Multifactorial analysis of radiochemical purity in high-activity ¹⁷⁷Lu-labeled theranostics: impact of precursor source, ¹⁷⁷Lu form, and production parameters

Abstract

Background

Lutetium-177 (¹⁷⁷Lu) theranostics have revolutionized personalized cancer treatment, particularly with FDA-approved therapies like [¹⁷⁷Lu]Lu-DOTA-TATE for neuroendocrine tumors and [¹⁷⁷Lu]Lu-PSMA for prostate cancer. Despite growing clinical adoption, there is limited understanding of how different production variables affect radiochemical purity, especially when scaling to high activities for multi-patient batches. This study evaluates the impact of precursor sources, ¹⁷⁷Lu forms (carrier-added (C.A) vs. non- carrier-added (N.C.A)), and radiochemical concentration on product quality.

Results

We analyzed 355 clinical batches of [¹⁷⁷Lu]Lu-DOTA-TATE (n = 101), [¹⁷⁷Lu]Lu- PSMA-617 (n = 169), and [¹⁷⁷Lu]Lu-PSMA-I&T (n = 85) produced with standardized protocols using lutetium-177 from multiple suppliers in both carrier-added and non-carrier-added forms. All radiopharmaceuticals demonstrated consistently high yields (≥ 98%) and met release criteria regardless of starting materials. [¹⁷⁷Lu]Lu-DOTA-TATE and [¹⁷⁷Lu]Lu-PSMA-617 maintained radiochemical purity above 90% throughout 24 h, while [¹⁷⁷Lu]Lu-PSMA-I&T showed stability for 8 h but fell below specifications at 24 h. Negative correlations between bulk activity/concentration and radiochemical purity were observed across all preparations. The lutetium-177 products from various suppliers displayed notably distinct quality profiles. Some suppliers consistently provided higher radiochemical purities, irrespective of the carrier-added or non-carrier-added forms of lutetium-177. However, carrier- added formulations exhibited greater radiostability compared to non-carrier-added ones at elevated concentrations. Furthermore, differences in precursor quality among manufacturers were noted, with certain suppliers offering enhanced radiostability characteristics that may enhance product performance at high activity concentrations.

Conclusion

This comprehensive analysis reveals that hospital-based production can be automized resulting in high-quality and efficient multi-dose production. Small differences in radiochemical purity of ¹⁷⁷Lu -labeled theranostics depends on complex interactions between precursor source, ¹⁷⁷Lu supplier, and ¹⁷⁷Lu form, beyond simple activity-dependent radiolysis. These findings underscore the importance of optimizing production parameters for high- activity preparations and highlight the need to explore the various multifactorial variables that impact the quality of ¹⁷⁷Lu-theranostics.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41181-025-00372-5.

Background

Theranostics is a combined approach that integrates targeted therapies with diagnostic agents to optimize patient treatment in personalized medicine. This rapidly evolving strategy has demonstrated remarkable effectiveness in treating metastatic cancer, achieving significant clinical results with relatively low toxicity (Solnes et al. 2020). A key component of theranostics is the use of targeted ligands, often small peptides, which serve as vehicles for delivering radioactive payloads directly to cancer cells (Hall and Haskali 2022). These ligands can function as either diagnostic or therapeutic agents, depending on the specific radionuclide attached to them. Common diagnostic radionuclides include the positron emitters gallium-68 (⁶⁸Ga) and fluorine-18 (¹⁸F), which are crucial for imaging and locating tumors. In contrast, the primary therapeutic radionuclide used is lutetium-177 (¹⁷⁷Lu), a cytotoxic beta-emitting isotope that when incorporated into a targeted ligand destroys malignant cells. The precision with which targeted ligands deliver radioactive payloads is essential for enhancing treatment efficacy and improving patient outcomes in cancer therapy.

The use of ¹⁷⁷Lu-labeled theranostics has surged over the past two decades, particularly with [¹⁷⁷Lu]Lu-DOTA-TATE for neuroendocrine tumors and [¹⁷⁷Lu]Lu-PSMA-617 and [¹⁷⁷Lu]Lu-PSMA-I&T for prostate cancer (Hofman et al. 2018, 2021, 2024). These therapies are increasingly being assessed for earlier lines of use, expanding their applicability beyond late-stage diseases (Eapen et al. 2024; Azad et al. 2024; Emmett et al. 2024; Strosberg et al. 2021), and in various combinations (Kostos et al. 2022, 2025). In response to the growing global demand for ¹⁷⁷Lu-theranostics (Abdel-Wahab et al. 2024), FDA approval and commercialization of [¹⁷⁷Lu]Lu-PSMA-617 have enabled pharmaceutical companies to scale up manufacturing, allowing for the direct shipment of individual doses to hospitals for patient treatment in some parts of the world (Poschenrieder et al. 2024).

