Contribution of Auger/conversion electrons to renal side effects after radionuclide therapy: preclinical comparison of 161Tb-folate and 177Lu-folate

Background The radiolanthanide 161Tb has, in recent years, attracted increasing interest due to its favorable characteristics for medical application. 161Tb exhibits similar properties to the widely-used therapeutic radionuclide 177Lu. In contrast to 177Lu, 161Tb yields a significant number of short-ranging Auger/conversion electrons (≤50 keV) during its decay process. 161Tb has been shown to be more effective for tumor therapy than 177Lu if applied using the same activity. The purpose of this study was to investigate long-term damage to the kidneys after application of 161Tb-folate and compare it to the renal effects caused by 177Lu-folate. Methods Renal side effects were investigated in nude mice after the application of different activities of 161Tb-folate (10, 20, and 30 MBq per mouse) over a period of 8 months. Renal function was monitored by the determination of 99mTc-DMSA uptake in the kidneys and by measuring blood urea nitrogen and creatinine levels in the plasma. Histopathological analysis was performed by scoring of the tissue damage observed in HE-stained kidney sections from euthanized mice. Results Due to the co-emitted Auger/conversion electrons, the mean absorbed renal dose of 161Tb-folate (3.0 Gy/MBq) was about 24 % higher than that of 177Lu-folate (2.3 Gy/MBq). After application of 161Tb-folate, kidney function was reduced in a dose- and time-dependent manner, as indicated by the decreased renal uptake of 99mTc-DMSA and the increased levels of blood urea nitrogen and creatinine. Similar results were obtained when 177Lu-folate was applied at the same activity. Histopathological investigations confirmed comparable renal cortical damage after application of the same activities of 161Tb-folate and 177Lu-folate. This was characterized by collapsed tubules and enlarged glomeruli with fibrin deposition in moderately injured kidneys and glomerulosclerosis in severely damaged kidneys. Conclusions Tb-folate induced dose-dependent radionephropathy over time, but did not result in more severe damage than 177Lu-folate when applied at the same activity. These data are an indication that Auger/conversion electrons do not exacerbate overall renal damage after application with 161Tb-folate as compared to 177Lu-folate, even though they result in an increased dose deposition in the renal tissue. Global toxicity affecting other tissues than kidneys remains to be investigated after 161Tb-based therapy, however.


