Radionuclide generator-based production of therapeutic ¹⁷⁷Lu from its long-lived isomer ^177mLu

Abstract

Background

In this work, a lutetium-177 (¹⁷⁷Lu) production method based on the separation of nuclear isomers, ^177mLu & ¹⁷⁷Lu, is reported. The ^177mLu-¹⁷⁷Lu separation is performed by combining the use of DOTA & DOTA-labelled peptide (DOTATATE) and liquid-liquid extraction.

Methods

The ^177mLu cations were complexed with DOTA & DOTATATE and kept at 77 K for periods of time to allow ¹⁷⁷Lu production. The freed ¹⁷⁷Lu ions produced via internal conversion of ^177mLu were then extracted in dihexyl ether using 0.01 M di-(2-ethylhexyl) phosphoric acid (DEHPA) at room temperature. The liquid-liquid extractions were performed periodically for a period up to 35 days.

Results

A maximum ¹⁷⁷Lu/^177mLu activity ratio of 3500 ± 500 was achieved with [^177mLu]Lu-DOTA complex, in comparison to ¹⁷⁷Lu/^177mLu activity ratios of 1086 ± 40 realized using [^177mLu]Lu-DOTATATE complex. The ¹⁷⁷Lu-^177mLu separation was found to be affected by the molar ratio of lutetium and DOTA. A ¹⁷⁷Lu/^177mLu activity ratio up to 3500 ± 500 was achieved with excess DOTA in comparison to ¹⁷⁷Lu/^177mLu activity ratio 1500 ± 600 obtained when lutetium and DOTA were present in molar ratio of 1:1. Further, the ¹⁷⁷Lu ion extraction efficiency, decreases from 95 ± 4% to 58 ± 2% in the presence of excess DOTA.

Conclusion

The reported method resulted in a ¹⁷⁷Lu/ ^177mLu activity ratio up to 3500 after the separation. This ratio is close to the lower end of ¹⁷⁷Lu/^177mLu activity ratios, attained currently during the direct route ¹⁷⁷Lu production for clinical applications (i.e. 4000–10,000). This study forms the basis for further extending the liquid-liquid extraction based ^177mLu-¹⁷⁷Lu separation in order to lead to a commercial ^177mLu/¹⁷⁷Lu radionuclide generator.

Background

Radionuclide generators are known to have brought revolutionary opportunities in the development of nuclear medicine (Knapp & Dash, 2016; Knapp & Mirzadeh, 1994; Knapp et al., 2014; Knapp & Baum, 2012). The current state of the art of ^99mTc, ¹⁸⁸Re, ⁶⁸Ga pharmaceuticals owes their existence largely to the availability of their corresponding radionuclide generators (Roesch & Riss, 2010; Pillai et al., 2012). They offer continuous, on-site and on-demand isolation of a short-lived daughter radionuclide from its longer-lived mother radionuclide. Lutetium-177 (¹⁷⁷Lu) is a radionuclide that could also benefit from the advantages of a generator vastly. ¹⁷⁷Lu is well-known for its theranostic potential and is expected to play a crucial role in fulfilling the global demand of radionuclides for many targeted radionuclide therapy applications (Das & Banerjee, 2016; Das & Pillai, 2013). The [¹⁷⁷Lu]Lu-DOTATATE has already been FDA approved for the application in neuroendocrine tumour therapy (https://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm594043.htm, n.d.). Currently, other ¹⁷⁷Lu radiopharmaceuticals have also entered the clinic in the treatment of prostate cancer, lung cancer, non-Hodgkin lymphoma, bone pain palliation and others (Banerjee et al., 2015; Emmett et al., 2017; Hofman et al., 2018; Repetto-Llamazares et al., 2018; Dho et al., 2018). Clearly, the demand of ¹⁷⁷Lu is only going to increase and radionuclide generator can complement the current production routes. The long half-life of ^177mLu (160.44 days) can potentially lead to on-site and on-demand ¹⁷⁷Lu supply for a long period of time without the need of weekly irradiations in nuclear reactor (De Vries & Wolterbeek, 2012; Bhardwaj et al., 2017). However, the development of ^177mLu/¹⁷⁷Lu radionuclide generator needs to tackle the great challenge of separating the physically and chemically alike nuclear isomers ¹⁷⁷Lu and ^177mLu.

