Qualification of NCI-Designated Cancer Centers for Quantitative PET/CT Imaging in Clinical Trials

Abstract

The National Cancer Institute (NCI) developed the Centers for Quantitative Imaging Excellence (CQIE) initiative in 2010 to pre-qualify imaging facilities at all of the NCI-designated Comprehensive and Clinical Cancer Centers for oncology trials using advanced imaging techniques, including positron emission tomography (PET). This paper reviews the CQIE PET/CT (Computed Tomography) scanner qualification process and results in detail. Methods: Over a period of approximately 5 years, sites were requested to submit a variety of phantom, including uniform and ACR (American College of Radiology) phantoms, PET/CT images, as well as examples of clinical images. Submissions were divided into 3 distinct time points: initial submission (T0), followed by two requalification submissions (T1 and T2). Images were analyzed using standardized procedures and scanners received a pass or fail designation. Sites had the opportunity to submit new data for failed scanners. Quantitative results were compared: across scanners within a given time point and across time points for a given scanner. Results: 65 unique PET/CT scanners across 42 sites were submitted for CQIE T0 qualification, with 64 passing qualification. 44 (68%) of the scanners from T0 had data submitted for T2. From T0 to T2 the percentage of scanners passing the CQIE qualification on the first attempt rose from 38% in T1 to 67% in T2. The most common reasons for failure were: standardized uptake value (SUV) out of specifications, incomplete data submission and uniformity issues. Uniform phantom and ACR phantom results between scanner manufacturers are similar. Conclusion: The results of the CQIE process show that periodic requalification may decrease the frequency of deficient data submissions. The CQIE project also highlighted the concern within imaging facilities about the burden of maintaining different qualifications and accreditations. Finally, we note that for quantitative imaging-based trials the relationships between the level of the qualification (e.g., bias or precision) and the quality of the image data, accrual rates, and study power needs further evaluation.