Lori Henderson, Presenter: Nothing to Disclose

Erich Huang PhD, Abstract Co-Author: Nothing to Disclose

Frank I. Lin MD, Abstract Co-Author: Nothing to Disclose

Lalitha K. Shankar MD, PhD, Abstract Co-Author: Nothing to Disclose

Patient accrual in imaging clinical trials has historically been problematic. To address this, factors that can potentially indicate good or poor ultimate accrual are explored qualitatively and analyzed by statistical methods for associations. The results of this study provide insight into factors that can predict accrual issues and aid in the design of successful studies.  

An analysis, including a qualitative description of the portfolio, was performed on R01-funded research grants which supported Phase 0, I, or II clinical trials that investigated novel or novel uses of imaging agents, modalities, interventions, or methodologies. The data included accrual information and trial characteristics of 109 grants which had 3 or 5 years of open enrollment from 2005 to 2010. Patient accrual was tabulated annually, most recently in 2014. Studied factors included clinical indication (e.g. screening, diagnostic, therapeutic assessment), trial design features (e.g. numbers of required scans), study objectives, imaging agent/modality, treatment regimens, and primary disease location. Univariate associations of accrual at the time of project completion with trial features and percent of target accrued (PTA) at 2 years were explored. The 2 year PTA association was derived from likelihood ratio tests based on Cox regression models whereas accrual performance vs. trial features associations were evaluated using log-rank tests.

Accrual performance demonstrated strong positive association with 2 year PTA (p<0.0001) and negative associations with two trial characteristics: primary liver disease site (p=0.0034) and use of X-rays (p=0.0073). Accrual performance demonstrated weaker positive association with use of computed tomography (CT) (p=0.093). There was no evidence of association of accrual performance with a study’s primary objective.

Higher 2 year PTA and use of CT are associated with greater likelihood of accrual success, while lower likelihoods are associated with X-ray use and primary liver disease site. No association between accrual and the primary objective was found.

Evidence-based medicine depends on data from clinical trials.  This study identifies factors which improve the successful completion rate of clinical trials with imaging components.

Henderson, L, Huang, E, Lin, F, Shankar, L, Predictors of Accrual Success in Oncologic Imaging Trials.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14006782.html