Sample Size Re-Estimation for Confirmatory Two-Stage Multi-Arm Trials With Normal and Binary Outcomes