Prediction of the Clinical SUV Ratio in Amyloid PET Imaging Using a Biomathematic Modeling Approach Toward the Efficient Development of a Radioligand

Abstract

Our study aimed to develop a method to mathematically predict the kinetic parameters K 1 (influx rate constant), k 2 (efflux rate constant), and BP ND (nondisplaceable binding potential) of amyloid PET tracers and obtain SUV ratios (SUVRs) from predicted timeactivity curves of target and reference regions. Methods: We investigated 10 clinically applied amyloid PET radioligands: 11 CPittsburgh compound B, 11 C-BF-227, 11 C-AZD2184, 11 C-SB-13, 18 F-FACT, 18 F-florbetapir, 18 F-florbetaben, 18 F-flutemetamol, 18 F-FDDNP, and 18 F-AZD4694. For each tracer, time-activity curves of both target and reference regions were generated using a simplified 1-tissue-compartment model, with an arterial plasma input function and the predicted kinetic parameters. K 1 , k 2 , and BP ND were derived from the lipophilicity (logP), apparent volume, free fraction in plasma, free fraction in tissue, dissociation constant, and density of amyloid b using biomathematic modeling. Density was fixed at 3 nM to represent healthy control conditions and 50 nM to represent severe Alzheimer disease (AD). Predicted SUVRs for the healthy and AD groups were then obtained by dividing the integrated time-activity curve of the target region by that of the reference region. To validate the presented method, the predicted K 1 , k 2 , BP ND , and SUVR for the healthy and AD groups were compared with the respective clinically observed values. Results: The correlation between predicted and clinical kinetic parameters had an R 2 value of 0.73 for K 1 in the healthy group, 0.71 for K 1 in the AD group, 0.81 for k 2 in the healthy group, 0.85 for k 2 in the AD group, and 0.63 for BP ND in the AD group. The regression relationship between the predicted SUVR (y) and the clinical SUVR (x) for the healthy and the AD groups was y 5 2.73x -2.11 (R 2 5 0.72). Conclusion:The proposed method showed a good correlation between predicted and clinical SUVR for the 10 clinically applied amyloid tracers.

Biomathematical screening of amyloid radiotracers with clinical usefulness index

“…The proposed screening methodology ( Fig. 1 ) consisted of the previously developed amyloid biomathematical model [8] ( Fig. 1 A), with population and noise simulations, and tracer evaluation based on CUI ( Fig.…”

“…Details of the biomathematical model used in this study are found elsewhere [8] . We briefly summarize the model in predicting SUVR in the following sections.…”

“…Molecular volume and lipophilicity of each radiotracer were represented by McGowan volume ( V x , cm 3 /mol/100) and Moriguchi log P (M log P , unitless), which were generated based on the chemical structure of the radiotracer using commercial software, dproperties (Talete, Italy). In silico free fractions of the radiotracer in tissues ( f ND , unitless) and in plasma ( f p , unitless) were calculated using M log P , from the following relationships [8] : …”

“…Previously, we had developed a biomathematical model to predict the in vivo binding capability of amyloid PET radiotracers in terms of cortical SUVR, under healthy control (HC) and AD conditions, representing Aβ-negative and Aβ-positive diagnosis [8] . In this study, we proposed a screening methodology using amyloid PET radiotracers by extending the previous model to include the noise level of the imaging modality, population variation, and clinical usefulness evaluation [8] . As SUVR was used to measure amyloid load in clinical studies, it was chosen as the outcome parameter of interest [8] .…”

“…In this study, we proposed a screening methodology using amyloid PET radiotracers by extending the previous model to include the noise level of the imaging modality, population variation, and clinical usefulness evaluation [8] . As SUVR was used to measure amyloid load in clinical studies, it was chosen as the outcome parameter of interest [8] .…”