Abstract
86Y (half-life = 14.74 h, 33% β+) is within an emerging class of positron-emitting isotopes with relatively long physical half-lives that enables extended imaging of biologic processes. We report the synthesis and evaluation of 3 low-molecular-weight compounds labeled with 86Y for imaging the prostate-specific membrane antigen (PSMA) using PET. Impetus for the study derives from the need to perform dosimetry estimates for the corresponding 90Y-labeled radiotherapeutics. Methods: Multistep syntheses were used in preparing 86Y-4–6. PSMA inhibition constants were evaluated by competitive binding assay. In vivo characterization using tumor-bearing male mice was performed by PET/CT for 86Y-4–6 and by biodistribution studies of 86Y-4 and 86Y-6 out to 24 h after injection. Quantitative whole-body PET scans were recorded to measure the kinetics for 14 organs in a male baboon using 86Y-6. Results: Compounds 86Y-4–6 were obtained in high radiochemical yield and purity, with specific radioactivities of more than 83.92 GBq/μmol. PET imaging and biodistribution studies using PSMA-positive PC-3 PIP and PSMA-negative PC-3 flu tumor-bearing mice revealed that 86Y-4–6 had high site-specific uptake in PSMA-positive PC-3 PIP tumor starting at 20 min after injection and remained high at 24 h. Compound 86Y-6 demonstrated the highest tumor uptake and retention, with 32.17 ± 7.99 and 15.79 ± 6.44 percentage injected dose per gram (%ID/g) at 5 and 24 h, respectively. Low activity concentrations were associated with blood and normal organs, except for the kidneys, a PSMA-expressing tissue. PET imaging in baboons reveals that all organs have a 2-phase (rapid and slow) clearance, with the highest uptake (8 %ID/g) in the kidneys at 25 min. The individual absolute uptake kinetics were used to calculate radiation doses using the OLINDA/EXM software. The highest mean absorbed dose was received by the renal cortex, with 1.9 mGy per MBq of 86Y-6. Conclusion: Compound 86Y-6 is a promising candidate for quantitative PET imaging of PSMA-expressing tumors. Dosimetry calculations indicate promise for future 90Y or other radiometals that could use a similar chelator/scaffold combination for radiopharmaceutical therapy based on the structure of 6.
The positron-emitting radionuclide 86Y (half-life [t1/2] = 14.74 h, β+ = 33%, energy of the positron [Eβ+] = 664 keV) is an attractive isotope for molecular imaging (1). 86Y can readily be prepared on a small biomedical cyclotron using the 86Sr(p, n)86Y nuclear reaction (2). The extensive use of the high-energy β−-emitter 90Y (t1/2 = 64.06 h, β− = 72%, β particle energy [Eβ−] = 2.288 MeV) for endoradiotherapy (3,4) makes 86Y ideal for dosimetry estimates of 90Y-labeled radiotherapeutics (5). Antibodies and peptides radiolabeled with 86Y have properties identical to those labeled with 90Y, enabling accurate absorbed dose estimates for 90Y for radiotherapeutics (1,6).
The prostate-specific membrane antigen (PSMA) is increasingly recognized as a viable target for imaging and therapy of prostate and other forms of cancer (7–9). We and others have demonstrated PSMA-targeted radionuclide imaging in experimental models of prostate cancer (10–12) and in the clinic (13–15) using functionalized cysteine-glutamate or lysine-glutamate ureas. For the attachment of large molecular fragments, such as radiometal (99mTc, 68Ga, 111In) complexes (16–18) and nanoparticles (19,20), a long linker was placed between the large molecule and the targeting urea to retain PSMA-targeted binding. On the basis of those initial positive results, we and others have reasoned that urea-based agents could also be used for radiotherapy of PSMA-containing lesions using radionuclides. In fact, clinical studies using that approach with 131I-MIP1095 ((S)-2-(3-((S)-1-carboxy-5-(3-(4-131I-iodophenyl)ureido)pentyl)ureido)pentanedioicacid) (15) and 177Lu-labeled PSMA-targeted agents (14) are under way for the treatment of castrate-resistant prostate cancer. Although 177Lu has a shorter β-particle range (t1/2 = 6.7 d, Eβ− = 0.5 MeV) than 90Y, because they have similar chelation chemistry, we proposed 86Y as a suitable imaging surrogate to investigate potential 177Lu-based radiotherapeutics as well as those radiolabeled with 90Y. A similar rationale has been applied to agents for neuroendocrine-targeted peptide receptor radionuclide therapy (21). The aim of this study was to prepare and investigate the biodistribution of three 86Y-labeled PSMA-binding ureas (Fig. 1) in a rodent experimental model and image the most pharmacokinetically favorable agent in nonhuman primates for radiation dosimetry in preparation for clinical trials with the corresponding 90Y- and 177Lu-labeled agents.
