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.
Combination data shown by Linda Liau on Dec 5.
I do not think the shown data is from the trial that everyone is posting about. That trial is designed to compare DCVax-L treatment arm to DCVax- L + Keytruda treatment arm.
The data she showed was for DCVax-L + Keytruda in both arms of the trial. In one arm Keytruda was given as the preadjuvant and in the other arm DCVax-L was given as the preadjuvant. I have not previously seen a description of this trial.
Biosectinvestor, thanks I had not seen those announcements.
Any chance NWBO collected samples from the patients to do DNA microarray studies to determine which patients were Mesenchymal, similar to the studies done by Linda Liau's group to show that this subgroup responded best to DCVax-L. That would be an exciting surprise to spring in the publication.
ATLnsider, that is fine I was not offended.
ATLnsider,
I wasn't saying one was better than the other, just that it is not fair to directly compare mOS between the 2 curves which have a different starting point..
As I pointed out before the data on record is in the paper.
Hope: as close as I can tell it is 20% at the true 36 months. Just type into Google 2018 Linda Liau paper and it comes up. Print the graph in question. Then take the curve out to 39 months to correct for it being from time of surgery. Do this by extrapolating the existing curve (which is pretty straight) with a ruler out to 39 months and read across to the Y axis and as close as I can tell it intersects at 20%.
The statistics say DCVax-L is better that SOC.
Dan88's arguments are good and are easier to buy if you are looking at the true numbers for the 2 curves.
I think expectations were inflated by showing the 2018 data plotted from time of surgery.
Dan88: I have been trying to get across to every body on the board that the bears are doing an unfair comparison between the 2018 data (plotted from time of surgery) and the TLD ( plotted from time of radomization). See my last few posts. When corrected the 2018 data drops from 28% to ~20%.
This is sort of funny because I remember the bears criticizing the 2018 data because it was plotted from time of surgery and they suggested this overestimated the vaccine's effect.
Oh well I guess they just forgot the 2018 data was from time of surgery.
Hope, it looks like it takes it down to ~40% estimating it on my screen. However, I can not seem to capture a copy of the figure from the paper to put in photoshop to cut out 3 months, slide the curve back to zero and then extrapolate it another 3 months (from 33 to 36 months).
The data of record is that published in the Journal.
Ex, you know by now that the 2018 curve is plotted from time of surgery and the TLD is plotted from time of radomization, so why do you insist on still putting this false narrative out there?
Hope, you wrote: "Also from the nwbo 2018 pr from SNO they showed blinded 3 year survival for methylated at 49.1%, trial patients from graph show about 30-35% 3 year survival."
I just wanted to point out that the 2018 plotted data from date of surgery and the TLD plotted data from time of the radomization. This has to be corrected for in the comparisons between these data. If you plot the 2018 data from time to radomization it will be significantly lower than 49.1% at 36 months.
Thanks Doc.
You are basing that statement on a comparison of the 2018 data versus the TLD data without pointing out that the 2018 data is plotted from time of surgery while the TLD data is plotted from time of radomization.
If you take out the first 3 months data from the 2018 (when no one is dying) and slide the curve to the left 3 months the TLD data does not get worse.
They do not censor patients that have progressed, they censor patients that they have lost to follow up.
Flipper thanks for answering my question. I look forward to the expansion on the topic tomorrow.
Hi Extremist223, in your defense of NWBO against the bears you stated
"If I understand this correctly, the shorts are already assuming that for some reason 35 people of 99 never had progression while starting in the placebo group, so they couldn't be used in the recurrent analysis. They are implying why is this so?"
If the bears had bothered to look at the progression curves in the TLD presentation they would have seen that 81 out of 99 placebos progressed. The rest were censored as the number of subjects at risk was zero at the 36 months.
FeMike,
The other problem with Bear Argument #4: Placebo outperformed the treatment group is that they are comparing apples and oranges.
The 2018 paper measured survival from time of surgery while the TLD survival data was from time of radomization. If you cut out the 1st 3 months from the 2018 survival curve and slide the curve backwards to zero then you get a curve that declines much faster because during the 1st 3 months no one dies. This gives you a curve from 0-33 months. The trend is very well established at 33 months which allows you to extrapolate the curve out 3 months more to 36 months. This gives a 36 mo survival of ~20% (not 28%) for the entire population.
Whether, the Bears are knowingly being deceitful or just sloppy in their research is unknown.
Flipper 44, that is an excellent analysis and puts into perspective the various subgroups that responded better than the entire population.
I also, have a question for you. Do you have an explanation for the anomaly in the nGBM MGMT methylated curve at 36 months. It appears that the curve is rapidly approaching the external control curve and then suddenly moves away from it and establishes a much slower decline. It appears as though there are two distinct populations of patients in this group. Any thoughts on this.
Exwannabe,
That is not the right way to correct the curves. You must cut out the first 3 months of the 2018 curve not the last 3 months of the TLD data curve to correct the curves. The first 3 months of the 2018 curve is when no one was dying as you can see because the curve stays around 100% for 3 months. This pushes the whole curve to the right and makes it look like everyone is living longer. When you cut out the first 3 months then the survival at 36 months drops to around 20% for the 2018 data.
Bears are comparing 2 different curves. The interim paper had the survival curve plotted from date of surgery. The survival curve shown in TLD is plotted from date of radomization. They are making a lot of inferences about the difference in the % survival at 36 months.
If you shift the interim curve to the left 3 months and extrapolate the curve from 33 months out to 36 months you will see that there is no difference in the survival at 36 months.
