Evaluation of ¹³⁴Ce/¹³⁴La as a PET Imaging Theranostic Pair for ²²⁵Ac α-Radiotherapeutics

Visual Abstract

Abstract

²²⁵Ac-targeted α-radiotherapy is a promising approach to treating malignancies, including prostate cancer. However, α-emitting isotopes are difficult to image because of low administered activities and a low fraction of suitable γ-emissions. The in vivo generator ¹³⁴Ce/¹³⁴La has been proposed as a potential PET imaging surrogate for the therapeutic nuclides ²²⁵Ac and ²²⁷Th. In this report, we detail efficient radiolabeling methods using the ²²⁵Ac-chelators DOTA and MACROPA. These methods were applied to radiolabeling of prostate cancer imaging agents, including PSMA-617 and MACROPA-PEG₄-YS5, for evaluation of their in vivo pharmacokinetic characteristics and comparison to the corresponding ²²⁵Ac analogs. Methods: Radiolabeling was performed by mixing DOTA/MACROPA chelates with ¹³⁴Ce/¹³⁴La in NH₄OAc, pH 8.0, at room temperature, and radiochemical yields were monitored by radio–thin-layer chromatography. In vivo biodistributions of ¹³⁴Ce-DOTA/MACROPA.NH₂ complexes were assayed through dynamic small-animal PET/CT imaging and ex vivo biodistribution studies over 1 h in healthy C57BL/6 mice, compared with free ¹³⁴CeCl₃. In vivo, preclinical imaging of ¹³⁴Ce-PSMA-617 and ¹³⁴Ce-MACROPA-PEG₄-YS5 was performed on 22Rv1 tumor–bearing male nu/nu-mice. Ex vivo biodistribution was performed for ¹³⁴Ce/²²⁵Ac-MACROPA-PEG₄-YS5 conjugates. Results: ¹³⁴Ce-MACROPA.NH₂ demonstrated near-quantitative labeling with 1:1 ligand-to-metal ratios at room temperature, whereas a 10:1 ligand-to-metal ratio and elevated temperatures were required for DOTA. Rapid urinary excretion and low liver and bone uptake were seen for ¹³⁴Ce/²²⁵Ac-DOTA/MACROPA. NH₂ conjugates in comparison to free ¹³⁴CeCl₃ confirmed high in vivo stability. An interesting observation during the radiolabeling of tumor-targeting vectors PSMA-617 and MACROPA-PEG₄-YS5—that the daughter ¹³⁴La was expelled from the chelate after the decay of parent ¹³⁴Ce—was confirmed through radio–thin-layer chromatography and reverse-phase high-performance liquid chromatography. Both conjugates, ¹³⁴Ce-PSMA-617 and ¹³⁴Ce-MACROPA-PEG₄-YS5, displayed tumor uptake in 22Rv1 tumor–bearing mice. The ex vivo biodistribution of ¹³⁴Ce-MACROPA.NH₂, ¹³⁴Ce-DOTA and ¹³⁴Ce-MACROPA-PEG₄-YS5 corroborated well with the respective ²²⁵Ac-conjugates. Conclusion: These results demonstrate the PET imaging potential for ¹³⁴Ce/¹³⁴La-labeled small-molecule and antibody agents. The similar ²²⁵Ac and ¹³⁴Ce/¹³⁴La-chemical and pharmacokinetic characteristics suggest that the ¹³⁴Ce/¹³⁴La pair may act as a PET imaging surrogate for ²²⁵Ac-based radioligand therapies.

Advances in targeted molecular imaging and radionuclide therapy have given rise to the field of targeted theranostics (1). In this paradigm, a molecular agent with a PET or SPECT imaging isotope (e.g., ⁶⁴Cu, ⁸⁹Zr, or ¹²³I) is paired with a cognate radionuclide therapy agent (e.g., ¹⁷⁷Lu, ²²⁵Ac, or ¹³¹I) (2). α-emitting radiotherapies with isotopes, including ²²⁷Th, ²²⁵Ac, ²¹³Bi, ²¹²Pb/²¹²Bi, ²¹¹At, and ¹⁴⁹Tb, have demonstrated promise in human trials (3,4). α-particles have a shorter range in tissue (40–100 μm) and higher linear energy transfer than β-particles (5).

