Whole-Body Biodistribution and Dosimetry of the Dopamine Transporter Radioligand ¹⁸F-FE-PE2I in Human Subjects

Participants

Five healthy volunteers (3 women, 2 men) were recruited for participation in this study via an announcement in the local newspaper. Median age was 69.6 y (range, 63–77 y). The relatively high mean age was purposely selected to match the mean age of most patients with Parkinson disease and atypical parkinsonian syndromes. The participants were in subjectively good health as confirmed by a thorough neurologic and physical examination at inclusion. The study was approved by the regional ethical review board at Umeå University, the local radiation safety committee at the University Hospital of Umeå, and the Swedish Medical Products Agency. All subjects signed an informed consent form before participation. The participants were not given any dietary restrictions.

The scanning procedures were supervised by authorized personnel, and the participants were asked to report any kind of discomfort during or after the scanning. Participants were encouraged to, if required, contact the research nurse after scanning and were followed up with a short interview by the research nurse after 14 d, to confirm the presence or absence of any possible late-onset adverse reactions.

Radiochemistry and Synthesis

¹⁸F-FE-PE2I was produced at the University Hospital of Umeå according to a previously published method (12). Quality control was performed on each batch, including identity, chemical purity, radiochemical purity, and pH. Endotoxin and sterility testing was performed on every 20th batch or at least once every third month. All methods complied with current good manufacturing practices according to the European Pharmacopoeia.

Administered Activity

The administered activity of the ¹⁸F-FE-PE2I was aimed to be 2.86 MBq/kg, with a maximum activity of 200 MBq. The dosage, 2.86 MBq/kg, was based on an administered activity of 200 MBq for a patient weighing 70 kg. The true mean administered activity per body weight of ¹⁸F-FE-PE2I in this study was 2.5 MBq/kg (range, 1.8–3.1 MBq/kg), deduced from measurements of the syringe and catheter before and after injection. The participants’ mean weight was 69.5 kg (range, 56–87 kg).

PET/CT Image Acquisition

The PET/CT scanning was performed at the University Hospital of Umeå on a Discovery 690 (GE Healthcare) (13) in 4 sessions, summarized in Figure 1. The first session, aimed to catch the fast tracer distribution, consisted of 3 head-to-thigh scans commencing at tracer injection, followed by a head-to-toe whole-body scan approximately 30 min after injection. The remaining 3 sessions were all head-to-toe scans, commencing at 2, 4, and 6 h after injection. The CT scans were acquired without x-ray modulation with 120 kV, 10 mA, and 0.5 s/revolution, except for the whole-body scan at the first session, which was acquired with 30 mA (Fig. 1) to provide good anatomic information. CT images were viewed at a 2.5-mm slice thickness. All PET scans were acquired using time-of-flight technique, with a duration of 60 s per bed position and a bed-position overlap of 11 slices. PET images, corrected for attenuation and scatter, were reconstructed using the scanner implementation of the 3-dimensional iterative ordered-subset expectation-maximization reconstruction including time-of-flight technique (VuePoint HD ViP) with 2 iterations, 24 subsets, a 6.4-mm postprocessing filter, and a 70-cm field of view, to a 128 × 128 matrix (5.5 × 5.5 × 3.27 mm voxels).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Illustration of PET/CT scanning protocol. All PET scans were performed with 60 s/bed position. Three head-to-thigh PET scans and one whole-body PET/CT scan, using low-dose CT (30 mA), were performed within the first 40 min, starting immediately after injection. At 2, 4, and 6 h after injection, an additional whole-body PET/CT scan with ultra-low-dose CT (10 mA) was performed at each time point.

Blood and Urine Sampling

All urine was collected up to 6 h after injection. Urine from each voiding was collected in a separate bottle and measured separately. The voided volume was weighed, and the activity concentration of a 1-mL sample was measured using a well counter (NaI(Tl) scintillator) combined with a scale, so that the total voided activity could be calculated.

Venous blood samples were drawn at 40 min, 2 h, 4 h, and 6 h after injection. The activity concentrations of these samples were measured in the above-mentioned well counter. No metabolite analysis of the samples was performed, because such an analysis was beyond the scope of this dosimetry study.

Image Analysis

For image analysis and activity measurement, imlook4d (14), an in-house–made MATLAB (MathWorks)-based noncommercial software program, was used. The advantage of this software is that it easily allows complex volume-of-interest (VOI) delineation on images from different modalities to be merged into the PET space. It is also easily extendable, so that specialized algorithms may be developed using the MATLAB language. Using imlook4d, full-organ VOIs were delineated by one physician, double-licensed in radiology and nuclear medicine (with >15 y of clinical experience in whole-body diagnostic imaging). By combining manual drawing and segmentation techniques, we could delineate the organs for each scan. The PET and CT volumes with the best contrast were used to delineate any given organ for each individual. For instance, Hounsfield-based automatic segmentation of the low-dose CT scan was used to create VOIs for the lungs and the spine, which were superimposed on the PET images. VOIs of some organs that could be visualized on only a few PET images were delineated using a thresholding algorithm and then transferred to the remaining PET series, after which their positions were carefully manually adjusted to anatomic markers. Transferring of VOIs between scans was done only within the same individual, followed by careful tailoring if needed. VOIs were not copied between individuals. Care was taken to avoid superposition of the VOIs with neighboring organs.

Absorbed and Effective Dose Calculation

The absorbed and effective doses were calculated using the biokinetics for each participant combined with standard phantoms (15). The PET images were decay-corrected to the start of the scan. The mean activity concentrations (Bq/mL) in different organs as obtained from the PET images were transformed to total activity content by multiplying by the organ masses, including blood content (g), for males or females from International Commission on Radiological Protection publication 133 (16). For lungs, the activity concentration was multiplied by the lung volume, which was calculated by dividing the mass by the density (0.38 g/mL). The activities were thereafter divided by the injected activity (Bq) for the specific participant to retrieve the percentage of injected activity (%IA). For bone marrow, the mean activity concentration (Bq/mL) in the spine was multiplied by the measured volume of the spine, to obtain the total amount of activity in the spine VOI. Assuming that the activity is present only in the bone marrow, we calculated the activity concentration by dividing the total activity by the literature value for the volume of active bone marrow in the spine (15).

For each participant, a mono- or biexponential function was fitted to the measured data using the computer program SAAM II (17), and the function was integrated to obtain the area under the curve, which represent the cumulated activity—that is, the mean residence time per injected activity. The software IDAC 2.0 (18) was used to calculate the effective dose according to International Commission on Radiological Protection publication 60 (19), as well as the absorbed doses to the different organs.

To obtain the residence time in the urinary bladder, the compartmental part of SAAM II was used. A 2-compartment model was created, with one compartment being the urine and the other the rest of the body. From the rest of the body, the activity could be transferred to the urine as well as out of the system. At time zero, 100% of the activity was inserted into the rest-of-body compartment, and a function was fitted to the data points of the cumulated %IA excreted with the urine. The urine compartment was thereafter adjusted to be emptied at predefined time points, as described below.

For the calculations of absorbed dose, the bladder was assumed to be voided at a 3.5-h interval, an interval used by the International Commission on Radiological Protection (20). To compare with a more realistic assumption, an alternative calculation was also made with the assumption that the bladder was voided 1 h after injection and thereafter at a 2-h interval (i.e., 1 h after injection, 3 h after injection, 5 h after injection, and so on).