First, one should point out to the bears that the 2 curves they are comparing are not the same. The data for the unblinded survival curve is plotted from date of surgery while the TLD survival curve is plotted from date of radomization.
In regard to the DCvax Direct trial still being listed as ongoing, perhaps some of (or all of)the 7 patients listed as being alive in the previous publication are still alive. If they are following these surviving patients is that considered to be an ongoing trial?
Thanks Swegen
Where do you find the meeting ID number? The only number on my proxy card was 16 digits while they are clearly looking for a 9 digit meeting ID number.
Where does one find the meeting ID number? All I see on my proxy card is a 16 digit number and it appears they are looking for a 9 digit number for the meeting ID number.
Just got an e-mail that AACR is cancelled (to be re-schedule later in the year). Can ASCO be far behind? That would put a kink in a planned big reveal at ASCO.
New twist in resistance to immune checkpoint inhibitors
New paper suggests tumors put out exosomes with PDL-1 expressed on their surface which interact with T-cells in lymph nodes to suppress their anti-tumor activity.
Not clear why ICIs don't block the PDL-1 on exosomes from interacting with PD1 on T-cells but that is what the authors are suggesting.
https://www.genengnews.com/news/cancer-immunotherapy-may-run-afoul-of-exosomal-checkpoint-proteins/
The trial has been ongoing for a decade and we know there are a number of long term survivors. Thus, if the plateau has been established and there are a number of tick marks at 4, 5 6 8 yrs then that would be even more impressive than the truncated 36 month curves.
I certainly have no clue whether there is a buyout by BP on the horizon. However, for those who are arguing BP would not buy on the basis of still blinded data, I would point out that there is some still blinded data that we have not been shown. To date we have only been shown data up to 36 mo post surgery. How much would the case for efficacy be strengthened by showing the entire curve (including the long time survivors)to BP? Would that be sufficient to assure them of the trials success?
Senti, I just counted the number of steps down without considering the length of the drop. When I get some time I will make another attempt to count taking the length of the drop into consideration.
Flipper, in order to be counted in the number surviving for >36 mos you have to be among the 182 patients that are 36 months or more out from surgery. However, to be counted among the number of deaths between 0 and 36 months, the patient can be any number (0 to 36) of months out from surgery.
Totaling the 11 apparent LTFU early (prior to 19.5 mo), the 182 deaths that I counted on the 0 to 36 month curve and the 65 censored gives 257 patients. Subtract that from 331 and that gives 74. There is the problem as the paper seems to say there are only 44 patients past the 36 month mark when the numbers were collected. So where are the other 30?
I should say that counting the drops in the curve and the tick marks is not an exact procedure. I have more confidence in counting steps down than ticks. If a tick is 3 times as wide as normal I assume they are plotting 3 events at virtually the same time point, so it may not be an exact count.
When the paper first came out, I did a careful count of the deaths and censored events and came to the same conclusion as you did. There were well over 44 patients that were not on the published curve. I could not reconcile this contradiction so I just set the numbers aside. My counting did not break it down into time increments as you did, so to compare with your numbers I went back and recounted. My numbers (see below) are very close to your counts.
First 12 mo (7,34) (censored, deaths)
12 to 18 mo (4,62)
18 to 24 mo (37,47)
24 to 30 mo (21,23)
30 to 36 mo (7,16)
11 LTFU
182 died
65 censored
to equal 258 patients
331 - 258 = 73 patients not on the chart
Yes, but now they are saying they have 7 long term survivors out of 39 patients. If they could instead say they have 7 long term survivors out of 20 patients treated with Method B that would have been more impressive.
I have gone through the DC direct paper and have a couple of comments. The first is that I was disappointed that there was no discussion of the two ways for activating DCs (method A and B) and the supposed better survival for patients treated with method B. Does this mean that the initial observation has not held up or is there some reason that prevents them from discussing this issue. I have thought for some time that the delay in publishing the DC Direct paper might be related to patent issues. A careful reading of the patent applications seems to indicate that they have worked out how to get the appropriately high cytokine production linked to high overall survival in most future patients. So its not just that method B is better than method A but that they can keep adjusting the procedure to get the right cytokine release from most patients. So to publish the comparison between method A and method B they would have to describe each method in the paper. They would be reluctant to do that prior to getting the patent coverage finalized.
The second point is that all 7 of the patients that were alive at the time of the ASCO 2017 presentation were still alive when the recent paper was submitted in its final revision.
Flipper,
Yes, after your post I re-read the paper and I agree with your interpretation. Great pick-up. You are truly an asset to this message board.
Dr Liau PET scan method was for differentiating tumor progression from pseudoprogression.
https://www.ncbi.nlm.nih.gov/pubmed/28874539
New PET scan method for measuring PDL-1 expression in tumors described by Bristol Meyers Sqibb. Might be useful for monitoring the PDL-1 expression in patients prior to and sometime after treatment with DCVax-L as a early indicator of whether the patient is responding to treatment.
https://immuno-oncologynews.com/2018/03/09/pet-scans-could-identify-those-likely-to-respond-to-pd-1-targeting-immunotherapy/?utm_source=IO+News&utm_campaign=677765d91b-RSS_WEEKLY_EMAIL_CAMPAIGN&utm_medium=email&utm_term=0_f04c303b86-677765d91b-71536941
Doc logic,
After a more in-depth reading of the article, I would agree with your suggestion that these are Phase 3 (or parallel study)patients who have progressed and are moved onto these combo treatments.
Just quickly scanning the paper I don't see a reference to a clinical trial number.