Background
The radiolanthanide terbium-161 ( 161 Tb) has garnered increasing interest in recent years due to its favorable properties for medical application [1]. 161 Tb has similar chemical properties to the clinically established lutetium-177 ( 177 Lu), which enables stable coordination using a DOTA-chelator and, consequently, its application with well-established targeting agents [2]. The physical properties of 161 Tb are also similar to those of 177 Lu (T 1/2 = 6.7 d, Eβ − av = 134 keV, linear energy transfer (LET) of about 0.34 keV/μm [3]). 161 Tb decays, with a half-life of 6.9 days, by the emission of low-energy β − -particles (Eβ − av = 154 keV [3]) with a maximal tissue range of 0.29 mm and a LET of around 0.32 keV/μm, which is suitable for the treatment of metastasized malignancies. Importantly, 161 Tb exhibits a co-emission of γ-radiation, enabling imaging using single photon emission computed tomography (SPECT) ( Table 1). The decay process of 161 Tb also releases a significant number of Auger/conversion electrons of an energy ≤50 keV (~12.4 e − , 46.5 keV per decay) [4], which are postulated to contribute to therapeutic anti-tumor effects [2]. This may be an advantage over 177 Lu which emits a negligible number of Auger/ conversion electrons in comparison (~1.11 e − , 1.0 keV per decay, Table 1) [4]. High-energy electrons emitted during β − -decay have a LET of about 0.2 keV/μm and a tissue range of 0.5-12 mm, suitable for larger metastases [5]. The high LET (~4-26 keV/μm) and short tissue range (~2-500 nm) of Auger/conversion electrons may be of particular value for the treatment of single tumor cells or tumor cell clusters [5]. This assumption has been corroborated by dosimetric calculations published by Uusijärvi et al., which clearly indicate the superiority of 161 Tb over 177 Lu for the treatment of small lesions (<200 μm) [6].
Another advantage of 161 Tb may be the fact that other Tb nuclides exist that emit diagnostic radiation only, potentially allowing pre-therapeutic dosimetry and monitoring therapy response with chemically identical imaging agents using SPECT ( 155 Tb) and positron emission tomography (PET) ( 152 Tb) [7].
The few studies in which 161 Tb was applied thus far, focused on basic in vivo behavior of 161 Tb-labeled biomolecules [8] and their effects on the malignant tissue in comparison to 177 Lu [2,9]. These preliminary therapy studies revealed that the therapeutic effect of 161 Tblabeled compounds was superior to the effect of their 177 Lu-labeled counterparts, if applied at the same activities as indicated by dosimetric calculations [2].
Potential differences between undesired side effects of 161 Tb and 177 Lu have not been elucidated, however. The most critical organs in terms of side effects after radionuclide therapy are the bone marrow, populated by the highly radiosensitive hematopoietic cells, and the kidneys, which accumulate radioactivity of small molecular weight targeting agent at a high extent [10]. In the present study, we focused on the investigation of renal damage after exposure to therapeutic doses of 161 Tb and 177 Lu using a folate-based small molecular targeting agent.
The folate receptor (FR) has been found expressed in a wide variety of different tumor types, including cancer of the ovaries, endometrium, lungs, and kidneys [11]. It has therefore been exploited (pre)clinically for targeting with folate-based radioconjugates for nuclear imaging purposes [12,13]. Following receptor binding on the cell surface, folic acid and its conjugates are internalized via FR-mediated endocytosis [14,15]. This uptake mechanism makes folic acid also attractive for FR-targeted chemotherapy [16]. Recently, FR-targeted radionuclide therapy was exemplified using a DOTA-folate conjugate which was labeled with 177 Lu [17]. It was shown that the application of ≥20 MBq of 177 Lu-folate resulted in complete tumor remission in mice bearing KB tumor xenografts.
Targeted tumor therapy with folic acid-based radioconjugates is, however, known to present a risk of damage to the kidneys, since radiofolates accumulate specifically in the proximal tubule cells where the FR is expressed [11,18]. This topic has been addressed recently, in a detailed investigation of radionephropathy after administration of 177 Lu-folates with different in vivo distribution profiles [19].
The purpose of the current long-term study was to explore undesired side effects to the kidneys of mice  after radionuclide therapy with high doses of 161 Tbfolate and compare its nephrotoxic profile with the  results obtained with 177 Lu-folate. Histopathological  changes of the renal tissue were examined after treatment with 161 Tb-folate and compared to the results  obtained with 177 Lu-folate. Time-dependent monitoring of kidney function was performed over the whole period of investigation by the determination of blood plasma parameters, which are indicative for renal damage. Moreover, renal uptake of 99m Tc-DMSA was quantified using SPECT. This has proven to be a reliable method to determine kidney function in mice non-invasively after radionuclide therapy. The results of plasma parameters and SPECT were correlated with the data obtained from histological evaluation.

Methods
Preparation of 161 Tb-folate and 177 Lu-folate The DOTA-folate conjugate (cm09), herein referred to as "folate," was kindly provided by Merck & Cie (Schaffhausen, Switzerland) [17]. No-carrier-added 161 Tb was produced at the spallation-induced neutron source (SINQ) at Paul Scherrer Institut (PSI) (Villigen-PSI, Switzerland) or at the high-flux reactor of the Institut Laue-Langevin (ILL, Grenoble, France) using various quantities of enriched 160 Gd targets. Separation of 161 Tb from the target material was carried out at PSI using cation exchange chromatography, as previously reported [1,7]. No-carrier-added 177 Lu was purchased from Isotope Technologies Garching (ITG GmbH, Garching, Germany). The radionuclidic purity of 161 Tb and 177 Lu, respectively, was ≥99 %. A HCl solution (0.05 M) of 161 TbCl 3 or 177 LuCl 3 (500 MBq, 100 μL) was mixed with sodium acetate (0.5 M, pH 8, 20 μL) to obtain a pH of~4.5. After the addition of the folate conjugate (25 μL, 1 mM), the reaction mixture was incubated for 10 min at 95°C. The resulting specific activity of the radiofolates was 20 MBq/nmol. Quality control was performed using HPLC as previously reported [17]. Before in vivo application, the radioconjugate solution was diluted with phosphate buffered saline (PBS) (pH 7.4) and Na-DTPA (10 μL, 5 mM, pH 5) was added for the complexation of potential traces of unreacted 161 Tb(III) or 177 Lu(III).

