In my previous blog, I reported on my analysis of outcomes and endpoints by disease for the top five Asian cancers. For the final blog in this series, I will present my analysis of the outcomes and endpoint by start year and then wrap up the series with some trends and correlations.
For this set of 440 phase II-III trials in the top five Asian cancers, 8% did not publicly disclose the start year (data not shown), 6% started during 1991-1998 (data not shown), 17% started during 1999-2003, 45% started during 2004-2008 and 24% started during 2009-2013. The drop-off in trial starts during the most recent five year period could be due in part to the under reporting of start dates in the public domain. Start dates and end dates are often reported in the public domain after a trial has been initiated and/or ended (CitelineAnalyst Tip #9).
As shown in the chart below, associations with positive or negative primary outcomes for the four later-stage efficacy primary endpoints of DFS, PFS, OS and TTP changed over successive five year time periods. The totals for all time periods are shown as a comparison. DFS was associated with more positive than negative outcomes in 1999-2003 but the opposite in 2004-2008. Use of DFS was minimal in 2009-2013. PFS was associated with more positive outcomes in 1999-2008 but more negative outcomes in 2009-2013. For any of the five year time periods, OS was consistently associated with more negative than positive outcomes. Finally, TTP was associated with more positive outcomes in 1999-2003 and 2009-2013, though use was minimal. Use of TTP was nonexistent in 2004-2008.
Overall, trials increasingly incorporated more pharmacogenomic biomarkers to select or stratify patients over successive time periods. But as shown in the table, the rates of incorporation vary by primary endpoint (Data is not shown if less than five trials existed during a time period). Progressive incorporation occurred most dramatically in PFS followed by OS. Incorporation for DFS seems to have peaked during 1999-2003 and incorporation for TTP was minimal during all time periods.
Trends and Correlations
- Overall, non-industry sponsored trials were associated with more positive primary outcomes. The opposite was true for industry sponsored trials.
- Sponsor type, disease and start year all influenced the association with positive or negative primary outcomes for DFS, PFS and TTP. OS was not associated with more positive than negative outcomes regardless of sponsor type, disease type or start year. At best, OS was associated equally with positive and negative outcomes for specific cancers.
- Incorporation of pharmacogenomic biomarkers to select or stratify patients was progressive over time for PFS and OS. Incorporation was most likely to be associated with positive primary outcomes if PFS was chosen as the primary endpoint, especially in Breast and Lung cancers.