To date, ²²⁵Ac is one of the most promising radionuclides for targeted α-therapy (6). However, an imaging isotope to match with ²²⁵Ac to measure pharmacokinetics and dosimetry has been elusive (7). Actinium has 2 short-lived daughter isotopes, ²²¹Fr and ²¹³Bi, that emit low-energy γ-rays, which are challenging to image with SPECT (8). Thus, ²²⁵Ac therapy is commonly paired with ⁶⁸Ga, ⁸⁹Zr, or ¹¹¹In for imaging-based pharmacokinetic or dosimetry information. However, because of substantial differences in half-life (t_1/2) (⁶⁸Ga) or chelation chemistry (⁸⁹Zr), these are imperfect PET imaging surrogates for ²²⁵Ac. To overcome these limitations, lanthanum-based PET imaging agents such as ¹³²La (t_1/2 = 4.8 h, 42% β⁺) and ¹³³La (t_1/2 = 3.9 h, 7% β⁺) have emerged as potential imaging surrogates for ²²⁵Ac (9,10). Unfortunately, the t_1/2 values of these isotopes are considerably shorter than for ²²⁵Ac, restricting their translation to longer-t_1/2 macromolecule-based PET imaging.

In this context, the Department of Energy isotope program (11) has recently initiated the production of ¹³⁴Ce, an isotope with a 3.2-d t_1/2 that decays by electron capture to ¹³⁴La with the emission of low-energy Auger electrons. The ¹³⁴La is a positron emitter (63% β⁺; endpoint energy, 2.69 MeV) with a t_1/2 of 6.45 min. The unique relationship between the t_1/2 values of ¹³⁴Ce and ¹³⁴La establishes a secular equilibrium (12). In pioneering work, ¹³⁴Ce cation in the +3 oxidation state has been shown to complex with diethylenetriamine pentaacetate (DTPA) (11) and DOTA (13) and to be used for in vivo PET imaging of the chelate as well as the antibody trastuzumab. It was suggested that the similar chemical characteristics between ²²⁵Ac³⁺ and ¹³⁴Ce³⁺ and the longer ¹³⁴Ce t_1/2 (3.2-d) might be advantageous for tracking in vivo pharmacokinetics, especially at later time points. However, DOTA and DTPA require higher molar ratios and elevated temperatures for isotope complexation. Alternatively, MACROPA has demonstrated superior chelate properties for ²²⁵Ac and a high stability (K_LnL = 15.1) for nonradioactive cerium (14), suggesting that it may function well for ¹³⁴Ce/²²⁵Ac theranostic development (15).

²²⁵Ac-based radiopharmaceutical therapy has recently attracted great interest in prostate cancer, particularly ²²⁵Ac-PSMA-617 in small trials, demonstrating great efficacy, especially in the context of resistance to ¹⁷⁷Lu-PSMA-617 (16,17). Our own laboratories have identified the antibody YS5, which targets a tumor-selective epitope, CD46, that is highly expressed in prostate cancer (18). An immuno-PET agent, ⁸⁹Zr-DFO-YS5, has successfully imaged both PSMA-positive and PSMA-negative tumor xenografts and patient-derived PDX models (19). Development of cognate ²²⁵Ac-YS5 radiopharmaceuticals for therapy is currently under way (20–22). These therapeutic approaches would significantly benefit from a companion imaging agent.

Here, we aim to evaluate the potential of positron-emitting ¹³⁴Ce/¹³⁴La as a PET imaging surrogate for ²²⁵Ac. We describe methods for efficient chelation of ¹³⁴Ce using the MACROPA and DOTA chelators and demonstrate the stability of the conjugates. The imaging and distribution characteristics of the ¹³⁴Ce-labeled tumor-targeting agents PSMA-617 and MACROPA-PEG₄-YS5 are evaluated in prostate cancer models. These studies demonstrate the feasibility and applicability of ¹³⁴Ce-based radiopharmaceuticals for cancer imaging.

MATERIALS AND METHODS

Radiolabeling of DOTA, MACROPA.NH₂, and PSMA-617 with ¹³⁴CeCl₃

¹³⁴Ce(NO₃)₃ in 0.1 M HCl was produced at the Isotope Production Facility of Los Alamos National Laboratory as previously described (11). Test batches were supplied by the Department of Energy isotope program for our studies. Radiolabeling reactions of DOTA, MACROPA.NH₂, and PSMA-617 at various ligand-to-metal molar ratios were performed using 2 M NH₄OAc buffer, pH 8.0, except when the product was used for animal injections (0.1 M NH₄OAc, pH 8.0). For radiolabeling, aliquots of ¹³⁴CeCl₃ in 0.1 M HCl (5.17 μL) were mixed with MACROPA.NH₂ (23 μL, 630 μg/mL in 2 M NH₄OAc buffer) or DOTA (20 μL, 375 μg/mL in 2 M NH₄OAc buffer) in 2 M NH₄OAc buffer, pH 8.0 (100 μL) at 25°C for 30 min and PSMA-617 (1.5 μL, 0.8 μg, 500 μg/mL) at 60°C for 1 h. The reaction solution was analyzed by radio–thin-layer chromatography (TLC) using C₁₈ TLC plates (Supelco; Sigma) eluted with 10% NH₄Cl:MeOH (1:1).

