Dendreon Corp fka DNDNQ

[rancherho]

ocyanblue:

1. The following references are relevant: (a) The analysis by Dr. Daniel Petrylak in the GU presentations at: http://www.chemotherapyfoundationsymposium.org/meeting_archives/meetingarchives_tcf2006_main.html
(b) DNDN’s PR at: http://investor.dendreon.com/ReleaseDeta
© the list of the 20 covariates used in the Provenge 9901 trial in Slides 19 and 20 at: http://www.asco.org/portal/site/ASCO/menuitem.34d60f5624ba07fd506fe310ee37a01d/?vgnextoid=76f8201eb
61a7010VgnVCM100000ed730ad1RCRD&vmview=abst_detail_view&confID=34&abstractID=31910
(d) the Provenge Advisory Committee statistical Briefing Materials at:http://www.fda.gov/ohrms/dockets/ac/07/briefing/2007-4291B1-00-index.htm
(e) the NEJM published TAX 327 Taxotere pivotal trial data at: Volume 351:1502-1512 October 7 Number 15 (2004) http://content.nejm.org/cgi/content/abstract/351/15/1502
(f) Updated Survival in the TAX327 Study at: Journal of Clinical Oncology, Vol 26, No 2 (January 10), 2008: pp. 242-245 http://jco.ascopubs.org/cgi/content/abstract/26/2/242

2. The citation that you take from page 72of DNDN’s Briefing Materials (d) says so little that as a standalone statement it appears meaningless. First,as you know, a Cox proportional hazards model adjusts for factors present at the time of randomization. Thus, a Cox proportional hazards model could not adjust for docetaxel (Taxotere) use after Provenge, nor does the citation make such claim. I believe statisticians refer to such after occurring factors as “instrumental variables”. The Briefing Materials also point out that the dosing and duration of subsequent docetaxel was not recorded, only the type of chemo and the date of initiation. See Section 2.3.6 of the Statistical Briefing Materials (d). “This analysis revealed that there was not a statistically significant imbalance in the administration of chemotherapies between treatment groups” . This suggests that any adjustment for the treatment effect adjusted only for simple percentage of use comparisons between the two arms as previously presented by Dr, Small for 9901 at the 2005 ASCO.

3. As the DNDN PR (b) states: "The median survival observed in the PROVENGE treated patients who subsequently received docetaxel was 34.5 months compared to 25.4 months for patients randomized to receive placebo who went on to receive docetaxel, a difference of 9.1 months (HR = 1.90; p-value = 0.023). Approximately 68 percent of the patients randomized to receive placebo also subsequently participated in a cross-over salvage protocol that allowed them to receive active cellular immunotherapy with APC8015F, a version of PROVENGE generated from cryopreserved cells. The median survival was 25.7 months for patients who received APC8015F followed by docetaxel. In contrast, the median survival was 20.2 months for patients who received placebo only and subsequent treatment with docetaxel, a 14.3 month difference compared to 34.5 month median survival seen in the patients who received initial treatment with PROVENGE followed by docetaxel." In other words, an unstated net percentage of the 68% of control patients receiving the 2/3 dosing (compared to the experimental arm) of Frovenge processed from cryopreserved cells subsequently received docetaxel and those patients had a median survival of 25.7 months compared to 34.5 months for the experimental arm patients who received docetaxel, a difference of 8.8 months. Dosing, of course, does make a difference as both the Total Nucleated Cell data and the recently published CD54 Provenge data points out. There should be little surprise that any simple adjustment for treatment effect between experimental and control arms that ignores dosing of Provenge and the net use of the combination therapy and its net impact means little.

4. Dr. Petrylak’s presentation (a) of the Provenge data analyzed how median survival was increased by the combination therapy use in the 147 patent experimental arms of its two Phase 3 9901 and 9902a trials using the Halabi nomogram.. He reported that the experimental arm patients receiving subsequent docetaxel had a actual survival of 34.6 months vs, that predicted by the Halabi nomogram of 20.9 months, a difference of 13.6 months. The crossover control patients who subsequently received doceataxel had an actual median survival of 25.4 months vs. a predicted 19.9 months for a 5.5 month difference. The median survival in 9901, the more successful of the two trials in AIPC, of the Provenge experimental arm was compared to that projected using the Halabi nomogram and bettered it by 5.8 months (reported by Dr. Small in a 10/06 presentation). If that median survival was conservatively extrapolated to all experimental arm patients in both trials, the total projected increase in median survival would be some 853 patient months over that predicted by the Halabi nomogram (147 X 5.8). According to Dr. Petrylak’s analysis, the 51 patients receiving the Provenge + Taxotere combination would represent a projected 694 patient months of that increase over what the Halabi would predict (51 patients X 13. 6 month increase in survival) , which suggests that the 96 patient (147-51) who received Provenge alone would account for an overall increase in projected median survival of 159 patient months over that predicted by the Halabi nomogram, or about 1.7 months each ((853-694)/96). Taxotere showed an increase in median survival as a monotherapy in AIPC of 2.4 months in its TAX 327 pivotal trial (e), where 45% of the enrollees were symptomatic.

5. The 36 month survival rate for the 333 patients in the Taxotere TAX327 pivotal Phase 3 trial who took the FDA approved dosing every three weeks was recently reported to be 18.6%.(f), again where almost half of the patients were symptomatic. If the 51 patients out of the 147 in the 9901 and 9902a experimental arms who received the Provenge + Taxotere combination had median survival of 34.5 months, it is a reasonable assumption that some 24 or 25 of this subgroup were alive at 36 months post randomization, a month and a half later. 33% of the 147 experimental arm patients or 49 patients in the integrated experimental arms survived 36 months. That would mean that 24 or 25 of the 96 integrated experimental arm patients who did not receive subsequent Taxotere, or about 26%, survived 36 months. This is better than the 19.3% 3 year survival of Taxotere treated TAX 27 patients, even allowing for some upward adjustment for more advanced symptomatic patients, but is obviously far lower than 9901 and 9902a experimental arm patients who received the combination therapy.

6. Thus, whether considering the median survival point or the 3 year survival point, IMHO, the DNDN / Petrylak supplied data conclusively demonstrates the significant advantage that the combination Provenge plus Taxotere therapy provides over either Provenge or Taxotere alone. Dr, Petrylak’s presentation suggests why this may be so from a clinical MoA standpoint. IMHO, the increased efficacy of using Taxotere after an immunotherapy demonstrated in the Provenge clinical trials will also be demonstrated in CEGE’s Vital 1 trial where GVAX patients taking subsequent Taxotere will also demonstrate substantially increased efficacy over both the Taxotere only control arm and those only using GVAX in the experimental arm.