53 and 48 patients in the palbociclib plus letrozole arm and the letrozole alone arm, respectively, were censored in the BICR PFS analysis. Of these, 14 and 4 patients, respectively, were still in follow-up for disease progression. 24% of the Phase II population (29 and 21% in Part 1 and 2, respectively) were censored for PFS due to reasons other than still being on the study drug. The most common reason in each treatment arm was due to treatment being discontinued without a BICR PFS event. There was an imbalance in censoring rates between the 2 treatment arms, with fewer patients being censored for this reason in the palbociclib plus letrozole arm (26 patients, 31.0%) compared with the letrozole alone arm (33 patients, 40.7%).
Comment: The data presented in Table 28 [not in this document] differ from those data presented for the BICR censoring in Table 26 [not in this document] of the FDA report on the website for Study A5481003. The sponsor is requested to provide an explanation for all differences in the data presented in the dossier versus the FDA report, including but not limited to, the higher AE rates, clinical progression, and withdrawal of consent; noting that the FDA table was generated in response to an FDA query on 28 Feb 2014. See Clinical Questions.
For the Ph2P1 Cohort:
median PFS in the palbociclib plus letrozole arm was 31.6 months (95% CI: 11.2-NR) and in the letrozole alone arm was 38.6 months (95% CI: 7.5-38.6);
observed HR was 0.731 (95% CI: 0.300-1.779; unstratified 1-sided p=0.2442).
For the Ph2P2 Cohort:
median PFS in the palbociclib plus letrozole arm was 20.3 months (95% CI: 12.2-NR) and in the letrozole alone arm was 14.6 months (95% CI: 8.1-20.0);
the observed HR was 0.576 (95% CI: 0.316-1.050; 1-sided p=0.0342).
Comment: While a large number of patients were censored in both the treatment and control arms, more were censored from analysis in the control arm, mostly due to discontinuation without evidence of disease progression. The Part 1 and Part 2 groups were not prespecified subgroups and the study is not powered to do subgroup analyses or assessments of HR, and together with the low number of PFS events, this precludes any conclusions being drawn. In the BICR, the Ph2P1 group, those receiving palbociclib were found to have an inferior median PFS which differs from the investigator findings. This raises two concerns:
about including this population in an overall combined analysis
that protocol amendments were made based on these initial assessments
The findings in the Ph2P2 groups are encouraging, but not sufficient to provide adequate data to support registration for the proposed first line indication, particularly for such a common cancer.
This supports that the use of this combination is promising in the proposed population, but requires confirmation from full evaluation of a well-designed, randomised, double-blind controlled trial, that is, Study 1003 is suitable as a supportive but not pivotal study for the proposed usage.
Figure 7: A5481003 Kaplan-Meier plot of progression-free survival by blinded independent central review for Part 1 and Part 2 cohorts of Phase II study (ITT population)
Discordance of Investigator Assessment and Blinded Independent Central Review of Progression-Free Survival (ITT Population)
The evaluation of the discordance of investigator and BICR assessment of PFS events is presented in Table 11.
Comment: Essentially, the greatest degree of discordance arose within the Ph2P1 group assessments, which is consistent with the different PFS outcomes reported by the two groups of assessors.
Table 11: Study A5481003 Discordance of Investigator assessment and BICR on progression-free survival: ITT population