3 years ago. This bodes very well for both PFS and OS>
To AVI77 (or anyone else who may have pertinent information), in Phase III approximately how much time does it take from surgery to the first baseline visit?