[225Ac]Ac-SibuDAB for Targeted Alpha Therapy of Prostate Cancer: Preclinical Evaluation and Comparison with [225Ac]Ac-PSMA-617

Simple Summary Prostate-specific membrane antigen (PSMA) radioligands have proven effective to treat patients with metastatic castration-resistant prostate cancer. Targeted α-therapy using actinium-225 has been used for patients with end-stage disease who no longer responded to βࢤ-therapy through the use of 177Lu-based radioligand therapy (RLT). In this study, we investigated and compared the therapeutic efficacy of [225Ac]Ac-SibuDAB and [225Ac]Ac-PSMA-617 and assessed potential undesired side effects in the preclinical setting. Due to a dedicated albumin-binding entity integrated into [225Ac]Ac-SibuDAB, this radioligand showed an enhanced blood circulation time and, hence, increased tumor uptake but also higher retention in normal tissues. The therapeutic efficacy of [225Ac]Ac-SibuDAB was enhanced as compared to that of [225Ac]Ac-PSMA-617, yet, undesired side effects were in the same range for both radioligands. Our data suggest that [225Ac]Ac-SibuDAB could be a powerful alternative to [225Ac]Ac-PSMA-617, however, the safe therapeutic window should be carefully defined in clinical dose escalation studies. Abstract In the present study, SibuDAB, an albumin-binding PSMA ligand, was investigated in combination with actinium-225 and the data were compared with those of [225Ac]Ac-PSMA-617. In vitro, [225Ac]Ac-SibuDAB and [225Ac]Ac-PSMA-617 showed similar tumor cell uptake and PSMA-binding affinities as their 177Lu-labeled counterparts. The in vitro binding to serum albumin in mouse and human blood plasma, respectively, was 2.8-fold and 1.4-fold increased for [225Ac]Ac-SibuDAB as compared to [177Lu]Lu-SibuDAB. In vivo, this characteristic was reflected by the longer retention of [225Ac]Ac-SibuDAB in the blood than previously seen for [177Lu]Lu-SibuDAB. Similar to [225Ac]Ac-PSMA-617, [225Ac]Ac-SibuDAB was well tolerated at 30 kBq per mouse. Differences in blood cell counts were observed between treated mice and untreated controls, but no major variations were observed between values obtained for [225Ac]Ac-SibuDAB and [225Ac]Ac-PSMA-617. [225Ac]Ac-SibuDAB was considerably more effective to treat PSMA-positive tumor xenografts than [225Ac]Ac-PSMA-617. Only 5 kBq per mouse were sufficient to eradicate the tumors, whereas tumor regrowth was observed for mice treated with 5 kBq [225Ac]Ac-PSMA-617 and only one out of six mice survived until the end of the study. The enhanced therapeutic efficacy of [225Ac]Ac-SibuDAB as compared to that of [225Ac]Ac-PSMA-617 and reasonable safety data qualify this novel radioligand as a candidate for targeted α-therapy of prostate cancer.


