TL32: Oncology: Imaging Endpoints in Clinical Trials
*Grace-Hyun Jung Kim, UCLA 

Keywords: Imaging, Imaging endpoints, RECIST, irRECIST, Two stages design

Oncology: Imaging Endpoints in Clinical Trials In the development of oncologic therapeutic clinical trial, imaging has been used for several years to determine the duration of progressive-free survival.

Several important points are involved in the procedure of collecting imaging outcome: Imaging modality, reader credential (the adjudication rate), response evaluation criteria for solid tumor (RECIST), imaging time-points including unscheduled visits, standardization of Study Data Tabulate Model (STDM) data structure and Clinical Data Interchange Standards Consortium (CDISC), and statistical inference.

Although statisticians are the last receivers of all the process above, there are many important things can be ignored and brought unintended missing or incorrectly driven data set. Because statisticians are not involved in the each step, the simplification of the process or the decision of the process can affect at the end formally and informally.

(1) We will discuss outcomes in detail from imaging studies such as RECIST (oncology criteria), irRECIST for immune-related response criteria, and other methods such as Prostate Working Group 2 (PCWG2) in prostate cancer or Response Assessment in Neuro-Oncology Criteria (RANO) criteria in glioblastoma multiforme, special type of brain cancer. (2) A developmental medical imaging outcome in early phases (i.e. phase 1 or 2). (3) Study designs –two stages in phase 2 or adaptive design in dose-escalation

We will address the general understanding of imaging modalities, the basic concept of determining progressive disease, data collection, data structure in C-DISC and what to ask to the people who involve in the process of data collection. Furthermore we will discuss critically watchful points in statistical inference in Part (1) and statistical methods in Part (2) and (3).