Radiolabeling of MACROPA-PEG₄-YS5 with ¹³⁴CeCl₃

MACROPA-PEG₄-YS5 (221.4 μg; 1:1 total metal-to-YS5 molar ratio) was incubated with an aliquot of ¹³⁴CeCl₃ (105 μL, 48.1 MBq) in 2 M NH₄OAc (pH 8.0) at 25°C for 1 h. The radiolabeling progress was monitored by instant thin-layer chromatography (iTLC) on Varian iTLC silica gel strips using 50 mM ethylenediaminetetraacetic acid, pH 5.5, as an eluent. The reaction mixture was purified over PD10 column gel filtration eluting with 0.9% saline solution.

Small-Animal PET Imaging

¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA reactions in 0.1 M NH₄OAc buffer were diluted in saline (1:1 ratio), and 4.81–5.92 MBq in 100 μL were administered via the tail vein to 5- to 6-wk-old wild-type C57BL/6 male mice under isoflurane anesthesia. The specific and molar activities were 19.24 GBq/mg and 20.4 GBq/μmol, respectively, for ¹³⁴Ce-MACROPA.NH₂ and 3.7 GBq/mg and 1.9 GBq/μmol, respectively, for ¹³⁴Ce-DOTA. Dynamic small-animal PET/CT (Inveon; Siemens Medical Solutions) was performed for 1 h simultaneously on 3 mice for both ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA. Free ¹³⁴CeCl₃ (∼4.81–5.92 MBq) in saline (100 μL) was injected similarly to the method described above, to a group of 2 mice for dynamic small-animal PET/CT and a group of 3 mice for static small-animal PET/CT (20-min PET acquisition) at 2 h and 24 h.

For tumor imaging studies, ¹³⁴Ce-PSMA-617 (∼4.3 MBq) in saline (100 μL) was injected via the tail vein into 22Rv1 tumor–bearing mice, and the mice were imaged at 1 h after injection using small-animal PET/CT. For ¹³⁴Ce-MACROPA-PEG₄-YS5 (∼4.44 MBq), the conjugate was injected intravenously into mice implanted with 22Rv1 xenografts and imaged at 4 h and then at 1, 2, 4, and 7 d after injection. Small-animal PET/CT was performed with 20 min of PET at earlier time points (4 h, 1d, and 2 d) and with 60 min of PET at later time points (4 and 7 d). The specific and molar activities were 2.58 GBq/mg and 2.67 GBq/μmol, respectively, for ¹³⁴Ce-PSMA-617 and 0.18 GBq/mg and 26.94 GBq/μmol, respectively, for ¹³⁴Ce-MACROPA-PEG₄-YS5.

RESULTS

Radiolabeling of Bifunctional Chelators DOTA and MACROPA.NH₂

We assessed the radiolabeling efficiencies of MACROPA.NH₂ and compared with DOTA at varying ligand-to-metal (L/M) ratios (Fig. 1 left). The L/M ratios were calculated using the stable cerium plus lanthanum present in the ¹³⁴CeCl₃ solution as per the certificate of analysis (Supplemental Fig. 1; supplemental materials are available at http://jnm.snmjournals.org). As posited, MACROPA.NH₂ complexed all the ¹³⁴Ce in greater than 95% yield from 0.5:1 to 10:1 L/M ratios. In contrast, DOTA complexed 94.2% ± 1.8% of the ¹³⁴Ce only at the 10:1 L/M ratio (Fig. 1; Supplemental Fig. 2). A slight increase in radiolabeling complexation was observed for DOTA using L/M ratios of 2:1 (32.6% vs. 23.3%) and 5:1 (88.2% vs. 72.53%) at an elevated temperature of 60°C (Supplemental Fig. 3). These studies demonstrate that MACROPA.NH₂ exhibited a radiolabeling yield superior to that of DOTA, notably allowing rapid, near-quantitative radiolabeling at a 1:1 L/M ratio at room temperature. The ¹³⁴Ce-MACROPA.NH₂ (1:1 ratio) radiocomplex was analyzed by reverse-phase radio–high-performance liquid chromatography, and the retention time was compared with the ^NatCe-MACROPA.NH₂ complex (Supplemental Figs. 4–8; Supplemental Scheme 1). However, the radio–high-performance liquid chromatogram showed a tailing behavior, likely due to the ejection of ¹³⁴La from the chelate after the decay by its parent, ¹³⁴Ce. The stability of the ¹³⁴Ce-MACROPA.NH₂ complex was evaluated in physiologic buffers and in human and rat serum. Over 7 d, more than 95% of the complex was intact in all buffers and serum (Supplemental Fig. 9).

