Enhanced Antitumor Efficacy of ²²⁵Ac-NM600 Compared to ¹⁷⁷Lu-NM600 in Murine Prostate Cancer Models

Abstract

Purpose:

Several individuals with prostate cancer (PCa) develop metastatic castration-resistant PCa (mCRPC) after current treatment, which has a death rate of more than 50%. Although many approaches target mCRPC and show promising results, mCRPC is still incurable. Therefore, we aimed to investigate the efficacy and dosimetry of α (²²⁵Ac) versus β (¹⁷⁷Lu) radiopharmaceutical therapy using NM600 in murine PCa models.

Methods and Materials:

NM600 was radiolabeled with ¹⁷⁷Lu and ²²⁵Ac for targeted radionuclide therapy and therapeutic efficacy. ¹⁷⁷Lu-NM600 single-photon emission computed tomography (CT)/CT imaging was conducted on syngeneic Myc-CaP and TRAMP-C1 PCa mouse models, and they were administered 7.4 MBq of ¹⁷⁷Lu-NM600 in the tail vein. We calculated the dosimetry of ¹⁷⁷Lu-NM600 therapy using the single-photon emission CT/CT imaging data and the biodistribution study. The complete blood count, comprehensive metabolic panel, and histology analysis were carried out to assess toxicity in Myc-CaP and TRAMP-C1-beering mice (n = 9), which were given 5.55 (low injected activity [IA]) or 18.5 MBq (high IA) of 177Lu-NM600, and 7.4 (low IA) or 18.5 KBq (high IA) of ²²⁵Ac-NM600. Finally, the overall survival and tumor growth rate were monitored for all groups.

Results:

Both ²²⁵Ac/¹⁷⁷Lu-NM600 demonstrated tumor-specific uptake and retention. ²²⁵Ac-NM600 exhibited superior anti-tumor effects and significantly improved overall survival compared to ¹⁷⁷Lu-NM600 at similar doses. The enhanced efficacy of ²²⁵Ac-NM600 was attributed to its higher relative biological effectiveness. Toxicity studies revealed transient, dose-dependent hematological changes for both agents, with no significant long-term adverse effects.

Conclusions:

²²⁵Ac-NM600 demonstrated enhanced antitumor efficacy compared to ¹⁷⁷Lu-NM600 in murine PCa models, with a favorable toxicity profile. These outcomes reveal a strong rationale for further developing α-emitting radiopharmaceutical therapy agents for PCa treatment.

Introduction

Prostate cancer (PCa) is a significant worldwide health burden, accounting for over 35, 250 deaths, and 299,010 new cases in the United States in 2024.¹ While androgen deprivation therapy is the cornerstone of PCa management, most individuals ultimately develop metastatic castration-resistant PCa (mCRPC), which has a median life expectancy of less than 3 years and a mortality rate of more than 50%.² Current approaches to treat mCRPC have shown promising outcomes, particularly via approved therapies such as radiopharmaceutical therapy (RPT) targeting prostate-specific membrane antigen (PSMA), and PARP blockades, but mCRPC remains incurable.^3–5 RPT is a powerful approach for treating metastatic disease by selectively delivering cytotoxic radiation to tumor cells.⁶ Recently, the FDA approved ¹⁷⁷Lu-PSMA-617 for mCRPC treatment, indicating the potential of RPT.⁷ However, ¹⁷⁷Lu-PSMA-617 only extended survival by 4 months in the best-responding patients, and 10%–20% of patients show low or negative PSMA expression, limiting its efficacy.^8,9 Additionally, concerns remain about hematologic toxicity in individuals with diffuse bone marrow involvement.⁷ Though optimizing established PSMA-targeted agents is valuable, novel treatment combinations and agents targeting alternative biomarkers may benefit specific patients and have complementary utility. Given the dosimetric advantages of α emitters in terms of dose rate, particle range, and relative radiobiological effectiveness compared to β emitters, there is an increased demand for targeted alpha therapy (TAT) for PCa. Those include several PSMA ligands radiolabeled with ²²⁵Ac, ²¹²Pb, and ²¹³Bi.¹⁰ However, few systematic dosimetry-based studies have evaluated the radiobiological effects of RPT with β- and α emitters in PCa.^11,12 Our group has developed NM600, an alkylphosphocholine analog (2-(trimethylammonio)ethyl(18-(4-(2-(4,7,10-tris(carboxymethyl)-1,4,7,10 tetraazacyclododecan1-yl)acetamido)phenyl)octadecyl) phosphate), which shows elevated tumor uptake and retention across multiple cancer types, including PCa.^13,14 The radiolabeled NM600 with β emitters like ⁹⁰Y and ¹⁷⁷Lu has demonstrated a significant antitumor efficacy in preclinical models. However, the comparative efficacy and biological effects of α versus β-emitting NM600 in PCa have not been thoroughly investigated.^15,16 Therefore, the goal of this work is to compare the biodistribution, dosimetry, toxicity profiles, and therapeutic efficacy of ²²⁵Ac-NM600 (α emitter) and ¹⁷⁷Lu-NM600 (β emitter) in syngeneic murine models of PCa. We hypothesized that the higher linear energy transfer and relative biological activity of ²²⁵Ac would translate to superior antitumor effects compared with ¹⁷⁷Lu when conjugated to NM600. Our findings provide important insights into the relative merits of α versus β RPT for PCa and inform future clinical translation efforts.

