Quality Assurance Methodology for Multicenter Clinical Trials using PET/CT: Experiences from the NCI National Clinical Trial Network (NCTN) Imaging and Radiation Oncology Core (IROC) Service

1293 Objectives PET/CT as a non-invasive molecular imaging methodology is widely used in multi-center clinical cancer Phase II and III trials enabled through the National Cancer Institute (NCI) National Clinical Trial Network (NCTN). The Imaging and Radiation Oncology Core (IROC) cooperative was formed with the reorganization of the NCTN and started to provide network wide services since 3/2014. IROC Ohio is one of six imaging core laboratories (ICL) within the cooperative and focuses on supporting and managing NCTN trials for the Alliance and SWOG network groups. This team has made it an innovation priority to help facilitate the increased use of PET/CT within such trials and to develop best practices for quality assurance (QC). While most quality assurance has been focused only on phantom based performance assessment of the PET/CT devices, we focus in this effort specifically of the patient imaging quality assurance aspects. Methods We are the responsible ICL team for a portfolio of innovative therapeutic clinical trials that use PET/CT for imaging based disease and therapy assessment. We specifically refer to the methodology development for PET/CT imaging based QC of a phase III randomized study of diffuse large B-cell lymphomas in a broad, multi-center setting. Imaging was based on a trial-specific protocol of which institutional credentialing was mandatory prior to patient enrollment. The team that focused on the methodology development consisted of a broad range of experiences from PET/CT physics to clinical oncology to clinical imaging trial experts. Results Prior to mapping out the methodology approach, an extensive review of potential learning points from prior imaging methodology quality assurance efforts such as the FNIH supported Oncology Biomarker Qualifying Initiative (OBQI) was performed. SOP driven QC workflow with more than 15 individual sub-components integrated from site credentialing to thin-client real-time image review were developed. As a key enabling methodology, a DICOM image tag based semi-automated QC approach appeared to be the most desirable, objective and cost-effective QC support methodology. QC standards in 15 items with 4-level categories (timing, imaging, data and patient) were established specifically for PET/CT imaging. (Semi)-automatic software were developed enabling metadata processing and database management in 10 essential steps and audit processes. These support use of a DICOM tag based QC assessment matrix. Within such a template, the desired compliant parameter range should be defined as well as the additional ranges that may be not desired but still acceptable and those that are beyond and as such not acceptable. All those components facilitate a highly standardized and parameter driven semi-automated QC approach. Using a color schema (Green, yellow, red) for parameter heat mapping enables a quickly assessable performance overview that then allow for cluster based analysis of parameter compliance and help to spot issues that may be patient characteristic or site performance dependent. Site specific mapping can be an efficient quality assurance tool as many clinical sites do not even realize their variability and protocol variations within their clinic practice. Conclusions The reliable collection of metadata and effective assessment of protocol compliance in clinical PET/CT trials within broad multi-center environments has been a major challenge. The demonstrated quality assurance methodology facilitates a semi-automated, highly structured approach adapted to the complexity of multi-institutional PET/CT clinical trials. The presented heat mapping of QC assessment parameter enables an effective and efficient visual assessment of overall and/or specific performance characteristics. RESEARCH SUPPORT: This work is supported by the IROC 5U24 CA-180803 grant.