Preclinical comparison of [177Lu]Lu-rhPSMA-10.1 and [177Lu]Lu-rhPSMA-10.2 for endoradiotherapy of prostate cancer: biodistribution and dosimetry studies

Background Radiohybrid PSMA-targeted ligands (rhPSMA) have been introduced as a novel platform for theranostic applications. Among a variety of rhPSMA-ligands developed for radioligand therapy, two stereoisomers [177Lu]Lu-rhPSMA-10.1 and -10.2 have been synthesized and initially characterized in preclinical experiments with the aim to provide an optimized binding profile to human serum albumin, a reduction of charge, and thus accelerated kidney excretion, and unaffected or even improved tumor uptake. As both isomers showed similar in vitro characteristics and tumor uptake at 24 h post injection in tumor bearing mice and in order to identify the isomer with the most favorable pharmacokinetics for radioligand therapy, we carried out in-depth biodistribution and dosimetry studies in tumor-bearing and healthy mice. Results rhPSMA-10.1 and -10.2 were radiolabeled with lutetium-177 according to the established procedures of other DOTA-based PSMA ligands and displayed a high and comparable stability in all buffers and human serum (> 97%, 24 h). Biodistribution studies revealed fast clearance from the blood pool (0.3–0.6%ID/g at 1 h) and other background tissues within 48 h. Distinctive differences were found in the kidneys, where [177Lu]Lu-rhPSMA-10.1 displayed lower initial uptake and faster excretion kinetics compared to [177Lu]Lu-rhPSMA-10.2 expressed by a 1.5-fold and ninefold lower uptake value at 1 h and 24 h in healthy animals, respectively. Tumor uptake was comparable and in the range of 8.6–11.6%ID/g for both isomers over 24 h and was maintained up to 168 h at a level of 2.2 ± 0.8 and 4.1 ± 1.4%ID/g for [177Lu]Lu-rhPSMA-10.1 and [177Lu]Lu-rhPSMA-10.2, respectively. Conclusion Our preclinical data on biodistribution and dosimetry indicate a more favorable profile of [177Lu]Lu-rhPSMA-10.1 compared to [177Lu]Lu-rhPSMA-10.2 for PSMA-targeted radioligand therapy. [177Lu]Lu-rhPSMA-10.1 shows fast kidney clearance kinetics resulting in excellent tumor-to-organ ratios over a therapy relevant time course. Meanwhile, [177Lu]Lu-rhPSMA-10.1 is currently being investigated in clinical phase I/II studies in patients with mCRPC (NCT05413850), in patients with high-risk localized PC (NCT06066437, Nautilus Trial) and after external beam radiotherapy (NCT06105918). Supplementary Information The online version contains supplementary material available at 10.1186/s41181-024-00246-2.

Recently radiohybrid PSMA-targeted ligands (rhPSMA) have been introduced as a novel platform with potentially improved characteristics for theranostic applications (Wurzer et al. 2020a).rhPSMA radiopharmaceuticals combine a silicon-fluoride acceptor for easy and fast 18 F-labeling and a chelator for complexation of a (radio) metal in one molecule.Respective ligand pairs of " 18 F/non-radioactive metal" and " 19 F/radiometal" are chemically identical and thus display identical pharmacokinetics, allowing their application for 18 F-based PET-imaging and radiotherapy in a truly theranostic manner (Wurzer et al. 2020a, b).As a very first rh-ligand, [ 18 F] Ga-rhPSMA-7.3(POSLUMA ® , flotufolastat F18, 18 F-rhPSMA-7.3)was recently FDAapproved for imaging of patients with suspected prostate cancer recurrence (Jani et al. 2023;Surasi et al. 2023).
Preclinical analyses established [ 177 Lu]Lu-rhPSMA-7.3as a potential alternative to [ 177 Lu]Lu-PSMA-I&T which has similar clearance kinetics and a similar radiation dose to healthy organs but superior tumor uptake and retention (Yusufi et al. 2021).In a small comparative clinical investigation, these differences were less pronounced (Feuerecker et al. 2021).With the aim of faster clearance from healthy organs, further investigations ultimately led to the development of the stereoisomeric pair, rhPSMA-10: rhPSMA-10.1 with a D-Dap (diaminopropionic acid) and rhPSMA-10.2with L-Dap branching unit, respectively (Fig. 1).Both ligands provide, reduced overall charge and lower binding strength to human serum albumin (HSA) potentially enhancing clearance from healthy organs (Wurzer et al. 2022).After initial preclinical studies, rhPSMA-10.1 was prioritized for further development of an improved theranostic candidate (Wurzer et al. 2022).
In the present preclinical study, we investigate the effect of stereoisomerism on the pharmacokinetics of [ 177 Lu]Lu-rhPSMA-10.1 and [ 177 Lu]Lu-rhPSMA-10.2 in healthy and PSMA-positive LNCaP-tumor bearing mice in order to identify the rhPSMA-10 isomer with the most favorable characteristics for radioligand therapy.

