Objective To compare Gehan two-stage design and Simon two-stage design in sample size calculations for phase Ⅱ clinical trials of anti-tumor drugs. Methods We explained the sample size calculation methods with a single-stage design, Gehan two-stage design, and Simon optimal two-stage and minimax two-stage designs in line with the principle of exact binomial probability. By setting up different parameters in SAS macro program, the advantages and disadvantages of these designs were compared. Results The minimax two-stage design does not increase the maximum sample size compared with the single-stage design. Compared with the Gehan two-stage design, the Simon two-stage design has the advantage of being able to determine an early termination of trials when no or low anti-tumor activities are evident. Conclusion Simon two-stage design is better than single-stage design and Gehan two-stage design. The minimax design is more popular than the optimal design.