Despite this progress, hospital-based batch production of ¹⁷⁷Lu-theranostics remains an option, particularly in regions where commercial supplies are unavailable or their supply is cost- prohibitive. Australia is one such region where hospital-based production has enabled implementation of theranostics including development of guidelines to support safe use (Lee et al. 2022). As demand for these therapies increases, hospitals often need to produce large multi-dose batches starting with high ¹⁷⁷Lu activities. However, there is a significant gap in our understanding of how these larger-scale preparations respond to radiolysis—the chemical decomposition caused by ionizing radiation. Literature on the production and stability of ¹⁷⁷Lu-theranostics at activities exceeding 40 GBq is scarce, if available at all (Aslani et al. 2015; Iorio et al. 2022; Kraihammer et al. 2023; Hooijman et al. 2022; Mukherjee et al. 2015; Boasa et al. 2021; Guleria et al. 2021; Martin et al. 2021; Weineisen et al. 2015; Sørensen et al. 2020; Schmitl et al. 2023).

This knowledge gap is further exacerbated by variations in starting materials, such as carrier- added (CA) and non-carrier-added (NCA) ¹⁷⁷Lu, which are offered in different volumes (0.2 —7 ml), radioactivity concentrations and formulations (0.04 M and 0.05 M HCl) from various suppliers (Table S1 summarizes the specifications of ¹⁷⁷Lu obtained from different sources). Literature reports often present these diverse sources of lutetium-177 and their different forms as equivalent, overlooking their potential impact on the quality of the final product. Additionally, sourcing essential production consumables, including precursors, from multiple global suppliers further complicates the assessment of how these variables influence product quality.

The literature on the production and quality control of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T cannot be meaningfully compared, even for the same agent, due to significant discrepancies in the source of starting materials, as well as in their production and quality control protocols. Table S2 provides a comprehensive, though not exhaustive, summary of key findings from various published reports, highlighting critical points of difference among them. Additionally, some of the reported production studies do not use well-established and comprehensively validated quality control methods, which calls into question the validity of their findings. For instance, Schmitl et al. reported an in-depth quality assurance investigation on [¹⁷⁷Lu]Lu-PSMA-I&T, revealing a likely overestimation of the radiochemical purity of this product, further complicating our ability to rely on published reports for accurate comparisons (Schmitl et al. 2023). The inherent differences in radioactive detector systems used across various laboratories also limit any direct comparison of results, making it extremely challenging to draw meaningful conclusions.

The high cost of ¹⁷⁷Lu limits the ability to conduct large-scale, systematic comparative studies to evaluate the effects of various sources and forms of ¹⁷⁷Lu, as well as other reagents and precursors, on the quality of the final product. This is especially important when utilizing high starting activities of ¹⁷⁷Lu. Our center operates one of the largest hospital-based theranostic production and treatment facilities globally, which allows us to retrospectively collect significant data (n > 350) on high-activity radiolabeling of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T, with starting activities reaching up to 90 GBq. By employing consistent production and quality control protocols, this data provides valuable and directly comparable insights into the effects of different sources and forms of ¹⁷⁷Lu, the impact of various precursor sources, and the significance of high-activity radiolabeling on the purity of the final product. This knowledge enhances our understanding of production processes and their implications for product quality, enabling us to better meet the needs of patients requiring these essential therapies.

Methods

The production and quality control of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T were carried out using validated methods outlined in a detailed published protocol (Hunt et al. 2025). Briefly, 2,2',2'',2'''-(1,4,7,10-tetraazacyclododecane-1,4,7,10- tetrayl)tetraacetic acid (DOTA) and 2-(4,7,10-tris(carboxymethyl)-1,4,7,10- tetraazacyclododecan-1-yl)pentanedioic acid (DOTAGA) conjugated to their corresponding radiopharmaceutical precursor molecules were labelled in sodium acetate buffer (0.4 M, pH 5.0) containing [¹⁷⁷Lu]LuCl₃ (up to 100 GBq), 2,5-dihydroxybenzoic acid (4 mg), sodium L-ascorbate (20 mg), and heated to 80 °C for 30 minutes (Figure 1). The ¹⁷⁷Lu-labeled mixture was then directly formulated with the addition of an aqueous solution of sodium L-ascorbate (480 mg), pentetic acid (DTPA; 1 mg), and water for injection (10 ml). This process was automated using an iPHASE MultiSyn radiochemistry module with sterile kits affording sterile and apyrogenic ¹⁷⁷Lu-labeled radiopharmaceuticals in non-decay corrected yields > 95%. When necessary, the final product was diluted with saline to obtain radiochemical concentrations below 3.5 GBq/ml. Before clinical use the prepared ¹⁷⁷Lu-radiopharmaceutical underwent validated prerelease quality control tests including activity reconciliation, HPLC (methods reported below), TLC, and bubble point testing, meeting the release criteria outlined in Table S3. Results from all batches of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T in 2023 and 2024 were processed to obtain data used in the manuscript. The peptide requirements based on activity for CA [¹⁷⁷Lu]LuCl₃ were 25 μg GBq⁻¹, yielding specific activities of 40.0 MBq/µg and molar activities of 57.5, 41.8, and 60.0 GBq/µmol for [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T, respectively, while for NCA [¹⁷⁷Lu]LuCl₃, 15 μg GBq⁻¹ was required, resulting in specific activities of 66.7 MBq/µg and molar activities of 95.7, 69.5, and 99.9 GBq/µmol for the same compounds.

