Assessment of quality control and effect of radiolysis on the in vivo biodistribution of [177Lu]Lu-PSMA-I&T Authors: Eline L. Hooijman, Carolline M. Ntihabose, Stijn L. W. Koolen, Maryana Handula, Savanne Beekman, Debra Stuurman, Corrina de Ridder, Jan R. de Jong, Jordy R. J. P. van de Merbel, Nyah Rook, Else A. Aalbersberg, Tessa F. F. T. Ververs, Peter Laverman, Jeroen J. M. A. Hendrikx, Erik de Blois Published: 2025 ABSTRACT BackgroundOne of the challenges in optimizing PSMA-based therapies for the treatment of prostate cancer is the stability of the radiopharmaceutical, defined by the degradation of radiolabeled compounds due to exposure to ionizing radiation (radiolysis). Radiolysis of a radiopharmaceutical is difficult to measure and can influence the biodistrubion of radiopharmaceuticals in vivo, possibly leading to toxicity or suboptimal outcomes. Therefore, an inter-laboratory study was performed to confirm accurate detection of radiolysed [177Lu]Lu-PSMA-I&T, and the biodistribution of radiolysed radiopharmaceutical was investigated in vivo, and HPLC methods are compared.ResultsMultiple HPLC-methods were compared to evaluate the analysis of degraded radiopharmaceutical and an inter-laboratory comparison was conducted to evaluate the consistency of current analytical methods across various institutions and with current methods from literature. Additionally, for the biodistribution experiments, a radiolabeling of [177Lu]Lu-PSMA-I&T was adjusted to achieve a radiochemical purity of 50%, 70%, and 97% of the final product. A PC3-PIP patient-derived xenograft mouse model was used to investigate in vivo behavior of the formed impurity, followed by SPECT/CT imaging and biodistribution. When evaluating radio-HPLC methods for the impurity detection, the European Pharmacopoeia method (with a phosphate buffered eluent) shows no significant difference in radiochemical purity, an increased peak separation but decreased recovery (< 75%). The inter-laboratory comparison demonstrated the ability to measure the radiolysed radiopharmaceutical across multiple hospitals, with a maximum standard deviation of 6.6%. The in vivo SPECT/CT and biodistribution results show that as ingrowth of impurities caused by radiolysis increasing the tumor-to-kidney ratio is decreasing (97%/50% RCP, p < 0.05).ConclusionThe inter-laboratory study confirms that impurities caused by radiolysis can be measured accurately across different institutions, ensuring consistency and reliability in clinical diagnostics and therapy, contributing to improved patient safety. Phosphate buffered HPLC-methods showed no significant benefits in quality control. Increased impurities caused by radiolysis lead to lower tumor-to-kidney ratios, with a decrease in specific tumor binding.Supplementary InformationThe online version contains supplementary material available at 10.1186/s41181-025-00414-y. Introduction Prostate cancer (PCa) is globally one of the most prevalent cancers among men, with significant implications for mortality and morbidity (Merrill 2024). Over the past few decades, advancements in diagnosis and treatment have been critical in managing the disease (Posdzich et al. 2023; Le et al. 2023). For example, prostate-specific membrane antigen (PSMA) is becoming increasingly recognized and utilized in the field of nuclear medicine, particularly in the treatment of metastatic castration-resistant prostate cancer (mCRPC) (Golan et al. 2023). The ability to target PSMA has resulted in both imaging (diagnostics) and therapy of PCa, offering a high degree of specificity (Chatalic et al. 2016). PSMA ligands can be radiolabeled with various radioisotopes, such as gallium-68 (Ga-68) or lutetium-177 (Lu-177), to serve either as diagnostic or therapeutic purposes. When labeled with Lu-177, the radiopharmaceutical can deliver beta-radiation to PCa cells, thereby sparing surrounding healthy tissues. Although uptake in the salivary glands and intestinal organs can be observed, this approach has shown promising results, and the use of PSMA-targeting radiopharmaceuticals is increasing (Sadaghiani et al. 2022a). During the production and administration of radiopharmaceuticals, particularly those labeled with therapeutic isotopes such as Lu-177, ionizing radiation can alter the chemical composition of the compound by a process defined as radiolysis (Hooijman et al. 2021). Among the PSMA-targeting radiopharmaceuticals, [177Lu]Lu-PSMA-I&T ([177Lu]Lu-zadavotide gurexetan) and [177Lu]Lu-PSMA-617 ([177Lu]Lu-vipivotide tetraxetan) are frequently compared. Although [177Lu]Lu-PSMA-I&T shows promising results, it appears to be more