It has been previously shown that ¹⁷⁷Lu can be separated from ^177mLu due to the chemical effects occurring as a consequence of internal conversion decay of ^177mLu (Bhardwaj et al., 2017). Internal conversion decay may result in the emission of multiple Auger electrons, often accompanied with the loss of valence electrons and leaving the atom in a highly positively charged state which can result in bond rupture (Cooper, 1942). This principle presents a possibility to separate two isomers, provided that a separation process that can quickly & selectively capture the freed ions is feasible. Additionally, from a radionuclide generator perspective, the separation process should also allow the periodic extraction of the produced daughter radionuclide during the lifetime of the generator.

Previously, a column chromatography based ¹⁷⁷Lu-^177mLu separation process has been reported, where the ^177mLu complexed with DOTATATE has been immobilized on a tC-18 silica and the freed ¹⁷⁷Lu ions produced after the decay have been separated using a mobile phase flow (Bhardwaj et al., 2017). The ¹⁷⁷Lu/^177mLu activity ratio of 250 has been reached after separation compared to the equilibrium ¹⁷⁷Lu/^177mLu activity ratio of 0.25. However, in order to fulfil the clinical demand the separation method should provide ¹⁷⁷Lu having minimum breakthrough of ^177mLu. The current direct production route delivers ¹⁷⁷Lu with ¹⁷⁷Lu/^177mLu activity ratio ranging from 4000 to 10,000 (Dvorakova et al., 2008; Pawlak et al., 2004; Knapp FFJA et al., 1995; Das et al., 2007; Chakraborty et al., 2014), while the indirect production route affords the no-carrier added ¹⁷⁷Lu with almost negligible amount of ^177mLu (Watanabe et al., 2015).

In this work, a radionuclide generator for the production of ¹⁷⁷Lu based on the pair of nuclear isomer^177mLu-¹⁷⁷Lu is presented. The ^177mLu-¹⁷⁷Lu separation has been performed using liquid-liquid extraction (LLE). LLE has been explored several times before in the development of other radionuclide generators, such as ⁹⁹Mo/^99mTc, ⁶⁸Ge/⁶⁸Ga, ¹⁸⁸Re/¹⁸⁸W, and ⁹⁰Y/⁹⁰Sr radionuclide generators (Le Minh & Lengyel, 1989; Fikrle et al., 2010; Bhatia & Turel, 1989; Boyd, 1982; Ehrhardt & Welch, 1978; Mushtaq et al., 2007; Dutta & Mohapatra, 2013). The present work demonstrates the application of LLE in ¹⁷⁷Lu-^177mLu separation which can potentially lead to a commercial ^177mLu/¹⁷⁷Lu radionuclide generator. The metastable isomer, ^177mLu, was complexed with the chelating agents (DOTA and DOTATATE) and the freed ¹⁷⁷Lu ions was extracted in dihexyl ether using Di-(2-ethylhexyl) phosphoric acid (DEHPA) as the cation extracting agent.

Materials and methods

Materials

Lutetium chloride hexahydrate, LuCl₃.6H₂O (≥99.99%), di (2-ethylhexyl) phosphoric acid, DEHPA (97%), di-n-hexyl ether, DHE (97%), sodium acetate (≥99%), chelex resin (chelex-100, 50–100 mesh) and acentonitrile (99.3%) were purchased from Sigma Aldrich. 1,4,7,10-tetraazacyclododecane N, N′, N″, N″’-tetraacetic acid, DOTA (98%) was purchased from ABCR GmBH & Co. KG Germany. DOTATATE was obtained as a kind gift from Erasums Medical Centre (Rotterdam) and was produced by Biosynthema, MO, USA. The lutetium-177 (¹⁷⁷Lu) used in the optimization studies was produced by irradiating around 1 mg of natural LuCl₃.6H₂O in the Hoger Onderwijs Reactor Delft (HOR) with a thermal neutron flux of 4.72*10¹² neutrons∙s^− 1∙cm^− 2 (less than 1.5% epithermal contribution) and an irradiation time of 10 h. The solid sample was weighed inside polyethylene capsule and sealed, packed inside polyethylene rabbits. After irradiation, the samples were left for a cooling period of 3 days, resulting in the production of around 17 MBq of ¹⁷⁷Lu. The capsules were opened and transferred into a plastic vial containing 2.5 mL, pH -3, HCl solution, resulting in a 1 mM [¹⁷⁷Lu]LuCl₃ solution.