Fig. 1.

Generalized reaction scheme for the chelation of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T

HPLC chromatographic conditions

[¹⁷⁷Lu]Lu-PSMA-617 Chromatographic separation was performed using a Phenomenex Jupiter 4 µm Proteo column (150 × 4.6 mm) at 25°C with a flow rate of 1.0 mL/min. The mobile phase consisted of 0.1% trifluoroacetic acid (TFA) in water (A) and 0.1% TFA in acetonitrile (B). Gradient elution over 17 minutes: 80 A/20% B initially, transitioning to 65 A/35% B at 9 minutes, shifting to 10 A/90% B at 9.1 minutes (maintained until 11 minutes), then returning to 80 A/20% B at 11.1 minutes. Detection: UV at 220 nm and scintillation (PMT at 1 volt = 1,000 counts); injection volume: 40 µL. Retention times: 2.5 minutes ([¹⁷⁷Lu]Lu/[¹⁷⁷Lu]Lu-DTPA) and 8.5 minutes ([¹⁷⁷Lu]Lu-PSMA-617).

[¹⁷⁷Lu]Lu-PSMA-I&T Separation utilized a Phenomenex Jupiter 4 µm Proteo column (150 × 4.6 mm) at 35°C with a flow rate of 1.0 mL/min. Mobile phases: 0.1% TFA in water (A) and 0.1% TFA in acetonitrile (B). Gradient elution over 18 minutes: 82 A/18% B initially, transitioning to 60 A/40% B at 10 minutes, shifting to 10 A/90% B at 10.1 minutes (maintained until 13 minutes), returning to 82 A/18% B at 13.1 minutes. Detection: UV at 220 nm and scintillation (PMT at 1 volt = 1,000 counts); injection volume: 40 µL. Retention times: 2.5 minutes ([¹⁷⁷Lu]Lu/[¹⁷⁷Lu]Lu-DTPA) and 9.0 minutes ([¹⁷⁷Lu]Lu-PSMA-I&T).

[¹⁷⁷Lu]Lu-DOTA-TATE Separation was achieved using a Phenomenex Kinetex 5 µm XB-C18 column (150 × 4.6 mm) at 30°C with a flow rate of 1.0 mL/min. Mobile phases: 0.1% TFA in water (A) and 0.1% TFA in acetonitrile (B). Gradient elution over 16 minutes: 85% A/15% B initially, transitioning to 65% A/35% B at 9 minutes, shifting to 10 A/90% B at 9.1 minutes (maintained until 11 minutes), returning to 85 A/15% B at 11.1 minutes. Detection: UV at 220 nm and scintillation (PMT at 1 volt = 1,000 counts); injection volume: 40 µL. Retention times: 1.8 minutes ([¹⁷⁷Lu]Lu/[¹⁷⁷Lu]Lu-DTPA) and 7.5 minutes ([¹⁷⁷Lu]Lu-DOTA-TATE).

Shelf-life stability assessment

The remaining amounts of large batches of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu- PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T, after patient doses were drawn, were resampled by extracting approximately 200 µL from the bulk dose vial, which was stored at room temperature. Resampling occurred at 4, 6, 8, and 24 h after calibration. The material was then analyzed using the validated quality control methods detailed in our published protocol (Hunt et al. 2025).

Statistical analysis

Analysis was conducted on batches of radiopharmaceuticals, defined by precursor and supplier, to determine the correlations between bulk activity and concentration with radiochemical purity (Table S4). Pearson's correlation coefficient was calculated along with bootstrapped confidence intervals (1000 replications). Means and standard deviations described the distribution of activity and concentration. This analysis was limited to products that comprised at least 10 batches with the same combination of precursor supplier, ¹⁷⁷Lu supplier, and ¹⁷⁷Lu form.

Results

Production yields and activity ranges of clinical ¹⁷⁷Lu-labeled radiopharmaceuticals

A total of 101 batches of [¹⁷⁷Lu]Lu-DOTA-TATE were produced, with formulated activities ranging from 8.3 to 65.3 GBq, among these, 14 (14%) batches exceeded 50 GBq. Radiochemical concentrations ranged from 0.74 to 3.02 GBq/ml. The yield for this radiopharmaceutical was 98 ± 3%.