sensitive to degradation, compared to the [177Lu]Lu-PSMA-617 (Ruigrok et al. 2021). In regards to the degradation products, the monograph of [177Lu]Lu-PSMA-I&T published in the European Pharmacopoeia (Ph. Eur.) (becoming in place in version 11.7 in 2025) specify limits for each impurity (Europe E.E.P.S.C 2025). However, the degradation products are not identified or further specified. Within this research, the radiopharmaceutical [177Lu]Lu-PSMA-I&T is considered the intact product and the impurities are defined as the free Lu-177/[177Lu]Lu-DTPA and radiolysed species of [177Lu]Lu-PSMA-I&T. Presently, among the release criteria of radiopharmaceuticals, two radiochemical criteria are monitored namely the radiochemical yield (RCY) and radiochemical purity (RCP), by thin layer chromatography (radio-TLC) and high performance liquid chromatography (HPLC) analysis, respectively (Blois et al. 2012). Understanding the influence and the effects of radiolysis is important, as the radiolysed radiopharmaceutical may show an altered biodistribution and might influence the treatment efficacy (Hooijman et al. 2022). Previously, studies showed that induced radiolysed radiopharmaceutical in cell experiments, result in altered cellular uptake and kinetics (Hooijman et al. 2021). These changes can potentially impact the therapeutic outcomes and may increase the risk of adverse effects, making it essential to ensure that the radiopharmaceuticals administered to patients, are of the highest quality and stability (Kratochwil et al. 2019; Sadaghiani et al. 2022b; Kalmthout et al. 2020). To gain insight into the impact of radiolysis on [177Lu]Lu-PSMA-I&T, an inter-institutional comparison was first carried out to explore potential variability in analytical outcomes between laboratories. This is followed by a comparison of multiple HPLC methods to evaluate which approach most reliably detects the degradation by radiolysis and their impurities. Beyond assessing analytical performance, it is also relevant to explore whether undetected impurities might influence the biological behavior of the product. Therefore, the biodistribution of radiolysed [177Lu]Lu-PSMA-I&T was investigated as a follow-up to previously published in vitro studies (Hooijman et al. 2022). To further investigate the impact of radiolysis on [177Lu]Lu-PSMA-I&T, it is important to consider the biological behavior of the radiolysed [177Lu]Lu-PSMA-I&T, when introduced into in vivo models. If the biodistribution of those impurities are altered, it is of high importance that the analytical methods can detect these impurities. This research is focused on whether the impurities caused by radiolysis can be effectively separated and quantified by the analytical methods. Materials and methods Inter-laboratory comparison The interlaboratory comparison involved 4 Dutch institutes that routinely prepare [177Lu]Lu-PSMA-I&T as an in-house preparation. All HPLC-systems and methods of the collaborating institutes were validated according to the EANM guidelines (Kratochwil et al. 2019; Gillings et al. 2020). A summary of the validation results and overview of the used HPLC equipment is presented in Table S1 and S2 of supplementary material. Each institute reported their HPLC-results of the samples, and corresponding chromatograms were sent to the comparative center. HPLC-methods inter-laboratory comparison The HPLC-analyses were performed within each center, and Table 1 describes the gradient and standardized sample parameters that were applied. Figure 1 gives an outline of the applied methods and gradients, further information is also available in Table 1 & S4 of the supplementary material.Table 1Interlaboratory HPLC-conditions and gradients of the collaborating centersMobile phaseGradient*HPLC sample preparationComparative center (1)(A) 95% 0.1% TFA/water with 5% 0.1% TFA/ACN(B) 95% 0.1% TFA/ACN with 5% 0.1% TFA/water0–2 min 100% A, 2–10 min 20% A to 80% B, 10–15 min 100% B, 15–20 min 100% A6 MBq/100 μL, diluted in 25% EtOH and co-injection of 10 µg PSMA-I&TCenter 2(A) 0.1% TFA/water(B) 0.1% TFA/ACN0–16 min 85–65% A, 16–20 min to 85% A10 MBq/20 μL with 7.5 µg PSMA-I&T co-injection solution (Aalbersberg et al. 2022)Center 30–12.5 min 100% A to 100% B, 12,5–18 min 100% B, 18–20 min 100% A10 MBq/20 μL, diluted in 0.9% NaClCenter 40–5 min 5% B, 5–15 min 95% B, 15–20 min 95% B, 20–25 min 5% B, 25–30 min 5% B5 MBq/10 μL undiluted sample*The flow is 1 mL/min for the comparative center, center 3 and 4, for center 2 it is 1.5 mL/minFig. 