Introduction
Prostate-specific membrane antigen (PSMA)-targeted radioligand therapy (RLT) has emerged as an efficient treatment strategy for metastatic castration-resistant prostate cancer (mCRPC) [1]. Based on the positive outcome of the clinical Phase III trial (VISION; NCT03511664 [2]), [ 177 Lu]Lu-PSMA-617 was recently approved by the U.S. food and drug administration (FDA) under the trade name Pluvicto TM for the treatment of patients with mCRPC. Despite these overall positive data for 177 Lu-based RLT, up to 30% of the patients have shown inherent or acquired resistance over time [3]. In these cases, targeted α-therapy may present an effective option, since α-particles are characterized by a short tissue range (≤100 µm) and a substantially higher linear energy transfer (LET; 50-300 keV/µm) [4].
RLT using [ 225 Ac]Ac-PSMA-617 and [ 225 Ac]Ac-PSMA-I&T was applied to patients that did not respond to β -therapy or suffered from multiple bone lesions, which presents a contraindication for 177 Lu-based RLT due to the increased risk of hematologic toxicity [5,6]. Both, [ 225 Ac]Ac-PSMA-617 and [ 225 Ac]Ac-PSMA-I&T revealed to be effective in the clinical setting to treat end-stage prostate cancer [7]. A clinical Phase I dose escalation study to evaluate the safety of [ 225 Ac]Ac-PSMA-617 (AcTION: NCT04597411) in prostate cancer patients with extensive skeletal metastases is currently ongoing. At the same time, a clinical Phase II study (TATCIST: NCT05219500) is in progress to investigate the therapeutic efficacy of [ 225 Ac]Ac-PSMA-I&T using a de-escalation scheme. 225 Ac-based RLT led to significant reduction of the tumor burden and biochemical response with >50% prostate-specific antigen (PSA) decline in the majority of mCRPC patients [7]. Bone marrow toxicity, damage to the salivary glands and, on a longer term, also to the kidneys was observed frequently in treated patients as a result of the high absorbed radiation dose in these organs and tissues [8,9]. In particular, irreversible xerostomia due to salivary gland damage significantly compromised the patients' quality of life [7,10]. Lower activities applied at shorter intervals were, therefore, recommended to alleviate potential side effects regarding the salivary glands [8]. Furthermore, specific measures to reduce salivary gland uptake, such as the cooling during radioligand application or the injection of botulinum toxin prior to 225 Ac-based RLT showed variable efficacy [11].
The development of innovative PSMA radioligands that show reduced salivary gland accumulation relative to the tumor uptake would certainly present the smartest solution to address the challenge of salivary gland toxicity. In this context, the question arises whether albumin-binding radioligands, which show enhanced blood retention and increased tumor accumulation, would be advantageous in this regard. So far, only few long-circulating PSMA ligands were investigated in combination with actinium-225 [12,13], but preclinical data are not predictive for salivary gland accumulation in patients. Initial clinical application of [ 177 Lu]Lu-EB-PSMA-617, a radioligand modified with Evans Blue as an albumin-binding entity, showed a significantly increased salivary gland accumulation compared to that of [ 177 Lu]Lu-PSMA-617 [14]. In contrast, a first-in-human application of [ 177 Lu]Lu-PSMA-ALB-56, a PSMA radioligand modified with a p-tolyl-butanoyl entity as an albumin binder, revealed similar salivary gland accumulation as [ 177 Lu]Lu-PSMA-617 and, therefore, a favorable tumor-to-salivary gland dose ratio [15].
Recently, we have developed a new class of PSMA ligands comprising ibuprofen as an albumin-binding entity [16], whereof SibuDAB emerged as the most promising candidate [17]. The calculation of the area under the curve (AUC) values revealed higher tumor uptake for [ 177 Lu]Lu-SibuDAB than for [ 177 Lu]Lu-PSMA-617 but reduced retention in blood and kidneys as compared to [ 177 Lu]Lu-PSMA-ALB-56. Based on the tissue distribution profile of [ 177 Lu]Lu-SibuDAB, the application of SibuDAB in combination with actinium-225 appeared, thus, feasible and reasonable to be investigated.
The aim of this study was to investigate 225 Ac-based RLT using SibuDAB and compare the antitumor efficacy with that of [ 225 Ac]Ac-PSMA-617 (Figure 1). In a first step, [ 225 Ac]Ac-SibuDAB was evaluated in vitro for comparison with [ 225 Ac]Ac-PSMA-617 and their 177 Lu-labeled counterparts using PC-3 PIP (PSMA-positive) and PC-3 flu (PSMAnegative) prostate cancer cells. In a second step, the two 225 Ac-based radioligands were compared with regard to potential early side effects in immunocompetent mice followed by investigations of their pharmacokinetic profiles and therapeutic efficacy in PC-3 PIP tumor-bearing mice.
Cancers 2022, 14, x FOR PEER REVIEW 3 of 18 their 177 Lu-labeled counterparts using PC-3 PIP (PSMA-positive) and PC-3 flu (PSMAnegative) prostate cancer cells. In a second step, the two 225 Ac-based radioligands were compared with regard to potential early side effects in immunocompetent mice followed by investigations of their pharmacokinetic profiles and therapeutic efficacy in PC-3 PIP tumor-bearing mice.  [17]; (B) PSMA-617 as a reference compound [18].

Preparation of the Radioligands
The PSMA ligands used in this study were synthesized using solid-phase chemistry methods as previously reported [17,18]. The labeling of the PSMA ligands with actinium-225 was performed at ITM Medical Isotopes GmbH, Munich, Germany (Supplementary Material: Section S1). The radioligands (~50 kBq/nmol; ~0.5 MBq/mL) were shipped to the Paul Scherrer Institute (PSI) overnight followed by assessing their integrity using thinlayer chromatography. It was demonstrated that both radioligands, [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617, were stable over a period of 48 h (>99% intact radioligand) and could, therefore, be used for preclinical experiments on the two following days upon arrival. Depending on the experiment, the molar activity of the radioligands was adjusted by adding unlabeled ligand and the final formulation was prepared by adding vehicle (saline with 0.05% bovine serum albumin (BSA)) to obtain the desired activity concentration. Radiolabeling of SibuDAB with lutetium-177 (5-10 MBq/nmol) was performed as previously reported [17].