86Y-labeled inhibitors of PSMA.
MATERIALS AND METHODS
Detailed chemical and radiochemical syntheses of 86/89Y-4, 86/89Y-5, and 86/89Y-6 (Fig. 1) are provided in the supplemental materials (available online at http://jnm.snmjournals.org). PSMA inhibitory activities were determined using a fluorescence-based assay (17). Enzyme inhibitory constants (Ki values) were generated using the Cheng–Prusoff conversion (22). Sublines of the androgen-independent PC-3 human prostate cancer xenograft were used (17). Those sublines have been modified to express high (PC-3 PIP) or naturally produce low (PC-3 flu) levels of PSMA and were generously provided by Dr. Warren Heston (Cleveland Clinic). Details related to cell culture and animal models are included in the supplemental materials. Six- to 8-wk-old male, nonobese diabetic/severe-combined immunodeficient mice (Charles River Laboratories) were implanted subcutaneously with PSMA-positive (PSMA+) PC-3 PIP and PSMA-negative (PSMA−) PC-3 flu cells (2 × 106 in 100 μL of Matrigel [BD Biosciences]) at the cephalad right and left flanks, respectively. Mice were imaged or used in biodistribution assays when the xenografts reached 5–7 mm in diameter. Details of the biodistribution assay are included in the supplemental materials.
Animal Imaging
Small-Animal PET and CT
Dynamic, whole-body PET and CT images were acquired on an eXplore VISTA small-animal PET system (GE Healthcare) and an X-SPECT small-animal SPECT/CT system (Gamma Medica Ideas), respectively, with details presented in the supplemental materials.
Papio Anubis (Baboon) PET Imaging of 86Y-6
A male Papio anubis (8 y, 27.1 kg) was used to study the biodistribution of 86Y-6. Nine static PET images were acquired at 5, 10, 15, 20, and 35 min as well as at 1, 2, 3.5, and 23 h after intravenous administration of 80.7 MBq (2.2 mCi) of 86Y-6 as a bolus. Images were acquired in 2-dimensional mode on a Discovery RX VCT scanner (GE Healthcare). Details related to imaging and analyses are provided in the supplemental materials.
Radiation Dosimetry
Related equations, explanation, and assumptions for dosimetry calculation can be found in the supplemental materials. Measured activity concentration (in Bq/cm3) values per time point per organ were decay-corrected and divided by the baboon organ mass, determined by the CT density and volume from the drawn contours, and the injected radioactivity to obtain the fraction of initial radioactivity per gram (FIA/g) for each time point and each organ. The baboon FIA/g values were then converted to human FIA (per organ) using the related equation (23,24). The resulting human FIA values were then plotted as a function of time and fit to a biexponential expression, and the value for the time-integrated activity coefficient (previously known as residence time (25)) for each source organ was calculated. Radiation absorbed doses were obtained by converting time-integrated activity to absorbed doses according to the MIRD absorbed-fraction methodology (25) through the use of the OLINDA/EXM software (26).
RESULTS
The chemical structures of the 86Y-labeled PSMA-targeting compounds 86Y-4, 86Y-5, and 86Y-6 are shown in Figure 1. Radiolabeling of the target compounds proceeded in high yield (∼90%–97%) and radiochemical purity (>98%), with a high specific radioactivity (>83.92 GBq/μmol [2.27 Ci/μmol]). All compounds displayed high binding affinity, with Ki values ranging from 0.10 to 4.69 nM (Table 1).