1HPLC-gradients applied by different centers (a-d), expressed as percentage of mobile phase B Table 1Interlaboratory HPLC-conditions and gradients of the collaborating centersMobile phaseGradient*HPLC sample preparationComparative center (1)(A) 95% 0.1% TFA/water with 5% 0.1% TFA/ACN(B) 95% 0.1% TFA/ACN with 5% 0.1% TFA/water0–2 min 100% A, 2–10 min 20% A to 80% B, 10–15 min 100% B, 15–20 min 100% A6 MBq/100 μL, diluted in 25% EtOH and co-injection of 10 µg PSMA-I&TCenter 2(A) 0.1% TFA/water(B) 0.1% TFA/ACN0–16 min 85–65% A, 16–20 min to 85% A10 MBq/20 μL with 7.5 µg PSMA-I&T co-injection solution (Aalbersberg et al. 2022)Center 30–12.5 min 100% A to 100% B, 12,5–18 min 100% B, 18–20 min 100% A10 MBq/20 μL, diluted in 0.9% NaClCenter 40–5 min 5% B, 5–15 min 95% B, 15–20 min 95% B, 20–25 min 5% B, 25–30 min 5% B5 MBq/10 μL undiluted sample*The flow is 1 mL/min for the comparative center, center 3 and 4, for center 2 it is 1.5 mL/min Interlaboratory HPLC-conditions and gradients of the collaborating centers (A) 95% 0.1% TFA/water with 5% 0.1% TFA/ACN (B) 95% 0.1% TFA/ACN with 5% 0.1% TFA/water (A) 0.1% TFA/water (B) 0.1% TFA/ACN *The flow is 1 mL/min for the comparative center, center 3 and 4, for center 2 it is 1.5 mL/min HPLC-gradients applied by different centers (a-d), expressed as percentage of mobile phase B Samples for inter-laboratory comparison Each institute received a sample of [177Lu]Lu-PSMA-I&T from the different supplying centers, and each institute has served as a supplying institute (except center 4). The sample originated from a [177Lu]Lu-PSMA-I&T batch administered within the validated shelf-life. The RCY (> 97%) and RCP (> 95%) were confirmed using validated methods. A sample of ± 300 MBq was split into four batches (± 70 MBq each) and stored in a 1.5 mL maximum recovery HPLC-vial (Waters, Etten-Leur, The Netherlands). Each institute performed an HPLC-injection at 24(± 1) and 48(± 2) hours after radiolabeling. These time points that are dependent on the institutes and after expiration for clinical use, were intentionally selected to ensure that degradation by radiolysis has occurred, allowing for a critical evaluation. Statistical analysis The evaluation of the obtained results from the imaging and biodistribution studies was first performed using a Grubbs' outlier test to identify and exclude the presence of outliers (no outliers present). Subsequently, a two-tailed t-test was conducted to compare the data. To evaluate the HPLC-data, the standard deviation was calculated to measure the dispersion of data points from the mean. Additionally, the relative standard deviation (coefficient of variation) was computed to assess the variability of the data relative to the mean, expressed as a percentage. All data is processed using GraphPad version 9.0.0. HPLC processing methods After integration the chromatograms were reintegrated by 1 trained operator from the comparative center to assess operator variability in comparison to the original analysis performed by the institutes themselves. The parameters were manually selected on background, selection in region of interest of the expected peaks and selection of the main peak. Each of the initial measurements were compared with the reintegration of the chromatograms by the trained operator from the comparative center, thereby determining the mean and the standard deviation. Center 2, which utilized standardized processing methods (Empower, Waters), was excluded from the variation analysis due to the uniform analysis conditions. Comparison of HPLC method(s) Various HPLC-methods for analysis of [177Lu]Lu-PSMA-I&T are described in literature (Hooijman et al. 2022; Aalbersberg et al. 2022; Kraihammer et al. 2023). One approach involves the use of phosphate buffers as an eluent, in contrast to the more commonly employed TFA-based systems and is incorporated into the Ph. Eur. monograph that is published in 2025 (Europe, E.E.P.S.C. 2025). These methods were also compared to the comparative method (Table 2) including recovery measurements (Gillings et al. 2020). A standard activity of 3 MBq/100 µL was injected and for all methods a flow of 1 mL/min was used. For comparison of the radiolysed impurity detection, samples of [177Lu]Lu-PSMA-I&T were injected at 24 (± 1 h) and 48 (± 2 h) hours after synthesis to ensure the presence of radiolysed radiopharmaceutical. Additionally, the recovery and carry-over were also determined for both methods (Hooijman et al. 2024) on the same HPLC system.Table 2Overview of applied HPLC-methods for literature comparison, including used columns, analytical parameters and sample preparation details at the comparative centerColumnMobile phaseGradientHPLC sample preparationComparativemethodLiChrospher® 100 