General information
rhPSMA-10.1 and rhPSMA-10.2were obtained by Almac Sciences Scotland (Penicuik, UK).All reagents were purchased from Sigma-Aldrich (Merck Group, St. Louis, Missouri, United States).Phosphate buffered saline (PBS) pH 7.4 was prepared by the Hospital Pharmacy of the University Hospital rechts der Isar according to the following recipe: sodium chloride 66 mM, potassium chloride 1.3 mM, sodium hydrogenphospate dihydrate 4.0 mM, and potassium hydrogen phosphate 0.6 mM.The 0.9% NaCl solution was as well prepared by the Hospital Pharmacy.Human serum was obtained from PAN Biotech (Aidenbach, Regensburg, Germany).

Cell culture
The human prostate cancer cell line LNCaP (DSMZ German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany) was used for the establishment of tumor xenograft models.Cells were routinely screened for mycoplasma and authenticity was confirmed by short tandem repeat analysis.The cells were maintained as monolayer cultures in RPMI (Gibco, Carlsbad, US) containing 10% fetal bovine serum and Penicillin-Streptomycin (Gibco, Carlsbad, US) at 37 °C in a humidified CO 2 atmosphere (5%).

Animals and tumor xenograft model
Six-week old male CB17 SCID (severe combined immunodeficiency, Charles River Laboratories) mice were used for the animal experiments according to guidelines for the welfare and use of animals in cancer research experimentation.All experiments were approved by local authorities (animal license 55. 2-1-54-2532-216-2015), and mice were maintained in the animal facility within the Department of Nuclear Medicine at the University Hospital rechts der Isar according to institutional guidelines.For the tumor implantation, mice were anesthetized under isoflurane flow 3-5% (v/v).Each animal was injected subcutaneously with 5 × 10 6 LNCaP cells in 100 µl serum-free RPMI medium mixed with 100 µl matrigel (BD Biosciences, Germany) in the right shoulder region.Animals with tumors above a diameter of 0.5 cm were used for the experiments.
Radiolabeling procedures and quality controls rhPSMA-10.1 and rhPSMA-10.2were radiolabeled with non-carrier-added lutetium-177 according to established procedures (Benesova et al. 2015;Weineisen et al. 2015).Briefly, 12.5 nmol of the precursor was dissolved in DMSO to a concentration of 1 mg/mL and filled up to 400 μL using 0.4 M acetate buffer and gentisic acid (10 mg/mL) at pH 5.5.500 MBq of [ 177 Lu]LuCl 3 in 0.04 M HCl (ITM Medical Isotope GmbH, Garching, Munich, Germany) were added to this solution and incubated at 90 °C for 30 min.
Radionuclide incorporation (RNI) was determined via radio-thin layer chromatography (TLC), using glass microfiber paper impregnated with silicic acid (Agilent Technologies, Santa Clara, US) and 0.1 M sodium citrate at pH 5.0 as mobile phase.Samples were read-out on the MiniScan scanner from Bioscan (Eckert and Ziegler, Brussels, Belgium).Radiochemical purity (RCP) was further evaluated via radio high-performance liquid chromatography (HPLC) on a Prominence system with a Photo Diode Array detector (Shimadzu, Kyoto, Japan) and a GABI Star detector (Raytest, Straubenhardt, Germany).Eluents for all HPLC operations were water (solvent A) and acetonitrile (solvent B), both containing 0.1% trifluoroacetic acid.An XTerra MS C18 OBD column (Waters, Germany) was used with a linear gradient of 15-95% B in 20 min, followed by 95% B for 10 min.

