Initial investigation of single-scan FDG+FLT PET tumor imaging techniques

Abstract

Rapid multi-tracer PET aims to image two or more tracers in a single scan, thereby characterizing multiple aspects of function. Using dynamic imaging with tracers staggered in time, constraints on the kinetic behavior of each tracer can be applied to predict the makeup of the multi-tracer PET signal. Signal-separation algorithms can then be applied to recover estimates of each individual tracer. The effectiveness of the approach depends critically upon the tracers used, their kinetic behavior, the injection timing, and dynamic scanning protocol. A key question is what imaging endpoints (e.g. static images, SUVs, kinetic macroparameters or individual rate parameters) can be reliably recovered for each tracer. The ability to rapidly and reliably image both <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F-fluorodeoxyglucose (FDG) and <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F-fluorothymidine (FLT) is of great interest, as characterization of tumor metabolism and proliferative activity provides complementary and relevant information for oncologic treatment decisions. However, it also provides one of the most challenging signal-separation problems for single-scan multi-tracer PET—since both tracers are labeled with relatively long-lived <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F, differences in radioactive decay cannot be utilized to aid in the multi-tracer signal-separation process. This work presents initial investigations into rapid dual-tracer FDG+FLT tumor imaging, characterizing the feasibility and main technological limitations of the approach. Simulation studies show that longer injection delays provided better signal-separation of the single-tracer measures, and markedly better performance was observed when FLT was administered first. Injection delays of approx. 30min led to good recovery (R>0.93) of SUVs, K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">net</inf> , and K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</inf> ; however, recovery of higher-order rate parameters (k <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> , k <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> ) was poor (R=0.5-0.6 at 30min delay), indicating that information regarding those parameters was effectively lost by the rapid dual-tracer technique.