⁶⁴Cu-TETA-Octreotide as a PET Imaging Agent for Patients with Neuroendocrine Tumors

Preparation of ⁶⁴Cu-TETA-OC

⁶⁴Cu-TETA-OC was prepared in compliance with guidelines established by the Washington University Radioactive Drug Research Committee. ⁶⁴Cu was produced on a biomedical cyclotron (5). The specific activity of ⁶⁴Cu ranged from 222,000 to 1,850,000 GBq/mmol. TETA-OC was prepared as previously described (9). ⁶⁴Cu was labeled to TETA-OC as previously described (8). Briefly, ⁶⁴Cu-acetate (370 MBq [10 mCi]) was labeled to 10 μg TETA-OC in 0.1 mol/L ammonium acetate buffer, pH 5.5. Gentisic acid (1 mg) was added to the radiolabeled conjugates to reduce the effects of radiolysis. The preparations were purified by a C₁₈ SepPak (Waters, Milford, MA) to remove uncomplexed ⁶⁴Cu-acetate. ⁶⁴Cu-TETA-OC was sterile filtered with a 0.22-μ Millex GV (Millipore, Bedford, MA) before injection. Radiochemical purity was assessed by radio–thin-layer chromatography (Bioscan, Inc., Washington, DC), using reversed-phase C₁₈ plates developed in 70:30 methanol:5% ammonium acetate, and by reversed-phase high-performance liquid chromatography (Waters). Patients received an injection of 107–130 MBq (2.9–3.5 mCi) (3–5 μg) ⁶⁴Cu-TETA-OC.

¹¹¹In-DTPA-OC was prepared using the OctreoScan kit provided by the manufacturer (Mallinckrodt, Inc., St. Louis, MO). Patients received an injection of 204–233 MBq (5.5–6.3 mCi) ¹¹¹In-DTPA-OC labeled to 10 μg DTPA-OC.

Patients

Patients who were scheduled for routine imaging with ¹¹¹In-DTPA-OC were eligible for ⁶⁴Cu-TETA-OC and PET. Eight patients (six men, two women; age range, 45–70 y) with a history of neuroendocrine tumors who were being evaluated with ¹¹¹In-DTPA-OC as part of their routine clinical evaluation also underwent PET with ⁶⁴Cu-TETA-OC at our institution. This investigation was approved by the Human Studies Committee and the Radioactive Drug Research Committee at Washington University School of Medicine. Each patient gave informed consent before participating in the study. In all patients but one, the ¹¹¹In-DTPA-OC was administered before ⁶⁴Cu-TETA-OC. The time between administration of ¹¹¹In-DTPA-OC and ⁶⁴Cu-TETA-OC ranged from −1 to 40 d (mean time, 14 d).

Imaging

PET imaging was performed with an ECAT EXACT scanner (Siemens/CTI, Knoxville, TN). Single-photon imaging was also performed on each patient using a dual-head Genesys scanner (ADAC Laboratories, Milpitas, CA) fitted with medium-energy collimators. Conventional scintigraphy was performed after injection of 204–233 MBq (5.5–6.3 mCi) ¹¹¹In-DTPA-OC. All patients underwent whole-body planar scintigraphy at 4 and 24 h. The whole-body planar study consisted of anterior and posterior views proceeding from head to toe at 6 cm/min. All but one patient underwent SPECT (imaging of the chest and abdomen at 24 h in one patient and at 4 and 24 h in one patient; imaging of only the abdomen at 24 h in two patients, at 4 h in one patient, and at 4 and 24 h in two patients). Full-field SPECT was performed without attenuation correction on the chest and abdomen for 20 min each. On a separate day, PET was performed after intravenous injection of approximately 111 MBq (3.0 mCi) ⁶⁴Cu-TETA-OC in an antecubital vein. Delay intervals ranging from 0 to 24 h between injection and imaging were evaluated, from which an optimal delay of 4–5 h was selected as providing the best combination of image quality and lesion-to-background contrast. Five of the eight PET studies were performed to estimate the radiation dosimetry of ⁶⁴Cu-TETA-OC. In these patients, the entire torso was scanned to allow complete determination of the activity biodistribution throughout the body. In the other three patients, only a single bed position was used, covering 16 cm axially around the suspected tumor region. At each bed position, the study consisted of static (if delayed) or dynamic (if at injection) data acquisition, 20 min in duration, followed by a 2- to 3-min transmission scan. From the transmission scan, segmented attenuation maps were derived (10) and used to apply full attenuation correction to the emission image volume. PET and SPECT images were reconstructed using filtered backprojection (Hanning filter, cutoff of 0.8 Nyquist) and were volume rendered for cinegraphic viewing.