The Lutetium-177 m (^177mLu) source was provided by IDB- Holland as a 1 mM [^177mLu]LuCl₃ solution with about 5 MBq ^177mLu per g of solution.

Methods

γ ray spectroscopy analysis

All the activity measurements were performed on a well-type HPGe detector for counting time up to 5 h to reduce the error from the counting statistics to less than 5%. The measurement of the samples obtained at the end of LLE was repeated after 3–4 half-lives of ¹⁷⁷Lu to decrease the background and measure the ^177mLu activity with less than 5% uncertainty. The efficiency calibration for different peaks was performed using a known activity of ¹⁷⁷Lu source supplied by IDB Holland. The obtained gamma ray spectra were analysed using an in-house software to calculate the activity of each fraction (Blaauw, 1993). In order to minimize the error, all the vials were weighed before and after the fraction collection.

Preparation of aqueous phase

The ^177mLu containing LuCl₃ solution (1 mM) was used to prepare [^177mLu]Lu-DOTA complex in three different molar ratios (1:1, 1:2, 1:4). Typically, 1 mM [^177mLu]LuCl₃ solution (0.150 mL, 0.150 μmoles) was mixed with 0.01 M DOTA in different molar ratios (1:1, 1:2 & 1:4) in the presence of 0.150 mL, 1 M sodium acetate- acetic acid buffer at pH 4.3. The reaction mixture was heated at 80 °C for 30 min. The [^177mLu]Lu-DOTATATE complex was synthesized as reported previously in a Lu:DOTATATE molar ratio of 1:4 (Bhardwaj et al., 2017). Typically, 1 mM [^177mLu]LuCl₃ solution (0.050 mL, 0.050 μmoles) was mixed with 0.200 μmol DOTATATE solution in the presence of 0.150 mL, 1 M sodium acetate- acetic acid buffer (pH- 4.3). The reaction mixture was heated at 80 °C for about 1 h followed by incubation at room temperature for about 1 h.

The complex formation was confirmed using instant thin layer chromatography. Free ^177mLu ions traces were removed using a cation exchange resin (chelex-100). (Details in S1, Additional file 1).

Liquid-liquid extraction (LLE) procedure

The schematic representation of LLE to separate the freed ¹⁷⁷Lu ions from the complexed ^177mLu ions is shown in Fig. 1.

Fig. 1

Schematic representation of liquid-liquid extraction to extract the freed ¹⁷⁷Lu ions

All the LLE experiments were performed in 2 mL Eppendorf by placing them in a shaking incubator at room temperature. The aqueous and the organic phases were mixed in volumetric ratio (1:1) at 1400 rpm for a stirring time of about 10 min. The stirring time of 10 min was optimised by studying the ¹⁷⁷Lu extraction efficiency as a function of extraction time (see Additional file 1 Figure S1(b), S2, supplementary information). At the end of stirring, the layer separation was achieved after a settling time of about one minute. In order to avoid any contamination of the aqueous layer in the organic layer, only the upper 2/3rd organic layer was taken out using a 20- 200uL pipette in all the LLE experiments. The pipetted organic layer was transferred to a pre-weighed vial to know the exact amount of organic phase removed in each extraction.

