Secondary analysis of PFS based on blinded independent central review (BICR)
Comment: Although this retrospective BICR requested by the FDA was not prespecified and was a secondary analysis, for reasons of flow it is included here.
A blinded independent central review (BICR) was undertaken at the request of the FDA of radiographic images for 161/165 patients (97.6%) obtained retrospectively and read by the independent third party - missing scans for 4 patients were equally distributed across the 2 treatment arms.
median PFS in the palbociclib plus letrozole arm was 25.7 months (95% CI: 17.7-NR) and in the letrozole alone arm was 14.8 months (95% CI: 9.3-20.4)
observed HR was 0.621 (95% CI: 0.378-1.019; stratified 1-sided p=0.0286)
Figure 6: A5481003 Kaplan-Meier plot of progression-free survival by blinded independent central review in the combined Phase II study (ITT population)
Comment: Compared with investigators, this represents 8 PFS events fewer than reported for the palbociclib and letrozole arm and 26 fewer PFS events in the letrozole alone arm. This independent analysis also indicates a 10-month improvement in median PFS, but the lower number of progression events identified in the control arm have resulted in the 95% confidence intervals overlapping between the two arms, and the HR now crosses 1, indicating a loss of reported statistical significance reported from the investigator-assessed progression. While it is possible for clinical progression without radiological confirmation to result in censoring, the large number and particularly the imbalance affecting the control arm, in an open label trial mean investigator bias cannot be excluded.