Biodistribution studies
For each radiotherapeutic agent two cohorts of animals were used, one with nontumor-bearing SCID mice (n = 2, each time point) and one with LNCaP tumor-bearing SCID mice (n = 4-5, each time point).Mice were anesthetized by isoflurane inhalation and maintained under continuous isoflurane anesthesia (1.75-3% v/v), while 1.6 MBq (~ 40 pmol) of radioligand were injected in the tail vein using a catheter in a total volume of 200 µl saline solution.Non-tumor bearing animals were sacrificed at 1, 12, 24, 48, and 168 h post injection (p.i.), while the tumor-bearing animals were sacrificed at 1, 24, and 168 h p.i. Tumor, blood, and other selected tissues were excised, collected, weighted, and then measured in a gamma-counter (2480 WIZARD2, PerkinElmer, Waltham, US).Radioactivity uptake was calculated as a percentage of the injected dose per gram of tissue (%ID/g) using a 1% (v/v) standard of the injected activity.

Dosimetry calculations
Five time points were employed to calculate the Time-Integrated Activity Coefficients (TIAC) for each radiopharmaceutical.The activity accumulation in significant source organs was determined using both numerical integration and physical decay, as per Yuan et al. (1993).We estimated radiation doses to normal organs in a standard 70 kg adult model by using time-dependent organ activity concentrations (expressed as %ID/g) and total-body activities.Tissue activity concentrations in mice were converted to fractional activities in the standard 70 kg adult by considering the relative organ masses between the standard adult and the 25 g mouse.The time-dependent total-body activity was modelled using an exponential function and assumed that the difference between the injected activity and total-body activity was excreted through urine, as liver and gastrointestinal tract activity concentrations remained consistently low throughout.Residence times were computed as the difference between the total-body residence time and the sum of the organ and urine residence times.Finally, the absorbed doses (in mGy/MBq) in organs of a standard adult were calculated using OLINDA/EXM1.0(Stabin et al. 2005).For estimating tumor doses the unit density sphere model described by Stabin and Konijnenberg (2000) was utilized.

Statistics
All data are presented as mean ± standard deviation (SD).Statistical analyses were performed with GraphPad Prism 4.0 software using Student's t-test for unpaired data.Twosided significance levels were calculated and P < 0.05 values were considered statistically significant.

Radiochemistry and stability
After radiolabeling of the rhPSMA-10 isomers according to the previously published procedures for DOTA-conjugated ligands (Benesova et al. 2015;Weineisen et al. 2015), analysis via radio-TLC revealed a high RNI of > 99.5%.RCP as determined by radio-HPLC was 94.2 ± 0.2% for [ 177 Lu]Lu-rhPSMA-10.1 and 94.5 ± 0.2 [ 177 Lu]Lu-rhPSMA-10.2 with retention times of 14.2 min and 13.8 min, respectively.A hydrophilic impurity eluting approximately 2.3 min before the product was identified as the main side product (relative area under the curve: 5%).Representative HPLC and TLC chromatograms have been included into the supporting information (Additional file 1: Figures S1 and S2).Both ligands were produced in a molar activity of 38.0 GBq/µmol.