First, free ¹⁷⁷Lu cations were extracted from a 0.3 mL, pH -4, 1 mM [¹⁷⁷Lu]LuCl₃ solution as the aqueous phase. The organic phase consists of 0.3 mL dihexyl ether containing different DEHPA concentrations, namely 0.01, 0.05, 0.1, 0.15, 0.2, 0.4, 0.6, 1.0, 1.2 and 1.6 M. At the end of LLE, the ¹⁷⁷Lu activity in the organic and the aqueous layer was measured using γ ray spectroscopy to obtain the ¹⁷⁷Lu extraction efficiency (EE). The EE is defined as the percentage of the ¹⁷⁷Lu activity moving from the aqueous phase in to the organic phase after the extraction. All the experiments were performed in triplicate.

Subsequently, the LLE was performed to extract the freed ¹⁷⁷Lu ions from the aqueous phase containing [^177mLu]Lu-DOTATATE, [^177mLu]Lu-DOTA complex. For [^177mLu]Lu-DOTATATE complex, the ¹⁷⁷Lu extraction was performed successively at varying ¹⁷⁷Lu accumulation periods for a total time period of up to 60 days. For, [^177mLu]Lu-DOTA complex, the freed ¹⁷⁷Lu ions were extracted successively at every 7 days for a total time period of 35 days. In between the extractions, the [^177mLu]Lu-DOTA and [^177mLu]Lu-DOTATATE complexes were left in a liquid N₂ tank to allow for the accumulation of freed ¹⁷⁷Lu ions. The ¹⁷⁷Lu separation was performed by bringing the vial out of the liquid N₂ tank and quickly adding the 0.01 M DEHPA in DHE in a 1:1 volumetric ratio (0.3 mL: 0.3 mL), at room temperature and 10 min of stirring time, as shown schematically in Fig. 1. At the end of LLE, the ¹⁷⁷Lu and ^177mLu activity in the organic layer was measured using γ ray spectroscopy to calculate the amount of ¹⁷⁷Lu and ^177mLu ions extracted in the organic phase and the ¹⁷⁷Lu/^177mLu activity ratio.

The ¹⁷⁷Lu extraction efficiency is defined as the amount of ¹⁷⁷Lu ions that were extracted into the organic phase divided by the theoretically produced ¹⁷⁷Lu ions (see section S3, eq. S2 in Supplementary Information). The percentage of ^177mLu extracted is defined as the activity of ^177mLu ions measured in organic phase after the LLE divided by the starting activity of the ^177mLu ions in the aqueous phase.

Results

¹⁷⁷Lu/ ^177mLu separation using [^177mLu]Lu-DOTATATE complex

The ¹⁷⁷Lu/ ^177mLu separation was performed using [^177mLu]Lu-DOTATATE complex synthesized in the presence of an excess of DOTATATE (Lu:DOTATATE molar ratio of 1:4). The ¹⁷⁷Lu ions production via the decay of ^177mLu is represented by eq. S1, Supplementary Information, S3 and the expected growth of ¹⁷⁷Lu ions with the increase in the ¹⁷⁷Lu accumulation period is shown in Additional file 1 Figure S2, Supplementary Information. The amount of ¹⁷⁷Lu ions produced increases with an increase in ¹⁷⁷Lu accumulation period and reaches a maximum after 32 days of ¹⁷⁷Lu accumulation. In the presented results, the freed ¹⁷⁷Lu ions were extracted from [^177mLu]Lu-DOTATATE complex by performing LLE successively after different ¹⁷⁷Lu accumulation intervals. Figure 2 (a)&(b) show the ¹⁷⁷Lu extraction efficiency and percentage of the ^177mLu ions extracted in the organic phase at the end of the LLE at different time intervals, respectively. An average ¹⁷⁷Lu extraction efficiency of 60 ± 10% was obtained at the end of LLE. This is 40% less than the 99 ± 2% ¹⁷⁷Lu extraction efficiency observed during the LLE of ¹⁷⁷Lu ions from a 1 mM [¹⁷⁷Lu]LuCl₃ solution using 0.01 M DEHPA in DHE (see Additional file 1 Figure S1, supplementary information S2). Additionally, along with the ¹⁷⁷Lu ions, 0.0085 ± 0.0015% of the starting ^177mLu activity was also extracted in the organic phase. Figure 2(b), shows the ¹⁷⁷Lu/^177mLu activity ratios obtained after different extractions. An increase in the ¹⁷⁷Lu/^177mLu activity ratio is observed with an increase in the time interval between the extractions. The maximum ¹⁷⁷Lu/^177mLu activity ratio of 1086 ± 40 is obtained during the LLE at 43 days after a ¹⁷⁷Lu accumulation period of 26 days. A decrease in the ¹⁷⁷Lu accumulation period leads to a decrease in the ¹⁷⁷Lu/ ^177mLu activity ratios. The ¹⁷⁷Lu/^177mLu activity ratios 600 ± 100 was obtained for ¹⁷⁷Lu accumulation periods between 6 and 10 days.