RP-18 end-capped column(250 mm × 4 mm, 5 µm)(A) 95% 0.1% TFA + 5% 0.1% TFA in ACN(B) 95% 0.1% TFA ACN + 5% 0.1% TFA in water0–2 min 100% A, 2–10 min 20% A to 80% B, 10–15 min 100% B, 15–20 min 100% ASample diluted in 25% EtOH and co-injection of 10 µg PSMA-I&TMethod APhenomenex Jupiter Proteo C18 column(250 mm × 4.6 mm, 4 µm,)(A) 0.03 M phosphate buffer, pH 2.3(B) ACN0–2 min: 23% B, 2–10 min: 23–30% B, 10–12 min: 30–60% B, 12–14 min: 60% B, 14–18 min: 23% B80% A/20% B + non-radioactive Lu-PSMA-I&T standard*Method BPhenomenex Kinetex C18 column(150 mm × 4.6 mm, 5 µm)0–2 min: 20% B, 2–10 min: 20–25% B, 10–12 min: 25–80% B, 12–14 min: 80% B, 14–18 min 20% B*ICP [175Lu]Lu-PSMA-I&T standard, 100 µg/mL, UV confirmed Table 2Overview of applied HPLC-methods for literature comparison, including used columns, analytical parameters and sample preparation details at the comparative centerColumnMobile phaseGradientHPLC sample preparationComparativemethodLiChrospher® 100 RP-18 end-capped column(250 mm × 4 mm, 5 µm)(A) 95% 0.1% TFA + 5% 0.1% TFA in ACN(B) 95% 0.1% TFA ACN + 5% 0.1% TFA in water0–2 min 100% A, 2–10 min 20% A to 80% B, 10–15 min 100% B, 15–20 min 100% ASample diluted in 25% EtOH and co-injection of 10 µg PSMA-I&TMethod APhenomenex Jupiter Proteo C18 column(250 mm × 4.6 mm, 4 µm,)(A) 0.03 M phosphate buffer, pH 2.3(B) ACN0–2 min: 23% B, 2–10 min: 23–30% B, 10–12 min: 30–60% B, 12–14 min: 60% B, 14–18 min: 23% B80% A/20% B + non-radioactive Lu-PSMA-I&T standard*Method BPhenomenex Kinetex C18 column(150 mm × 4.6 mm, 5 µm)0–2 min: 20% B, 2–10 min: 20–25% B, 10–12 min: 25–80% B, 12–14 min: 80% B, 14–18 min 20% B*ICP [175Lu]Lu-PSMA-I&T standard, 100 µg/mL, UV confirmed Overview of applied HPLC-methods for literature comparison, including used columns, analytical parameters and sample preparation details at the comparative center Comparative method LiChrospher® 100 RP-18 end-capped column (250 mm × 4 mm, 5 µm) (A) 95% 0.1% TFA + 5% 0.1% TFA in ACN (B) 95% 0.1% TFA ACN + 5% 0.1% TFA in water Phenomenex Jupiter Proteo C18 column (250 mm × 4.6 mm, 4 µm,) (A) 0.03 M phosphate buffer, pH 2.3 (B) ACN Phenomenex Kinetex C18 column (150 mm × 4.6 mm, 5 µm) *ICP [175Lu]Lu-PSMA-I&T standard, 100 µg/mL, UV confirmed Biodistribution of radiolysed [177Lu]Lu-PSMA-I&T in mice Radiolabeling The radiolabeling conditions of PSMA-I&T with Lu-177 at a clinical scale (30 MBq/nmol) as previously described (Hooijman et al. 2022), were adjusted to be able to obtain increased radiolysed [177Lu]Lu-PSMA-I&T product by reducing the amount of added quencher (radical scavenger) and/or increasing exposure time to allow for radiolysis to form. The quality control (QC) was performed using radio-TLC (RCY) and HPLC-analysis (RCP), as previously described (Hooijman et al. 2022). The three solutions with decreased RCP’s (RCY > 97%, radio-TLC confirmed) were stabilized after the radiochemical purity was confirmed by HPLC with an excess of quencher(s) (Table 3) to a final volume of 60 μL.Table 3Labeling conditions of [177Lu]Lu-PSMA-I&T for obtaining a decreased RCPRadiochemical purity (%)Excipients added during labellingWaiting time after labellingExcipients added after heating500.01 M sodium acetate2.5 hExcess DTPA (50x)1 mM gentisic acid3.5 mM ascorbate70No waiting step involved971 M sodium acetate0.1 mM gentisic acid0.4 mM ascorbate Table 3Labeling conditions of [177Lu]Lu-PSMA-I&T for obtaining a decreased RCPRadiochemical purity (%)Excipients added during labellingWaiting time after labellingExcipients added after heating500.01 M sodium acetate2.5 hExcess DTPA (50x)1 mM gentisic acid3.5 mM ascorbate70No waiting step involved971 M sodium acetate0.1 mM gentisic acid0.4 mM ascorbate Labeling conditions of [177Lu]Lu-PSMA-I&T for obtaining a decreased RCP Excess DTPA (50x) 1 mM gentisic acid 3.5 mM ascorbate 1 M sodium acetate 0.1 mM gentisic acid 0.4 mM ascorbate Imaging and biodistribution studies All animal experiments were conducted with the approval of the Erasmus Medical Center’s animal welfare committee, and adhered to the established ethical guidelines (application number SP2300235/1). The experiments are based upon studies performed by Ruigrok et al. (2021). The animal experiments were performed with 6–8 week old male mice, with a PC3-PIP PDX xenograft (prostate cancer PSMA-positive cell line, Uppsala University, Sweden). The cells for the xenograft (PC3-PIP) were cultured in RPMI 1640, Glutamax medium, supplemented with 10% fetal calf serum, penicillin (100 units/mL) and streptomycin (100 units/mL, Gibco), and in presence of puromycin (10 µg/mL) (Invivogen). The cells were cultured at 37 °C with 