PSMA Inhibitory Activities
Small-Animal PET Imaging
Whole-body PET/CT images were obtained for 86Y-4, 86Y-5, and 86Y-6 (Figs. 2–4). All 3 radiotracers enabled visualization of PSMA+ PC-3 PIP tumor and kidneys (Fig. 2), a known PSMA-expressing organ, at 2 h after injection. Renal uptake of the radiotracers is partially due to the route of excretion of these agents and to specific uptake from the expression of PSMA in mouse proximal renal tubules (27). Agent 86Y-5 demonstrated nonspecific accumulation in the gastrointestinal tract, presumably due to the increased hydrophobicity from the 3 phenylalanine residues on the linker moiety. PET/CT images of 86Y-4 were acquired at 1, 4, and 18 h after injection considering the short biologic half-life of this class of low-molecular-weight compounds. The presence of the radiotracer in PSMA+ PC-3 PIP tumor, kidneys, and urinary bladder was observed up to 4 h (Fig. 3A). Radioactivity in the bladder and kidneys cleared significantly by 18 h, although the PSMA+ PC-3 PIP tumor retained some activity. As a further test of in vivo binding specificity, we performed a blocking study of 86Y-4 by pretreating the animal with the potent, selective PSMA inhibitor ZJ43 (50 mg/kg) (28). Figure 3B demonstrates that ZJ43 was capable of blocking the binding of 86Y-4 not only within tumor but also within the renal cortex, another PSMA-expressing tissue (27). Figure 4 displays PET/CT images of 86Y-6 to 12 h after injection. Significantly, 86Y-6 exhibited faster clearance of radioactivity from normal tissues, and by 12 h after injection radioactivity was largely cleared from the kidneys, producing clear tumor-to-background contrast. Clear delineation of PSMA+ PC-3 PIP tumor was achieved as early as at 15 min after injection. Notably, 86Y-6 does not contain the additional phenylalanine moieties of 86Y-4 and 86Y-5 and uses a p-isothiocyanatobenzyl DOTA chelator, which adds an additional carboxylate to hold the metal strongly and decreases lipophilicity.
Whole-body PET/CT images of 86Y-4, 86Y-5, and 86Y-6 in mice bearing PSMA+ PC3 PIP and PSMA− PC3 flu tumors at 2 h after injection. Mice were injected with approximately 3.3 MBq (90 μCi) of radiotracer intravenously. Images are decay-corrected and scaled to the same maximum value. GB = gallbladder; GI = gastrointestinal tract; K = kidney.
PET/CT images of 86Y-4 in mice bearing PSMA+ PC3 PIP and PSMA− PC3 flu tumors. Images obtained without (A) and with (B) blockade of PSMA using the potent, selective PSMA inhibitor ZJ43 as blocking agent (50 mg/kg). Reduction of radiotracer uptake in both tumor and kidneys (another PSMA+ site) on cotreatment with ZJ43 provided a further check on PSMA-specific binding. Mice were injected with approximately 6.2 MBq (168 μCi) of radiotracer intravenously. Images are decay-corrected and scaled to the same maximum value. B = bladder; K = kidney.
PET/CT images of 86Y-6 in mice bearing PSMA+ PC3 PIP and PSMA− PC3 flu tumors. Mice were injected with approximately 6.2 MBq (160 μCi) of radiotracer intravenously. Images are decay-corrected and scaled to same maximum value. K = kidney.
Biodistribution in Mice
On the basis of the results of imaging, compounds 86Y-4 and 86Y-6 were further assessed in a standard biodistribution assay (17). Tables 2 and 3 show the percentage injected dose per gram (%ID/g) uptake values in selected organs at 1, 2, 5, and 24 h after injection. Both radiotracers showed PSMA-dependent binding in PSMA+ PC-3 PIP tumor xenografts, with 86Y-4 demonstrating high tumor uptake at as early as 1 h after injection (29.3 ± 8.7 %ID/g) with relatively slow clearance to 15.7 ± 1.7 %ID/g at 5 h and to 5.9 ± 0.8 %ID/g at 24 h after injection. PSMA+ PC-3 PIP tumor to PSMA− PC-3 flu tumor uptake ratios ranged from 89 at 1 h to a high of 229 at 24 h. Blood and normal tissues such as the heart, liver, stomach, and pancreas did not show significant uptake (∼1 %ID/g) and decreased below 0.02 %ID/g after 24 h. PSMA+ PC-3 PIP tumor-to-muscle ratios were also high, achieving a maximum value of 1,046 at 24 h. Kidney uptake was found expectedly high and peaked at 244.9 ± 8.8 %ID/g at 1 h and decreased to 1.5 ± 0.7 %ID/g by 24 h.
