Predicting survival outcomes in Myeloma using surrogate markers

Abstract

BACKGROUND: OS should be considered the gold standard before adopting a particular treatment strategy as standard of care in Multiple Myeloma Phase III randomized clinical study but the use of OS as trial endpoint results in long trial duration. The primary objective of our project was to investigate whether there was a required minimum PFS difference between two arms in phase III randomized controlled trials (RCTs) that can be used as a predictor of benefit in overall survival (OS). Secondary objectives were to explore if there was a minimum threshold for VGPR rate and CR rate difference between two arms that will predict OS benefit in RCTs. DESIGN AND METHOD: We performed a PUBMED search to identify potentially relevant randomized controlled trials (RCTs) between January 1992 to January 2012. We also scanned references of abstracts presented at the American Society of Hematology (ASH) between January 2005 to August 2012; this was supplemented by manual searches of others clinical trials. We used both absolute differences in the survival improvement (in months) and response rates between the two arms, as well as proportional improvements for the purpose of analysis. Descriptive statistics were used to summarize the minimum threshold PFS, CR AND VGPR median differences respectively. RESULTS: Assessment of all publications resulted in identification of 75 RCTs. Of the 75 RCTs studied, 17 (22%) had statistically significant improvement in OS on intent to treat analysis (p-value ¡Ü 0.05) .We found that the minimum improvement in median PFS/TTP required to produce a significant improvement in OS was at least 2.5 months or more. This number varied depending on the stage of the disease and the type of treatment . CR improvements appeared to be widely variable, ranging from -5% (arm with survival improvement having worse CR rate by -5%) to 36%, with no particular pattern relative to type of therapy administered and the minimum threshold needed for survival benefit. VGPR rates were reported only in 5 of the 18 trials and therefore could not be accurately computed. CONCLUSION: The current data is still immature to consider PFS improvement a pivotal surrogate marker of OS. We are limited by lack of data on Multiple Myeloma clinical trials showing OS significance.