Preparation of the Radioligands
The PSMA ligands used in this study were synthesized using solid-phase chemistry methods as previously reported [17,18]. The labeling of the PSMA ligands with actinium-225 was performed at ITM Medical Isotopes GmbH, Munich, Germany (Supplementary Material: Section S1). The radioligands (~50 kBq/nmol;~0.5 MBq/mL) were shipped to the Paul Scherrer Institute (PSI) overnight followed by assessing their integrity using thin-layer chromatography. It was demonstrated that both radioligands, [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617, were stable over a period of 48 h (>99% intact radioligand) and could, therefore, be used for preclinical experiments on the two following days upon arrival. Depending on the experiment, the molar activity of the radioligands was adjusted by adding unlabeled ligand and the final formulation was prepared by adding vehicle (saline with 0.05% bovine serum albumin (BSA)) to obtain the desired activity concentration. Radiolabeling of SibuDAB with lutetium-177 (5-10 MBq/nmol) was performed as previously reported [17].
Relative albumin-binding affinities were determined using an ultrafiltration assay according to an adapted version of a recently published protocol (Supplementary Material: Section S6) [20].

In Vivo Experiments
All applicable international, national, and/or institutional guidelines for the care and use of laboratory animals were followed and all animal experiments were carried out according to the guidelines of Swiss Regulations for Animal Welfare. The preclinical studies were ethically approved by the Cantonal Committee of Animal Experimentation and permitted by the responsible cantonal authorities (license N • 75668). Female mice were used as they can be housed in groups over a long period without the risk of fighting with each other as it is commonly seen with male mice. This is relevant, in particular for studies performed over longer periods. Female mice were also chosen for the tumor therapy study, which is feasible when using an androgen-independent prostate cancer model.

Design of the Tolerability Study
Five-to six-week-old immunocompetent, female FVB mice were purchased from Charles River Laboratories (Sulzfeld, Germany). After an acclimatization period of at least 7 days, the mice (n = 5 per group) were intravenously injected with 30 kBq [ 225 Ac]Ac-SibuDAB or 30 kBq [ 225 Ac]Ac-PSMA-617 and additional mice (n = 7 per group) remained untreated as controls. Mice were monitored by measuring their body mass and potential signs of pain or unease three times a week (Supplementary Material: Section S7). The mice were euthanized on Day 10, Day 28 or on Day 56, respectively, and variable parameters were assessed from blood and tissue samples.

Blood Sampling and Assessment of Parameters Indicative for Tolerability
Hematological analysis was performed using blood samples taken by cardiac puncture immediately after euthanasia. Blood and bone marrow smears were prepared and stained using the Pappenheim protocol [21]. The assessment was carried out by a board-certified veterinary anatomic pathologist (AnaPath Services GmbH, Liestal, Switzerland). Blood plasma parameters were determined in plasma after centrifugation of blood samples obtained shortly before euthanasia (Supplementary Material: Section S7).

Histopathology
Tissue sections of formalin-fixed paraffin-embedded (FFPE) spleen, salivary glands and kidneys were prepared and stained with hematoxylin and eosin (H&E) by AnaPath Services GmbH (Liestal, Switzerland). The tissues were analyzed by a board certified veterinary pathologist using a pre-defined scoring system (Supplementary Material: Section S7).

Biodistribution Studies
Five-to six-week-old female BALB/c nude mice were purchased from Charles River Laboratories (Sulzfeld, Germany). Approximately 2 weeks after subcutaneous inoculation of PC-3 PIP and PC-3 flu tumor cells (6 × 10 6 cells and 5 × 10 6 cells, respectively, in 100 µL Hanks' balanced salt solution (HBSS)) on the right and left shoulder, respectively, biodistribution studies were performed. [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617 (40 kBq; 1 nmol, 100 µL saline containing 0.05% BSA per mouse) were intravenously injected into n = 4-5 mice per timepoint. Collected tissues and organs were weighed and counted for activity using a γ-counter (PerkinElmer Wallac Wizard 1480) the day after dissection. Biodistribution data were reported as the percentage of the injected activity per gram of tissue mass (% IA/g) obtained from decay-corrected data (Supplementary Material: Section S8).

