18F-fluoride PET is a promising noninvasive method for measuring bone metabolism and bone blood flow. The purpose of this
study was to assess the performance of various clinically useful simplified methods by comparing them with full kinetic analysis.
In addition, the validity of deriving bone blood flow from K1 of 18F-fluoride was investigated using 15O-H2O as a reference.
Methods: Twenty-two adults (mean age ± SD, 44.8 ± 25.2 y), including 16 patients scheduled for bone surgery and 6 healthy
volunteers, were studied. All patients underwent dynamic 15O-H2O and 18F-fluoride scans before surgery. Ten of these patients
had serial PET measurements before and at 2 time points after local bone surgery. During all PET scans, arterial blood was
monitored continuously. 18F-fluoride data were analyzed using nonlinear regression (NLR) and several simplified methods (Patlak
and standardized uptake value [SUV]). SUV was evaluated for different time intervals after injection and after normalizing
to body weight, lean body mass, and body surface area, and simplified measurements were compared with NLR results. In addition,
changes in SUV and Patlak-derived fluoride influx rate (Ki) after surgery were compared with corresponding changes in NLR-derived
Ki. Finally, 18F-fluoride K1 was compared with bone blood flow derived from 15O-H2O data, using the standard single-tissue-compartment
model. Results: K1 of 18F-fluoride correlated with measured blood flow, but the correlation coefficient was relatively low
(r = 0.35, P < 0.001). NLR resulted in a mean Ki of 0.0160 ± 0.0122, whereas Patlak analysis, for the interval 10-60 min
after injection, resulted in an almost-identical mean Ki of 0.0161 ± 0.0117. The Patlak-derived Ki, for 10-60 min after injection,
showed a high correlation with the NLR-derived Ki (r = 0.976). The highest correlation between Ki and lean body mass-normalized
SUV was found for the interval 50-60 min (r = 0.958). Finally, changes in SUV correlated significantly with those in Ki (r
= 0.97). Conclusion: The present data support the use of both Patlak and SUV for assessing fluoride kinetics in humans. However,
18F-fluoride PET has only limited accuracy in monitoring bone blood flow.