Fig. 2

a The ¹⁷⁷Lu extraction efficiency (y axis, left) and the % ^177mLu extracted (y axis, right) at different extraction time during the successive LLE of free ¹⁷⁷Lu ions from [^177mLu]Lu-DOTATATE complex using 0.01 M DHEPA in DHE. b The ¹⁷⁷Lu/^177mLu activity ratio obtained in the organic phase at different extraction time. The error bars represent the error in the individual measurements due to counting statistics

¹⁷⁷Lu/^177mLu radionuclide separation using [^177mLu]Lu-DOTA complex

The results obtained when the LLE was performed to extract the freed ¹⁷⁷Lu ions from the [^177mLu]Lu-DOTA complex are shown in Fig. 3&4. The LLE was performed successively at time intervals of 7 days. Figure 3(a) shows the effect of Lu: DOTA molar ratios on ¹⁷⁷Lu extraction efficiency. Figure 3(b) displays the percentage of initial ^177mLu activity extracted in the organic phase at the end of LLE for the different Lu: DOTA molar ratios.

Fig. 3

a The ¹⁷⁷Lu extraction efficiency and b the percent ^177mLu extracted at different extraction time during the successive LLE of free ¹⁷⁷Lu ions from [^177mLu]Lu-DOTA complex using 0.01 M DHEPA in DHE. The experiments were performed for three different Lu: DOTA molar ratios, (1:1) in black, (1:2) in red and (1:4) in blue. The data points represent the average ± STD of three experiments, the individual error in measurements due to counting statistics is less than 5%

It can be seen from Fig. 3(a) that the ¹⁷⁷Lu extraction efficiency reaches a maximum value of 95 ± 4% when Lu & DOTA were present in 1:1 M ratio and decreases to 58 ± 2% for 1:4 Lu:DOTA molar ratio. Further, the ¹⁷⁷Lu extraction efficiency remains almost constant for the first three extractions followed by a slight increase during the 4th and 5th extraction for all the three Lu:DOTA molar ratios. Figure 3(b) shows that 0.0061 ± 0.0015% of ^177mLu activity was extracted in the first extraction when Lu and DOTA were present in 1:1 M ratio, which got reduced to 0.0020 ± 0.0010% for the Lu:DOTA molar ratio 1:4. The percentage of ^177mLu activity extracted remains almost constant during the successive extractions in the presence of excess DOTA, and increases from 0.0061 ± 0.0015% to 0.0095 ± 0.0015% in the presence of 1:1 Lu:DOTA molar ratio. The error bars in Fig. 3 represent the standard deviation in the results of three experiments performed in parallel.

Figure 4 shows the ¹⁷⁷Lu/^177mLu activity ratios observed in the organic phase at the end of LLE for the three different Lu:DOTA molar ratios. It reveals that the ¹⁷⁷Lu/^177mLu activity ratio increases with an increase in the molar quantities of DOTA. The highest ¹⁷⁷Lu/^177mLu activity ratio of 3500 ± 500 was obtained when DOTA was present in excess (1:4) and decreases to around 1500 ± 600 in the presence of 1:1 Lu:DOTA molar ratio. Further, a slight decrease in the ¹⁷⁷Lu/^177mLu activity ratios was observed in every successive LLE performed during the 35 days of experiments. The fifth ¹⁷⁷Lu extraction performed at the end of the experiments resulted in a 40 ± 5% decrease in the ¹⁷⁷Lu/^177mLu activity ratios compared to the ¹⁷⁷Lu/^177mLu activity ratio obtained in the first extraction.