For [¹⁷⁷Lu]Lu-PSMA-617, 169 batches were produced, with formulated activities ranging from 5.2 to 88.9 GBq. Notably, 56 (33%) batches surpassed 50 GBq, while the radiochemical Concentrations ranged from 0.43 to 2.96 GBq/ml, yielding 99 ± 5%.

In the case of [¹⁷⁷Lu]Lu-PSMA-I&T, 85 batches were prepared, with activity levels ranging from 8.1 to 77.0 GBq. Seven (8.2%) batches exceeded 50 GBq and 28 (33%) batches fell between 30 and 50 GBq, with concentrations from 0.67 to 2.91 GBq/ml and a yield of 99% ± 3%. Some yields were reported to exceed 100% due to calibration discrepancies caused by varying wall thicknesses of vials supplied by different manufacturers compared to those used during the final formulation process. Importantly, all production batches passed the pre-release criteria outlined in Table S3 at the end of synthesis (EOS).

24-hour stability profile of [¹⁷⁷Lu]Lu-DOTATATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T

The stability of [¹⁷⁷Lu]Lu-DOTA-TATE (n = 3; activity range: 34–65.3 GBq; concentration: 2.9- 3.5 GBq/ml), [¹⁷⁷Lu]Lu-PSMA-617 (n = 6; activity range: 17–88.9 GBq; concentration: 1.5–3.3 GBq/ml), and [¹⁷⁷Lu]Lu-PSMA-I&T (n = 3; activity range: 34–77 GBq; concentration: 2.3–2.7 GBq/ml) were evaluated over a 24-h period (Table 1). Due to the demands of clinical workflows and patient administration schedules, the mid-point sampling window was adjusted to accommodate a 6–8 h timeframe rather than a fixed 8-h timepoint. While additional stability data exists for other production batches at various individual timepoints, the results presented here represent only those batches where complete sampling profiles were achieved across all specified timepoints.

Table 1.

24-Hour stability profile and quality control analysis of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T

		[177Lu]Lu-DOTA-TATE Activity range: 34–65.3 GBq, Concentration range: 2.9–3.5 GBq/ml; n = 3				[177Lu]Lu-PSMA-617 Activity range: 17–88.9 GBq, Concentration range: 1.5–3.3 GBq/ml; n = 6							[177Lu]Lu-PSMA-I&T Activity range: 34–77 GBq, Concentration range: 2.3–2.7 GBq/ml; n = 3
Parameter	Specification	0 h	4 h	6–8 h	24 h	0 h		4 h		6–8 h	24 h	0 h	4 h	6–8 h	24 h
Appearance	Clear and colourless (slight yellow tint)	Pass	Pass	Pass	Pass	Pass		Pass		Pass	Pass	Pass	Pass	Pass	Pass
pH	4—8	6	6	6	6	6		6		6	6	6	6	6	6
Radiochemical Identity (HPLC)	Reference Std ± 1 min	Pass	Pass	Pass	Pass	Pass		Pass		Pass	Pass	Pass	Pass	Pass	Pass
Radiochemical Purity (HPLC)	≥ 90	95.2 ± 1.3	95.5 ± 0.5	94.7 ± 0.5	93.7 ± 0.9	95.8 ± 1.6		95.7 ± 2.2		95.0 ± 1.4	91.8 ± 1.2	94.3 ± 0.9	92.3 ± 0.9	91.0 ± 1.4	85.5 ± 3.5
Radiochemical Purity (TLC)	≥ 98	100	100	100	100	100	100		100		100	100	100	100	100

Throughout the study period, all preparations consistently maintained a physical appearance that was clear and colorless, with a permissible slight yellow tint attributable to the use of sodium ascorbate stabilizer solutions. The pH remained stable at 6, which is well within the acceptable range of 4 to 8. The radiochemical identity of the preparations was verified at all time points using HPLC analysis. Additionally, radiochemical purity was assessed through two complementary analytical methods: HPLC and Thin Layer Chromatography (TLC). TLC analysis demonstrated consistent 100% purity across all compounds and time points. HPLC analysis revealed high radiochemical purity for [¹⁷⁷Lu]Lu-DOTA-TATE (93.7–95.5%) and [¹⁷⁷Lu]Lu-PSMA-617 (91.8–95.8%) throughout the 24-h period. While [¹⁷⁷Lu]Lu-PSMA-I&T demonstrated acceptable radiochemical purity through the clinically relevant 0–8 h window (91.0–94.3%), a decline to 85.5 ± 3.5% was observed at 24 h, falling below the release specification of ≥ 90%. In our practice, this was not clinically relevant as radiopharmaceutical administration is given on the same day as their production.

