Skip to main content
. Author manuscript; available in PMC: 2019 Apr 15.
Published in final edited form as: Clin Cancer Res. 2017 Dec 6;24(8):1785–1794. doi: 10.1158/1078-0432.CCR-17-1970

Figure 3.

Figure 3

Correlation between toxicities in phase I and in later trials.

Y axis is the types of toxicities described in phase I trials as a percent of the types of toxicities described in the registration trials.

Left panel: correlation between the number of patients included during the phase I trial and the description of all clinically significant toxicities (black diamond), as well as immune-related toxicities (gray square) from later trials. This panel shows that there is an increase in the ability to identify the immune-related toxicities as the number of patients in phase I increases. The number of patients in phase I did not, however, correlate with the ability to identify all types of clinically relevant toxicities.

Right panel: correlation between the ratio of patients included during phase I over patients included in registration trials and the description of clinically significant (black diamond) and immune-related toxicities (gray square) from later trials.

This panel shows that the ratio of the number of patients in phase I over the number of patients in the registration trial(s) correlated with the ability to identify either immune-related or all clinically relevant toxicities.