FIGURE 1.

(Left) Radiolabeling of MACROPA.NH₂ and DOTA with ¹³⁴CeCl₃. (Right) Percentage radiolabeling at increasing L/M ratios for MACROPA.NH₂ and DOTA (n = 2) at 25 °C, as assayed by radio-TLC.

In Vivo Stability of ¹³⁴Ce-MACROPA.NH₂ and DOTA Demonstrated by PET Imaging and Biodistribution Studies

After successful ¹³⁴Ce radiolabeling of MACROPA.NH₂ and DOTA, complex pharmacokinetics and stability were studied in healthy wild-type C57BL/6 mice via PET imaging and biodistribution compared with free ¹³⁴CeCl₃. ¹³⁴CeCl₃ showed a gradual increase in liver uptake, as well as in bladder and kidney uptake (Fig. 2A; Supplemental Fig. 10). In contrast, PET imaging of the ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA complexes demonstrated clearance from most organs, with accumulation in the kidneys and bladder at over 1 h after injection, consistent with renal excretion (Fig. 2A; Supplemental Figs. 11–13). The time–activity curves in Supplemental Figure 14 show the slow blood clearance of ¹³⁴CeCl₃ in comparison with ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA. The 1-h ex vivo biodistribution of ¹³⁴CeCl₃, ¹³⁴Ce-MACROPA.NH₂, and ¹³⁴Ce-DOTA are shown in Figures 2B–2D and Supplemental Table 1. High liver (71.5 ± 4.3 percentage injected dose [%ID]/g) and bone (15.54 ± 2.69 %ID/g) uptake was observed for free ¹³⁴CeCl_3, with similar results found at 2.5 and 24 h after injection (Supplemental Fig. 15). In contrast, ¹³⁴Ce-MACROPA.NH₂ (4.36 ± 2.54 %ID/g) and ¹³⁴Ce-DOTA (5.17 ± 2.33 %ID/g) were equally taken up in the kidney, with low accumulation in the liver and other organs, indicating low nonspecific accumulation and renal clearance. Overall, the PET imaging and biodistribution studies of ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA versus free ¹³⁴Ce demonstrated high complex in vivo stability.

FIGURE 2.

Evaluation of PET imaging of ¹³⁴Ce and chelated complexes in wild-type mouse studies. (A) Coronal small-animal PET/CT images of free ¹³⁴CeCl₃, ¹³⁴Ce-MACROPA.NH₂, and ¹³⁴Ce-DOTA in wild-type mice. (B–D) Ex vivo biodistribution of ¹³⁴CeCl₃ (n = 2) and ²²⁵AcCl₃ (n = 3) (B), ¹³⁴Ce/²²⁵Ac-MACROPA.NH₂ (n = 3) (C), and ¹³⁴Ce/²²⁵Ac-DOTA (n = 3) (D). Error bars represent SD. ***P < 0.0008. ****P < 0.0001.

The ex vivo biodistribution of ²²⁵AcCl₃, ²²⁵Ac-MACROPA.NH₂, and ²²⁵Ac-DOTA (Supplemental Fig. 16; Supplemental Table 2) was assessed and compared with the respective ¹³⁴Ce complexes. Free ²²⁵Ac accumulates primarily in the liver (38.33 ± 6.75 %ID/g) and bone (29.56 ± 2.40 %ID/g), similarly to ¹³⁴Ce (Fig. 2B). ²²⁵Ac-MACROPA.NH₂ (3.54% ± 1.07%) and DOTA (3.07 ± 0.99 %ID/g) complexes displayed a higher uptake in the kidney, with minimal uptake in the liver (0.74 ± 0.19 and 0.28 ± 0.008 %ID/g), similarly to ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA (Figs. 2C and 2D). Notable differences were observed in bone uptake for ²²⁵Ac-DOTA (2.26 ± 0.56 %ID/g) versus ¹³⁴Ce-DOTA (0.45 ± 0.24 %ID/g) and in blood uptake for ¹³⁴Ce-MACROPA.NH₂ (0.86 ± 0.19 %ID/g) and ¹³⁴Ce-DOTA (1.07 ± 0.80 %ID/g) versus ²²⁵Ac-MACROPA.NH₂ (0.32 ± 0.08 %ID/g) and ²²⁵Ac-DOTA (0.23 ± 0.02 %ID/g). Taken together, the data indicate that the biodistributions of the ¹³⁴Ce- and ²²⁵Ac-chelated complexes are largely similar.