PSMA-Specific Antitumor Efficacy of [ 225 Ac]Ac-SibuDAB
Five days after subcutaneous inoculation of four groups of mice with either PC-3 PIP or PC-3 flu tumor cells on the right shoulder (4 × 10 6 cells in 100 µL HBSS), the mice were intravenously injected with either vehicle only (saline with 0.05% BSA, n = 6 and n = 4, respectively) or with 30 kBq [ 225 Ac]Ac-SibuDAB (1 nmol per mouse; n = 4 per group) diluted in vehicle (Table 1). Table 1. Design of the therapy study including information about the number (n) of mice per group, the tumor type, the treatment, the injected activity and the average ± standard deviation (SD) of the tumor volume and body mass of mice of each group on Day 0. After tumor cell inoculation, mice bearing PC-3 PIP tumor xenografts were intravenously injected with vehicle only (n = 6) or with [ 225 Ac]Ac-SibuDAB or [ 225 Ac]Ac-PSMA-617, respectively, applied at 5 kBq or 10 kBq (1 nmol per mouse; n = 6 per group; Table 1). The control group consisted of mice from several studies (n = 12 mice) and the data from this group were published previously [17].

Monitoring of Treated Mice and Therapy Assessment
After administration of the radioligands, the mice were monitored over 8 weeks using the same measures as previously published by Umbricht et al. [22]. The relative body mass was defined as [BM x /BM 0 ], where BM x is the body mass in grams at a given Day x and BM 0 is the body mass in grams at Day 0. The tumor dimension was determined by measuring the longest tumor axis (L) and its perpendicular axis (W) with a digital caliper. The tumor volume (V) was calculated according to the equation [V = 0.5 × (LW 2 )]. The relative tumor volume was defined as [TV x /TV 0 ], where TV x is the tumor volume in mm 3 at a given Day x, and TV 0 is the tumor volume in mm 3 at Day 0. Endpoint criteria that required euthanasia were defined as (i) body mass loss of >15%, (ii) a tumor volume of >800 mm 3 , (iii) a combination of body mass loss of >10% and a tumor volume of >700 mm 3 or, (iv) signs of unease and pain or a combination thereof using a scoring system as previously reported [22]. The median survival of mice of each group was assessed with Kaplan-Meier curves. Potential early side effects for mice that survived until study end were determined based on blood chemistry and organ-to-brain mass ratios on Day 56 by comparison to values obtained from an additional group of therapy-naïve mice (n = 4) without tumors but of the same age (Supplementary Material: Section S9).

Analysis of Statistical Significance of the Data
The data of the in vitro and in vivo studies were analyzed using GraphPad Prism software (version 8) with a p-value < 0.05 considered as statistically significant. The cell uptake and internalization studies, the biodistribution studies, the body mass and tumor volume of mice on Day 0 of the therapy study, as well as parameters to determine adverse events during the tolerability and therapy studies were analyzed and compared using a one-way ANOVA test with a Tukey's multiple comparisons post-test. The efficacy of the therapy study was assessed using a log-rank test (Mantel Cox).

In Vitro Characteristics of the 225 Ac-and 177 Lu-Based PSMA Radioligands
In a first step, the aim was to investigate the in vitro features of [ 225 Ac]Ac-SibuDAB and to compare them with those of [ 225 Ac]Ac-PSMA-617. Furthermore, the question whether actinium-225 and lutetium-177 would be interchangeable without affecting the radioligands' physicochemical properties was addressed.

Serum Albumin Binding of [ 225 Ac]Ac-SibuDAB and [ 177 Lu]Lu-SibuDAB
The relative albumin-binding affinity of [ 225 Ac]Ac-SibuDAB determined in mouse and human blood plasma, respectively, was approximately 2.8-fold and 1.4-fold higher than for [ 177 Lu]Lu-SibuDAB ( Figure 2B). Possibly, the larger radius of actinium-225 changed the conformation of the molecule and, hence, the binding mode to serum albumin. The fact that this effect was much more pronounced in mouse than in human plasma lessens, however, its relevance in view of an application of [ 225 Ac]Ac-SibuDAB in human patients. It is also important to note that both [ 225 Ac]Ac-SibuDAB and [ 177 Lu]Lu-SibuDAB revealed a stronger binding to mouse serum albumin than to human serum albumin, hence, pharmacokinetic data acquired in mice may not precisely reflect the situation in humans as to what concerns the radioligands' blood circulation time.