We examined the bulk activity and radiochemical concentration of the various ¹⁷⁷Lu-labeled theranostics, and their correlation to the radiochemical purity at the end of synthesis (EOS) (Table 2). This analysis highlights the impact of different precursor sources—such as ABX and Huayi —paired with various ¹⁷⁷Lu sources and forms including ANSTO N.C.A, ITM N.C.A, Isotopia C.A, and Isotopia N.C.A (Fig. 2–4).

Table 2.

Mean (SD) and pearson correlation coefficient of bulk activity, radiochemical concentration with HPLC-derived radiochemical purity

	N	Bulk activity (GBq)		Concentration (GBq/ml)		Purity (%)
		Mean (SD)	Correlation to HPLC RP % (95% CI)	Mean (SD)	Correlation to HPLC RP % (95% CI)	Mean (SD)
[177Lu]Lu-PSMA-617
ABX—ANSTO N.C.A	52	31.7 (18.0)	− 0.46 (0.69 to− 0.22)	1.99 (0.60)	− 0.20 (− 0.45 to + 0.04)	95.2 (1.3)
ABX—Isotopia N.C.A	18	47.8 (17.2)	− 0.63 (0.97 to− 0.28)	2.29 (0.27)	− 0.67 (0.86 to− 0.49)	95.7 (0.8)
ABX—Isotopia C.A	80	46.8 (21.4)	− 0.43 (0.64 to− 0.23)	2.23 (0.54)	− 0.44 (0.64 to− 0.24)	95.6 (1.3)
ABX—ITM N.C.A	19	30.9 (18.0)	− 0.42 (0.76 to− 0.07)	1.84 (0.60)	− 0.59 (0.79 to− 0.39)	96.1 (0.7)
[177Lu]Lu-PSMA I&T
Huayi—ANSTO N.C.A	31	30.6 (17.2)	− 0.68 (0.89 to− 0.48)	1.78 (0.56)	− 0.60 (0.77 to− 0.42)	95.5 (1.8)
Huayi—Isotopia N.C.A	13	29.7 (17.8)	− 0.70 (0.99 to− 0.41)	1.79 (0.55)	− 0.56 (0.88 to− 0.24)	96.3 (1.2)
Pi Chem—ANSTO N.C.A	11	18.6 (7.9)	− 0.19 (− 0.89 to + 0.51)	1.53 (0.66)	− 0.17 (− 0.87 to + 0.54)	97.0 (1.3)
Huayi—Isotopia C.A	12	24.6 (11.6)	− 0.63 (0.90 to− 0.36)	1.76 (0.72)	− 0.44 (− 0.80 to− 0.07)	96.4 (1.3)
Huayi—ITM N.C.A	15	28.5 (12.8)	− 0.74 (1.00 to− 0.47)	1.92 (0.60)	− 0.80 (1.00 to− 0.59)	96.6 (0.9)
[177Lu]Lu-DOTATATE
Huayi—ANSTO N.C.A	23	27.9 (18.5)	− 0.63 (0.87 to− 0.39)	1.71 (0.74)	− 0.61 (0.87 to− 0.34)	93.5 (1.8)
Huayi—Isotopia N.C.A	14	31.0 (11.4)	− 0.27 (− 0.92 to + 0.37)	2.03 (0.43)	− 0.85 (1.00 to− 0.61)	95.6 (1.2)
Huayi—Isotopia C.A	47	35.8 (14.7)	− 0.48 (0.66 to− 0.30)	2.12 (0.54)	− 0.40 (0.59 to− 0.22)	95.4 (1.3)
Huayi—ITM N.C.A	11	24.2 (14.6)	− 0.63 (1.00 to− 0.06)	1.64 (0.70)	− 0.46 (− 1.00 to + 0.09)	95.3 (1.3)

(%) in various radiopharmaceuticals

Fig. 2.

Scatterplot of bulk activity (GBq) and radiochemical concentration (GBq/ml) versus radiochemical purity (%) in [.¹⁷⁷Lu]Lu-PSMA-617

Fig. 4.

Scatterplot of bulk activity (GBq) and radiochemical concentration (GBq/ml) versus radiochemical purity (%) in [¹⁷⁷Lu]Lu-DOTA-TATE

For [¹⁷⁷Lu]Lu-PSMA-617, the precursor was sourced entirely from ABX. By combining the ABX-sourced precursor with four different forms and sources of ¹⁷⁷Lu, moderate correlations were generally observed between bulk activity or concentration and radiochemical purity, with Pearson's correlation coefficients (ρ) ranging from -0.67 to -0.20 (Fig. 2). These were especially noted for the Isotopia supplied N.C.A product, (activity) ρ = − 0.63 (95%CI:0.97 to − 0.28), (concentration) ρ = − 0.67 (95%CI: 0.86 to − 0.49) though the number of batches analysed was relatively few (n=18). Mean bulk activity was appreciably higher for the Isotopia products than the ANSTO or ITM N.C.A. products (mean difference: 15.5, 95%CI: 9.5 to 21.5).