The ¹¹¹In-DTPA-OC and ⁶⁴Cu-TETA-OC images were examined qualitatively. The criteria for interpretation were those routinely used in scintigraphic imaging. The PET images were examined by a nuclear medicine physician, and the interpretation of the images was based on knowledge of normal biodistribution of ⁶⁴Cu-TETA-OC derived from animal models. Foci of abnormal radiotracer uptake were recorded on a four-point scale (0 = no uptake; + = mild uptake; ++ = moderate uptake; and +++ = intense uptake).

Measurement of ⁶⁴Cu Activity in Blood and Urine of Patients

⁶⁴Cu activity in the blood and urine was measured in a gamma counter (Beckman, Arlington Heights, IL) containing a NaI crystal. Two blood samples were drawn per patient at times ranging from 1.5 to 22 h after injection of ⁶⁴Cu-TETA-OC. Urine was collected from patients (from two to six samples per patient) at times ranging from 1.5 to 27 h after injection.

Human Absorbed Dose Measurements from Patient PET Images

For evaluation of biodistribution and calculation of dosimetry, five of the eight PET studies involved imaging the chest and abdomen at different times (0 and 4 h in two patients, 5 and 24 h in two patients, and 5 h in one patient) after administration of ⁶⁴Cu-TETA-OC. All but one of the patients were imaged twice.

Activity was observed mainly in the liver, spleen, kidneys, bladder, and tumors. Regions of interest (ROIs) were drawn on the PET images of the five dosimetry study patients to measure the total activity in these organs. For the liver, the average activity concentration was calculated from ROIs traced inside and encompassing most of the organ for five to six adjacent slices. The average activity concentration in the liver was then multiplied by the estimated liver volume. For the other visible organs, ROIs were drawn around the entire organ. The total activity in these organs was thus directly measured. In addition, ROIs were traced around the heart to evaluate the activity concentration in the blood. The blood activity concentration was measured by the average maximum value in the slices containing the lower left ventricle. Finally, elliptic ROIs were drawn in several adjacent slices of the lower abdomen beside the bladder to measure the average soft-tissue activity concentration. The red marrow activity concentration, A_rm, was calculated from the blood activity concentration, A_bl, in proportion to the mass of the red marrow, m_rm, and assuming a hematocrit value of 39%, as follows (11): A_rm = A_blm_rm0.19/(1 − 0.39).

Measurements were repeated for each time point. The percentage of injected activity in each organ immediately after the injection (at t_o⁺) was calculated from its estimated blood volume. This decision was based on the assumption that the bolus of activity distributes uniformly throughout the body immediately after the injection. The fractions of injected activity at t_o⁺ were 4.9% for the liver, 5.9% for the lungs, 1.5% for the spleen, 1.4% for the kidneys, 3.7% for the bones, and 10% for the heart.

A time–activity curve was drawn for each organ combining the data from all patients. The measured organ activities were corrected for decay to the time of the start of the scan to simplify the subsequent fitting procedure. The time–activity curves were fitted with a sum of one or two exponentials in which the parameters were determined by a least squares minimization procedure. The fitted functions were then integrated analytically (including physical decay) to give the residence times. Dose calculation was performed by the MIRD methodology in which the S values were calculated with the MIRDOSE3.0 program using the adult male model. Along with any unaccounted activity, the soft-tissue activity and 72% of the blood activity were assigned to the remainder of the body as a MIRD source organ. Although some activity was seen to accumulate in the tumors, this contribution could not be included under the MIRD scheme. Instead, the tumor activity was assumed to be distributed uniformly throughout the body. An uptake function was fitted to the time-cumulated urine data from all patients. The absorbed fraction and the filling half-time were used in the MIRDOSE3 bladder model (12) along with a voiding period of 4 h to obtain the residence time of the bladder contents.

