Skip to Content
Biometric Research Program (BRP)
Last Updated: 11/22/16

Sample Size Planning for an Integrated Phase II/III Study

The integrated phase II/III study design contains two portions. In the first portion of the study evidence of activity is gathered using progression free survival (PFS). If there is sufficient evidence of activity, accrual continues and enough patients are accrued until the phase III endpoint of overall survival can be assessed (patients from the first portion and the second portion are used in the final overall survival analysis) (Hunsberger, Zhao and Simon, Clin Cancer Res 2009).

This program calculates either simulated or an approximate expected time, expected sample size and final power for:

  • the integrated phase II/III design
  • a single randomized study with overall survival as the endpoint
  • a single study with overall survival as the endpoint with a futility analysis based on survival
  • separate randomized phase II followed by phase III

Approximations of sample size for the integrated approach and the futility design based on OS do not account for correction between OS and PFS or correction due to performing repeated analyzes on OS. The simulation option below calculates sample size for all designs. Data is simulated as the paper describes. The simulations account for the correlation in the data. For the simulated PFS data the assumption of an exponential distribution is no longer valid. Therefore, simulated sample size calculations for the sequential of 2 independent studies are also provided.

Choose one of the followings to calculate sample size for integrated design and futility:
based on OS using approximations
based on simulation (calculation will take ~5 minutes)

Input parameters below. Note all time entries should be in the same units.

Design parameters for overall survival study (with no futility analysis):

Design parameters for independent phase II study with progression as endpoint:

Interim stopping parameters: