Application of a new multinomial phase II stopping rule using response andearly progression

Purpose: A multinomial stopping rule had previously been developed that inc orporated both objective response and early progression into decisions to s top or continue phase II trials of anticancer agents. The purpose of this s tudy was to apply the multinomial rule to two independent sets of phase II data to assess its utility in appropriately recommending early trial closur e as compared with other stopping rules. Materials and Methods: Data from completed phase II trials of the National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) and European Or ganization for Research and Treatment of Cancer Early Clinical Studies Grou p (ECSG) formed the basis of the study. Based on observed results for each trial, the recommendation of the multinomial stopping rule was applied, as was the recommendation of the actual stopping rule used (Fleming or Gehan). The appropriateness of the recommendations was evaluated based on interpre tation of final study results. Results: The standard and multinomial rules disagreed on early stopping in one of 16 NCIC CTG trials and in seven of 23 ECSG trials. In all cases, the standard rule advised continuing to the second stage whereas the multinomi al rule advised stopping early because of excessive numbers of patients exp eriencing early disease progression. Final trial results indicated that the multinomial recommendation was appropriate, because in no study did final results lead to conclusions of activity. Conclusion: In this series of trials, the multinomial stopping rule perform ed more efficiently than the Fleming or Gehan rules in advising early stopp ing of trials. These results encourage continued exploration of this approa ch for phase II trials of cytotoxic and noncytotoxic anticancer agents. (C) 2001 by American Society of Clinical Oncology.