Methods and Materials

Radiochemistry

We performed radiolabeling of NM600 with ¹⁷⁷Lu or ²²⁵Ac as follows: (10 μg per mCi for ¹⁷⁷Lu and 100 μg per mCi for ²²⁵Ac) in 0.1 M NaOAc buffer (pH 5.5) for 30 minute at 90 °C. Next, the reverse-phase Waters Oasis HLB Light (Milford, MA) method was used to purify the labeled compounds, eluted by 100% ethanol, dried with a stream of N₂, and reconstituted in 0.9% NaCl with 0.4% v/v Tween 20. To determine the yield and purity of the radiolabeled compound, we employed instant thin-layer chromatography (iTLC) using silica-impregnated paper (Perkin Elmer) and 50 mM EDTA. ²²⁵Ac-NM600 Labeling stability in 100% human serum (under agitation at 37 °C) was carried out through iTLC. At predetermined time points, samples were analyzed using iTLC, with the plates allowed to reach secular equilibrium and then exposed simultaneously to ensure consistency between measurements. The cyclone phosphor imaging system enabled accurate quantification of radiolabel retention near the origin for the chelated compound versus migration to the solvent front for any free radioisotope, thereby providing robust data on both initial labeling efficiency and the extent of radiolabel dissociation over time.

Animal models

TRAMP-C1 (ATCC, CRL-2730), a transgenic C57BL/6 prostate adenocarcinoma cell line,² and Myc-CaP (ATCC, CRL-3255), a murine FVB epithelial-like PCa cell line,³ were both grown in Dulbecco’s Modified Eagle’s Medium, and augmented with 4500 mg/L glucose, 4 mM L-glutamine, 1500 mg/L sodium bicarbonate, and 1 mM sodium pyruvate. Finally, the cells were kept at 5% CO₂ and 37 °C.

We carried out all animal experiments per the instructions of the Institutional Animal Care and Use Committee at the University of Wisconsin. The 6-week-old male C57BL/6 and 6-week-old male FVB/NJ mice were ordered from Jackson Laboratories. After acclimation for 1 week, we inoculated TRAMP-C1 cells subcutaneously into the right flank of C57BL/6 mice at 1 × 106 cells/100 μL of sterile 1× PBS mixed 1:1 with Matrigel. Myc-CaP cells were inoculated s.q. into FVB/NJ mice at 1 × 106 cells/100 μL. Tumors were allowed to grow to a size of ~200 mm³before animal use in imaging or therapy studies. We assessed the tumor volumes of TRAMP-C1 and Myc-CaP 3 times weekly via digital caliper. The tumor volumes were estimated by the formula for ellipsoid volume $\left(V=L\times W^2\right)$ with L and W being the length and width of the tumor, respectively.