Animal studies
The in vivo experiments were approved by the local veterinarian department and conducted in accordance with the Swiss law of animal protection. Female athymic nude mice (Crl:CD1-Foxn1 nu , 5-to 7-week-old) were purchased from Charles River Laboratories (Sulzfeld, Germany). The mice received a folate-deficient rodent chow (ssniff Spezialdiäten GmbH, Soest, Germany), starting 1 week prior to the injection of the radiofolates.
The folate-free diet was replaced by a standard rodent diet (Kliba Nafag, Kaiseraugst, Switzerland) 3 weeks after application of the radiofolates.

Dosimetry
Dosimetric calculations were performed based on biodistribution data, which were previously obtained after the injection of 161 Tb-folate and 177 Lu-folate (2-3 MBq/ 0.5 nmol per mouse) using the same strain of mice (Crl:CD1-Foxn1 nu ) [2,17,19]. To estimate the mean absorbed radiation dose of 161 Tb-folate and 177 Lu-folate in the renal tissue, calculations were performed as follows: the cumulative radioactivity was calculated by fitting a bi-exponential curve to the non-decaycorrected biodistribution data (% IA/g). The area under the curve (MBq•s) was determined by integrating the bi-exponential function to infinity. The mean absorbed dose to the kidneys was assessed for a kidney mass of 125 mg, having an elliptical shape and uniform activity distribution [20]. For the calculation of the absorbed fraction, the Monte Carlo code PE-NELOPE was used, resulting in an absorbed fraction of 0.93. In the simulation, detailed simulation mode was employed and the electron-absorption energy was 100 eV [21]. The mean absorbed dose (Gy/MBq) was calculated by multiplying the area under the curve normalized to 1 MBq injected activity (IA), with the absorbed fraction and the emitted energy per decay for 161 Tb and 177 Lu [3] and multiplied with a conversion factor. with W X = body weight at day x and W 0 = body weight at day 0) and indicated as average RBW for each group.
Determination of renal uptake of 99m Tc-DMSA using SPECT The renal function of mice was monitored by the determination of the renal uptake of 99m Tc-DMSA using SPECT, as previously reported [22,23]. DMSA (Tech-neScan®, Mallinckrodt, Petten, The Netherlands) was radiolabeled with 99m Tc (radionuclidic purity >99.9 %), obtained from a 99 Mo/ 99m Tc-generator (Mallinckrodt, Petten, The Netherlands) at an activity concentration of 3 GBq in 5 mL. Quality control performed by TLC revealed a radiochemical purity of >95 %. SPECT acquisitions were performed 2 h after the injection of 99m Tc-DMSA (30-40 MBq per mouse) using a NanoSPECT/ CT™ (Mediso Medical Imaging Systems, Budapest, Hungary) and Nucline Software (version 1.02, Bioscan Inc., Poway, USA). For this purpose, the energy window of 99m Tc was set to 140.5 ± 14 keV. The acquired data were reconstructed using HiSPECT software (version 1.4.3049, Scivis GmbH, Göttingen, Germany), and the renal uptake of 99m Tc-DMSA was determined, as previously reported, using the VivoQuant post-processing software (version 1.23, inviCRO Imaging Services and Software, Boston, USA) [19]. The determined activities were decay-corrected and expressed as percentage of injected activity (% IA) per kidney.

Determination of blood plasma parameters
Blood urea nitrogen and creatinine were measured in weeks 8, 19, and 26 and at the day of euthanasia of the mice. Blood was taken from the sublingual vein of each mouse, collected in heparinized vials, and centrifuged. Blood plasma samples were analyzed using a Fuji Dri-Chem 4000i analyzer (Polymed Medical Center AG, Glattbrugg, Switzerland).

