Reproducibility of kinetic analysis of [f-18]3' -deoxy-3' -fluorothymidine PET in animal tumor xenograft model

The purpose of this study was to evaluate reproducibility of F-18 FLT dynamic PET data acquired in mouse tumor xenograft models. The A431 cell line and LLC cell line were injected subcutaneously into the shoulders of Balb-c/nu mice (n=10) and C57BL/6 mice (n=11), respectively. Two-hour dynamic microPET scans were acquired after injection of F-18 FLT of 0.2 mCi when the tumors achieved diameters of 6∼8mm. Twenty one mice were divided into 2 groups by cell line ( group 1: A431 and 0.2 mCi, n=10; group 2: LLC and 0.2 mCi, n=11). Intraobserver and interobserver variability were analyzed. A second scan was performed 1d later after reinjection of F-18 FLT to assess test-retest variability of group 1 (n=9) and group 2 (n=9). We measured mean standardized uptake value (SUVmean), percentage injected dose per gram (%ID/g), rate constants (K1, K1/k2, k3, k4), Vdm (Volume of distribution of FLT phosphorylated region), KFLT (Flux constant for FLT from plasma into phosphorylated state in tissue), Ki (FLT net influx constant). The reproducibility of these parameters was evaluated by mean and standard deviation of the difference of the two test (d), mean percent difference (D) and by intraclass correlation coefficient (ICC). In group 1, static parameters showed high reproducibility in three studies. KFLT, Ki, Vdm and k4 of kinetic parameters was reproducible in at least two studies. Especially, Vdm was reproducible in three studies. In group 2, F-18 FLT tumoral uptake was lower than group 1. Static parameters were highly reproducible than any other parameters in intra and interobserver studies. However, they showed high variability in test-retest study. In contrast, K1/k2 was highly reproducible. F-18 FLT dynamic microPET data provides reproducible static and some kinetic parameters of tumor proliferation. Serial studies may be performed to assess therapy related changes in specific steps of F-18 FLT metabolism