To improve efficiency and safety of phase I trials, various new designs have been proposed. We reported that rule-based design and model-based design were similar profile in efficiency and safety (Shimomura A, et al. TAT 2017). The limitation of the study was retrospective single institutional study. Then, we investigated the literature or presentation based analysis to compare model-based design with rule-based design of phase I trials.
Published article or presentation of phase I trials which conducted at our institute between Dec 1996 and Dec 2016 were reviewed. Trials which met to the following criteria were excluded; combination trials, trials without dose escalation, trials with expansion cohort only, uncompleted trials and dose-finding trials (to assess safety and pharmacokinetic profiles in two or three doses up to the maximum tolerated dose (MTD) previously determined in the West). Efficiency and safety were assessed in two design groups; rule-based design (conventional 3 + 3 design and accelerated titration design) and model-based design (CRM using escalation with overdose control). Number of patients who treated in MTD and around MTD (from under 20% to over 20% of MTD) was assessed to evaluate efficiency. Number of patients who treated over MTD or developed dose limiting toxicity (DLT) was assessed to evaluate safety.
Of 47 trials, 35 trials have published. A total of 670 patients were included to the analysis. Five-hundred sixty (83.6%) and 110 (16.4%) were assigned to rule-based and model-based trials, respectively. 183 (32.7%) patients in rule-based trials and 17 (15.5%) patients in model-based trials were treated in MTD (p=.0001). 256 (41.6%) patients in rule-based trials and 23 (20.9%) patients in model-based trials were treated around MTD (p=.0013). Number of patients who treated over MTD was 63 (11.3%) in rule-based design and 8 (7.3%) in model-based design (p=.196). Frequency of DLT was 74 (13.2%) in rule-based design and 7 (6.4%) in model-based design (p=.0308).
Trials with rule-based design tend to treat more number of patients not only at both MTD/around MTD but also with DLTs than model-based design equipping trials.
Akihiko Shimomura
Has not received any funding
S. Iwasa: Grants and personal fees from Eli Lily, personal fees from Taiho, grants and personal fees from Chugai, grants from Novartis, grants from Merck-Serono, grants from Daiichi-Sankyo, grants from Bristol-Myers Squibb, from null, outside the submitted work. Y. Fujiwara: Grants from AbbVie, grants and personal fees from AstraZeneca, grants from Chugai, grants from Daiichi-Sankyo, grants from Eisai, grants from Eli Lilly, grants from Incyte, grants from Merck Serono, grants and personal fees from MSD, grants from Novartis, personal fees from Taiho, grants and personal fees from BMS, personal fees from ONO, outside the submitted work. T. Shimizu: Research expenses from Bristol-Myers Squibb, Daiichi-Sankyo, PharmaMar, FivePrime, 3D Medicine, Takeda-Milleniam, personal fees (lecture fees) from Chugai Pharmaceutical Co., Ltd., Taiho Pharmaceutical Co., Ltd., Ono Pharmaceutical Co., Ltd., Ono Pharma Taiwan Co., Ltd., Boehringer Ingelheim, outside the submitted work. N. Yamamoto: Reports grant from Bristol-Myers Squibb, Chugai, Taiho, Eisai, Lilly, Quintiles, Astellas, Novartis, Daiichi-Sankyo, Pfizer, Boehringer ingelheim, Kyowahakko. Kirin, Bayer, ONO PHARMACEUTICAL Co. Ltd., and Takeda, other from Chugai., Ono Pharmaceutical Co., Ltd., AstraZeneca, Pfizer, Lilly and Bristol-Myers Squibb outside the submitted work. All other authors have declared no conflicts of interest.