Single-photon emission computed tomography/computed tomography imaging and ex vivo biodistribution studies

We administered 7.4 MBq of ¹⁷⁷Lu-NM600 in the lateral tail vein to the animals with subcutaneous TRAMP-C1 (C57BL/6) or Myc-CaP (FVB/NJ) tumors. The mice (n = 4) were then positioned prone into a MILabs U-SPECT6/CTUhr system (Houten, The Netherlands) with 2% isoflurane to perform imaging at 5, 24, 48, and 72 hours after injection. We performed computed tomography (CT) scans to be fused with the single-photon emission computed tomography (SPECT) scan, and also for anatomic reference and attenuation correction.

A similarity-regulated ordered-subset expectation maximization algorithm was used in the image reconstruction. To assess the injected activity (IA) percent per gram (%IA/g) for each tissue, we analyzed the images by determining volumes of interest over the organs of interest and tumors.

Serial ex vivo biodistribution studies (n = 3) in TRAMP-C1 and Myc-CaP tumor-bearing mice were carried out at 4, 24, 48, 72, and 192 hours after IV administration of 3.7 MBq of ¹⁷⁷Lu-NM600 or 7.4 kBq of ²²⁵Ac-NM600. Samples were wet-weighed, allowed to reach secular equilibrium, and counted on a gamma counter.

Dosimetry estimations

We calculated SPECT/CT-based ¹⁷⁷Lu-NM600 dosimetry per the previously described method.^4,5 Using the extrapolation of %IA/g at a specific time, we estimated each organ’s total IA. We relied on the standard mouse model to convert organ-specific cumulative activity into absorbed dose per IA (Gy/MBq). In the calculations, we also included the dose contributions from surrounding organs.

Toxicity evaluations

We conducted a complete blood count (CBC), comprehensive metabolic panel (CMP), and histology investigations in control and therapy mice to determine the toxicity of ¹⁷⁷Lu-NM600 and ²²⁵Ac-NM600. We gave groups of Myc-CaP bearings (FVB/NJ) and TRAMP-C1 bearings (C57BL/6) (n = 9), either 5.55 (low IA) or 18.5 MBq (high IA) of ¹⁷⁷Lu-NM600, 7.4 (low IA) or 18.5 KBq (high IA) of ²²⁵Ac-NM600, or vehicle IV. Then, we euthanized 3 mice from every cohort on days 7, 14, and 28 postinjection (p.i.), and 500 μL of blood was obtained by axillary bleed.

The CBC analysis was performed using whole blood and an Abaxis VetScan HM5 hematology analyzer. To separate the serum from RBCs, we centrifuged the remaining blood for 10 minutes at 4000 rpm. The serum was then run via an Abaxis VetScan VS2 analyzer. For baseline comparison, we conducted the toxicity tests in control mice. Organs with marked radiosensitivity or elevated ¹⁷⁷Lu/²²⁵Ac-NM600 uptake, including the liver, spleen, kidneys, femur, and small intestine, were collected and fixed for H&E staining.

Therapeutic studies

Mice with Myc-CaP (FVB/NJ) or TRAMP-C1 (C57BL/6) subcutaneous tumors were treated when tumors reached an approximate volume of 150 mm³. Mice were randomized into groups of n = 10 and injected with 5.55 (low IA) or 18.5 MBq (high IA) of ¹⁷⁷Lu-NM600, 7.4 (low IA) or 18.5 (high IA) kBq of ²²⁵Ac-NM600, or vehicle IV. Tumor volume, body weight, and survival were monitored 3 times weekly for 50 days. Humane endpoints were used based on the UW-Madison Institutional Animal Care and Use Committee guidelines, including tumor volume reaching 700% of the initial volume, significant weight loss (20% of initial body weight), and significant decline in general health. A schematic representation of therapeutic study groups can be found in Table E1. To ensure accurate interpretation of tumor growth and survival data, it is important to note that any apparent discrepancies between tumor volume plots and survival curves result from the application of humane endpoint criteria. Animals were censored from tumor volume analysis on reaching predefined endpoints, including excessive tumor burden, significant weight loss, or marked health decline, in accordance with institutional animal care guidelines. Consequently, average tumor volume plots represent only the cohort of mice still enrolled at each time point, whereas survival curves capture the complete time-to-event data for all animals, regardless of when they were removed from the study. To further enhance data transparency, Figure E1 provides individual tumor volume trajectories, allowing detailed assessment of tumor progression even after animals are censored from group averages.