Five [¹⁷⁷Lu]Lu-PSMA-I&T variations were analyzed: four combining Huayi precursor with different forms of ¹⁷⁷Lu, and one using piChem precursor with ANSTO N.C.A ¹⁷⁷Lu. Generally more pronounced negative correlations with purity were observed between both bulk activity and concentration compared to [¹⁷⁷Lu]Lu-PSMA-617, with no bootstrapped confidence intervals crossing zero. In contrast, the product with a piChem precursor exhibited relatively weak correlation coefficients (ρ = − 0.17, − 0.19) with a considerably lower mean bulk activity (Fig. 3). There was numerically greater variation observed in measured purity with the ANSTO. N.C.A. product (SD = 1.8) versus all others (SD: 0.9 to 1.3).

Fig. 3.

Scatterplot of bulk activity (GBq) and radiochemical concentration (GBq/ml) versus radiochemical purity (%) in [¹⁷⁷Lu]Lu-PSMA-I&T

For [¹⁷⁷Lu]Lu-DOTA-TATE, four variations were analyzed including precursor obtained from Huayi precursor in combination with different sources and forms of ¹⁷⁷Lu. In this instance, correlations generally ranged from moderate to strongly negative. Of note, the mean bulk activity and concentration was notably higher for the Isotopia ¹⁷⁷Lu (N.C.A) and (C.A.) batches (Fig. 4). Similar to [¹⁷⁷Lu]Lu-PSMA I&T, higher variation in purity was observed with the ANSTO N.C.A. ¹⁷⁷Lu and additionally for the [¹⁷⁷Lu]Lu-DOTA-TATE batches, the mean purity percentage for ANSTO N.C.A. was lower than the other three products, mean: 93.5% vs 95.4% (mean difference − 1.9%; 95%CI: 2.6 to − 1.2%).

Discussion

Our standardized production protocol demonstrated significant versatility in the preparation of [¹⁷⁷Lu]Lu-DOTA-TATE, [¹⁷⁷Lu]Lu-PSMA-617, and [¹⁷⁷Lu]Lu-PSMA-I&T. This protocol successfully accommodated varying volumes (0.5-4.5 mL) of both C.A and N.C.A ¹⁷⁷Lu from multiple suppliers (ANSTO, Isotopia, and ITM) using a single cassette-based platform. The protocol’s robustness is evidenced by consistently high radiochemical yields of ≥ 98% across all three radiopharmaceuticals and the successful preparation of over 300 high-activity batches for more than 2000 cycles of therapy.

The 24-hour stability assessment revealed that both [¹⁷⁷Lu]Lu-DOTA-TATE and [¹⁷⁷Lu]Lu-PSMA-617 maintained high radiochemical purity (> 90%) throughout the study period, while [¹⁷⁷Lu]Lu-PSMA-I&T showed acceptable purity within the clinically relevant 0–8 hour window but declined below specification at 24 hours. This finding aligns with Schmitl et al.'s work, which employed a risk-based approach to establish 90% radiochemical purity thresholds, noting that heat-induced cyclization by-products known to occur in PSMA ligands (Martin et al. 2021) represented only a small fraction (2.87 ± 0.85%) of total radioactivity, while the main radiolysis products were activity concentration-dependent and comprised approximately 2.8 ± 0.1% of the samples (Schmitl et al. 2023). This confirms the suitability of our formulations for clinical use within standard administration timeframes, though with limitations for extended storage of [¹⁷⁷Lu]Lu-PSMA-I&T.

Our analysis uncovers important insights into the relationship between bulk activity, radiochemical concentration, and radiochemical purity in various ¹⁷⁷Lu-labeled theranostics produced from different precursors and ¹⁷⁷Lu sources. A negative correlation between bulk activity/concentration and radiochemical purity was noted across the three radiopharmaceuticals ([¹⁷⁷Lu]Lu-PSMA-617, [¹⁷⁷Lu]Lu-PSMA-I&T, and [¹⁷⁷Lu]Lu-DOTA-TATE), which aligns with the expected increase in radiolytic breakdown of products at higher activities. However, the magnitude of this correlation varied considerably. The variability in precursor and ¹⁷⁷Lu sources produced nuanced results for each radiopharmaceutical, highlighting the complex interplay of factors that can affect the final product quality.