Discussion
The PSMA binding pocket is reported to tolerate a variety of large structural modifications, e.g., fluorescent dyes (Derks et al. 2019), albumin binding groups (Kuo et al. 2021;Deberle et al. 2020) or silicon-fluoride acceptors (Wurzer et al. 2020a).However, we previously demonstrated that the stereoconfiguration of the linker has a strong influence on the PSMA binding characteristics and in vivo behavior (Wurzer et al. 2020b).
The subsequent development of a 177 Lu-labeled rhPSMA ligand for radioligand therapy aimed to accelerate the clearance from the kidneys and maintain the tumor uptake (Wurzer et al. 2022).These efforts resulted in the development of rhPSMA-10.1 and -10.2, two DOTA-based analogues of rhPSMA-7 with a lower overall negative charge and a lower binding strength to HSA compared with the former lead.In order to carefully compare the in vivo behavior of [ 177 Lu]Lu-rhPSMA-10.1 and -10.2 we performed the present biodistribution study in healthy and tumor-bearing mice over a 168 h time course.The injected mass of each isomer (0.04 nmol, 0.06 µg for a 25 g mouse) corresponds to a typical dose of [ 177 Lu]Lu-PSMA-617 (150-250 µg, 144-240 nmol, 75 kg patient), administered during the VISION trial (Sartor et al. 2021).
After radiolabeling of rhPSMA-10.1 and -10.2 with lutetium-177, formation of a hydrophilic impurity (relative area: 5%) was observed by radio-HPLC.In the context of clinical development, the hydrophilic impurity was identified as the hydrolyzed silicon-fluoride acceptor (SiOH), whereupon an optimized radiolabeling procedure was developed, resulting in an improved RCP ≥ 97%.In different formulations a comparable and high stability was found for both isomers over 24 h, allowing their centralized production and distribution.The reason for the different stability at later time points warrants further investigation in future studies.In the comparative biodistribution studies, the two 177 Lu-labeled isomers displayed the typical distribution pattern of PSMAtargeted radioligands, with fast clearance kinetics from the blood pool and background tissues and high tumor uptake.The high uptake in murine spleen at 1 h p.i., has also been observed previously for a variety of other PSMA-targeted ligands and was found to be blockable by excess of a competitor (Benesova et al. 2015;Schottelius et al. 2019).
In PSMA-targeted radioligand therapy, the kidney and the bone marrow are still considered the main organs at risk, and the uptake in these tissues should be carefully considered (Yordanova et al. 2017).Because of the increasing interest in using [ 177 Lu] Lu-PSMA ligands in early disease stages (e.g.PSMAfore and PSMAddition trials), especially the tumor-to-kidney ratio is one of the most important selection criteria for novel radioligands in order to reduce potential renal side effects in men with several years of life expectancy.Despite the structural similarity of both ligands, [ 177 Lu]Lu-rhPSMA-10.1 was found to display significantly lower initial accumulation in the kidneys at 1 h p.i. and faster excretion kinetics, compared with the L-Dap-based [ 177 Lu]Lu-rhPSMA-10.2.This is demonstrated by the ninefold and fivefold lower kidney uptake at 24 h p.i. in non-tumor-and tumor-bearing mice, respectively (Fig. 5A, B).Our findings support the results from the previously performed in vivo single-time point comparison of [ 177 Lu]Lu-rhPSMA-10.1 and -10.2, which was conducted at 24 h p.i. (Wurzer et al. 2022).Here, [ 177 Lu]Lu-rhPSMA-10.1 showed a fourfold lower accumulation in the kidneys, compared to [ 177 Lu] Lu-rhPSMA-10.2,while tumor uptake was comparable (9.8 ± 0.3 vs. 10.5 ± 3.3%ID/g for 10.1 vs. 10.2) (Wurzer et al. 2022).