Statistical analysis

The GraphPad Prism (version 8.4.2) was employed for data analysis, which presented as mean ± standard deviation. The survival curves were displayed using the Kaplan-Meier test and tested for significance via the Log-rank method. Then, we used 2-way analysis of variance or an unpaired t-test to conduct comparisons between groups, where P < .01 is considered statistically significant.

Results

¹⁷⁷Lu/²²⁵Ac-labeled NM600 shows elevated and sustained tumor accretion in murine PCa models

Radiolabeling yield and serum stability of radiolabeled NM600 can be seen in Figure E2. Radiolabeling yield was above 90% for both isotopes, and in serum, the percentage of bound radioactivity remained high throughout the study period, with more than 88% retention observed even at the final timepoint (168 hours). Longitudinal SPECT/CT studies determined the tumor targeting (Fig. 1A) and biodistribution of ¹⁷⁷Lu-NM600 and enabled ¹⁷⁷Lu-NM600 radiation dosimetry estimations in mice bearing TRAMP-C1 or Myc-CaP allografts. We observed initially elevated blood pool activity of the radiotracer 5 hours p.i. (Fig. 1B), which gradually cleared with circulation half-lives of 29.6 ± 8.7 hours and 17.1 ± 4.1 hours for Myc-CaP and TRAMP-C1 mice, respectively (Fig. E3).

Fig. 1.

Longitudinal SPECT/CT demonstrates ¹⁷⁷Lu-NM600 targeting in murine PCa. (A) Maximum-intensity projections of TRAMP-C1 and Myc-CaP tumor-bearing mice at 24, 48, and 72 h after ¹⁷⁷Lu-NM600 (n = 3) administration. (B) Time-activity curves with quantitative region-of-interest analysis of SPECT imaging. Data are presented as %IA/g (mean ± SD). Abbreviations: CT = computed tomography; PCa = prostate cancer; %IA/g = injected activity percent per gram; SPECT = single-photon emission CT.

Due to its primarily hepatobiliary excretion, NM600’s liver uptake peaked at 24 hour p.i. and declined over time. Accumulation in off-target tissues such as kidneys, spleen, bone, and muscle remained under 3% IA per gram of tissue (%IA/g) at all measured time points. SPECT/CT also demonstrated elevated and persistent tumor accumulation with peak values of 7.9 ± 1.3 and 5.0 ± 1.3%IA/g in TRAMP-C1 and Myc-CaP tumors, respectively.

To better describe and contrast the pharmacokinetics and longitudinal biodistribution and clearance of ¹⁷⁷Lu/²²⁵Ac-NM600, we performed serial biodistribution studies out to 192 hour p.i. Uptake was similar between ¹⁷⁷Lu-NM600 and ²²⁵Ac-NM600 (Fig. E4), agreeing with SPECT/CT (Fig. 1). Tumor uptake was highest at 24 hour with 10.0 ± 2.5 and 8.3 ± 2.7 %IA/g for ¹⁷⁷Lu-NM600 and at 72 hour with 7.4 ± 3.3 and 4.2 ± 0.9 %ID/g for ²²⁵Ac-NM600 in TRAMP-C1 and Myc-CaP tumors, respectively. Notably, ²²⁵Ac-NM600 liver uptake was markedly higher than ¹⁷⁷Lu-NM600. This discrepancy in biodistribution could stem from differences in the injected mass of the radioligands or a relatively reduced in vivo stability of ²²⁵Ac-DOTA analogs,¹⁷ and should be investigated further. In fact, we have also performed radiolabeling stability of ²²⁵Ac-NM600 in human serum (Fig. E2) and found over 90% stability for up to 120 hours. At the 168-hour time point, stability decreased to 88.4 ± 1.0% bound activity, which might explain discrepancies in liver uptake.

