[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 34.204.52.4. Please contact the publisher to request reinstatement.
[Skip to Content Landing]
Views 158
Citations 0
Comment & Response
August 2018

Estimating and Interpreting the Overall Survival Benefit of Checkpoint Inhibitors via Meta-analysis—Reply

Author Affiliations
  • 1National Health and Medical Research Council Clinical Trials Centre, University of Sydney, Sydney, New South Wales, Australia
  • 2Graduate Institute of Oncology, National Taiwan University, Taipei, Taiwan
  • 3Department of Oncology, National Taiwan University Hospital, Taipei, Taiwan
  • 4Cancer Care Centre, St George Hospital, Sydney, New South Wales, Australia
JAMA Oncol. 2018;4(8):1138-1139. doi:10.1001/jamaoncol.2018.1105

In Reply We thank Uno and colleagues for their comments. While graphical displays, such as shapes of Kaplan-Meier curves or Schoenfeld residual plots, are commonly used to assess the proportional hazard (PH) assumption, the interpretation of these displays is nevertheless subjective. Formal tests to verify the PH assumption are available, with the simplest of these described by Harrell and Lee,1 but the results are not commonly presented. When examining the Kaplan-Meier curves, one should consider several issues: (1) for studies where there appears to be an apparent delayed benefit, early overlap but late curve separation could still be consistent with chance rather than suggesting a true difference in hazard rates between early and late periods; (2) late curve separation needs to be interpreted in the context of extensive censoring after 12 months for these trials; and (3) curves might appear more proportional with longer follow-up and less censoring.

×