Radiolabeling of Prostate Cancer–Targeting Agents PSMA-617 and MACROPA-PEG₄-YS5

Given the encouraging in vivo results in normal mice, we investigated the ¹³⁴Ce radiochemistry of cancer-targeting radiopharmaceuticals, including the small-molecule prostate-specific membrane antigen (PSMA)–targeting agent PSMA-617 (23) and the CD46-targeting antibody derivative MACROPA-PEG₄-YS5. For PSMA-617, higher L/M ratios were required for quantitative ¹³⁴Ce-labeling, as 24.3%, 81.0%, and 100% radiolabeling yields were noted by radio-TLC for 2:1, 5:1, and 10:1 L/M ratios, respectively (Fig. 3A; Supplemental Fig. 17). The radiolabeling yields were comparable to the similar ratios (10:1) of ²²⁵Ac-PSMA-617 based on the prior literature (24). After 1 h of incubation of ¹³⁴Ce with PSMA-617 (Fig. 3), iTLC showed 94.1% radiolabeling yield. Surprisingly, the radiolabeling yields were apparently reduced to about 53.2% when the reaction was diluted in saline. However, when the same TLC plate was allowed to decay and rescanned, quantitative labeling was again observed. Similarly, when the apparently 94.1% pure ¹³⁴Ce-PSMA-617 was analyzed on reverse-phase radio–high-performance liquid chromatography (Supplemental Fig. 18), a significant tailing behavior was observed between 4 and 9 min. These data are consistent with the release of ¹³⁴La due to the dechelation or recoil effect after the decay of the parent, ¹³⁴Ce.

FIGURE 3.

Radiolabeling of prostate cancer–targeting agents PSMA-617 and MACROPA-PEG₄-YS5. (A) Radiolabeling of PSMA-617 (left) and radiolabeling yields at increasing molar ratios of PSMA-617 (right). (B) Radio-iTLC of ¹³⁴Ce-PSMA-617 (left), same reaction mixture diluted in saline scanned without waiting for 1-h decay (middle), and same radio-iTLC scanned after 1-h decay showing quantitative radiochemical yield (right). (C) Radiolabeling of MACROPA-PEG₄-YS5. (D) Radio-iTLC of ¹³⁴Ce-MACROPA-PEG₄-YS5 (left), same reaction mixture after PD10 column purification immediately scanned without waiting for 1-h decay (middle), and same radio-iTLC after 1-h decay (right). (E) ¹³⁴La dechelation due to recoil effect. PBS = phosphate-buffered saline.

On the basis of the favorable model labeling studies, we hypothesized that MACROPA would be a superior chelator to enable ¹³⁴Ce immuno-PET imaging. To facilitate the bioconjugation of MACROPA to the YS5 antibody, we prepared a bifunctional chelator containing MACROPA with a short PEG₄ linker with an activated TFP ester. MACROPA-PEG₄-TFP (7 g) was synthesized over 7 steps in 56.3% overall yield (Supplemental Figs. 19–37; Supplemental Scheme 2) (25). MACROPA-PEG₄-TFP ester (7 g) was conjugated to lysine residues on YS5 (Supplemental Scheme 3), with an average of about 2.6 chelators per antibody as determined by MALDI-TOF MS (Supplemental Fig. 38). Optimized conditions for MACROPA ¹³⁴Ce-labeling were applied, and the radiochemical yield was 96.4% as confirmed by radio-iTLC, with 69.3% isolated yield after purification and a specific activity of 0.18 GBq/mg (Figs. 3C and 3D). In contrast, DOTA-YS5 was unable to complex ¹³⁴Ce even at higher molar ratios (L/M ratio, 2 or 4) and 40°C (Supplemental Fig. 39). Calculation of the ligand-to-metal ratios was based on the number of chelators per antibody YS5. Unexpectedly, the purified eluted fraction of ¹³⁴Ce-MACROPA-PEG₄-YS5 showed an apparent decrease in radiochemical purity to about 53.8% (Fig. 3D). As seen in the case of labeled PSMA-617, when the same TLC plate was scanned after decaying for 1 h, 100% radiochemical yield was observed (Fig. 3D). Size-exclusion chromatography demonstrated no evidence of aggregation, whereas an elevated baseline was noticed between the product peak at 9.65 to 25 min, indicating the possible dechelation of daughter isotope ¹³⁴La (Supplemental Fig. 40). The release of daughter ¹³⁴La was also evident when these reaction mixtures were diluted in saline either for purification or for mouse injections, irrespective of MACROPA or DOTA ligands (Fig. 3E).

