Perhaps a simpler way to look at things. Old and new insights.
I came back to post this. See you soon….have a nice weekend.
Linda Liau and other researchers determined, before they started the current phase [III] study on DCVAX-L, that complete resection without any trace of residual tumor recurrence after surgery/radiation and chemotherapy has a synergistic effect with dendritic cell therapy when DCVAX-L is given on the heels of the Stupp’s protocol.
The current phase III trial only accepts patients with no trace of tumor recurrence at the initiation of DCVAX-L therapy. (The patients who are not selected are placed in a separate study and/or compassionate use program) This means legitimate and ethical cherry picking happened in the phase III trial and not in the phase I trials.
All current DMC interim analysis in phase III trials that considers efficacy have a predetermined stopping rule/point. For some reason, many posters do not want to believe this can statistically occur with 66 events. Nor do many posters believe Phase 3 trials (as opposed to phase 2 trials) can be stopped early. I have one word for them – Idelalisib. While we can all debate what the stopping P Value, Confidence Index and the like might be for the first interim review, we cannot IMHO debate that one exists for the primary endpoint (PFS). (Note 66 events is 60% of a total possible 110 events in the DCVAX-L trial. Idelalisib was stopped for efficacy in stage 3 at approximately 50% events)
Linda Powers graciously shared with us that patients with various cancers in clinical trials and compassionate use, had a “consistent” 80% response rate to DCVAX-L treatment. Here is an insight not discussed since that information was released. The current phase III trial will almost certainly have a response rate above 90%. Why? Because the phase three trial excludes patients demonstrating any trace of residual/recurring tumor at the time DCVAX-L therapy is initiated. As stated by Linda Liau in previous scholarly articles, the response for DCVAX-L (and other dendritic therapy) is enhanced significantly when there is no bulk and or trace of tumor recurrence after the Stupp’s protocol when DCVAX-L treatment is initiated.
Linda Powers also graciously verified that GBM treatment with DCVAX-L in clinical and compassionate use settings consistently extended Overall Survival by one-and-a-half years compared to standard of care overall survival, and it extended Progression free survival by one-and-a-half years compared to standard of care progression free survival. Here is another insight not discussed since this information was shared. The current phase III trial is almost certain to extend PFS (as well as OS) by more than 2 years beyond current standard of care progression free survival. Why? Because the phase three trial excludes patients demonstrating any trace of residual/recurring tumor at the time DCVAX-L therapy is initiated. As found in Linda Liau’s prior research, DCVAX-L’s effect upon PFS and OS is synergistically enhanced significantly when there is no residual bulk and/or trace of tumor recurrence after the Stupp’s protocol when DCVAX-L treatment is initiated. In other words, while PFS for standard of care might be enhanced where no trace of tumor is found immediately following the Stupp’s protocol, this PFS enhancement does not compare to the synergistic impact on PFS extension when combined with DCVAX-L treatment.
Like most of you, I have taken some time to understand probability values and Kaplan-Meier’s curves. Please review the DCVAX-L Progression Free Survival curve for phase 1 DCVAX-L used in Linda Power’s December 2013 presentation.
Below is my hypothesis regarding how the PFS curves will change from phase 1 to phase 3.
Its easiest to visualize it like this:
Imagine you print off the curves on the PFS graph from phase 1. Next take a pair of scissors and cut the 75% horizontal line all the way across the page. Now Change the 70% line to 100%. Next, remark the percentages accordingly all the way down.
That is what the phase 3 curves will likely look like. Notice the standard of care curve will be somewhat enhanced, but it can only improve so much without synergy, whereas the treatment curve will become much 1. Flatter/extended in months and 2. higher in percentage.
The P Value is basically the overlap and/or distance between the 2 curves. What happens when you flatten the treatment curve more dramatically than the control curve is you significantly change the P Value. The P value for Phase 1 PFS was .00001. Or 1 chance in one-hundred thousand that the beneficial results from DCVAX-L treatment were a result of chance. I believe that the phase 3 curve will result in a P Value of .000001. In other words, I speculate it will show there is only a 1 in one-million chance the results from DCVAX-L treatment were the result of random chance.
(Note: These results with sufficient data points would not be possible so soon if it were not for the 33 patients carried over from when the trial began almost 7 years ago. The Kaplan Meier’s curve would otherwise have “censored” data points where the PFS data point is not yet known. In this case, we are very fortunate to have a great number of data points from the very old treatment group as well as new data points from the recent control group.)
