Subjects
Five healthy control subjects were recruited for the F-DEX imaging protocol. Table 1 presents details of all subjects and the F-DEX administration. Subjects were assessed by neuropsychological testing and physical examination. This study was approved by the Austin Health Human Research Ethics Committee, and written consent was obtained from all subjects before the imaging studies. No adverse events related to the study drug were observed or reported by the subjects following the F-DEX scan.
Tracer synthesis and purification
Production of F-DEX was fully automated using the iPhase Flexlab synthesis module. The radiosynthesis of F-DEX was achieved by reductive amination of (S)nordexetimide with 4-[ 18F]fluorobenzaldehyde. No-carrier-added 4-[ 18F]fluoride was produced from irradiation of [ 18O]water. [ 18F]Fluoride was then eluted to reactor 1 by a mixture of K2CO3 and kryptofix 2.2.2 in a 1:1 solution of acetonitrile/water. The [ 18F]fluoride was dried by azeotropic distillation at 90 ∘C using 1 mL of dry acetonitrile. A solution of 4-formyl- N,N,N-trimethylbenzenaminium trifluoromethanesulfonate (2.5 mg) in dimethylformamide (DMF) (0.5 ml) was added to the dried [18F]fluoride, and reactor 1 was then heated at 120 ∘C for 25 min to produce 4-[18F]Fluorobenzaldehyde. Using 600 μL of DMF containing 12 μL of acetic acid, the reaction mixture was transferred into reactor 2, which was loaded with NaBH3CN (4.5 mg), (S)nordexetimide (3 mg). Reactor 2 was then heated to 120 ∘C for 15 min to form F-DEX by one pot reductive amination between 4-[18F]fluorobenzaldehyde and (S)nordexetimide. The reaction mixture was purified by C18 Sep-pak and eluted with 1 mL of acetonitrile followed by high-performance liquid chromatography (HPLC) purification with a Gemini Phenomenex 250 ×10 mm semi-preparative HPLC column using gradient elution technique with ammonium formate/acetonitrile (0% acetonitrile–45% acetonitrile over 45 min) as mobile phase. F-DEX was collected at 48 min into 80 mL of water and reformulated in 10% ethanol/saline using the solid phase extraction technique. The reformulated F-DEX solution was then passed through a 0.22 μm filter and recovered in a sterile vial. The total synthesis time was 140 min.
PET CT protocol
A Philips Ingenuity TF-128 PET CT system was used to acquire five sets of whole body PET and CT images at post-injection time points of 0, 20, 60, 100 and 190 min with each image acquired from the top of the head to the mid-thigh, and each PET image acquired for 60 s per bed position with a total scan time of 10 min. Additional PET and CT brain acquisitions were performed at 120 min post administration for assessment of image quality. All subjects were instructed to void their bladder before entering the imaging room after which, they were positioned arms down supine. Whole body CT images were initially acquired followed immediately by an intravenous administration of F-DEX at the commencement of PET imaging. Each subject remained on the bed for the initial 30 min allowing for a single whole body CT image applicable to the first two PET images. Subsequent PET images were preceded by a whole body CT image. All whole body CT images were acquired at an X-ray tube voltage of 120 kV and a mean exposure of 22.8 mAs (range, 14–30 mAs) providing adequate image quality for delineating dosimetric source regions and providing attenuation correction μ-maps for PET images. The additional brain acquisition at 120 min post administration consisted of a low-dose head CT at an X-ray tube voltage of 80 kV and exposure of 30 mAs followed by a single bed position 20 min PET acquisition. Four of the five subjects voided their bladder in the 20 to 60-min and 100 to 190-min intervals, with the remaining subject voiding in the 60 to 100-min and 100 to 190-min intervals. PET data was reconstructed with the Philips BLOB-OS-TF reconstruction algorithm, CT attenuation correction, simulated scatter correction, randoms estimates and dead time corrections.
Image analysis for tracer uptake
The subjects PET images were coregistered with their corresponding CT images in the quantitative image analysis environment Pmod for the purposes of defining a volume of interest (VOI) over each source organ. The fourteen source organs considered in the dosimetric analysis are the brain, heart contents, kidneys, large intestine, liver, lungs, muscle tissue, red marrow, small intestine, spleen, stomach, thyroid, urinary bladder and remaining tissue. The left ventricular myocardium was also segmented to assess whether there is retention relative to blood pool in the heart chambers but was included in the definition of heart contents for the dosimetric analysis. The average activity uptake in a source organ was determined from the volume weighted activity concentration of the corresponding VOI for eleven of the fourteen source organs with the exceptions being the red marrow, muscle tissue and remaining tissue. Activity in the red marrow was inferred from the average activity concentration inside the vertebrae of the lumbar spine, assuming a red marrow density of 1 g/cm3 and considering the total mass of red marrow for ICRP 133 standard man and woman. Uptake in the muscle tissue and remaining tissue was determined by defining a VOI over the entire body at each time point, subtracting the activity measured in each of the other VOI to determine a residual activity and assigning the fraction of residual activity to the muscle tissue or remainder tissue according to their mass. The blood clearance rate was estimated by segmenting the aortic arch and determining the normalised activity concentration in the delineated volume. The mean and standard deviation of uptake values in the organs considered were determined from all five subjects at each time point to determine the pharmacokinetic properties of F-DEX.
Activity uptake models
For all VOI other than the bladder, the activity uptake as a function of time was fitted with a least squares regression to either a sum of two exponentials
$$ A(t) = C_{0} e^{-\alpha t} + C_{1} e^{-\beta t} $$
(1)
or a delayed uptake and washout model
$$ A(t) = C_{0}\big(1-e^{-\alpha t}\big)e^{-\lambda_{\text{phys}}t} + C_{1} e^{-\beta t} $$
(2)
where for both models A(t) describes the total activity A in the VOI at time t, \(\lambda _{\text {phys}} = \frac {ln(2)}{t_{1/2}}\) where t1/2 is the half life of 18F and C0, C1, α, β are free parameters determined by the regression. Activity in the bladder was assumed to accumulate as
$$ A(t) = A_{0,\text{full}}\big(1 - e^{-\lambda_{\text{fill}} t}\big)e^{-\lambda_{\text{phys}}t} $$
(3)
where A0,full represents the activity at t=0 of a full bladder and λfill is a constant describing the rate at which the bladder fills. The concentration of tracer in the urine was assumed to be constant. Upon voiding, the model assumed instantaneous and total emptying with instantaneous refilling. The activity inside each source organ was assumed to decay according to the physical half life of 18F after the last time point to ensure the most conservative estimate to the dosimetric calculations.
Dosimetry
Internal dosimetry was performed according to the formulation presented in the Committee on Medical Internal Radiation Dose pamphlet 21 [13] which specifies the time dependent absorbed dose to a target organ as
$$ D(r_{t}, T) = \sum_{r_{s}} \tilde{A}(r_{s},T) S(r_{t} \leftarrow r_{s}) $$
(4)
where the term \(\tilde {A}(r_{s},T)\) is the cumulative activity inside source organ rs after time T and S(rt←rs) is the S factor describing the absorbed dose delivered to the target organ rt per unit of cumulative activity in source organ rs. S factors for 18F were derived for standard man and standard woman from the specific absorbed fraction (SAF) data complementing ICRP publication 133. The photon contributions were calculated by linear interpolation of the ICRP 133 photon SAF data to 511 keV, and the beta contributions were calculated by numerically integrating the ICRP 133 electron SAF data over an integral normalised probability distribution of the 18F beta spectrum.
