Is the Median PFS Difference Meaningful for Assessing Treatment Effect?

• Activity Number: 412
• Type: Topic Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Steven Snapinn*
• Companies: Amgen
• Keywords: Absolute risk reduction ; Clinical meaningfulness ; Hazard ratio ; Overall survival ; Progression-free survival

Abstract

The magnitude of a treatment's effect on PFS is a key consideration for assessing the clinical meaningfulness of that effect. Various metrics can be used for this assessment, such as the hazard ratio or the absolute risk reduction. One metric in common use, particularly among clinicians, is the difference between groups with respect to median PFS. In this presentation I will discuss some limitations of this metric, including the fact that the difference in medians cannot be interpreted as the median causal effect of the treatment. I will also contrast the difference in medians with the absolute risk reduction and show that, for a given hazard ratio, they are in conflict with each other; based on that, I will argue that the hazard ratio is the most appropriate metric for assessing clinical meaningfulness.