Dosimetry of the Dopamine Transporter Radioligand ¹⁸F-FPCIT in Human Subjects

Subject Population

Thirteen dosimetry evaluations were performed on 12 human subjects. One subject had both bladder and thorax or abdomen scans performed. Nine men and 3 women (age range, 23–75 y; mean ± 1 SD, 59.2 ± 14.5 y) were studied. With the exception of 1 young person, the age range matches the patient population normally studied for PD.

Urinary Bladder Scans

Six subjects had dynamic PET scans of the bladder. There was some variation among subjects with regard to how data were acquired but a typical scan session went as follows. The subject was injected intravenously with approximately 185 MBq ¹⁸F-FPCIT. The subject was positioned in the Advance PET scanner (General Electric Medical Systems) and a 10-min transmission scan of the lower abdomen was obtained using a rotating pin ⁶⁸Ge source. The transmission scan was used for attenuation correction of the subsequent emission data. Emission scanning commenced 1 h after injection without moving the subject. Twelve 10-min dynamic scans of the bladder were obtained over the next 2 h. At 3 h after injection, the subject was removed from the scanner and sent to void. The urine was collected for assay of ¹⁸F activity. The void volume was measured and recorded. Approximately 0.3 mL of urine was aliquoted and counted in a Canberra Series 20 well counter. Urine counts were background subtracted and corrected for well counter efficiency for ¹⁸F (≈25%). The counts were then scaled to the void volume and the fractional injected activity in the void was determined. The patient was repositioned in the scanner and 6 additional 10-min scans were obtained.

The total study duration was approximately 4 h 15 min. The PET scans were reconstructed and analyzed by making a composite image set of all transaxial sections that contained the bladder. A region of interest (ROI) was drawn around the bladder and a time–activity curve was generated. The counts in the ROI were scaled to fractional injected activity using the measured data from the urine void (11). Time–activity curves were fit to a model of exponential uptake and elimination by voiding and physical decay. Figure 1 shows an example of a bladder time–activity curve from 1 subject. Residence times were calculated from the numeric integration of the curve. Only physical decay was considered beyond 6 h because no significant additional activity was assumed to enter the bladder.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Bladder time–activity curve for subject 7. Data points reflect actual radioactivity measured by PET scanner scaled by measured urine concentration at approximately 180 min. Data are then further normalized by injected radioactivity. Solid line is fit to experimental data using exponential uptake and elimination by voiding and physical decay.

Thorax or Abdomen Scans

Five subjects had dynamic PET scans of the thorax or abdomen. There was some variation among subjects with regard to how data were acquired, but a typical scan session went as follows. The subject was placed in the scanner such that the lungs, heart, liver, spleen, and kidneys could be visualized. This required operating the PET in scanning mode using 2 fields of view (FOVs). The axial extent of each 2 FOV scans was approximately 30 cm. The subject was injected with approximately 185 MBq ¹⁸F-FPCIT and 14 dynamic scans were acquired over 2 h for both FOVs. The first 9 scans were acquired for 3 min in each FOV and the last 5 scans were acquired for 5 min in each FOV. Each FOV was acquired sequentially but there was approximately a 1-min gap between each dynamic scan. Emission data were corrected for attenuation using a 2-FOV transmission scan acquired without moving the subject. Emission data were reconstructed and composite transaxial images of each organ visualized were generated.

ROIs were drawn over parts of each organ. The ROIs were manually drawn by 2 of the investigators in such a way as to limit contributions from adjacent structures. For example, lung regions were drawn strictly on the right lung to avoid contribution from heart activity. The Advance scanner is calibrated to yield activity concentration directly from the scans. These values were converted to fractional organ uptake using Equation 1: Eq. 1 where UPT is fractional uptake, C_PET is activity concentration (in kBq/mL), ρ is organ tissue density (in g/cm³), M is organ mass (in g), and A is the activity injected into the patient (in kBq). Organ densities and masses are taken from the report of the task group on Reference Man (13). Time–activity curves were generated for the fractional uptake values for each organ for the 2-h period after injection. The tails of the curves were fit to a model of exponential elimination and the half-lives were calculated. Residence times were computed by numeric integration of the time–activity curves and the integrals were extrapolated to infinity using the calculated half-lives. Figure 2 shows examples of lung and liver time–activity curves from 1 subject.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

(A) Lung time–activity curve for subject 7. (B) Liver time–activity curve for subject 7.

Brain Scans

Two subjects had dynamic PET scans of the brain. There was slight variation between the 2 subjects with regard to how data were acquired, but a typical scan session went as follows. The subject was placed in the scanner such that the brain above the orbitomeatal line could be visualized. The subject was injected with approximately 185 MBq ¹⁸F-FPCIT and 33 dynamic scans were acquired over 4 h. Emission data were corrected for attenuation using a transmission scan acquired without moving the subject.

Emission data were reconstructed and summed to create a composite transaxial image. An ROI was drawn over the entire brain. Average tissue concentration values were converted to fractional organ uptake using Equation 1. Time–activity curves were generated for the fractional uptake values for the 4-h period after injection. The curves were fit to a model of exponential uptake, and elimination and residence times were computed by integration of the time–activity curves (14).

Dosimetry Calculations

The average residence time was computed for each organ from among the subjects studied for that organ. These residence times were input into the MIRDOSE3 program to compute organ doses. The ICRP 30 gastrointestinal (GI) tract model was used to estimate residence times for the intestines with input to the small intestine from liver clearance (15).