Biodistribution of 86Y-4 in Mice (%ID/g)
Biodistribution of 86Y-6 in Mice (%ID/g)
Table 3 shows the organ %ID/g uptake values for 86Y-6. Compound 86Y-6 quickly accumulated within the PSMA+ PC-3 PIP tumor within 1 h after injection, with an uptake value of 26.6 ± 1.9 %ID/g. The radiotracer concentration continuously increased within PSMA+ PC-3 PIP tumor to exhibit the highest uptake of 32.2 ± 8.0 %ID/g at 5 h after injection. Tumor uptake remained high until 24 h after injection. Normal organs such as the blood, heart, liver, spleen, stomach, and pancreas exhibited low uptake at 1 h, which decreased to below 0.4 %ID/g by 5 h. Renal uptake for 86Y-6, 86.5 ± 13.6 and 54.0 ± 9.2 %ID/g at 1 and 2 h, respectively, was much lower than for 86Y-4.
Baboon PET Imaging and Pharmacokinetics of 86Y-6
Figure 5 depicts the baboon PET study, for which radiotracer is seen in the liver, salivary glands, kidney, and bladder. For whole kidney, renal cortex, and prostate, contours were drawn on each PET image for quantification. All organs showed 2-phase (rapid and slow) biologic clearance. The kidneys had the highest uptake at about 25 min after injection (8 %ID/g). Sixty-eight percent of the radioactivity seen in the kidneys was cleared with a biologic half-life of about 1 h (0.84 h), and the remaining radioactivity was cleared with a biologic half-life of 16.6 h. Most (66%) of the radioactivity in the renal cortex was cleared with a biologic half-life of 1.1 h, and the remaining radioactivity was cleared with a biologic half-life of about 19 h. Significant uptake and retention were seen in the liver and salivary glands, although milder compared with PET scans of patients imaged with 68Ga-labeled PSMA-targeted agents and 124/131I-MIP-1095 (15). Supplemental Table 1 gives the summary of the biologic clearance kinetics of all organs. The time-integrated activity coefficients used in the dose calculations are listed in Supplemental Table 2.
Three-dimensional time course maximum-intensity reprojection display of 86Y-6 PET in baboon. To enhance visualization, bladder radioactivities were segmented semiautomatically using a thresholding method and subsequently removed. Maximum-intensity reprojection 3-dimensional rendering was used to provide an overview of whole-body radiotracer distribution. Little radiotracer was observed in most normal tissues except for bladder (not shown) and kidney (K). Animal was catheterized for this study. Mild uptake in lacrimal glands, parotids, and salivary glands was noted (short, long, and unfilled arrows, respectively).
Organ-Absorbed Doses
Table 4 provides a detailed list of the organ-absorbed doses, expressed in units of mGy/MBq, for 86Y and 90Y/177Lu. For all isotopes, the renal cortex received the highest absorbed dose per unit activity. Accordingly, it is likely that the renal cortex would be the dose-limiting organ for therapeutic radiometals in the context of patient-specific absorbed dose treatment planning (29,30), followed by the bladder. For the diagnostic isotope 86Y, an effective dose of 0.099 mSv/MBq was also calculated in OLINDA/EXM.
Organ-Absorbed Doses in Reference Adult Male Based on Baboon PET Imaging Data
DISCUSSION
We have synthesized and evaluated three 86Y-labeled, PSMA-targeted agents to undertake nonhuman primate dosimetry. Those compounds contain a DOTA or DOTA mono-amide chelated radiometal attached to the targeting urea similar to others we have published (12,17). We have focused on DOTA and its derivatives because they can be used both for PET (86Y) and for radiopharmaceutical therapy (90Y). It has been documented that pharmacokinetics are dependent on the radiometal chelator used, including those for compounds specifically designed to bind to PSMA. That is primarily attributed to the overall charge of the radioligand and the stability of the metal chelate complexes. Specifically, in our previous report of 68Ga-labeled PSMA-binding DOTA-conjugated agents, 68Ga-4 demonstrated the fastest clearance from normal tissues, including the kidneys (12). However, in the current study we observed that 86Y-4 exhibited unexpectedly higher renal uptake and may not be a suitable candidate for radiotherapy. The evaluation of 86Y-6 demonstrated the desired lower kidney uptake and higher tumor retention required for radiotherapy and was subsequently selected for quantitative PET imaging in a baboon for dosimetry measurements.