Histological analysis
All animals were euthanized with CO 2 , followed by exsanguination. The kidneys were removed and fixed in 4 % buffered formalin (Formafix buffered 4 %, Switzerland AG, Hittnau, Switzerland). The tissue was trimmed, dehydrated, and embedded in paraffin wax. Sections of 3-5 μm thickness were prepared, mounted on glass slides, deparaffinized in xylene, and rehydrated through graded alcohols, before staining with hematoxylin and eosin (HE) for the histological examination. Renal injury was assessed by light microscopy. Glomerular, tubular, and interstitial lesions were evaluated separately using pre-defined criteria ( Table 2), adapted from previously published studies [19,24]. The values obtained for each compartment were summed to obtain a cumulative score. The cumulative score was then converted to a final score ranging from 0 to 5 (Table 3). No damage (score 0) indicated normal histology, while variably severe renal injury ranged from minimal damage (score 1), characterized by mildly reduced numbers of capillaries within the glomerular tuft and patchy tubular collapse, to severe renal damage (score 5), dominated by widespread glomerulosclerosis and prominent cortical Tubules: <10 % of cortical tubules collapsed or degenerated.
Interstitium: few small mononuclear cell infiltrates and/or minimal fibrosis in the cortex.
2 Glomeruli: mild/moderate fibrin deposition and reduced number of capillaries in numerous glomeruli.
Interstitium: multifocal small mononuclear cell infiltrates and/or mild fibrosis in the cortex. 3 Glomeruli: marked fibrin deposition and reduced number of capillaries in numerous glomeruli.
Interstitium: multifocal larger mononuclear cell infiltrates and/or moderate fibrosis in the cortex. 4 Glomeruli: marked fibrin deposition and reduced number of capillaries in numerous glomeruli; glomerulosclerosis in a few to several glomeruli.
Interstitium: multifocal larger mononuclear cell infiltrates and/or moderate fibrosis in the cortex and medulla; mild/moderate parenchymal collapse.
Interstitium: severe parenchymal collapse, affecting cortex and medulla. Table 3 The partial scoring of each compartment (glomeruli, tubules, and interstitium) was summed to obtain the cumulative score The cumulative score was then converted to a final score indicating the degree of renal injury parenchymal shrinkage. A phosphotungstic acidhematoxylin (PTAH) staining was performed on tissues of selected mice for the identification of fibrin.

Statistical analysis
Data are presented as mean ± standard deviation. Statistics were conducted by using one-way ANOVA with Bonferroni's multiple comparison post-test (GraphPad Prism, version 5.01).

Radiolabeling
The radiolabeling of the folic acid conjugate was successfully achieved with 161 Tb and 177 Lu, respectively, at a specific activity of 20 MBq/nmol. Quality control performed by HPLC revealed a radiochemical yield >97 %.

Body weight and survival
All mice gained weight over the first month of the experiment, as a consequence of normal growth. Mice which were injected with 10 or 20 MBq of 161 Tb-folate and 177 Lu-folate, respectively, showed body weight gain within the first 6 weeks, followed by stabilization of the body weight in the range of 1.1 to 1.3 RBW (Fig. 1a). The majority of these mice (>50 %), as well as the control mice (groups A, B, C, E, and F), survived until the end of the study in week 34, and consequently, the survival time remained undefined (Fig. 1b) (Fig. 1b). This difference was determined not to be significant.
Renal uptake of 99m Tc-DMSA When the study reached week 3, all treated mice showed comparable renal uptake of 99m Tc-DMSA, which was in the same range as for the control mice of group A (data not shown). The mice injected with low activities of 161 Tb-folate (group B; 10 MBq) and 177 Lu-folate (group E; 10 MBq), respectively, showed reduced 99m Tc-DMSA uptake only 30 weeks after radiofolate injection, whereas no significant deviation from controls was observed at earlier time points (Fig. 2). In mice which were treated with 20 MBq of 161 Tb-folate and 177 Lu-folate, respectively, the uptake of 99m Tc-DMSA was 7.6 ± 0.5 % IA/kidney (group C) and 8.5 ± 1.2 % IA/kidney (group F) in week 15 and, thus, slightly lower than the value obtained for control mice (11 ± 1.8 % IA/kidney). In these mice (groups C and F), renal uptake of 99m Tc-DMSA decreased further in week 22 and week 30 (Fig. 2). The reduction of 99m Tc-DMSA was, however, comparable in mice injected with 161 Tb-folate and 177 Lu-folate (10 and 20 MBq), respectively. Mice injected with the highest activity of 161 Tb-folate (group D; 30 MBq) and 177 Lu-folate (group G; 30 MBq) showed a significantly reduced renal uptake of 99m Tc-DMSA when compared to controls (group A) already in week 15 (Fig. 2). The determination of 99m Tc-DMSA uptake in week 15 revealed a significantly reduced value for mice injected with 30 MBq 161 Tb-folate, compared to mice, which received 30 MBq 177 Lu-folate. At later time points, the obtained values were, however, always comparable between these mice (groups D and G).

