Back to Search Start Over

Surrogacy of progression free survival for overall survival in metastatic breast cancer studies: Meta-analyses of published studies.

Surrogacy of progression free survival for overall survival in metastatic breast cancer studies: Meta-analyses of published studies.

Authors :
Kundu MG
Acharyya S
Source :
Contemporary clinical trials [Contemp Clin Trials] 2017 Feb; Vol. 53, pp. 20-28. Date of Electronic Publication: 2016 Dec 08.
Publication Year :
2017

Abstract

Purpose: PFS is often used as a surrogate endpoint for OS in metastatic breast cancer studies. We have evaluated the association of treatment effect on PFS with significant HR <subscript>OS</subscript> (and how this association is affected by other factors) in published prospective metastatic breast cancer studies.<br />Methods: A systematic literature search in PubMed identified prospective metastatic breast cancer studies. Treatment effects on PFS were determined using hazard ratio (HR <subscript>PFS</subscript> ), increase in median PFS (ΔMED <subscript>PFS</subscript> ) and % increase in median PFS (%ΔMED <subscript>PFS</subscript> ). Diagnostic accuracy of PFS measures (HR <subscript>PFS</subscript> , ΔMED <subscript>PFS</subscript> and %ΔMED <subscript>PFS</subscript> ) in predicting significant HR <subscript>OS</subscript> was assessed using receiver operating characteristic (ROC) curves and classification tree approach (CART).<br />Results: Seventy-four cases (i.e., treatment to control comparisons) from 65 individual publications were identified for the analyses. Of these, 16 cases reported significant treatment effect on HR <subscript>OS</subscript> at 5% level of significance. Median number of deaths reported in these cases were 153. Area under the ROC curve (AUC) for diagnostic measures as HR <subscript>PFS</subscript> , ΔMED <subscript>PFS</subscript> and %ΔMED <subscript>PFS</subscript> were 0.69, 0.70 and 0.75, respectively. Classification tree results identified %ΔMED <subscript>PFS</subscript> and number of deaths as diagnostic measure for significant HR <subscript>OS</subscript> . Only 7.9% (3/39) cases with ΔMED <subscript>PFS</subscript> shorter than 48.27% reported significant HR <subscript>OS</subscript> . There were 7 cases with ΔMED <subscript>PFS</subscript> of 48.27% or more and number of deaths reported as 227 or more - of these 5 cases reported significant HR <subscript>OS</subscript> .<br />Conclusion: %ΔMED <subscript>PFS</subscript> was found to be a better diagnostic measure for predicting significant HR <subscript>OS</subscript> . Our analysis results also suggest that consideration of total number of deaths may further improve its diagnostic performance. Based on our study results, the studies with 50% improvement in median PFS are more likely to produce significant HR <subscript>OS</subscript> if the total number of OS events at the time of analysis is 227 or more.<br /> (Copyright © 2016 Elsevier Inc. All rights reserved.)

Details

Language :
English
ISSN :
1559-2030
Volume :
53
Database :
MEDLINE
Journal :
Contemporary clinical trials
Publication Type :
Academic Journal
Accession number :
27940185
Full Text :
https://doi.org/10.1016/j.cct.2016.12.004