Tolerability of [ 225 Ac]Ac-SibuDAB in Immunocompetent Mice
Common undesired side effects of RLT with [ 225 Ac]Ac-PSMA-617 in clinics include hematological changes, xerostomia and impairment of kidney function [26]. Hematotoxicity up to Grade 3/4 was observed in about 30% of the patients in a clinical study reported by Feuerecker et al. [8]. Xerostomia was sometimes associated with reduced appetite and body mass loss, and a decreased quality of life has been reported in the majority of patients treated with [ 225 Ac]Ac-PSMA-617 [26]. The topic of kidney radiotoxicity is controversially discussed among clinicians in the community. A clear conclusion about how 225 Ac-based RLT affects long-term kidney function is still difficult to draw as [ 225 Ac]Ac-PSMA-617 has been used for the treatment of end-stage prostate cancer patients with a limited life expectancy.
In this preclinical study, we investigated the tolerability of [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617 in immunocompetent mice. According to recently reported preclinical studies [12,27,28], an activity of 20-40 kBq [ 225 Ac]Ac-PSMA-617 was effective to control tumor growth without causing major side effects. [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617 were, therefore, applied at 30 kBq per mouse. Indeed, both radioligands were well tolerated based on the average body mass of mice of each group, which increased similarly over time until the end of the study on Day 56 ( Figure S2).

Hematological Effects of [ 225 Ac]Ac-SibuDAB and [ 177 Lu]Lu-SibuDAB
The complex decay chain of actinium-225 with several α-particle-emitting daughter nuclides could potentially present a risk of side effects [29]. In this regard, the efficient internalization of PSMA upon binding of PSMA radioligands is certainly an advantage, whereas the enhanced circulation time of [ 225 Ac]Ac-SibuDAB may increase the number of decays outside the target cell. Furthermore, a longer blood circulation time would certainly increase the risk of bone marrow toxicity. Most likely, the applicable activity of [ 225 Ac]Ac-SibuDAB would, therefore, have to be adjusted.
The hemograms of mice treated with [ 225 Ac]Ac-SibuDAB or [ 225 Ac]Ac-PSMA-617 did not differ significantly at any of the investigated timepoints (p > 0.05). Differences in hematology were, however, observed between treated mice and controls, although most values were still in the normal range based on the values reported for female FVB mice (Figures 3 and S3) [30]. A trend of lower leukocyte counts due to a drop in lymphocyte counts was observed on Day 10 and Day 28 in treated mice as compared to control mice (p > 0.05; Figure 3A,B). On Day 56, the difference was only statistically significant for mice that received [ 225 Ac]Ac-PSMA-617 (p < 0.05) but not for those that received [ 225 Ac]Ac-SibuDAB (p > 0.05). Erythrocyte counts declined to (8.7 ± 0.3) × 10 12 /L on Day 28 after administration of either radioligand as compared to control mice ((9.3 ± 0.2) × 10 12 /L; p < 0.05) ( Figure 3C). In line with this, the hematocrit and hemoglobin levels also decreased on Day 28 after application of [ 225 Ac]Ac-SibuDAB (43 ± 1% and 11.2 ± 0.5 g/dL, respectively) and [ 225 Ac]Ac-PSMA-617 (43 ± 3% and 11.3 ± 0.5 g/dL, respectively) as compared to those of untreated controls (46 ± 2% and 12.1 ± 0.5 g/dL, respectively; p < 0.05; Figure S3). As a result of compensatory mechanisms, the erythrocyte counts, hematocrit and hemoglobin showed higher values than control mice on Day 56. No significant differences between thrombocyte counts of treated and untreated mice were observed at any of the investigated timepoints ( Figure 3D). To compensate for the decreased number of leukocytes determined on Day 10, the majority of mice (three out of five in each group) that had received an 225 Ac-based RLT showed an increase in myleopoietic precursor cells in relation to erythropoietic precursor cells in the bone marrow smears as compared to control mice at that timepoint (Table S2-S4). On Day 28, the ratio tended to be shifted towards the erythorpoietic precursor cells to compensate for the lower number of erythrocytes determined in peripheral blood at that timepoint. No morphological cell abnormalities were observed in blood and bone marrow smears which would indicate severe hematological toxicity.
Histopathological analysis of the spleen, an organ involved in hematopoiesis, revealed a minimal to mild hyperplasia visible on Day 28 in 225 Ac-treated mice but showed otherwise no major differences between mice that received 225 Ac-based RLT and mice of the control group ( Figure S4, Tables S5 and S6).
Under the given experimental conditions, hematological differences between treated and untreated mice were observed, however, it can be concluded that no signs of an increased risk of hematological toxicity were identified for the use of [ 225 Ac]Ac-SibuDAB as compared to [ 225 Ac]Ac-PSMA-617. To compensate for the decreased number of leukocytes determined on Day 10, the majority of mice (three out of five in each group) that had received an 225 Ac-based RLT showed an increase in myleopoietic precursor cells in relation to erythropoietic precursor cells in the bone marrow smears as compared to control mice at that timepoint (Tables S2-S4). On Day 28, the ratio tended to be shifted towards the erythorpoietic precursor cells to compensate for the lower number of erythrocytes determined in peripheral blood at that timepoint. No morphological cell abnormalities were observed in blood and bone marrow smears which would indicate severe hematological toxicity.