In Vitro Analysis and In Vivo Distribution of Prostate-Targeting Agent PSMA-617

The cell-binding assay of ¹³⁴Ce-PSMA-617 was performed with different concentrations using the 22Rv1 cell line. The percentage of cell-bound activity was significantly higher for all the concentrations than for blocking controls. A decrease in cell-bound activity percentage for a higher concentration (0.8 nM) was observed because of the cold mass effect (Supplemental Fig. 41) (26). Small-animal PET/CT was performed on a 22Rv1 tumor–bearing mouse at 1 h after injection. As shown in Figure 4, most of the activity was in the bladder and kidney at 1 h after injection, with low uptake in the tumor, whereas almost all the activity was eliminated from the other organs. This pattern of tumor uptake is similar to that found using other PSMA-targeting agents in 22Rv1 tumors, which express moderate levels of PSMA (27,28).

FIGURE 4.

Small-animal PET imaging of ¹³⁴Ce-PSMA-617 in 22Rv1 xenograft at 1 h after injection.

In Vitro and In Vivo Analysis of ¹³⁴Ce-MACROPA-PEG₄-YS5

The properties of ¹³⁴Ce-MACROPA-PEG₄-YS5 for immuno-PET imaging of prostate cancer were evaluated. A magnetic bead–based radioligand-binding assay revealed a 80.5% ± 4.6% target binding fraction for ¹³⁴Ce-MACROPA-PEG₄-YS5 (Fig. 5A), whereas approximately 16.25% ± 4.4% for blocking and approximately 6.7% ± 2.6% for no CD46 were observed (n = 3). In a saturation binding assay, the dissociation constant of MACROPA-PEG₄-YS5 was 3.7 nM, similar to that previously reported for ⁸⁹Zr-DFO-YS5 (6.7 nM) (Fig. 5B) (19). These data demonstrate that ¹³⁴Ce-MACROPA-PEG₄-YS5 could be synthesized effectively with 1:1 ligand-to-metal ratios, with little or no loss of CD46 binding affinity.

FIGURE 5.

In vitro and in vivo analysis of radioimmunoconjugate ¹³⁴Ce-MACROPA-PEG₄-YS5. (A) Left: Magnetic bead–based radioligand assay for ¹³⁴Ce-MACROPA-PEG₄-YS5 (n = 3). Right: Saturation binding assay of ¹³⁴Ce-MACROPA-PEG₄-YS5 on 22Rv1 cells (dissociation constant, 3.7 nM) (n = 3). (B) Maximum-intensity-projection PET/CT and transverse small-animal PET/CT images obtained up to 7 d after ¹³⁴Ce-MACROPA-PEG₄-YS5 injection in mouse bearing 22Rv1 xenografts, demonstrating gradual increase in tumor uptake over time (n = 4). MIP = maximum-intensity projection.

Encouraged by the promising radiolabeling studies, we evaluated the PET imaging properties of ¹³⁴Ce-MACROPA-PEG₄-YS5 in prostate cancer xenografts. Figure 5C and Supplemental Figure 42 show representative small-animal PET/CT images after intravenous administration of ¹³⁴Ce-MACROPA-PEG₄-YS5 in athymic nude mice bearing 22Rv1 tumors over 7 d. The ex vivo biodistribution confirmed the elevated uptake in the tumor (37.16 ± 8.17 %ID/g) and liver (21.60 ± 1.70 %ID/g). Persistent high tumor uptake (33.11 ± 9.27 %ID/g) was seen 14 d after administration (Fig. 6; Supplemental Table 3).

FIGURE 6.

Ex vivo biodistribution analysis of ¹³⁴Ce/²²⁵Ac-MACROPA-PEG₄-YS5 in mouse bearing 22Rv1 xenografts at 7 d after injection. Higher tumor and liver uptake was obtained. Error bars represent SD (n = 5 at 7 d and 2 at 14 d for ¹³⁴Ce; n = 4 for ²²⁵Ac at 7 d).

²²⁵Ac-MACROPA-PEG₄-YS5 was radiolabeled, and in vivo biodistribution studies were conducted to compare with the ¹³⁴Ce-labeled YS5 (Supplemental Fig. 43). The imaging and ex vivo biodistribution results for ¹³⁴Ce-MACROPA-PEG₄-YS5 were similar to those for ²²⁵Ac-MACROPA-PEG₄-YS5 for tumor and most tissues (Fig. 6; Supplemental Table 3). High ²²⁵Ac-MACROPA-PEG₄-YS5 uptake in the tumor (34.75 ± 9.07 %ID/g) was observed on day 7 after injection, similar to the ¹³⁴Ce-MACROPA-PEG₄-YS5 uptake (37.16 ± 8.17 %ID/g). However, significant differences in liver (P < 0.0001) and spleen (P = 0.0109) uptake were observed.