Renal plasma parameters
Values of blood urea nitrogen and creatinine were not significantly different for control mice and mice treated with 10 MBq 161 Tb-folate (groups B) over the entire period of investigation (Table 5). Mice which received 20 MBq 161 Tb-folate (group C) showed slightly, but not significantly, increased levels of blood urea nitrogen as compared to control mice from week 26 on. Animals that received higher activities of 161 Tb-folate (30 MBq, group D), however, showed significantly increased levels of blood urea nitrogen at terminal bleeding (42 ± 8.1 mM (n = 4) and > 50 mM (n = 2)) as compared to control mice (group A: 8.4 ± 1.4 mM). In the 177 Lu-folate-treated groups, increasing levels of blood urea nitrogen were also observed in mice injected with 30 MBq (>50 mM). Moreover, significantly increased levels of blood urea nitrogen were also determined at terminal  Table 5). Application of high activities of 161 Tb-folate (group D, 30 MBq) resulted in a significant increase of creatinine plasma levels (77 ± 17 μM), as compared to control mice of group A (24 ± 5.1 μM (n = 6), <18 μM (n = 3)) at terminal state and the same observation (creatinine level of 107 ± 35 μM) was also made for 177 Lu-folate treated mice (group G, 30 MBq). In mice which received 20 MBq 161 Tb-folate (group C) and 177 Lu-folate (group F), respectively, slightly, but not significantly, increased levels of creatinine were determined in Week 26 and at terminal bleeding.

Renal histopathology
The application of 161 Tb-folate and 177 Lu-folate resulted in a dose-dependent renal damage consistent with radiation nephropathy ( Table 6). The kidneys of the control animals did not show histological abnormality and received a final score of 0. The renal tissue of mice injected with increasing activities of radiofolates showed progressively more severe kidney damage, which was graded with final scores between 1 and 5 (Fig. 3). For the mice treated with 161 Tb-folate, final scores of 1 (group B, 10 MBq), 3 (group C, 20 MBq), and 4 (group D, 30 MBq) were determined. In the case of 177 Lu-folate application, renal damage was scored with 2 (group E, 10 MBq), 3 (group F, 20 MBq), and 4 (group G, 30 MBq). Lower scores were characterized by localized tubular collapse, high proportions of viable tubules, and enlarged glomeruli with reduced numbers of capillaries. These changes became more prominent in the severely injured kidneys, where occasional to frequent glomerulosclerosis was also observed (Fig. 3). Moreover, a gradual thinning of the cortex with increasing amounts of applied radiofolate was observed during the investigation of renal cross sections (Fig. 3).

Correlation of kidney function tests and renal histopathology
Retrospectively, renal function parameters determined shortly before euthanasia were correlated with the histology scores on an individual basis (Fig. 4). Increasing blood urea nitrogen levels correlated (r 2 = 0.87) well with rising scores of radiation nephropathy (Fig. 4a). The creatinine levels indicated, however, only a slight trend in the same direction as they showed poor correlation with the scoring results (r 2 = 0.52). A relatively good correlation (r 2 = 0.68) was found between the determination of renal damage using 99m Tc-DMSA/SPECT and the overall score determined by histopathological evaluation of the renal damage.