Dosimetry corroborates selective radiation delivery to tumors

We estimated ¹⁷⁷Lu-NM600 dosimetry based on longitudinal SPECT/CT by a Monte Carlo voxel-based dosimetry approach,¹⁸ and data were presented as Gy/MBq of IA. Apart from the liver, which received the highest off-target dose among all organs, tumors got the highest dose of all other tissues in both models, with 1.1 and 0.8 Gy/MBq to TRAMP-C1 and Myc-CaP tumors, respectively (Table E2). Table E3 summarizes ²²⁵Ac-NM600 tumor and normal tissue absorbed dose estimates (Gy/kBq) for either Myc-CaP or TRAMP-C1-tumors bearing mice, including doses received by each IA. Estimated absorbed dose rates for Myc-CaP and TRAMP-C1 tumors were 0.34 and 0.66 Gy/kBq, respectively. Similarly, the liver received the largest dose of all normal tissues with 1.44 and 1.69 Gy/kBq in Myc-CaP and TRAMP-C1-bearing mice, respectively. The remaining normal tissues, such as the spleen and kidney, had markedly lower absorbed dose values under 1.0 Gy/kBq.

²²⁵Ac-NM600 affords a robust antitumor effect and prolongs survival compared with ¹⁷⁷Lu in murine PCa

¹⁷⁷Lu-NM600 treatment had minimal effect on survival and tumor growth in Myc-CaP-bearing mice (Fig. 2A, C). In contrast, in TRAMP-C1-bearing mice, it resulted in only modest tumor growth inhibition following high-dose (18.5 MBq) treatment (Fig. 2B), but no significant survival advantage compared with controls (Fig. 2D). In contrast, significant (P < .01) tumor growth inhibition was observed at either ²²⁵Ac-NM600 administered activity (7.4 or 18.5 kBq) in both Myc-CaP and TRAMP-C1 tumors (Fig. 3A, B).

Fig. 2.

¹⁷⁷Lu-NM600 therapeutic efficacy in syngeneic prostate cancer. Tumor growth curves of mice (n = 10) bearing Myc-CaP (A) or (B) allografts receiving vehicle (control) or ¹⁷⁷Lu-NM600 (5.55 or 18.5 MBq) intravenously. Compared with controls, no significant tumor growth inhibition was observed at either the administered dose of ¹⁷⁷Lu-NM600 in Myc-CaP tumors or TRAMP-C1 tumors. Overall survival was not improved, as seen by Kaplan-Meier curves, in Myc-CaP (C) or TRAMP-C1 (D) tumor models given low or high ¹⁷⁷Lu-NM600 IA. Values are reported as mean ± SD. Abbreviations: IA = injected activity; RPT = radiopharmaceutical therapy.

Fig. 3.

²²⁵Ac-NM600 is efficacious in syngeneic PCa. Tumor growth curves of mice (n = 10) bearing Myc-CaP (A) or TRAMP-C1 (B) allografts receiving vehicle (control) or ²²⁵Ac-NM600 (7.4 or 18.5 kBq) intravenously. Significant tumor growth inhibition was observed at either the administered dose of ²²⁵Ac-NM600 in both Myc-CaP and TRAMP-C1 tumors. The Kaplan-Meier curve demonstrates significantly improved overall survival in Myc-CaP (C) or TRAMP-C1 (D) mice given ²²⁵Ac-NM600 IA. Abbreviations: IA = injected activity; PCa = prostate cancer. **P < .01, ***P < .001, ****P < .0001.