Transient Salivary Gland Lesions after Application of 225 Ac-Based Radioligands
Histopathological analysis of the spleen, an organ involved in hematopoiesis, revealed a minimal to mild hyperplasia visible on Day 28 in 225 Ac-treated mice but showed otherwise no major differences between mice that received 225 Ac-based RLT and mice of the control group ( Figure S4, Tables S5 and S6).
Under the given experimental conditions, hematological differences between treated and untreated mice were observed, however, it can be concluded that no signs of an increased risk of hematological toxicity were identified for the use of [ 225 Ac]Ac-SibuDAB as compared to [ 225 Ac]Ac-PSMA-617.

Transient Salivary Gland Lesions after Application of 225 Ac-Based Radioligands
The high salivary gland accumulation of [ 225 Ac]Ac-PSMA-617 in humans can cause irreversible Grade 1/2 xerostomia in patients already after the first therapy cycle [8,31]. It is still unclear whether albumin-binding radioligands would show less or more salivary gland uptake in patients. While albumin-binding PSMA radioligands showed consistently increased tumor uptake in patients as compared to conventional radioligands, Zang   The data of our study may indicate that albumin-binding radioligands are advantageous in view of salivary gland toxicity even though a prediction for human application is not possible. The salivary gland cells and their organization are quite different in mice and humans [34]. This is probably the reason why radioligand uptake in humans is significantly higher as compared to the salivary gland uptake in mice, where it was mostly negligible [17]. The data of our study may indicate that albumin-binding radioligands are advantageous in view of salivary gland toxicity even though a prediction for human application is not possible. The salivary gland cells and their organization are quite different in mice and humans [34]. This is probably the reason why radioligand uptake in humans is significantly higher as compared to the salivary gland uptake in mice, where it was mostly negligible [17].

Assessment of Potential Kidney Damage after Application of [ 225 Ac]Ac-SibuDAB
The risk of kidney toxicity after 225 Ac-based RLT is controversially discussed among physicians and contradictive statements can be found in the literature. Feuerecker et al. [35] reported Grade 1/2 impairment of kidney function in about 19% of the patients of the respective study 4-8 weeks after treatment, however, similarly to previously published data [8], it was not of clinical relevance. Pelletier et al. reported two cases of progressive kidney disease in mCRPC patients with impaired baseline kidney function that received treatment with [ 225 Ac]Ac-PSMA-617 [9]. A kidney biopsy performed for one of the patients 34 weeks after first exposure showed significant tubular atrophy and ongoing tubular injury.
In our preclinical study, we observed similar (p > 0.05) blood urea nitrogen levels in mice treated with [ 225 Ac]Ac-SibuDAB (9.1 ± 1.1 mmol/L) and [ 225 Ac]Ac-PSMA-617 (8.7 ± 1.1 mmol/L) on Day 28. In the case of [ 225 Ac]Ac-SibuDAB these values were significantly elevated as compared to those of control mice (6.7 ± 0.7 mmol/L, p < 0.05), however, this was no longer the case on Day 56 ( Figure S5). Histopathological evaluation did not reveal any changes other than background lesions for the kidneys of treated mice and untreated controls (Tables S5 and S6). An increased risk of acute nephrotoxicity after application of [ 225 Ac]Ac-SibuDAB as compared to [ 225 Ac]Ac-PSMA-617 was not identified under the given experimental conditions. It has to be critically acknowledged, however, that radionephropathy in patients is commonly observed later than two months after administration of the radioligand [9,36]. Although the manifestations may occur earlier in mice than in humans, delayed radiotoxic effects cannot be excluded based on the data presented in this study. In an ideal case, the observation period of mice should be much longer to detect potential kidney toxicity as it was done by Banerjee et al. who assessed side effects of [ 225 Ac]Ac-L1 one year after application [33]. It is obvious that, based on the increased kidney retention of [ 225 Ac]Ac-SibuDAB, the injected activity would have to be reduced as compared to that of [ 225 Ac]Ac-PSMA-617 in order not to increase the risk of radionephrotoxicity.

Tissue Distribution Profiles of [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617
In a next step, we investigated the tissue distribution profile of the radioligands using PC-3 PIP tumor-bearing mice in view of a therapeutic application of the radioligands that should allow an estimation of their antitumor efficacy.

