The QIBA Profile for FDG PET/CT: Improving the Value of PET

See also the article by Kinahan et al in this issue.

Dr Gary Ulaner is an associate member (tenure track) in the Department of Radiology at Memorial Sloan Kettering Cancer Center. He is currently on the editorial boards of Radiology, Journal of Nuclear Medicine, European Journal of Nuclear Medicine and Molecular Imaging, and Clinical Nuclear Medicine. He is the principal investigator of four prospective clinical trials of novel PET radiotracers for patients with breast cancer and myeloma, including funding from the National Institutes of Health and the Department of Defense. He founded and directs the MSK Oncologic PET/CT course and authored Fundamentals of Oncologic PET/CT, a textbook geared toward those learning the modality.

PET, PET/CT, and PET/MRI are widely used in oncology, most commonly with the glucose analog 2-[fluorine 18]fluoro-2-deoxyglucose (FDG), which allows visualization of many types of tumors. PET has long been referred to as only “semiquantitative.” If the same patient undergoes an FDG PET/CT examination on two consecutive days, then the standardized uptake values (SUVs) measured at FDG PET may vary between scans, despite the lesions being identical. This is because the amount of FDG uptake in a lesion depends on multiple biologic and technical factors. These include the blood glucose and insulin levels, FDG uptake time, partial volume effects in small lesions, variability in different scanners, reconstruction methods, and patient motion. Variation in SUVs and lack of standardization in PET imaging protocols has been a major obstacle to the use of FDG PET in solid tumor clinical trials and in clinical practice (1).

The Quantitative Imaging Biomarker Alliance (QIBA) was started by the Radiological Society of North America in 2007 to address such obstacles. Its mission is “to improve the value and practicality of quantitative imaging biomarkers by reducing variability across devices, sites, patients and time” (2). The QIBA organization comprises volunteers from academia, industry, and government. In QIBA, there are four modality-based coordinating committees, including CT, MR, nuclear medicine, and US. In this issue of Radiology, Kinahan et al (3) provide impactful guidance for improving the value of FDG PET as an imaging biomarker for assessing response to cancer therapy. The QIBA committee provides substantial data to support the following two claims:

First, maximum SUV (SUV_max) is measurable from FDG PET/CT with a within-subject coefficient of variation of 10%–12%.

Second, a measured increase in SUV_max of 39% or more, or a decrease of 28% or more, indicates that a true change has occurred with 95% confidence.

The first claim addresses the variability of SUVs seen in published test-retest experiments. The second claim is derived from the first and provides information about the expected statistical variability of SUV measurements, assuming nothing has changed. In essence, the second claim asserts that: (a) increases in SUV_max of greater than or equal to 39% or decreases in SUV_max of greater than or equal to 28% are unlikely to be caused by the known variability in SUV measurements; and (b) thus, they reflect true biologic changes in a lesion (progression of disease or response to treatment). This provides an evidence-based metric for variability of FDG SUV measurements on PET scans and a quantitative limit for the accuracy of measured SUVs.

The devil is in the details. There are several requirements that must be met to meet this claim of true change in SUV. First, the same PET/CT scanner must be used for both examinations to look for change in SUV. When different PET/CT scanners are used, the within-subject coefficient of variation may be substantially greater (4), resulting in more variability in SUV_max and thus less confidence in predicting a true change in the lesion. Using two scanners of the same model likely mitigates this problem. Second, the FDG uptake time must be kept within 10 minutes for both PET/CT scans. Differences in uptake time may result in substantial effects on SUV measurements (5). Third, lesions must be greater than 2 cm in diameter. In small lesions, partial volume effects may result in considerable differences between measured FDG avidity and actual FDG avidity (6). Fourth, lesions must have an initial SUV_max greater than or equal to 4 g/mL. Measurements of lesions with low FDG uptake show less reproducibility. These limitations for the QIBA claims may give pause to clinicians.

Although the QIBA claims are evidence based, they are not the result of prospective clinical trials. Along the QIBA profile levels of development (2), this article represents achievement of stage 3, technically confirmed. Experts and manufacturers have found the QIBA claims to be practical and “expect it to achieve the claimed performance.” The next stage, claim confirmed, will require multicenter testing of the claims given the restrictions.

Despite these limitations, there is much to celebrate. Although criteria for interpreting FDG PET to determine tumor response have been available for 20 years (eg, European Organization for Research and Treatment of Cancer [7] and Positron Emission Tomography Response Criteria in Solid Tumors [8]), few clinical trials of solid tumors utilize FDG PET. The standard for solid tumor trials has been Response Evaluation Criteria in Solid Tumors (RECIST), a system for quantifying tumor burden that focuses primarily on tumor size (9). The role of FDG PET is often limited to complementing CT and MRI to confirm disease progression. But RECIST is not without its own limitations. Patients with advanced solid tumors may never develop disease that is measurable by using RECIST, particularly those with bone-predominant disease (osseous lesions are not measurable with RECIST). A recent trial using FDG PET to supplement RECIST allowed recruitment of patients with disease not measurable with RECIST who would have otherwise been denied access to the trial; this action facilitated patient accrual and provided adequate response assessment (10). Thus, there are untapped advantages to using FDG PET in trials of solid tumors. I believe that this statement from the Radiological Society of North America QIBA Committee will help provide rigor to the use of SUVs in FDG PET. This achievement from QIBA should overcome some of the obstacles that FDG PET has faced in terms of its integration into solid tumor trials and could ultimately lead to increased usage of PET FDG in these trials and clinical practice.

In summary, Kinahan et al have succeeded in their goals for this PET/CT statement. The principles articulated in this QIBA statement can be applied not only to PET tracers, but also to diverse QIBA projects in CT, MRI, US, and nuclear medicine. Although still early, QIBA has already begun to be recognized by manufacturers of imaging equipment and the imaging community at large. This will not be the last we hear of QIBA in Radiology.