Dendreon Corp fka DNDNQ

[alizumo]

Thanks Wall, iwfal for your interest

Here is another simulation

Here is another distribution (to be conservative enough, the curve has a very artificial look) and the subsequent PSADT for an unblinding on 08/17/2006 with 131 patients (75%) as the threshold or an unblinding on the 12/01/2006 if it occurs when DNDN says it will release it. I used 175 patients as specified by iwfal:

juin-01 1
juil-01 0
août-01 1
sept-01 0
oct-01 1
nov-01 0
déc-01 1
janv-02 1
févr-02 1
mars-02 1
avr-02 1
mai-02 1
juin-02 1
juil-02 1
août-02 1
sept-02 1
oct-02 1
nov-02 1
déc-02 1
janv-03 1
févr-03 1
mars-03 1
avr-03 2
mai-03 2
juin-03 2
juil-03 2
août-03 3
sept-03 3
oct-03 3
nov-03 4
déc-03 4
janv-04 4
févr-04 4
mars-04 5
avr-04 5
mai-04 5
juin-04 5
juil-04 6
août-04 6
sept-04 7
oct-04 7
nov-04 7
déc-04 8
janv-05 8
févr-05 8
mars-05 11
avr-05 11
mai-05 12
juin-05 12
175


I used a Chi-2 test comparing the number of patients having doubled or not in the two arms to get a p value. This is not calculated using the log rank method but it gives a feeling of whether the results will be significant.

PSADT for unblinding on 08/17/2006:

Combined: 10.01 months
If Control 6.7 months then Provenge 14.08 months p value 0.00042
If Control 9 months then Provenge 11.07 months and p value 0, 31

PSADT for unblinding on 12/01/2006:

Combined: 11.94 months
If Control 6.7 months then Provenge 18.85 months p value 0.0000011
If Control 9 months then Provenge 15.37 months and p value 0, 0079

If results are similar to the Avastin + Provenge trial and DNDN doesn’t say when the unblinding occurs:

Control 6.7 months
Provenge 12.7 months
P value 0.00247

Then the unblinding should have occurred on 07/17/2006

It is obvious that there are not enough details especially regarding the threshold for unblinding. Nevertheless I hope it helps figure out which hypotheses might be correct

If the enrolment curve towards the end is not as steep as this one, then the unblinding according to the Avastin + Provenge hypothesis should have occurred earlier or the PSADT should augment for a given unblinding date.

iwfal, I am not familiar with what you mean by r^2 or r^4. Is it a doubling (quadrupling) rate of enrolment every year or a doubling (quadrupling) between the first year and last year of a trial or something like that?