PET-Based Radiation Dosimetry in Man of ¹⁸F-Fluorodihydrotestosterone, a New Radiotracer for Imaging Prostate Cancer

Preparation of ¹⁸F-FDHT

¹⁸F-FDHT was prepared as previously described (12), starting with cyclotron-produced aqueous ¹⁸F-fluoride, purified by normal-phase high-pressure liquid chromatography, evaporated to dryness, redissolved in ethanolic normal saline, and sterile-filtered for injection. The total radiochemistry synthesis time was 100 min, the specific activity greater than 22,200 GBq/mmol, the decay-corrected radiochemical yield nearly 30%, and the radiochemical purity, as determined by reverse-phase high-pressure liquid chromatography, greater than 95%.

Patients

A total of 10 ¹⁸F-FDHT PET studies were performed on 7 patients with progressive prostate cancer. In all cases, the diagnosis of prostate cancer had been confirmed histologically and the progression of disease was confirmed by a standard bone scan, an abnormal CT or MRI scan, or rising serum titers of prostate-specific antigen. At the time of their respective studies, all subjects had serum levels of dihydrotestosterone, the active androgen in man, typical for prostate cancer patients after castration. Patients were injected intravenously with a bolus of 111–407 MBq of ¹⁸F-FDHT.

Blood Sampling and Counting

Serial blood samples were obtained 0–60 min after injection via an indwelling venous catheter, weighed, and counted in a scintillation well counter calibrated for ¹⁸F. Although exclusively blood-counting data were used for the dosimetric analysis, aliquots of some blood samples were also centrifuged and plasma decanted, weighed, and counted.

PET

A whole-body PET scan was obtained for each patient beginning at 66–154 min after injection using a dedicated PET scanner, the Advance (General Electric Medical Systems). The Advance is a bismuth germanate-based, multislice whole-body PET scanner with fields of view of 55 and 15 cm in the transaxial and axial directions, respectively. In the current study, all PET scans were obtained in 2-dimensional mode, yielding 35 transverse sections 4.25 mm in thickness over the 15-cm axial field of view. The nominal spatial resolution for ¹⁸F is 4.2 mm in full-width at half-maximum in the transaxial plane (13). The Advance uses ⁶⁸Ge rods for a transmission-based attenuation correction; the whole-body emission scan was obtained first, requiring up to 7 bed positions per patient, and the ⁶⁸Ge transmission scans for the segmented attenuation correction were obtained immediately thereafter. Scatter and randoms corrections were applied using software provided with the scanner. The PET data were reconstructed by the ordered-subset expectation maximization iterative method to minimize the obscuring starburst artifact around intense foci of radioactivity (in this case, the urinary bladder) characteristic of filtered backprojection reconstruction.

Data Analysis

In the following analysis, all activities and activity concentrations were corrected for radioactive decay of ¹⁸F from the time of the measurement/image (i.e., the start of the whole-body emission scan) back to the time of ¹⁸F-FDHT injection. In addition, it was assumed that, as a first approximation, organ masses and compartment sizes (i.e., the extracellular fluid space) are directly proportional to total-body mass.

The maximum blood activity concentration was conservatively used to estimate the red marrow total activity. Assuming ¹⁸F-FDHT instantaneously equilibrated between plasma and the red marrow extracellular fluid spaces, the maximum plasma and bone marrow activity concentrations were estimated from the maximum activity concentration among the serial blood samples using the following equation (14): Eq. 1 where maximum marrow [activity] = the maximum marrow activity concentration (in MBq/g) and maximum blood [activity] = the maximum blood activity concentration (in MBq/g). In Equation 1, the standard man red marrow extracellular fraction was taken as 0.40 (15) and plasmacrit as 0.58 (16). The maximum total marrow activity was then estimated as follows: Eq. 2

In Equation 2, the standard man total-body mass was taken as 70 kg and the standard man red marrow mass as 1,500 g (16).

Region-of-interest analysis was performed on the quantitative PET images, yielding the mean activity concentrations (in kBq/mL) in the liver, spleen, and total body. The total activity in organ i was then estimated using the following equation: Eq. 3 The ¹⁸F residence time τ_i in organ i (i.e., the liver, spleen, and red marrow) was conservatively calculated assuming elimination of activity only by physical decay in situ: Eq. 4 where T_p = the physical half-life of ¹⁸F, 1.8 h.