Tumor doubling times for control groups were 4.2 and 6.0 days in Myc-CaP and TRAMP-C1 mice, respectively. In comparison, doubling times for Myc-CaP and TRAMP-C1 administered ²²⁵Ac-NM600 18.5 kBq were 15.4 and 14.8 days, respectively. Notably, a significant (P < .0001) extension in overall survival (Fig. 3C, D) was achieved in all ²²⁵Ac-NM600-treated animals. The median survival of Myc-CaP mice was extended from 12 days in control animals to 43 and 51 days in the ²²⁵Ac-NM600 7.4 and 18.5 kBq groups, respectively.

In TRAMP-C1 mice, median survival was extended from 27 days (control) to 63 days in the ²²⁵Ac-NM600 7.4 kBq group, whereas median survival was not reached for the 18.5 kBq treatment arm. Spaghetti plots of individual tumor volumes of treatments can be found in Figure E1.

Toxicity studies evidenced ¹⁷⁷Lu/²²⁵Ac-labeled NM600 tolerability

Administration of ¹⁷⁷Lu-NM600 or ²²⁵Ac-NM600 did not negatively impact animal weight (Fig. E5) or survival (Figs. 2 and 3), suggesting the high tolerability of the treatments. We further characterized ¹⁷⁷Lu-NM600 and ²²⁵Ac-NM600 safety through a panel of cellular, histologic, and functional tests evaluating hematological, renal, and hepatic toxicity.

Results of CBC (n = 5; Fig. 4) revealed transient, dose-dependent cytopenia indicative of mild bone marrow toxicity. Treatments were generally well tolerated, even at the highest IA level. Further, based on CMP analysis, no major signs of liver and kidney toxicity were observed in Myc-CaP or TRAMP-C1 tumor-bearing animals following ¹⁷⁷Lu-NM600 (Fig. 5A) or ²²⁵Ac-NM600 treatments (Fig. 5B).

Fig. 4.

Complete blood count panels of Myc-CaP and TRAMP-C1 tumor-bearing mice (n = 5) on days 7, 14, and 28 after administration of a high or low IA of ²²⁵Ac-NM600 or ¹⁷⁷Lu-NM600. Compared with controls, animals that received RPT presented transient, dose-dependent reductions in hematological parameters, including white blood cells, lymphocytes, and platelets, indicative of mild bone marrow toxicity. Values returned to baseline by day 28 and did not result in animal mortality. Values are normalized to controls. Abbreviations: IA = injected activity; RPT = radiopharmaceutical therapy.

Fig. 5.

Comprehensive metabolic panel of Myc-CaP and TRAMP-C1 tumor-bearing mice at days 7, 14, and 28 after administration of 5.55 MBq or 18.5 MBq of ¹⁷⁷Lu-NM600 or 7.4 kBq or 18.5 kBq of ²²⁵Ac-NM600. Values are normalized to controls.

Additional histologic (H&E) examinations of the liver, kidneys, and spleen presented normal morphology with no signs of overt tissue degeneration in any of the treatment arms (Fig. E6-E9). H&E staining of the bone marrow revealed slightly reduced cellularity for both doses of ²²⁵Ac-NM600 or ¹⁷⁷Lu-NM600 in both tumor models on day 7, but normal morphology on day 28.

Discussion

Most studies have focused on maximizing the radiation dose delivered to prostate tumor cells using a maximum tolerable dose approach. However, this paradigm assumes that a tumor cannot be overdosed, generally ignoring the presence and importance of the tumor microenvironment.¹⁸ Overwhelming preclinical and clinical data indicate that the antitumor effects of radiation are partly mediated by radiation-induced immunologic effects.^19–24

Furthermore, RPT studies have rarely leveraged a theragnostic approach to enable dosimetry calculations; therefore, knowledge gaps exist surrounding the absorbed dose and radioisotope dependencies of RPT toxicity.

A few studies have investigated the dosimetry, toxicity, and therapeutic effects of ²²⁵Ac on tumor models and TME. Due to the impact of α-emitting RPT agents in the TME of PCa being relatively unknown, this study was conducted to compare the approved isotope, ¹⁷⁷Lu, and the α-emitting isotope, ²²⁵Ac, and to determine how different absorbed doses and emissions have the ability to trigger antitumor response in the TME of 2 syngeneic PCa tumor models.

