Consideration of Optimal Time Window for Pittsburgh Compound B PET Summed Uptake Measurements

standardized uptake value ratio (SUVR, or summed tissue ratio) has been used effectively in Pittsburgh compound B (PiB) PET studies to distinguish subjects who have significant amyloid-beta deposition in their brain from those who do not. Relative to quantitative measurements, advantages of the SUVR are improved study feasibility and low test-retest variation; disadvantages include inherent bias (PiB retention overestimation) and potential for time-varying outcomes. The PiB SUVR has proven to be highly correlated with quantitative outcomes and to allow reliable detection of significant group differences (or effective contrasts). In this work, regional PiB SUVRs were examined across 9 time windows to select the window that provided the best trade-offs between bias, correlation, and effective contrast. Methods: A total of 40 dynamic PiB PET studies were performed on controls (n = 16), patients with Alzheimer disease (AD; n = 11), and patients with mild cognitive impairment (MCI; n = 13) (555 MBq [15 mCi], 90-min scan, and arterial blood sampling). The SUVR was computed for five 20-min and four 30-min windows that spanned the 30- to 90-min postinjection period. The SUVRs were compared with Logan graphical distribution volume ratio (DVR) measurements (35-90 min), determined with arterial blood as input and without arterial blood as input (cerebellum as reference). Results: Greater correlation and more bias were generally observed for the SUVR measurement at later times than at earlier times (relative to DVR). The effective contrast between the control and AD PiB SUVRs was slightly better for earlier data than for later data. The temporal dynamics of the SUVR measurement indicated greater stability in the measurement at 40 min after injection. Conclusion: The 50- to 70-min time window provided a good compromise between physiologic validity, stability, sensitivity, and clinical feasibility across the control, MCI, and AD subject data examined in this study. The 40- to 60-min period A demonstrated many advantages and should be used in studies limited by low injected dose. Although more biased than the 40- to 60-min SUVR, the 50- to 70-min SUVR was thought to be optimal because of greater measurement stability, which may prove to be important for longitudinal multisite studies perA formed in control, MCI, and AD subjects that are not dose-limited.