Fig. 4

The ¹⁷⁷Lu/ ^177mLu activity ratio obtained during the successive LLE of free ¹⁷⁷Lu ions from the [^177mLu]Lu-DOTA complex. The experiments were performed with three Lu: DOTA molar ratios, (1:1) in black, (1:2) in red and (1:4) in blue. The data points represent the average ± STD of three experiments, the individual error in measurements due to counting statistics is less than 5%

Overall, the ¹⁷⁷Lu/^177mLu activity ratios obtained using DOTA as chelating agent were about 5 times higher when compared with ¹⁷⁷Lu/^177mLu activity ratios obtained using DOTATATE for a ¹⁷⁷Lu accumulation period of around 7 days. Also, the percentage of ^177mLu activity extracted in the organic phase was about 5 times higher with DOTATATE than that observed with DOTA as the ^177mLu complexing agent.

Discussion

The separation of the isomers ¹⁷⁷Lu and ^177mLu based on the nuclear decay after effects is achieved using liquid-liquid extraction (LLE) as the separation method and the [^177mLu]Lu-DOTA, [^177mLu]Lu-DOTATATE complexes. The ¹⁷⁷Lu production at 77 K resulted in negligible dissociation of the starting [^177mLu]Lu-DOTA based complexes, and increases the quality of extracted ¹⁷⁷Lu remarkably. The freed ¹⁷⁷Lu ions were extracted in the organic phase by performing the LLE at room temperature. The separation was done sufficiently fast resulting in production of limited quantities of free ^177mLu ions.

In the present work, the ¹⁷⁷Lu/^177mLu activity ratio of 1086 ± 40 is achieved using [^177mLu]Lu-DOTATATE complex which is about 4 times higher than the previously reported ¹⁷⁷Lu/ ^177mLu activity ratio of 250 realized using the same [^177mLu]Lu-DOTATATE complex (Bhardwaj et al., 2017). In the previously reported method, the ¹⁷⁷Lu ion accumulation was performed at 10 °C and the temperature could not be decreased further because of experimental limitations. In contrast, the present LLE based separation allows the ¹⁷⁷Lu accumulation at 77 K. At 77 K, the rate constants for the chemical reactions (i.e. association-dissociation kinetics) are extremely low making the ^177mLu contribution coming from the dissociation of the [^177mLu]Lu-DOTATATE complex negligible during the ¹⁷⁷Lu accumulation period. The ^177mLu contribution observed in the present work can be accounted to the dissociation of the [^177mLu]Lu-DOTATATE complex during the LLE at room temperature. After the dissociation, the ^177mLu and ¹⁷⁷Lu ions are indistinguishable and they will go into the organic phase with equal rate.

The LLE of ¹⁷⁷Lu ions from [^177mLu]Lu-DOTATATE complex resulted in co-extraction of 0.0085 ± 0.0015% of initial ^177mLu activity in the organic phase. This leads to an estimated dissociation rate constant of 1.3*10^− 7 ± 0.3*10^− 7 s^− 1. For Lu-DOTATATE complex, a dissociation constant rate 2*10^− 8 s^− 1 has been reported at pH -4.3 and 20 °C (van der Meer et al., 2013). It has also been shown that the Lu-DOTATATE complex is accompanied by the presence of short-lived unstable, mono- and di-protonated (MHL, MH₂L) complex species (van der Meer et al., 2013). These species have a dissociation rate constant of 8*10^− 5 s^− 1 (MHL) & 2*10^− 4 s^− 1 (MH₂L) at pH -4.3 and 20 °C (van der Meer et al., 2013). Therefore, the presently estimated dissociation rate constant does not represent the dissociation of single species, but is rather a combination of the dissociation contribution from three different species i.e. ML, MHL, & MH₂L. Overall, the [^177mLu]Lu-DOTATATE complex behavior clearly highlights the fact that a careful consideration of all the possible species at a certain pH should be given while assessing the role of any complexing agent in ¹⁷⁷Lu-^177mLu separation.

