Abstract 959
Background
The Adaptive phase (ph) 1/2 trial design with larger-sized expansion cohorts (ExC) (abbreviated, ‘SUMO’ design) has gained popularity in oncology. We examined the perceptions of ph 1 investigators toward this design.
Methods
Subjects were identified by expert nomination or via PubMed search of authors who have published > 3 ph1 or ph 1/2 cancer drug trials from 2010-17. Invitations were sent by emails containing an e-link to a survey with 4 domains: (1) ph 1 experience; (2) perception on utility, (3) safety & scientific validity, & (4) operational challenges. Responses were graded as Likert scores.
Results
534 invitations were sent & 98 responses (18.3%) were returned. Of the 80 respondents who completed all questions, 76.5% are male & 92.8% are academic physicians in university teaching hospitals from North America (31.7%), Asia (30.5%), Europe (25.6%) & Australia (12.2%). They are mainly experienced principal investigators (PI): 70% are PI of > 5 active ph 1 trials, 60% have been PI for > 10 yrs, 53% graduated > 20yrs. Over 80% felt that SUMO design can expedite drug development, but opinions are divided over if it should be preferred over other designs (46.3% Yes, vs 53.7% otherwise, o/w). When selecting ExC, 57.1% felt that strong scientific rationale based on preclinical data is the key, but 31.2% felt it was not necessary. Divergent opinions are reported over the appropriateness of waiving standard statistical methods like pre-defining type 1/2 error (45.5% OK to waive, vs 54.5% o/w), early stopping rule (52% OK to waive, vs 48% o/w) & a ‘control arm’ (26.7% OK to waive vs 73.3% o/w) in randomized cohorts. Most respondents (68%) felt that the SUMO design for first-in-human trials does not put patients at greater risk. Most agreed with increasing size & number of ExC, yet, 74.7% reported ‘some’ to ’significant’ operational challenges at their respective institutions with SUMO trials. Detailed correlations will be presented.
Conclusions
Although the SUMO design is perceived to be a safe & efficient way to develop cancer drugs by most experienced Ph 1 trialists, this survey has highlighted some pressing issues in its application: the need to optimize the scientific & statistical justification of ExCo selection, to recognize & improve the operational challenges posed by such design.
Editorial acknowledgement
Clinical trial identification
Legal entity responsible for the study
Brigette Ma.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.