The binding specificity study (Fig. 3B) indicated that at 1 h nearly all renal binding of 86Y-4 was specific rather than due to excretion. Evidence suggests that more organized and rapid blood flow in renal parenchyma, compared with tumors, may account for longer tumor rather than renal retention for many of these agents. Although PSMA-binding affinity is 1 factor that likely determines tumor versus renal uptake, other factors, such as lipophilicity, charge, plasma protein binding, and molecular weight, likely also play significant roles. The estimated renal cortex doses of 1.19 mGy/MBq for 90Y and 0.245 mGy/MBq for 177Lu compare favorably with the values of 1.97 mGy/MBq for 90Y and 0.45 mGy/MBq for 177Lu calculated in a report involving peptide receptor radiation therapy (29), for which the renal cortex was the dose-limiting organ. However, several caveats to the absorbed dose calculations must be made. First, the organ uptake measurements from PET are predominantly at early time points (median, 35 min; final time point, 3.5 h), and the time points through numerous organs are mostly short (median, 35 min; eighth time point, 3.5 h), raising questions regarding the accuracy of translating the results to the longer-lived isotopes such as 90Y (t1/2 = 64 h) and particularly 177Lu (t1/2 = 6.72 d). It is a problem faced by all theranostics that the surrogate has a significantly shorter half-life than the therapeutic. Second, the chelation stability of nonidentical therapeutic radionuclides, 177Lu, for example, relative to that of 86Y and the fate of the therapeutic radionuclide if the agent is internalized must also be taken into consideration. Cellular retention and residualization of chelated 90Y after internalization has also been well demonstrated (31).
The commonly used and clinically implemented chelating agent DOTA was used for all 3 radioligands because DOTA, and many DOTA derivatives, is known to form kinetically and thermodynamically stable complexes. The corresponding Y(III) complex has been shown in many cases to be stable in vivo, a desirable trait for a chelator. Significantly, DOTA is also reported to form stable complexes with an array of trivalent metal ions including lanthanides, for example, 177Lu(III), and actinides, for example, 225Ac(III), which are chemically disparate to 86Y(III). Moreover, PSMA-binding urea-based agents are stable under the radiolabeling conditions used for DOTA, so we have not pursued other chelators such as cyclohexyl-diethylenetriaminepentaacetic acid (CHX-A″-DTPA) (32) presently.
Recently, 90Y- or 177Lu-labeled versions of the PSMA-targeted monoclonal antibody J591 demonstrated promising results in phase I and II clinical trials (33–35). In those cases, 111In-labeled antibody was used for dosimetry calculations (36). Although those radiolabeled monoclonal antibodies hold potential for tumor detection and therapy, their modest tumor targeting and a relatively high absorbed dose to red marrow militate against routine clinical use. As an alternative approach, early clinical results using 131I-labeled PSMA-targeted, urea-based small molecules exhibited high dose delivery to malignant foci (15). In those published studies, the salivary glands showed the highest absorbed doses (4.62 mGy/MBq), followed by both liver (1.47 mGy/MBq) and kidneys (1.45 mGy/MBq) (15). It is probable that a significant contributor to the salivary gland absorbed dose is free iodine uptake, as also evidenced by the relatively high (0.91 mGy/MBq) thyroid absorbed dose, which does not occur in the current study. In general, the clearance rates from normal organs are more rapid for 86Y-6 than for the published results (15), with the exception of the kidneys.
CONCLUSION
Biodistribution and dosimetry results suggest that 86Y-6 may provide a suitable imaging surrogate for planning and monitoring PSMA-targeted 90Y- or 177Lu-based radiopharmaceutical therapy.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. We are grateful for the following sources of support: CA148901 (SRB), CA151838 (Peter Searson, MGP), CA134675 (MGP), CA184228 (GS, MGP), and CA116477 (GS). No other potential conflict of interest relevant to this article was reported.
Acknowledgments
We thank the NCI Cyclotron Research team for providing 86YNO3. We also thank James Fox and Gilbert Green for expert technical assistance.
Footnotes
Published online Feb. 26, 2015.
- © 2015 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication September 25, 2014.
- Accepted for publication January 21, 2015.