The ¹⁷⁷Lu/^177mLu activity ratio obtained during the LLE of ¹⁷⁷Lu ions from [^177mLu]Lu-DOTATATE complex was found to be influenced by the length of the ¹⁷⁷Lu accumulation period. The highest ¹⁷⁷Lu/^177mLu activity ratio of 1086 ± 40 was obtained after ¹⁷⁷Lu accumulation period of 26 days and decreased to 600 ± 200 for accumulation periods of 5 to 10 days. This was expected as the amount of ¹⁷⁷Lu ions produced from the internal conversion of ^177mLu ions grows as the ¹⁷⁷Lu accumulation period increases. In contrast, the ^177mLu contribution is only due to dissociation of the complex taking place during the extraction. Additionally, a ¹⁷⁷Lu extraction efficiency of 60 ± 10% was observed which can be associated to the loss of free ¹⁷⁷Lu ions due to their re-association with the excess complexing agent, as reported before by Bhardwaj et al. (Bhardwaj et al., 2017).

The crucial role of association kinetics on ¹⁷⁷Lu- ^177mLu separation is further emphasised by studying the ¹⁷⁷Lu-^177mLu separation in the presence of varying amounts of DOTA as the complexing agent. The ¹⁷⁷Lu extraction efficiency obtained during the LLE of freed ¹⁷⁷Lu ions was affected by the applied ratio of complexing agent. The ¹⁷⁷Lu extraction efficiency of 58 ± 2% was achieved in the presence of excess DOTA (Lu:DOTA molar ratio, 1:4), and it increases to 95 ± 4% when Lu:DOTA was present in the molar ratio 1:1, confirming that the association kinetics of freed ¹⁷⁷Lu and the excess of DOTA play an important role in the process. Similarly, the extracted ^177mLu activity decreases from 0.0060 ± 0.0015% to 0.0020 ± 0.0010% with the increase in the Lu:DOTA molar ratios from (1:1) to (1:4) respectively, due to the re-association of ^177mLu ions with the excess of DOTA.

The ¹⁷⁷Lu/^177mLu activity ratios obtained during the LLE of ¹⁷⁷Lu ions from [^177mLu]Lu-DOTA complex were also found to be effected by the starting Lu:DOTA molar ratio. A ¹⁷⁷Lu/^177mLu activity ratio up to 3500 ± 500 was achieved when the LLE was performed using aqueous [^177mLu] LuDOTA complex with Lu:DOTA present in the molar ratio 1:4. Remarkably, the obtained ¹⁷⁷Lu/^177mLu activity ratios are very close to the ¹⁷⁷Lu/^177mLu activity ratios of 4000–10,000 associated to the direct-route production of ¹⁷⁷Lu supplied to the clinics (Das et al., 2007; Chakraborty et al., 2014). These ratios were found to decrease with the decrease in the amount of DOTA, i.e. an activity ratio of 1500 ± 600 was observed when Lu and DOTA were present in the molar ratio 1:1. The presence of excess DOTA leads to a proportional decrease in the amount of both ¹⁷⁷Lu and ^177mLu ions due to re-association. However, the ¹⁷⁷Lu production from internal conversion of ^177mLu ions adds to a constant positive contribution in the amount of ¹⁷⁷Lu ions, which leads to an overall increase in the ¹⁷⁷Lu/^177mLu activity ratios.