5% CO2. Thereafter, NMRI nu/nu -nude mice were subcutaneously inoculated with 5 × 106 PSMA-expressing PC3-PIP cells suspended in Matrigel (Corning, NY, USA). The tumors were grown for 10 days, according to pilot-data to an average tumor volume of 280 mm3 (Ruigrok et al. 2021). After reaching the desired tumor volume, the animals were injected in the tail vein with either 20 MBq/0.7 nmol/100 µL (imaging) or 5 MBq/0.7 nmol/100 µL (biodistribution experiments) of [177Lu]Lu-PSMA-I&T (RCP 50%, 70% or 97%) (diluted in phosphate buffered saline (PBS) with Kolliphor® HS 15 (2.0 mg/mL)) which is added to prevent sticking of the radiopharmaceuticsl. SPECT/CT imaging The imaging study includes single-photon emission computed tomography and computed tomography (SPECT-CT). Of each group (50%, 70% and 97% RCP), one animal was imaged at 4 or 24 h. Whole body images of the euthanized mice were acquired using the VECTor + /CT (MILabs, Netherlands) within 40 min, using the extra ultrahigh sensitivity (XXUHS) collimator (trans axial FOV 8 cm) in list mode using 94 bed positions. The SPECT data was reconstructed using SROSEM with 9 iterations and maximally 128 subsets and a voxel size of 0.8 mm3 (MILabs REC 8.0 software). The photopeaks 113 and 208 keV were used to visualize Lu-177, scatter and background were corrected for with the triple-energy-window method. Furthermore, the CT was reconstructed using filtered back-projection. The SPECT and CT were registered; SPECT data was corrected for attenuation using the CT data. The resulting imaging data was analyzed using PMOD (Bruker Software). After the scan, the organs of these imaging mice were collected, and the data was added to the biodistribution groups of 4 and 24 h, respectively. Biodistribution studies The biodistribution study was performed (including the imaged mice) at 4 h (RCP 50%, 70% and 97%, n = 4 per RCP/group) and 24 h (RCP 50%, 70% and 97%, n = 4) post injection (p.i.). The organs were collected including blood (via orbital punction), stomach, heart, skin, tumor, pancreas, liver, spleen, small intestine, colon, adrenals, kidneys, prostate (anterior), lungs, salivary glands, muscle, bone (femur) and tail. To verify the specificity of the [177Lu]Lu-PSMA-I&T uptake, an additional group was co-injected with a 100-fold molar excess of unlabeled PSMA-I&T (block) and sacrificed at 4 p.i.. The tissues were weighed, and radioactivity was measured using a gamma counter (2480 Wizard-2 (Perkin Elmer, Waltham, MA, USA)) and the Lu-177 activity was measured (calibrated energy at 113 keV and 208 keV). The results are expressed in %IA/g (Fig S1 and S2 of in supplementary material). Results Inter-laboratory comparison HPLC-method(s) and statistical analysis All participating institutions performed the HPLC-analysis (Fig. 1) using the supplied [177Lu]Lu-PSMA-I&T sample, the analyses of each institute showed ingrowth of radiolysis over time, and no increase in non-labeled Lu-177/[177Lu]Lu-DTPA. An example chromatogram from each institute is added in Table S4 of supplementary material. The product ([177Lu]Lu-PSMA-I&T) in the HPLC-chromatograms of all institutes were compared and show a standard deviation (SD) varying from 2.1 to 6.6 (Table 4). Notable for all methods is that if the impurities over time increase, so does the level of uncertainty of the analysis.Table 4The RCP of [177Lu]Lu-PSMA-I&T including the SD, measured at different institutes (24 and 48 h) after productionMeasurement 1RCP (%)Measurement 2RCP (%)Measurement 3RCP (%)24 h48 h24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.7 ± 0.90.7 ± 0.80.4 ± 0.60.3 ± 0.30.3 ± 0.30.3 ± 0.2Radiolysed [177Lu]Lu-PSMA-I&T15.7 ± 1.627.8 ± 4.03.7 ± 1.07.0 ± 1.917.1 ± 3.927.2 ± 6.4[177Lu]Lu-PSMA-I&T83.6 ± 2.571.5 ± 3.895.9 ± 1.592.7 ± 2.182.7 ± 4.072.5 ± 6.6 Table 4The RCP of [177Lu]Lu-PSMA-I&T including the SD, measured at different institutes (24 and 48 h) after productionMeasurement 1RCP (%)Measurement 2RCP (%)Measurement 3RCP (%)24 h48 h24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.7 ± 0.90.7 ± 0.80.4 ± 0.60.3 ± 0.30.3 ± 0.30.3 ± 0.2Radiolysed [177Lu]Lu-PSMA-I&T15.7 ± 1.627.8 ± 4.03.7 ± 1.07.0 ± 1.917.1 ± 3.927.2 ± 6.4[177Lu]Lu-PSMA-I&T83.6 ± 2.571.5 ± 3.895.9 ± 1.592.7 ± 2.182.7 ± 4.072.5 ± 6.6 The RCP of [177Lu]Lu-PSMA-I&T including the SD, measured at different institutes (24 and 48 h) after production For the inter-laboratory comparison, a significant difference can be expected for the relative standard deviation (RSD) (Henderson et al. 2014; Letinski et al. 2021), thereby the RSD is focused on the main peak ([177Lu]Lu-PSMA-I&T). The RSD from the mean is within the range of 