A consistent pattern emerged across all three radiopharmaceuticals regarding the influence of ¹⁷⁷Lu source on product quality. When using identical precursor sources, ITM N.C.A. ¹⁷⁷Lu consistently delivered higher radiochemical purities compared to ANSTO's N.C.A. ¹⁷⁷Lu. Moreover, both C.A and N.C.A ¹⁷⁷Lu from Isotopia provided consistent purities across all formulations, closely matching the high-quality results achieved with ITM's N.C.A ¹⁷⁷Lu. This consistent pattern points to inherent qualities in these ¹⁷⁷Lu sources that influence radiochemical stability regardless of which specific precursor is being labeled. Nevertheless, all sources and forms of ¹⁷⁷Lu led to products that passed acceptance testing and are therefore suitable for clinical use.

For [¹⁷⁷Lu]Lu-PSMA-I&T, the piChem precursor demonstrated slightly higher radiochemical purities and enhanced stability at elevated activity concentrations compared to the Huayi precursor, despite being used at lower mean bulk activity. However, this result should be interpreted with caution given the small sample size for the piChem precursor. A larger number of production runs would be required to confirm this observation. The observed difference is unexpected given that both precursor materials exhibited similar purity specifications (98–99%; certificates of analysis for both suppliers are provided in the supplementary materials). Should this trend prove to be genuine with additional data, it would suggest that radiochemical purity determinants extend beyond simple precursor purity and may include factors such as counter ion composition, trifluoroacetic acid content, total peptide content, and packaging methodology. Systematic investigation of these variables would be necessary to establish whether the observed differences reflect true variations in precursor performance characteristics.

A particularly interesting finding emerged when comparing C.A versus N.C.A ¹⁷⁷Lu from Isotopia across different formulations. When using Isotopia's N.C.A ¹⁷⁷Lu for both [¹⁷⁷Lu]Lu-PSMA-617 and [¹⁷⁷Lu]Lu-DOTA-TATE, we observed steeper radiochemical degradation with increasing concentrations compared to formulations using C.A ¹⁷⁷Lu. This suggests either that the higher amount of precursor typically used with C.A ¹⁷⁷Lu or some intrinsic property of C.A ¹⁷⁷Lu itself confers a protective effect against radiolytic degradation.

Molar activity is well-established as a critical factor in diagnostic radiopharmaceutical efficacy and safety (Luurtsema et al. 2021). Theranostic formulations commonly use significantly larger amounts of precursor compared to diagnostic applications. This is most apparent in preparations that use C.A ¹⁷⁷Lu, which typically use larger amounts of peptide precursor compared to N.C.A preparations. Despite these formulation differences, both C.A and N.C.A preparations are often used interchangeably in clinical practice, and the impact of varying molar activity on target protein binding has been poorly studied.

High molar activity formulations are generally considered important for radiopharmaceuticals targeting the central nervous system, where target protein density is low (Pike 2016). In contrast, theranostic applications in oncology may present different considerations. Cancer typically involves high-volume disease often with high target protein expression, which may influence molar activity requirements. The impact of molar activity on therapeutic responses may also vary depending on the target protein type, whether receptors, enzymes, or other oncoproteins, adding another layer of complexity that requires investigation.

In some clinical protocols, unlabeled peptide or antibody is intentionally added to formulated doses before administration to reduce off-target radiation exposure to healthy tissues by saturating normal clearance mechanisms (Zhou et al. 2022; Pickford and Turner 2012; Hegi-Johnson et al. 2022). This practice further complicates the relationship between molar activity and therapeutic outcomes. Prospective clinical studies comparing absorbed doses and therapeutic responses across different molar activities of specific theranostics would help clarify these relationships and may inform future guidelines for molar activity requirements in theranostic applications.

An important consideration with C.A ¹⁷⁷Lu is the presence of inseparable ^177mLu, a metastable nuclear isomer co-produced during the ¹⁷⁷Lu production process. Due to the significantly different half-lives of ¹⁷⁷Lu (t½ = 6.65 days) and ^177mLu (t½ = 160.4 days), the ^177mLu ratio increases over time as ¹⁷⁷Lu decays faster. ^177mLu poses additional radiation safety considerations for waste management due to its longer half-life. The European Pharmacopoeia establishes strict radionuclidic purity requirements for ¹⁷⁷Lu for radiolabeling, mandating greater than 99.9% purity with ^177mLu limited to < 0.07%. In our practice, all C.A ¹⁷⁷Lu batches are used within their expiry period, ensuring ^177mLu levels remain below this limit. Our waste management strategy minimizes ^177mLu waste generation by ordering C.A ¹⁷⁷Lu quantities within 5% of required doses. Waste materials are stored in designated radiation storage areas until radiation monitoring confirms the waste no longer meets the regulatory definition of radioactive material requiring controlled disposal. This approach ensures that C.A and N.C.A formulations are managed similarly, with potentially longer storage periods for C.A waste having minimal operational impact.