Finally, the observed decrease in the ¹⁷⁷Lu/^177mLu activity ratio with the increase in time are well in agreement with the theoretically expected ratios based on the ^177mLu and ¹⁷⁷Lu extracted shown in Fig. 3 and incorporating the effect of incomplete organic phase removal on every successive extraction (see Additional file 1 Figure S3, supplementary information). The reported separation method suffers from the drawback of incomplete organic phase removal during the LLE. The residual 1/3rd of the organic phase left unrecovered after every LLE contains un-extracted ¹⁷⁷Lu and ^177mLu ions. The ¹⁷⁷Lu ions will reduce to about a half after accumulation time of 7 days, but the ^177mLu ions will remain almost unchanged as they have a half-life of 160.44 days. They will add to the total amount of free ^177mLu ions in the successive extraction and correspondingly to a decrease the ¹⁷⁷Lu/ ^177mLu activity ratio. In case of a complete organic phase removal, the separation method could lead to a constant value of ¹⁷⁷Lu/^177mLu activity ratio of around 3500 on performing periodic ¹⁷⁷Lu extraction every 7 days. Additionally, the use of longer ¹⁷⁷Lu accumulation period of 32 days will lead to 1.7 times more ¹⁷⁷Lu production compared to 7 days ¹⁷⁷Lu accumulation period. This can potentially lead to an activity ratio of 7000 on considering a constant 0.0020 ± 0.0010% ^177mLu contribution due to dissociation and 58 ± 2% ¹⁷⁷Lu extraction efficiency. In such a case, the extracted ¹⁷⁷Lu would contain a ^177mLu contribution as low as 0.01% and would be comparable to the direct route ¹⁷⁷Lu production.

It should be pointed that the specific activity of the produced ¹⁷⁷Lu is not a discussed parameter since the starting ^177mLu source has very low specific activity and therefore also the extracted ¹⁷⁷Lu. Consequently, the values would not represent a fair comparison with the commercially available ¹⁷⁷Lu. Additionally, the extracted ¹⁷⁷Lu ions have not been stripped from the organic phase back into the aqueous phase considering that it is a well-reported process in literature (Trtic-Petrovic et al., 2010).

Overall, the presented work is an important milestone towards the development of a ^177mLu/¹⁷⁷Lu radionuclide generator for clinical application. It also establishes the possibility of employing other separation techniques such as micro-fluidic separation (Davide et al., 2014), membrane based liquid-liquid extraction (Pedersen-Bjergaard & Rasmussen, 2008) or an automatized LLE separation devices that can allow the commercialization of LLE based ^177mLu/ ¹⁷⁷Lu radionuclide generator. However, there are several aspects that needs further investigation and optimization. Firstly, the back extraction of ¹⁷⁷Lu from the organic phase and the complete removal of any traces of organic solvents will be crucial for its potential commercialization. Secondly, this work has been performed at lab-scale with low activity levels and excludes the effect of radiolysis on the proposed ^177mLu-¹⁷⁷Lu separation method. The radiolysis can impact the quality of the produced ¹⁷⁷Lu and should be carefully evaluated in the future investigations. Lastly, the described method can be further optimized in terms of shorter extraction time, use of lower temperature to perform the ¹⁷⁷Lu extraction improve the produced ¹⁷⁷Lu quality.

Conclusion

A novel ^177mLu-¹⁷⁷Lu separation method is developed that allows the ¹⁷⁷Lu production via internal conversion of ^177mLu at low temperatures (77 K) and the use of ultra-stable ^177mLu complexes with liquid-liquid extraction. For the best conditions, the [^177mLu]Lu-DOTA complex and LLE provides a ¹⁷⁷Lu/^177mLu activity ratio of 3500 ± 500. A value that is close to the ¹⁷⁷Lu/^177mLu activity ratio 4000–10,000 obtained during the ¹⁷⁷Lu production via the direct route and exemplifies the potential applicability of the ^177mLu/¹⁷⁷Lu generator in clinical studies. Future research will be focused on further optimization of novel ¹⁷⁷Lu-^177mLu separation technologies aimed to ultimately lead to a clinically applicable ^177mLu/¹⁷⁷Lu radionuclide generator. The around the clock availability of ¹⁷⁷Lu via a ^177mLu/ ¹⁷⁷Lu radionuclide generator can significantly accelerate the research on ¹⁷⁷Lu based radiopharmaceuticals and help in realizing its full potential in nuclear medicine.