If I am interpreting the slide correctly, I believe the figure on the right shows patients with disease control (CR+PR+SD). Note how nearly every single one of those plots lands above 20% PD-1-positive T Cells. There may be two that are just below 20% (potentially the stable disease patients?).
Why is this important?
If these are the responders in the combination trial, it demonstrates that the combination can significantly increase partially exhausted T Cell phenotypes; it provides more validation for the use of the biomarker (below 20%=no response, above 30%=response, between 20%-30%=response/no response); and it establishes an intratumoral immune target/threshold for achieving responses in advanced melanoma (and potentially in many other cancers).