Biodistribution of [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617
Biodistribution data obtained at 1 h, 4 h, 24 h and 48 h after injection of the radioligands showed the expected differences between [ 225 Ac]Ac-SibuDAB and [ 225 Ac]Ac-PSMA-617 that were previously found with the 177 Lu-labeled counterparts ( Figure 5; Tables S7-S10). Blood retention of [ 225 Ac]Ac-SibuDAB (23 ± 2% IA/g and 3.3 ± 0.7% IA/g at 1 h and 24 h post injection (p.i.), respectively) was enhanced as compared to that of [ 225 Ac]Ac-PSMA-617 and the retention of [ 225 Ac]Ac-SibuDAB in the kidneys was also higher (19 ± 1 % IA/g vs. 3.5 ± 0.6 % IA/g at 4 h p.i.). The tumor accumulation of [ 225 Ac]Ac-SibuDAB was significantly higher than for [ 225 Ac]Ac-PSMA-617 with a maximum value reached at 24 h p.i. (80 ± 8% IA/g). At 48 h p.i., a 2-fold higher tumor uptake was observed for [ 225 Ac]Ac-SibuDAB as compared to [ 225 Ac]Ac-PSMA-617 (64 ± 11% IA/g vs. 31 ± 3% IA/g). Unspecific accumulation in PSMA-negative PC-3 flu tumors was not observed.  Figure 5; Tables S7-S10). In line with the in vitro findings of an enhanced plasma protein-binding affinity of [ 225 Ac]Ac-SibuDAB as compared to [ 177 Lu]Lu-SibuDAB, the former showed an increased blood retention (16 ± 2% IA/g vs. 2.9 ± 1.9% IA/g at 4 h p.i.; Figure 5A). The altered clearance profile could potentially explain the observed differences in the renal uptake of [ 225 Ac]Ac-and [ 177 Lu]Lu-SibuDAB ( Figure  5B). Differences in the tumor uptake were also visible at early timepoints after injection of [ 225 Ac]Ac-SibuDAB as compared to the 177 Lu-labeled counterpart (45 ± 5% IA/g vs. 66 ± 11% IA/g at 4 h p.i.; Figure 5C). Nevertheless, the activity retention in the tumor was similar at later timepoints (64 ± 11% IA/g vs. 69 ± 12% IA/g at 48 h p.i.). It has to be noted that the distribution profiles of [ 225 Ac]Ac-SibuDAB and [ 177 Lu]Lu-SibuDAB are likely more similar in humans because the albumin-binding affinities of the two radioligands were found to be more similar in human blood plasma than in mouse blood plasma.  Figure 5; Tables S7-S10). In line with the in vitro findings of an enhanced plasma protein-binding affinity of [ 225 Ac]Ac-SibuDAB as compared to [ 177 Lu]Lu-SibuDAB, the former showed an increased blood retention (16 ± 2% IA/g vs. 2.9 ± 1.9% IA/g at 4 h p.i.; Figure 5A). The altered clearance profile could potentially explain the observed differences in the renal uptake of [ 225 Ac]Ac-and [ 177 Lu]Lu-SibuDAB ( Figure 5B). Differences in the tumor uptake were also visible at early timepoints after injection of [ 225 Ac]Ac-SibuDAB as compared to the 177 Lu-labeled counterpart (45 ± 5% IA/g vs. 66 ± 11% IA/g at 4 h p.i.; Figure 5C). Nevertheless, the activity retention in the tumor was similar at later timepoints (64 ± 11% IA/g vs. 69 ± 12% IA/g at 48 h p.i.). It has to be noted that the distribution profiles of [ 225 Ac]Ac-SibuDAB and [ 177 Lu]Lu-SibuDAB are likely more similar in humans because the albumin-binding affinities of the two radioligands were found to be more similar in human blood plasma than in mouse blood plasma.