Human Absorbed Dose Estimates from Baboon PET Images

The biodistribution of ⁶⁴Cu-TETA-OC was also determined in a 25-kg male baboon by PET imaging. This study was performed in compliance with the Guidelines for the Care and Use of Research Animals established by the Animal Studies Committee of Washington University School of Medicine. Images of the animal’s torso were acquired at approximately 30-min intervals from 0 to 3 h and then again at 24 h after injection. The baboon was anesthetized with isoflurane during the first 3-h imaging session and then was allowed to recover. The baboon was placed in approximately the same position the following day, anesthetized with isoflurane, and imaged for 30 min. Activity concentration values were derived from the PET images as previously described (13). The absorbed fraction and filling half-time were used in the MIRDOSE3 bladder model (12) along with a voiding period of 4 h to obtain the residence time of the bladder contents.

Blood samples were collected from the baboon at 6, 20, 40, 60, and 120 min after injection of ⁶⁴Cu-TETA-OC. The percentage injection dose (%ID) in the blood at those times was determined, and the data were used to determine blood clearance.

Radiochemistry

⁶⁴Cu-TETA-OC was prepared in greater than 95% radiochemical purity as shown by radio–thin-layer chromatography and radio–high-performance liquid chromatography.

Absorbed Dose Estimates from Baboon Imaging

PET imaging of a nonhuman primate was performed to estimate the human absorbed doses before beginning human PET imaging studies. The estimated human absorbed doses from the baboon PET images, in comparison with estimated human absorbed doses from rat biodistribution data (9), are listed in Table 1. The baboon PET data suggest that the dose-limiting organ is the bladder wall (0.17 mGy/MBq [0.62 rad/mCi]), followed by the kidneys (0.13 mGy/MBq [0.49 rad/mCi]). The largest discrepancy between dosimetry calculated from rats to baboons is in the intestinal absorbed doses. Rodents show excretion patterns significantly different from those of primates, and this fact is indicated by the data.

Patient Imaging Studies

The clinical and pathologic features of the eight patients are outlined in Table 2. All patients had histologically proven neuroendocrine tumors. Five patients had proven carcinoid tumors of the gastrointestinal tract, and three had pancreatic islet cell tumors.

Abnormal foci of increased ⁶⁴Cu-TETA-OC and ¹¹¹In-DTPA-OC accumulation were seen in seven of the eight patients. These foci of increased activity were detected in several locations: small or large bowel (three patients), stomach (one patient), liver (four patients), pancreas (two patients), skin (one patient), abdominal and pelvic nodes (four patients), lung (one patient), and bone (two patients). The results of ⁶⁴Cu-TETA-OC PET were, in general, similar to those of ¹¹¹In-DTPA-OC and conventional scintigraphy (Figs. 1 and 2). One patient (patient 6, Fig. 2) showed osseous involvement in addition to the small-bowel lesions on ⁶⁴Cu-TETA-OC PET. In two patients, more than one lesion was observed in the small bowel and bone with both agents (patients 2 and 6), and ⁶⁴Cu-TETA-OC PET detected more lesions than did ¹¹¹In-DTPA-OC imaging (Figs. 1 and 2). Although image contrast was better on ⁶⁴Cu-TETA-OC PET, in some cases the image quality was better with ¹¹¹In-DTPA-OC because of absence of intense activity in the kidneys and bladder. ⁶⁴Cu-TETA-OC PET was falsely negative in a patient with a lung lesion, which showed uptake on the ¹¹¹In-DTPA-OC scan. Disease seen with ⁶⁴Cu-TETA-OC PET and ¹¹¹In-DTPA-OC conventional scintigraphy was confirmed by a combination of histologic, radiologic, and clinical follow-up of the patients.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Anterior and posterior volume-rendered maximum pixel activity reprojected images of ⁶⁴Cu-TETA-OC PET (A) vs. anterior and posterior planar images of ¹¹¹In-DTPA-OC (B) throughout neck, chest, and abdomen of patient 2, 45-y-old man with numerous metastatic foci of carcinoid tumor. Images were obtained 4 h after injection.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Anterior and posterior volume-rendered maximum pixel activity reprojected neck, chest, and abdomen images of ⁶⁴Cu-TETA-OC PET (A) vs. anterior and posterior whole-body planar images of ¹¹¹In-DTPA-OC (B) of patient 6, both obtained at 4 h after injection. Arrows in anterior image of (A) show abdominal carcinoid lesions, whereas posterior view of (A) shows osseous lesions. Arrows in anterior and posterior views of (B) indicate abdominal carcinoid tumors.