The radioactive decay of ¹⁷⁷Lu produces gamma rays, which allowed researchers to use SPECT/CT imaging to detect the biodistribution of ¹⁷⁷Lu-NM600. This imaging technique, combined with an in-house computational platform called RAPID, allowed us to conduct detailed, voxel-based Monte Carlo estimations.

These provided calculations estimate the radiation dose delivered to both the tumor and surrounding healthy tissues, which is essential to estimate the exact dose required for each case and accordingly push in the direction of personalized therapy and precision medicine.¹⁸

This approach allowed us to compare various treatments and dosing schedules based on absorbed dose. These comparisons are critical for unraveling the underlying mechanisms that drive the anticancer effects of RPT.

SPECT/CT imaging and biodistribution studies show semiselective tumor targeting and prolonged retention of ¹⁷⁷Lu-NM600 and ²²⁵Ac-NM600 agents in Myc-CaP and TRAMP-C1 tumor models. Treatment with ¹⁷⁷Lu-NM600 did not lead to significant survival benefits in either tumor model, despite TRAMP-C1 tumor-bearing mice showing a significant reduction in tumor volume at the highest ~20 Gy dose.

Although Myc-CaP and TRAMP-C1 tumors have been shown to respond to similar doses of external beam radiation therapy,^25,26 the relative radioresistance of prostate tumors makes them susceptible to dose rate effects, explaining the relatively low effectiveness of protracted ¹⁷⁷Lu-NM600 radiation.^27,28

In contrast, both administered activity levels of ²²⁵Ac-NM600 significantly (P < .001) inhibited tumor growth and extended mean overall survival (P < .0001) by at least 3-fold in both animal models. In addition, antitumor effects were more pronounced in TRAMP-C1 than in Myc-CaP-bearing mice, in line with dosimetry findings showing a ~2× higher absorbed dose per unit of IA for TRAMP-C1 versus Myc-CaP tumors in the ²²⁵Ac-NM600 group.

Importantly, the lower ²²⁵Ac-NM600 IA resulted in greater tumor growth delay than the high ¹⁷⁷Lu-NM600 IA at day 35 posttreatment. These results evidenced ²²⁵Ac-NM600’s higher relative biological effectiveness (>5) compared with ¹⁷⁷Lu-NM600, corroborating well-known radiobiological advantages of α emitters.^29,30 This may be attributed to the higher energy transfer, greater cytotoxicity, and shorter range (precise targeting) of ²²Ac compared with ¹⁷⁷Lu.

To verify this claim, further investigation is required to compare α and βemitters with different half-lives and emitting energies.

The agreement of tumor uptake in both ¹⁷⁷Lu and ²²⁵Ac may be attributed to the ability of novel NM600 to target tumors selectively, increasing tumor uptake and retention. NM600’s ability to interact with lipid raft regions on the plasma membrane, which are abundant in cancer cells more than in normal cells, allowed us to reduce off-target effects on healthy tissues. The toxicity analysis results by CBC indicated that neither low nor high doses of treatments have significant concerns in our TRAMP-C1 and Myc-CaP tumor-bearing mice, which might be because of the precise target achieved by NM600 compared with other molecules, for example, PSMA or fibroblasts.