1.5%-9.1%, and increases when the percentage of radiolysis also increases (Table 5).Table 5The RSD (%) of the main product ([177Lu]Lu-PSMA-I&T) measured in different institutes at 24 and 48 h after productionMeasurement 1RSD (%)Measurement 2RSD (%)Measurement 3RSD (%)24 h48 h24 h48 h24 h48 h[177Lu]Lu-PSMA-I&T3.05.41.52.34.99.1 Table 5The RSD (%) of the main product ([177Lu]Lu-PSMA-I&T) measured in different institutes at 24 and 48 h after productionMeasurement 1RSD (%)Measurement 2RSD (%)Measurement 3RSD (%)24 h48 h24 h48 h24 h48 h[177Lu]Lu-PSMA-I&T3.05.41.52.34.99.1 The RSD (%) of the main product ([177Lu]Lu-PSMA-I&T) measured in different institutes at 24 and 48 h after production HPLC processing methods The inter-observer variability is assessed by 1 analyst from the comparative center*, by re-analyzing the chromatograms of the two other centers (Table 6), showing a maximum deviation of 2.7%.Table 6Deviation after reintegration of the HPLC-chromatograms at 24 and 48 h after productionCenter 3SD (%)Center 4SD (%)24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.1 ± 0.10.2 ± 0.30.4 ± 0.32.8 ± 0.1Radiolysed [177Lu]Lu-PSMA-I&T0.9 ± 1.40.5 ± 0.31.3 ± 1.50.1 ± 0.1[177Lu]Lu-PSMA-I&T0.9 ± 1.30.6 ± 0.61.1 ± 0.92.7 ± 0.2The trained operator performing the re-integration was from the comparative center and the comparative center is therefore excluded. Center 2 has a standardized procedure independent of a trained operator and this center is therefore also excluded Table 6Deviation after reintegration of the HPLC-chromatograms at 24 and 48 h after productionCenter 3SD (%)Center 4SD (%)24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.1 ± 0.10.2 ± 0.30.4 ± 0.32.8 ± 0.1Radiolysed [177Lu]Lu-PSMA-I&T0.9 ± 1.40.5 ± 0.31.3 ± 1.50.1 ± 0.1[177Lu]Lu-PSMA-I&T0.9 ± 1.30.6 ± 0.61.1 ± 0.92.7 ± 0.2The trained operator performing the re-integration was from the comparative center and the comparative center is therefore excluded. Center 2 has a standardized procedure independent of a trained operator and this center is therefore also excluded Deviation after reintegration of the HPLC-chromatograms at 24 and 48 h after production The trained operator performing the re-integration was from the comparative center and the comparative center is therefore excluded. Center 2 has a standardized procedure independent of a trained operator and this center is therefore also excluded Comparison of HPLC-method(s) The suitability of phosphate buffered HPLC-methods for the quality control of radiolysed [177Lu]Lu-PSMA-I&T, was evaluated by comparing the comparative method at the comparative center with two methods from literature (method A and method B, with both samples n = 3, and injection n = 3). In method A (Fig. 2a, Table 7), the results correspond with the data as described (Kraihammer et al. 2023), with a good baseline separation of the Lu-177/[177Lu]Lu-DTPA, radiolysed product and main product [177Lu]-Lu-PSMA-I&T. However, for method B (Fig. 2b, Table 7), a broad [177Lu]Lu-PSMA-I&T peak with visual tailing was observed. When comparing method A and B to the comparative method (see table S4 supplementary material), a maximum of 2.5% variation in RCP was observed.Fig. 2Chromatogram of [177Lu]-Lu-PSMA-I&T after 48 h, method A (a) and method B (b) with the gradient (%B) on the right y-axisTable 7[177Lu]Lu-PSMA-I&T analyzed with phosphate buffer HPLC-methods versus methods A and BComparative method (%)Method A(%)Method B(%)24 h48 h24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.4 ± 0.20.4 ± 0.11.0 ± 0.81.4 ± 1.61.3 ± 1.42.0 ± 2.6Radiolysed [177Lu]Lu-PSMA-I&T4.2 ± 0.26.4 ± 0.13.0 ± 1.15.9 ± 1.71.9 ± 1.13.9 ± 2.1[177Lu]Lu-PSMA-I&T95.5 ± 0.493.1 ± 0.292.9 ± 0.289.6 ± 0.994.0 ± 2.594.1 ± 1.0Tailing > 1 < 1.5 > 1 < 1.5 > 1 < 1.5 > 1 < 1.5 > > 1.5 > > 1.5Recovery96.5 ± 1.161.3 ± 25.173.3 ± 38.454.5 ± 10.866.7 ± 37.7Recovery + PSMA-I&T103.1 ± 2.264.5 ± 29.8x63.3 ± 8.5xCarry-over0.9 ± 0.1 < 0.1 < 0.11.2 ± 0.11.4 ± 0.4Carry-over + PSMA-I&Tx0.6 ± 0.1x1.5 ± 0.5x Chromatogram of [177Lu]-Lu-PSMA-I&T after 48 h, method A (a) and method B (b) with the gradient (%B) on the right y-axis Table 7[177Lu]Lu-PSMA-I&T analyzed with phosphate buffer HPLC-methods versus methods A and BComparative method (%)Method A(%)Method B(%)24 h48 h24 h48 h24 h48 hLu-177/[177Lu]Lu-DTPA0.4 ± 0.20.4 ± 0.11.0 ± 0.81.4 ± 1.61.3 ± 1.42.0 ± 2.6Radiolysed [177Lu]Lu-PSMA-I&T4.2 ± 0.26.4 ± 0.13.0 ± 1.15.9 ± 1.71.9 ± 1.13.9 ± 2.1[177Lu]Lu-PSMA-I&T95.5 ± 0.493.1 ± 0.292.9 ± 0.289.6 ± 0.994.0 ± 2.594.1 ± 1.0Tailing > 1 < 1.5 > 1 < 1.5 > 1 < 1.5 > 1 < 1.5 > > 1.5 > > 1.5Recovery96.5 ± 1.161.3 ± 25.173.3 ± 38.454.5 ± 10.866.7 ± 37.7Recovery + PSMA-I&T103.1 ± 2.264.5 ± 29.8x63.3 ± 8.5xCarry-over0.9 ± 0.1 < 0.1 < 0.11.2 ± 0.11.4 ± 0.4Carry-over + PSMA-I&Tx0.6 ± 0.1x1.5 ± 0.5x [177Lu]Lu-PSMA-I&T analyzed with phosphate buffer HPLC-methods versus methods A and B Consequently, the recovery of both methods was determined at 24 and 48 h, showing low recovery which did not meet the required > 90% (Table 7). Additionally, as described previously by Aalbersberg et al. (Aalbersberg et al. 2022), addition of unlabeled peptide to the HPLC-sample may improve the recovery. However, after addition of unlabeled PSMA-I&T, the recovery did not increase for both methods. Biodistribution of degraded [177Lu]Lu-PSMA-I&T Radiolabeling The radiolabeling to obtain a decreased RCP was performed with 3.3 nmol PSMA-I&T and 100 MBq of Lu-177 as described in Table 3. These adapted conditions resulted in a successful degradation of [177Lu]Lu-PSMA-I&T down to RCP of 50%, 70% and 97%, (supplemental material, Table S3) which was confirmed stable up to 24 h after HPLC-injection. SPECT-CT of radiolysed [177Lu]Lu-PSMA-I&T The SPECT-CT scan of the murine model at 4 h p.i. showed a higher tumor uptake with higher RCP values, starting from 50% RCP (Fig. 3a), followed by 70% RCP (Fig. 3b), and reaching the highest uptake at 97% RCP (Fig. 3c). No difference could be observed between the 50% and 70% at 4 h p.i.. Furthermore, a high kidney uptake (Fig S1, supplementary material) was observed for all three conditions at 4 h p.i., likely due to ongoing excretion (Ruigrok et al. 2021). At 24 h p.i., a clear difference was observed when the RCP is increased from 50% (Fig. 3d) to 97% (Fig. 3f), whereby the kidney uptake was drastically decreased. No notable off-target uptake was observed of the (radiolysed) radiopharmaceutical in any other organs.Fig. 3SPECT-CT scans of 1 mouse of each group after injection of [177Lu]Lu-PSMA-I&T with decreased RCP at 4 (a–c) and 24 h (d–f) p.i., the PC3-PIP cell line tumor(s) (blue arrow) and kidney(s) (red arrow) SPECT-CT scans of 1 mouse of each group after injection of [177Lu]Lu-PSMA-I&T with decreased RCP at 4 (a–c) and 24 h (d–f) p.i., the PC3-PIP cell line tumor(s) (blue arrow) and kidney(s) (red arrow) Biodistribution of radiolysed [177Lu]Lu-PSMA-I&T Following the data at 4 h p.i. (Fig. 4a), a higher RCP corresponds to an increased tumor-to-kidney ratio. However, likely due to ongoing clearance, the results are not significant. The tumor-to-kidney ratios at 24 h p.i. (Fig. 4b), showed an increased tumor-to-kidney ratio when increased from 50 to 97% RCP (p = 0.01).Fig. 4Tumor-to-kidney ratios at 4 (a) and 24 h (b) p.i. for a 50%, 70% and 97% RCP (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001), n = 4 Tumor-to-kidney ratios at 4 (a) and 24 h (b) p.i. for a 50%, 70% and 97% RCP (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001), n = 4 The tumor selectivity was evaluated by blocking studies, suggesting that the higher the RCP, the more specific the [177Lu]Lu-PSMA-I&T is for the PSMA target (50% RCP; p = 0.18, 50% RCP; p = 0.05 and 97% RCP; p = 0.04 respectively). Additionally, adding an excess of PSMA-I&T showed lower retention for the kidneys for a high RCP (97%) than a low RCP (50%) (Fig. 5, 50%; p = 0.05, 70%; p = 0.02 and for 97%; p = 0.005).Fig. 5Tumor and kidney uptake of 50%, 70% and 97% RCP at 4 h, with and without block (100 × excess, *p ≤ 0.05, **p ≤ 0.01), n = 4 Tumor and kidney uptake of 50%, 70% and 97% RCP at 4 h, with and without block (100 × excess, *p ≤ 0.05, **p ≤ 0.01), n = 4 Discussion In this study, the impact on the biodistribution of radiolysed radiopharmaceutical ([177Lu]Lu-PSMA-I&T) is explored within a murine mouse model. Detection of radiolysed radiopharmaceutical is one of the main challenges in quality control (Hooijman et al. 2022), and the impact in vivo needed to be further investigated. Within this study, an inter-laboratory analysis was performed, showing that when a method is both optimized at each institute individually and validated according to the EANM criteria (Gillings et al. 2020; EDQM xxxx), the deviation does not exceed 6.6% of SD and 9.1% of RSD. The RSD measurements are focused on the main product of interest [177Lu]Lu-PSMA-I&T. The overall deviation may be caused by multiple factors; Firstly, the performance of radio-detectors coupled to an HPLC for RCP measurement is crucial and might be affected by factors such as geometry, size, solvent, flow and volume. The type, size, and energy range of the detector also influences the detection accuracy. Detectors with lower background noise and higher signal-to-noise ratios improve the detection sensitivity, enhancing