DISCUSSION

In the design of theranostic agents, it is essential to match the structure and biodistribution of the imaging molecule to that of the radiotherapeutic. Recently, lanthanides have been proposed as nonradioactive surrogates for actinium because of similar chemical properties. ¹³²La (t_1/2 = 4.8 h) and ¹³³La (t_1/2 = 3.9 h) have been studied as complementary PET imaging isotopes for targeted α-therapy with ²²⁵Ac (t_1/2 = 9.9 d) (9,10). Aluicio-Sarduy et al. reported cyclotron-produced ¹³²La-labeled alkyl phosphocholine (NM600) in a 4T1 tumor and showed in vivo uptake characteristics similar to those of ²²⁵Ac (9). Similarly, Nelson et al. described a high-yield cyclotron method to produce ¹³³La using natural barium and isotopically enriched ¹³⁵BaCO₃ targets (10). Potential limitations of ¹³²La and ¹³³La include shorter t_1/2 values than for ²²⁵Ac (t_1/2 = 9.92 d) and elevated temperatures (80°C–90°C) required for higher radiochemical conversions (>95%). Although these may be more suitable for fast-clearing small molecules, antibody fragments, or small peptides, their t_1/2 values limit the ability to monitor the pharmacokinetics of macromolecules such as antibodies.

¹³⁴Ce has emerged as an isotope that may be complexed by the same chelates as actinium and thorium. Its decay to ¹³⁴La provides an in situ generator of a positron-emitting isotope with the apparent t_1/2 of its parent. The pioneering study by Bailey et al. highlighted the cyclotron production of ¹³⁴Ce/¹³⁴La from a natural lanthanum target and established the radiochemistry with ligands DTPA (as a potential surrogate for ²²⁵Ac) and hydroxypyridinone (as a potential surrogate for ²²⁷Th) (11). Later, the same group demonstrated the in vivo distribution of ¹³⁴Ce-DOTA-trastuzumab, an internalizing antibody (13). In the present study, imaging and biodistribution of a small-molecule conjugate, PSMA-617, and the antibody YS5 conjugated with MACROPA (MACROPA-PEG₄-YS5) were conducted on prostate cancer xenografts. Similar tumor uptake was observed between the ¹³⁴Ce- and ²²⁵Ac-labeled MACROPA-PEG₄-YS5. The ¹³⁴Ce/¹³⁴La pair allows lengthy in vivo monitoring of molecules because of its extended t_1/2 of 3.2 d, which is not possible with ^132/133La radioisotopes.

Broadly speaking, the radiolabeling findings and stability using MACROPA and DOTA chelators with ¹³⁴Ce recapitulate the prior reports using the same chelators with ²²⁵Ac (15). Radiolabeling efficiency of greater than 95% was achieved with 1:1 ligand-to-metal ratios for MACROPA.NH₂ and 10:1 for DOTA at room temperature. Dynamic PET imaging and ex vivo biodistribution studies of both ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA confirm in vivo stability and a biodistribution similar to that of ²²⁵Ac-MACROPA.NH₂ and DOTA complexes. Overall, the radiolabeling methodologies show that MACROPA.NH₂ was more efficient than DOTA and that both complexes showed excellent overall stability.

After radiolabeling and purification into saline of the tumor-targeting agents PSMA-617 and MACROPA-PEG₄-YS5 for mouse administration, we chromatographically observed the release of the daughter radionuclide ¹³⁴La from the chelate. In the reaction mixture, before dilution or purification, the ¹³⁴La may be rechelated after recoil effect if excess ligand is present (Fig. 3E). However, the rechelation may not occur in vivo even if the excess ligand is present, leading to possible ¹³⁴La redistribution. Though the stability constants were high for ^NatLa-MACROPA (14.91) and ^NatCe-MACROPA (15.11) (14), the ¹³⁴Ce bond dissociation occurs because of the nuclear recoil effect through electron capture decay and subsequent Auger electron emission (29). A similar phenomenon was seen by Severin et al. for another in vivo PET generator, ¹⁴⁰Nd (t_1/2 = 3.4 d, Electron capture (EC)/¹⁴⁰Pr (t_1/2 = 3.4 m, β⁺), with DOTA-LM3 (small peptide) and DTPA-ATN 291 (antibody). In their work, small differences in tissue distribution were noted via pre- and postmortem imaging—differences that were attributed to redistribution of the daughter. The differences were greater for noninternalizing agents (30,31). Our imaging findings are also consistent with these prior reports.