Ex vivo biodistribution results showed a relatively high accumulation of NM600 in the liver due to its clearance profile. ²²⁵Ac-NM600 presented higher liver accumulation than ¹⁷⁷Lu-NM600, warranting further investigation into biodistribution differences between the 2 compounds and potential isotope detachment. An important aspect of our findings is the observed discrepancy between increased hepatic accumulation of ²²⁵Ac-NM600 and the absence of liver pathology. Although ²²⁵Ac-NM600 exhibited consistently higher liver uptake than ¹⁷⁷Lu-NM600, absolute retention was modest and aligned with the expected hepatobiliary clearance of NM600. In vitro stability assays revealed that over 90% of ²²⁵Ac remained chelated to NM600 in serum for up to 120 hours, with chelation declining only slightly to 88.4% at 168 hours, which supports strong—though not absolute—in vivo stability. Minor chelation loss could partially explain the greater hepatic uptake; however, the similarity in overall tissue biodistribution between ²²⁵Ac- and ¹⁷⁷Lu-NM600 indicates largely comparable in vivo behavior for both constructs. Importantly, comprehensive metabolic analyses and histologic review of liver tissues up to 28 days posttreatment showed no evidence of liver dysfunction or histopathological injury, suggesting that hepatic ²²⁵Ac exposure did not reach hepatotoxic levels in our models. These findings do not rule out the possibility of subtle or delayed hepatic effects that might be missed within the observation window or using standard assays; therefore, extended monitoring and more sensitive evaluations will be considered in future studies to fully characterize the hepatic safety profile of ²²⁵Ac-NM600. To investigate the function abnormalities in the kidney, spleen, and liver, our findings of CMP analysis confirmed that no major signs were detected in Myc-CaP or TRAMPC-1 tumor-bearing mice, which indicated that the accumulation in both liver and kidney would not remain, causing toxicity, but might be cleared over time. Additional investigations are necessary to fully unravel the molecular basis of the enhanced efficacy observed with ²²⁵Ac-NM600.

To elucidate the molecular basis of the enhanced efficacy observed with 225Ac-NM600, we recognize the critical importance of incorporating mechanistic endpoints—including immunologic and DNA damage markers—in ongoing and future studies for the continued development and optimization of this and related TATs. In this regard, preliminary immunologic analyses from our group³¹ show that ²²⁵Ac-NM600 treatment modulates the tumor immune microenvironment by increasing infiltration of immune effector cells (eg, CD8+ T cells, NK cells), elevating proinflammatory cytokines such as IFN-γ and TNF-α, and shifting the effector-to-regulatory immune cell ratio in favor of tumor rejection. Collectively, these findings indicate that, in addition to direct cytotoxicity, 225Ac-NM600 may prime the tumor microenvironment for enhanced antitumor immune responses. Ongoing work in our laboratory is actively evaluating these and related immunologic parameters across different isotopes and tumor models to more fully define the therapeutic potential of NM600-based TAT. We also acknowledge that, due to study timeline constraints, dosimetry analyses and treatment administration were performed concurrently, which limited our ability to prospectively match absorbed doses (equidosimetric dosing) across treatment groups. Although retrospective dosimetry allowed for meaningful biological comparisons, this approach may have introduced variability in tumor and organ dose delivery between arms. To address this limitation, we plan to conduct prospective dosimetry in future studies, enabling preselection of equidosimetric dosing and thereby supporting more rigorous assessment of therapeutic efficacy and safety.

Conclusion

²²⁵Ac-NM600 demonstrated superior antitumor efficacy compared with ¹⁷⁷Lu-NM600 in murine PCa models, while maintaining a favorable toxicity profile. The enhanced therapeutic effects of ²²⁵Ac-NM600 were evident through significant tumor growth inhibition and prolonged overall survival in both Myc-CaP and TRAMP-C1 tumor models. This improved efficacy can be attributed to the higher relative biological effectiveness of α particles compared with β particles. Key findings include (1) selective tumor targeting and retention of NM600 in PCa models, (2) superior antitumor effects of ²²⁵Ac-NM600 at lower administered activities compared with ¹⁷⁷Lu-NM600, (3) transient, dose-dependent hematological changes for both agents, with no significant long-term adverse effects, and (4) dosimetry calculations corroborating selective radiation delivery to tumors. These results provide a strong rationale for further development of α-emitting RPT agents for PCa treatment. Future investigations should focus on elucidating the molecular mechanisms underlying the enhanced efficacy of ²²⁵Ac-NM600 and exploring potential synergies with other treatment modalities, such as immunotherapy.

Data Sharing Statement:

All raw data are available with the corresponding author upon request.