the precision of RCP measurements (Saldarriaga Vargas et al. 2020). Additionally, a maximum deviation of 2.8% occurs when different trained operators analyze 1 chromatogram. Current software often lacks the flexibility in selection of different algorithms needed for analysis standardization, specifically for radio-HPLC. Consequently, when combining the deviation of 6.6% and 2.7%, a maximum of 7.2% deviation in the overall process was observed. Thereby, it should be noted that radiolysis increases the variability in analytical results, making HPLC evaluation of [177Lu]Lu-PSMA-I&T at later timepoints more challenging. Degradation can alter peak profiles and reduce the accuracy. Therefore, the radiopharmaceutical stability should be optimized before analysis. Additionally, the short shelf-life of the product introduces further complexity and variability, as radiolysis occurs randomly and continuously, which makes it challenging to maintain consistent sample conditions and thereby introducing unsystematic impurities (see Sect. "Inter-laboratory comparison") (Ramos-Méndez et al. 2018). This is further supported by the chromatogram analysis from center 2 (see table S4 supplementary material) (Kraihammer et al. 2023; Iudicello et al. 2021), thereby underlining the need for further investigation of the structural differences between the derivatives of [177Lu]Lu-PSMA-I&T. Furthermore, HPLC QC methods from the literature (Kraihammer et al. 2023) were compared, one of them will be implemented in the Ph. Eur. monograph (2025) (Europe E.E.P.S.C. 2025). In this case of phosphate-buffered methods, although peak separation of the non-labeled Lu-177/[177Lu]Lu-DTPA and radiolysed product can be obtained, no better results than the TFA-based methods could be obtained. While phosphate buffers may offer a sufficient peak resolution for HPLC analysis, caution regarding salt formation in the used system is important, because of system blockages. Additionally, not commonly used eluents, lead to frequent system changes and an increased workload in clinical care (EDQM xxxx; Ioppolo et al. 2023). Additionally, recovery optimization poses a notable challenge, where in this example addition of unlabeled PSMA-I&T did not enhance recovery (< 90%). The biodistribution studies showed that higher RCP correlates with a higher—and thus better—tumor-to-kidney ratio. Since reactive radicals are form randomly and damage the radiopharmaceutical unpredictably, the variation in biodistribution increases as the RCP decreases (see Fig S2 supplementary material section 7.1) (Ramos-Méndez et al. 2018). To make a comparison of the biodistribution of solely the radiolysed radiopharmaceutical, a purification step is necessary to obtain the pure radiolysed product. However, this might introduce a potential matrix effect, and result in a difference in mass, as the unlabeled PSMA-I&T-fraction is also removed during purification, therefore no purification was performed. Additionally, a low RCP corresponds to a decreased specificity to the tumor site, and a higher retention in the kidneys, which may increase the toxicity and decrease the treatment efficacy posing a concern for patient safety. Reduced tumor uptake at lower RCP likely reflects reduced binding of radiolysed radiopharmaceutical, reducing competition with the unlabeled compound. In the kidneys, strong blocking may result from non‑specific binding combined with mass‑dose effects that interfere with reabsorption. Radiolysis radiopharmaceutical therefore could alter the biodistribution, and have to be further investigated. Therefore, a low RCP is associated with reduced tumor specificity and increased renal retention, which may elevate toxicity and compromise therapeutic efficacy, raising concerns regarding patient safety. Conclusion In conclusion, due to the challenges of radiolysed radiopharmaceutical detection, caution is required when optimizing HPLC-methods. However, when working according to the most recent guidelines in regards to equipment and method validation, the inter-laboratory study confirms that radiolysed radiopharmaceutical measurements can be consistently performed across various institutions. Additionally, when assessing the impact, a lower RCP of [177Lu]Lu-PSMA-I&T is associated with lower tumor uptake resulting in reduced tumor-to-kidney ratios, which can impact treatment efficacy. By recognizing and addressing the challenges of measuring radiolysed radiopharmaceuticals, patient safety can be enforced, and the optimal therapeutic outcomes can be realized. Supplementary Information Supplementary Material 1. Supplementary Material 1.