This detailed analysis provides a unique evidence base that can inform ¹⁷⁷Lu suppliers, radiopharmaceutical manufacturers, and regulatory bodies investigating the precise composition and characteristics of their products to maintain or further improve the quality of their material for clinical applications. As global demand for radioligand therapy continues to rise, especially in regions with limited access to commercial supplies, the variability observed in our study offers vital evidence-based data for practitioners and regulatory authorities. This information supports the development of corresponding monographs that favor radiochemical purities of 90% or greater as the cut-off when utilizing high-activity, multi-dose productions involving the full range of precursor and ¹⁷⁷Lu sources and forms needed to meet the growing demand for radioligand therapy. While all products in our study maintained clinically acceptable quality, the subtle yet consistent variations based on supplier and formulation highlight the need for a nuanced approach to establishing regulatory standards that balances theoretical ideals with practical clinical realities. These insights are particularly valuable as international initiatives work to expand global access to radiopharmaceutical therapies, allowing for the development of realistic quality parameters that can support equitable access to these life-extending therapies while maintaining appropriate safety and efficacy standards across diverse healthcare settings (Abdel-Wahab et al. 2024).

Hospital radiopharmaceutical production environments are often staffed by professionals with specialized academic and clinical expertise who are well positioned to investigate and interpret such subtle differences in radiopharmaceutical performance. The dismissive characterization of hospital-based radiopharmaceutical preparation as “home brew” or “ad hoc hospital-based compounding” is unfounded and disregards the foundational contributions these hospital-based departments have made to the field of nuclear medicine. While we recognize that certain practices in hospital-based theranostic production environments operate within evolving regulatory frameworks, we believe it’s essential to appreciate the significant, nuanced, and specialized contributions that these environments have made and continue to make to the advancement of the entire discipline. The most effective approach is for industry, academia, and clinical environments to collaborate and learn from each other, integrating pharmaceutical industry capabilities with hospital-based expertise to ensure equitable patient access and effectively address the complex logistical and technical challenges inherent in theranostic production.

Conclusion

Our comprehensive analysis of over 350 batches of ¹⁷⁷Lu-labeled theranostics reveals complex relationships between precursor sources, ¹⁷⁷Lu sources, and radiochemical purity. While all preparations maintained quality standards sufficient for clinical release, the observed variations in radiochemical purity associated with different precursor-¹⁷⁷Lu combinations highlight that radiopharmaceutical quality depends on a complex interplay of factors not fully recognized previously in the literature. The pronounced differences in stability profiles between C.A and N.C.A ¹⁷⁷Lu, along with precursor-specific effects on radiostability, emphasize the importance of optimizing production parameters beyond simply considering radiolytic effects.

This research highlights the highly nuanced territory of ¹⁷⁷Lu-theranostic production, where multiple variables and their interactions influence final product quality in subtle but measurable ways. While our data reveals consistent patterns across different combinations, drawing definitive practical recommendations would be premature given the multifactorial nature of these relationships. Rather than advocating for specific source choices, we believe the responsible approach is to continue systematically collecting data and analyzing these nuanced differences before making prescriptive suggestions to the radiopharmaceutical community. Our primary contribution is providing detailed, standardized data to help build a more complete understanding of these complex relationships. Only through collaborative data sharing and continued mechanistic investigations to elucidate the chemical and physical characteristics of different ¹⁷⁷Lu sources, precursor molecules, and stabilizing agents can the field develop evidence-based guidelines that optimize production protocols for enhanced patient care.

List of Abbreviations

C.A.

Carrier-added

N.C.A

Non Carrier-added

DOTA

2,2',2'',2'''-(1,4,7,10- Tetraazacyclododecane-1,4,7,10-tetrayl)tetraacetic acid

DOTAGA

2-(4,7,10- Tris(carboxymethyl)-1,4,7,10-tetraazacyclododecan-1-yl)pentanedioic acid

DTPA

Pentetic acid

EOS

End of synthesis

FDA

Food and Drug Administration

GBq

Gigabecquerel

HPLC

High-performance liquid chromatography

ITM

ITM Medical Isotopes

PSMA

Prostate-specific membrane antigen

SD

Standard deviation

TLC

Thin-layer chromatography

¹⁸F

Fluorine-18

⁶⁸Ga

Gallium-68

¹⁷⁷Lu

Lutetium-177

Author contributions

MBH developed and implemented the production protocol and was the major contributor in writing the manuscript. MSH assisted with setting the clinical specifications for the products and helped contribute to the writing of the paper. NP performed the statistical analysis and helped write the corresponding statistical sections of the paper. WH, ML, UK, SK, WN, PDR, and BE helped in the implementation of the protocol, data collection, and contributed to the writing of the paper. All authors read and approved the final manuscript.

Funding

Not applicable.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable

Competing interests

The authors declare that they have no competing interests.