The ¹⁸F residence time τ_TB in the total body was calculated from the total-body effective half-life of ¹⁸F-FDHT, assuming monoexponential biologic clearance from the time of injection to the time of the PET scan: Eq. 5 Eq. 6 where (T_b)_TB = the biologic half-time of FDHT-derived ¹⁸F in the total body.

The activity excreted up to the time of each PET scan was calculated from the total-body activity using the following equation: Eq. 7

Assuming the activity in the liver at the time of each PET scan would be eliminated via hepatobiliary excretion, the fraction of the excreted activity excreted in the urine was estimated using the following equation: Eq. 8

The ¹⁸F residence times in the intestines (i.e., the small intestine, upper large intestine, and lower large intestine) were calculated on the basis of the assumption that the activity in the liver was eliminated entirely via hepatobiliary excretion, equating the fraction of the administered activity in the liver at the time of each PET scan with the fraction of the administered activity entering the small intestine and using the International Commission on Radiological Protection (ICRP) 30 gastrointestinal tract model (17) as implemented with the MIRDOSE3 software program (18). The ¹⁸F residence time in the urinary bladder contents was calculated using the dynamic bladder model as implemented with the MIRDOSE3 software program (18), with the biologic half-time of FDHT-derived ¹⁸F in the total body and the fraction of the excreted activity excreted in the urine (Eq. 7). Because not all of the activity in the liver may be excreted via the intestines, that is, some of the liver activity may ultimately be eliminated in the urine, the foregoing approach may overestimate the radiation dose to the intestines while underestimating that to the bladder wall.

Mean organ absorbed doses (in cGy/MBq administered) were calculated using S-factor values for the 70-kg standard man anatomic model and the MIRD formalism as implemented with the MIRDOSE3 software program (18). The effective dose equivalent was calculated assuming a radiation weighting factor of unity for ¹⁸F radiation and using the ICRP 60 tissue weighting factors (19,20).

Imaging and Biodistribution Data

A PET image (coronal section) showing the typical whole-body distribution of ¹⁸F-FDHT in a patient with metastatic prostate cancer is presented in Figure 1C. For comparison, a planar γ-camera image of ^99mTc-methylene diphosphonate and a coronal-section PET image of ¹⁸F-FDG are shown in Figures 1A and 1B, respectively.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Comparative whole-body images of a 75-y-old man with progressive prostate cancer metastatic to bones of the thoracic spine, left rib cage, and scapula: planar γ-camera image of ^99mTc-methylene diphosphonate (A); 1-pixel-thick coronal PET image of ¹⁸F-FDG (B); and 1-pixel-thick coronal PET image of ¹⁸F-FDHT (C).

Based on region-of-interest analysis of the PET image data (10 studies in 7 patients), the decay-corrected percentage injected dose (mean ± SD) in the liver, spleen, and total body at 66–154 min after injection was 9.0% ± 4.3%, 0.82% ± 0.75%, and 78% ± 16%, respectively. The activity concentration in plasma rapidly reaches a maximum by ∼2 min after injection and decreases by less than 30% up to 60 min after injection (Fig. 2). The decay-corrected maximum activity concentration in plasma through 60 min after injection was 0.0061% ± 0.0015%/mL.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Typical plasma time-activity concentration curve for ¹⁸F-FDHT through 60 min after injection. The activity concentrations, expressed in units of MBq/mL, have been decay-corrected back to the time of injection.

Residence Times and Radiation Doses

The ¹⁸F residence times in the liver, spleen, red marrow, and total body ranged from 0.15 to 0.46 h, 0.0077 to 0.063 h, 0.11 to 0.24 h, and 1.0 to 2.6 h, respectively. The mean normal-tissue absorbed doses (in cGy/MBq) and the effective dose equivalent (in cSv/MBq), with the respective SDs and percentage SDs, are presented in Table 1 for all 10 ¹⁸F-FDHT studies. Except for the liver (0.00333 cGy/MBq) and the urinary bladder wall (0.00868 cGy/MBq), the mean absorbed dose to each tissue was less than 0.003 cGy/MBq. The effective dose equivalents ranged from 0.00152 to 0.00201 cSv/MBq. These absorbed doses, which provide the basis for the recommended administered activity of ¹⁸F-FDHT (21), are comparable to those of other ¹⁸F-labeled radiopharmaceuticals such as ¹⁸F-FDG (22).