So if you look at how they draw the survival curves, and if you imagine drawing a straight line from 0 to say year 5, the curve is UNDER the line, meaning "more events" in the curve than on the line.
So my point about the 133 and 208... according to the right curve, those should be 206 and 266 but, instead those are on the line (so to speak)
\ not L
And that is dramatic to me, in graphic terms.
From what I can see, it looks like the "signal of effectiveness" converges into year 5 (starts to fade into the background)