Figure 1.
Nomogram for known or expected HRs of a risk model, fractions of patients identified as high-risk, and numbers of events needed for a power of 0.9. The practical implications of these relationships can be exemplified as follows: validation of the m7-FLIPI in a R-CVP–treated cohort (HR of ∼3.6) requires fewer FFS events compared with a R-CHOP–treated cohort (HR of ∼2).41 Similarly, a smaller number of events will be required if the fraction of high-risk patients is approaching 0.5, for example, as expected when analyzing only patients with high-risk FLIPI.5