The imaging properties of ¹³⁴Ce/¹³⁴La have been evaluated in prostate cancer models using PSMA-617 and MACROPA-PEG₄-YS5. Low to moderate tumor uptake of ¹³⁴Ce-PSMA-617 was observed at 1 h after administration. High kidney uptake of PSMA-based targeting vectors is known, as they tend to excrete through renal elimination and the mouse kidneys express PSMA (27,28). In contrast, ¹³⁴Ce-MACROPA-PEG₄-YS5 showed elevated tumor uptake. Our findings are consistent with our prior report demonstrating elevated uptake of ⁸⁹Zr-DFO-YS5, compared against ⁶⁸Ga-PSMA-11 in the 22Rv1 xenograft model (19).

Remarkably, biodistribution studies of ¹³⁴Ce-MACROPA-PEG₄-YS5 showed tissue distribution almost identical to that of ²²⁵Ac-MACROPA-PEG₄-YS5 except for the liver and spleen. The high liver uptake observed in early images at 24 h (Fig. 5B) may be due to redistribution of daughter ¹³⁴La after ejection from the chelate. This possibility will be further investigated in future studies by conducting pre- and postmortem imaging and comparing it with ²²⁵Ac more systematically.

One notable advantage to using ¹³⁴Ce is that it allows facile imaging of conjugates bearing the MACROPA chelate, which was previously limited to therapeutic radionuclides. The similar chemical properties of these radionuclides (¹³⁴Ce/²²⁵Ac) may allow a single molecular platform by complexing with the ligands DOTA or MACROPA. This complexation could facilitate predicting the tumor distribution of ²²⁵Ac-labeled targeting vectors (²²⁵Ac-PSMA-617 or MACROPA-PEG₄-YS5) based on the (¹³⁴Ce-PSMA-617 or MACROPA-PEG₄-YS5) PET imaging results. Hence, this methodology addresses an important challenge in radiopharmaceutical sciences, namely the study of the biodistribution of ²²⁵Ac radiopharmaceuticals. Overall, these studies support our premise that ¹³⁴Ce/¹³⁴La may serve as an imaging radionuclide to pair with ²²⁵Ac.

CONCLUSION

MACROPA.NH₂ showed exceptional radiolabeling efficiency with ¹³⁴Ce at room temperature. PET imaging of ¹³⁴Ce-MACROPA.NH₂ and ¹³⁴Ce-DOTA revealed that both tracers are highly stable in vivo. The ex vivo biodistributions of both ¹³⁴Ce-DOTA and MACROPA.NH₂ were almost identical to the respective ²²⁵Ac complexes. ¹³⁴Ce-PSMA-617 shows high binding affinity and uptake in prostate cancer 22Rv1 xenografts. A bifunctional analog for MACROPA was synthesized, conjugated with antibody YS5, and radiolabeled with ¹³⁴Ce and ²²⁵Ac. Both the PET imaging and the biodistribution of ¹³⁴Ce-MACROPA-PEG₄-YS5 demonstrate elevated tumor retention in 22Rv1 prostate cancer xenografts. The ex vivo biodistribution is consistent with the ²²⁵Ac-MACROPA-PEG₄-YS5 distribution in most tissues, including the tumor. These studies support the future development of ¹³⁴Ce-radiopharmaceuticals for cancer imaging as a companion to paired α-particle radiotherapeutics.

DISCLOSURE

Kondapa Naidu Bobba and Robert Flavell have filed a patent application, “Radioimmunoconjugates and Therapeutic Uses Thereof” provisional patent application number 63/344537. This study was supported by U.S. Department of Energy, Office of Science, Office of Isotope R&D and Production, DOE Isotope program under Award Number DE-SC-0023467 and Department of Defense grant W81XWH2110792. No other potential conflict of interest relevant to this article was reported.

KEY POINTS

QUESTION: Are the radiochemistry and in vitro/in vivo characteristics of ¹³⁴Ce/¹³⁴La chelates similar to those of ²²⁵Ac?

PERTINENT FINDINGS: ¹³⁴Ce/¹³⁴La efficiently forms stable complexes with ²²⁵Ac-chelates, DOTA, and MACROPA. These may allow a single molecular platform for imaging and radiotherapy. The ex vivo tissue biodistribution was largely similar between ²²⁵Ac- and ¹³⁴Ce-labeled antibody YS5, with the exception of liver and spleen.

IMPLICATIONS FOR PATIENT CARE: Identification of an imaging surrogate for ²²⁵Ac may aid in the development of targeted α-radiotherapeutics and enable visualization of their distribution. Imaging with ¹³⁴Ce-labeled radiopharmaceuticals may guide therapeutic dosing of the concomitant ²²⁵Ac-labeled molecule.