PSMA-Specific Treatment Effect of [ 225 Ac]Ac-SibuDAB
PC-3 PIP and PC-3 flu tumor xenografts of untreated control mice rapidly increased in size so that all mice of these groups reached an endpoint between Day 14 and Day 28 or between Day 18 and Day 22, respectively, with median survival times of 18 and 21 days, respectively ( Figure 6, Table 2). Application of 30 kBq [ 225 Ac]Ac-SibuDAB eradicated PSMA-positive PC-3 PIP tumors entirely so that all mice of this group survived until the end of the study. Application of the same activity of [ 225 Ac]Ac-SibuDAB (30 kBq/mouse) did, however, not significantly delay the growth of PSMA-negative PC-3 flu tumors resulting in a survival time that was not much longer than in untreated control mice (p > 0.05, median survival of 25 days) ( Figure 6A,B, Table 2). This unambiguously confirmed that the therapeutic efficacy of [ 225 Ac]Ac-SibuDAB was dependent on PSMA expression and not due to an unspecific effect of actinium-225 or its daughter nuclides. These data further suggested that possibly even lower activities would be sufficient to obtain tumor control with the albumin-binding radioligand. The treatment of mice with 5 kBq or 10 kBq [ 225 Ac]Ac-SibuDAB effectively eradicated tumors of all mice of these groups, which survived until the end of the study on Day 56 ( Figure 6C,D, Table 2).
On the other hand, the application of 10 kBq [ 225 Ac]Ac-PSMA-617 per mouse resulted in the survival of four out of six mice until the end of the study. Application of 5 kBq  The treatment of mice with 5 kBq or 10 kBq [ 225 Ac]Ac-SibuDAB effectively eradicated tumors of all mice of these groups, which survived until the end of the study on Day 56 ( Figure 6C,D, Table 2).
On the other hand, the application of 10 kBq [ 225 Ac]Ac-PSMA-617 per mouse resulted in the survival of four out of six mice until the end of the study. Application of 5 kBq [ 225 Ac]Ac-PSMA-617 was even less effective. After initial regression, tumor re-growth was observed and, consequently, five out of six mice of this group reached an endpoint within the 56 days of investigation. Nevertheless, even this group showed still a significantly prolonged survival (p < 0.05) with a median survival of 46 days as compared to control mice. Other research groups used commonly higher activities of [ 225 Ac]Ac-PSMA-617 than 10 kBq per mouse, however, in most of these cases LNCaP and C4-2 tumors were used, which express PSMA naturally, and, thus, at much lower expression levels than it is the case for the transduced PC-3 PIP tumors [12,37,38]. Similar therapeutic efficacy was, however, reported by Banerjee et al. who used the PC-3 PIP tumor mouse model for the investigation of the therapeutic efficacy of [ 225 Ac]Ac-L1 [33].
The body mass of mice that were effectively treated with [ 225 Ac]Ac-SibuDAB was increasing over time, indicating that the therapy was well tolerated ( Figure S6). In order to further assess the therapy outcome, we compared several parameters of mice, which had a tumor initially but were cured after the treatment with [ 225 Ac]Ac-SibuDAB, with treatment-naïve, non-tumor-bearing BALB/c nude mice of the same age. No obvious adverse events were observed in cured mice at Day 56 after injection of 30 kBq [ 225 Ac]Ac-SibuDAB as compared to healthy mice (Tables S11-S13). The only exception was a lower kidney-to-brain mass ratio (0.62-0.68) determined in cured mice as compared to healthy mice (0.73 ± 0.04) which could potentially be a sign of a radiation induced impairment of the kidneys in mice that received [ 225 Ac]Ac-SibuDAB. 3 [39]. Our study confirmed this observation by demonstrating a better effect using 5 kBq or 10 kBq [ 225 Ac]Ac-SibuDAB than a 1000-fold higher activity of [ 177 Lu]Lu-SibuDAB (5 MBq and 10 MBq per mouse) ( Figure S7). Only in two out of six mice treated with 5 MBq [ 177 Lu]Lu-SibuDAB and in five out of six mice treated with 10 MBq [ 177 Lu]Lu-SibuDAB the tumor xenografts disappeared completely without relapse until Day 56. Tumors of the other mice of these groups started to regrow after~4 and~7 weeks, respectively. In comparison, no tumors were visible at Day 56 after injection of 5 kBq [ 225 Ac]Ac-SibuDAB.

Conclusions
The increased therapeutic efficacy of [ 225 Ac]Ac-SibuDAB as compared to the clinically investigated [ 225 Ac]Ac-PSMA-617 indicated an advantage of this novel albumin-binding PSMA radioligand for targeted α-therapy. Despite the differences regarding the biodistribution profile and the longer blood circulation time of [ 225 Ac]Ac-SibuDAB as compared to [ 225 Ac]Ac-PSMA-617, the preliminary data regarding tolerability in immunocompetent mice did not show an obvious increased risk when using [ 225 Ac]Ac-SibuDAB instead of [ 225 Ac]Ac-PSMA-617. Nevertheless, the safe therapeutic window for the application of [ 225 Ac]Ac-SibuDAB should be defined in a carefully designed clinical dose-escalation study.

Patents
A patent application on PSMA ligands with ibuprofen as an albumin-binding entity has been filed by ITM Medical Isotopes SE, Munich, Germany.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.