While no statistically significant difference in the tumor uptake was found at early time points for both radiopharmaceuticals, [ 177 Lu]Lu-rhPSMA-10.2showed improved tumor retention, expressed by an approximately twofold higher uptake after 168 h p.i. Due to the fast kidney clearance of [ 177 Lu]Lu-rhPSMA-10.1, the tumor-to-kidney values were found to be superior for [ 177 Lu]Lu-rhPSMA-10.1 after 24 h and similar for both ligands after 168 h.In the previous in vitro characterization of [ 177 Lu]Lu-rhPSMA-10.1 versus -10.2, both displayed a similar PSMA binding affinity (2.8 nM vs. 3.6 nM) and internalization rate (177% vs. 206%; expressed as a percentage of the reference ligand [ 125 I]IBA-KuE) (Wurzer et al. 2022).Moreover, the isomers were found to have an identical lipophilicity (logP: − 3.8 for both; distribution coefficient in octanol and PBS pH 7.4), and binding strength to HSA was in a similar range (Wurzer et al. 2022).Based on these characteristics, it remains challenging to explain the observed differences in the kidney accumulation of both ligands and the higher tumor retention of [ 177 Lu] Lu-rhPSMA-10.2 at 168 h p.i.. Given that both isomers exhibited comparable complex stability in human serum over a period of 168 h, the observed differences in biodistribution are likely attributed to distinct pharmacokinetics rather than different amounts of unbound lutetium-177.
These findings elucidate once again that there are potentially other parameters, influencing the biodistribution of radioligands which are currently not assessed in routine screening experiments, like for example differences in magnitude and binding affinities to plasma proteins other than HSA, such as α-1-acid glycoprotein (Smith and Waters 2018), transthyretin (Buxbaum and Reixach 2009), or lipoproteins (Wasan et al. 2008).Moreover, species differences between mouse serum albumin and HSA need to be considered (Roopenian et al. 2015).
In a recently performed preclinical biodistribution study of [ 177 Lu]Lu-PSMA I&T in the same xenograft models, the reference shows a similar clearance from background tissues but slower renal clearance kinetics, compared to the radiohybrid ligands in the present study (Yusufi et al. 2021).While initial kidney uptake is in a similar range at 1 h p.i. (166-224%ID/g), the accelerated clearance kinetics of radiohybrid ligands result in a 26-fold and fivefold lower kidney uptake at 24 h p.i. for [ 177 Lu]Lu-rhPSMA-10.1 and -10.2, respectively, compared to the reference.The tumor uptake of both isomers is superior at all time points, compared to [ 177 Lu]Lu-PSMA I&T: 4.4 ± 1.5%ID/g at 1 h, 6.2 ± 0.1%ID/g at 24 h and 1.0 ± 0.2%ID/g at 168 h.Despite the promising pharmacokinetics of [ 177 Lu]Lu-rhPSMA-10.1, it is in general questionable to what extent these data, especially murine kidney uptake, is transferable to humans.Data reported for the recently FDA-approved [ 177 Lu]Lu-PSMA-617 perfectly illustrate the pitfalls of murine kidney uptake associated with PSMA-ligands.While [ 177 Lu]Lu-PSMA-617 showed a 25-fold lower preclinical kidney uptake compared with [ 177 Lu]Lu-PSMA I&T (1.4 ± 0.4 vs. 34.7 ± 17.2%ID/g at 24 h p.i.) (Wurzer et al. 2022), Baum et al. impressively demonstrated similar renal half-lives, resulting in nearly identical absorbed radiation doses (0.8 vs. 0.9 Gy/GBq for 617 vs. I&T) in patients (Schuchardt et al. 2022).Further evaluations will be required to fully understand species differences of PSMA-targeted ligands.Until the availability of improved preclinical methods to predict human pharmacokinetics more accurately, especially in organs at risk, murine uptake values are still the most important selection criteria for clinical translation of novel radiopharmaceuticals. Data from ongoing clinical investigations will be valuable to retrospectively judge the results obtained in preclinical studies in mice.

Conclusion
Our preclinical data on biodistribution and dosimetry indicate a more favorable profile of [ 177 Lu]Lu-rhPSMA-10.1 compared to [ 177 Lu]Lu-rhPSMA-10.2 for PSMA targeting RLT.[ 177 Lu]Lu-rhPSMA-10.1 shows fast kidney clearance kinetics resulting in excellent tumor-to-organ ratios over a therapy relevant time course.[ 177 Lu]Lu-rhPSMA-10.1 is currently being investigated in clinical phase I/II studies in patients with mCRPC (NCT05413850), in patients with high-risk localized PC (NCT06066437, Nautilus Trial) and after external beam radiotherapy (NCT06105918).