Mild to moderate blood-pool activity was present on images obtained immediately after tracer administration, with only faint blood-pool activity seen thereafter. Adrenal glands were visualized in three of eight patients (patients 1, 2, and 7) with ⁶⁴Cu-TETA-OC PET. SSRs have been reported to be present in normal human adrenal glands as well as in malignant pheochromocytomas (14). No activity was seen in the gallbladder. Bowel activity was not present on early images but appeared faintly after 2 h and became more intense (mild to moderate) at 21–24 h after injection. Hepatic uptake was less than splenic uptake on the early images and became more intense with time, with the liver exhibiting activity similar to that in the spleen on delayed images. The renal collecting systems and bladder had high uptake at early times but became less intense on delayed images. Lesions were more clearly visualized on the early images than on the delayed images. The best imaging time appeared to be approximately 4–6 h after tracer injection.

Pharmacokinetic Analysis

Blood samples were obtained from patients during their imaging sessions. ⁶⁴Cu-TETA-OC cleared rapidly from the blood. The amount in the blood varied from patient to patient; however, by 4 h after injection, 7.9 ± 3.7 %ID remained (range, 3.2–13.5 %ID). From 6 to 22 h after injection, the activity in the blood decreased further, with the amount in the blood ranging from 0.8 to 6.6 %ID (mean, 3.3 ± 2.3 %ID). The large differences in blood-pool data are most likely related to patient-to-patient variation, possibly because of differing tumor burden. Urine excretion also varied between patients. During the first 4 h after injection, the amount excreted in the urine was 44.6 ± 10.6 %ID, and the amount during 27 h was 63.7 ± 18.6 %ID.

Human Dosimetry from Patient PET

Using PET data from five patients, the absorbed doses of ⁶⁴Cu-TETA-OC to the normal organs were measured. Time–activity curves are shown in Figure 3. Residence times were determined for seven tissues in which most of the ⁶⁴Cu activity was localized, and they are listed in Table 3. From these data, the absorbed doses listed in Table 1 were determined. Residence times measured in the organs listed in Table 3 account for 35% of the maximum residence time, whereas 47% of activity was excreted in the urine. As predicted from the estimated absorbed doses from the rodent and baboon PET data, the human PET data indicate that the bladder wall is the dose-limiting organ, with an absorbed dose of 0.25 mGy/MBq (0.94 rad/mCi). This dose was calculated assuming emptying of the bladder at 4 h after injection. Other organs with higher absorbed doses include the liver (0.092 mGy/MBq [0.34 rad/mCi]), kidneys (0.078 mGy/MBq [0.29 rad/mCi]), and spleen (0.070 mGy/MBq [0.26 rad/mCi]). The intestinal dose was relatively low, as was the dose to the bone marrow.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Cumulated patient time–activity curves. Unless noted, all values were determined from PET imaging data. (A) Urine data fitted with uptake function (percentage injected activity [%IA] = A_f (1 − exp(−F_ht × t)), A_f = 63.6, F_ht = 0.35). These parameters were used in MIRDOSE3.0 bladder model with voiding period of 4 h to obtain average residence time of bladder contents. Urine data presented here were collected from all eight patients, and individual patient data are not indicated. (B–F) Data for blood clearance (B), kidney uptake (C), liver uptake (D), spleen uptake (E), and soft-tissue uptake (F). ▴ = values determined from PET data; • = patient 1; ○ = patient 2; ▪ = patient 6; □ = patient 7; and ▵ = patient 8.