Discussion
A cumulative renal dose of >25 Gy as a result of 177 Luor 90 Y-based radionuclide therapy comprises a high risk of renal damage in patients [25]. In contrast, radionuclide therapy based on Auger/conversion electrons using 111 In-octreotide did not impair kidney function up to a renal dose of 45 Gy [26]. It was hypothesized that, because of their short tissue range, Auger/conversion electrons are unable to damage the radiosensitive glomeruli and potential cell damage would, therefore, be limited to the more radioresistant tubular cells, in which the radioconjugates accumulate [26]. Thus, it was speculated that the accumulation of activity into specific organ structures, yielding a local energy deposit for short-ranged electrons, might be the reason for the reduced impairment of renal function after therapy with shortranged Auger/conversion electrons, as compared to treatment with the long-ranged β − -particles.
The same activity of 161 Tb-folate and 177 Lu-folate did not cause significant differences regarding renal uptake of 99m Tc-DMSA and rising levels of plasma parameters, indicative for kidney function, over the whole period of investigation. This observation indicates that the impairment of renal function runs in parallel for 161 Tb-folate and 177 Lu-folate. A significantly lower renal accumulation of 99m Tc-DMSA was only found in week 15 for mice treated with 30 MBq 161 Tb-folate (Group D) compared to mice injected with 30 MBq 177 Lu-folate (Group G). Other than that, the determination of blood urea nitrogen and creatinine plasma levels did not confirm a significant difference of the renal function in mice of groups D and G at any time of the study. In contrast, at terminal state, there was even a tendency of increased blood urea nitrogen and creatinine levels in 177 Lu-folatetreated mice (Group G) as compared to the levels determined in 161 Tb-folate-treated mice (Group D).
Histopathological changes were investigated with renal tissue of euthanized mice, when an end-point criterion was reached or after termination of the study in Week 34. A reduced renal mass and thinning of the renal cortex was observed along with increasing tissue damage in the renal cortex, where radiofolates accumulate because of FR-expression in the proximal tubular cells [11,18]. Evaluation of the partial and cumulative scores in the tubules, glomeruli, and interstitium showed equal distribution and location of renal damage independent on the radionuclide which was employed (Table 6). This means Tc-DMSA uptake in the kidneys that the application of 161 Tb-folate resulted in renal damage that was comparable to that observed after the application of 177 Lu-folate at the same activity, in spite of the 1.3-fold higher dose which was deposited in the case of 161 Tb-folate. These data are in line with the hypothesis that the additional dose of 161 Tb-folate, due to Auger/conversion electrons, has no adverse effect on renal function and morphology. This theory appears plausible for radiopharmaceuticals that accumulate primarily in tubular cells, as it is the case for folateand peptide-based radioconjugates [18,[28][29][30]. Auger/ conversion electrons deposit energy over subcellular dimensions only and, hence, no other parts of the kidneys will be reached other than the tubules. Our observation is in agreement with the theory that in the case of kidneys which is defined as flexible tissue (in contrast to hierarchical tissues [31]), the irradiated volume and affected suborgan unit are critical factors for the final renal injury [32,33].
The above reported results are corroborated by the fact that values determined for blood urea nitrogen levels correlated well with the histopathological analysis which was performed in an independent setting (Fig. 4). The value of creatinine levels for predicting renal damage appears to be questionable due to high variability and poor correlation with the histopathological changes. The unreliable value of creatinine as an indicator for renal damage was previously reported in the literature [34]. From a methodological point of view, 99m Tc-DMSA/SPECT is undoubtedly the most sensitive method for non-invasive monitoring of renal function.

Conclusions
The functional and histopathological analysis of the kidneys after application of 161 Tb-folate revealed a dosedependent damage which was comparable to the damage caused by 177 Lu-folate applied using the same activity. These observations are in line with the hypothesis that Auger electrons and low energy conversion electrons do not result in additional renal injury, as previously observed during therapeutic application of 111 In-octreotide in patients. The confirmation of these results, in more detailed future studies, will support the clinical application of 161 Tb-based therapy, which could contribute to increase the success rate of the treatment without causing additional renal side effects.

Compliance with ethical standards
This article does not contain any studies with human participants performed by any of the authors.
All applicable international, national, and/or institutional guidelines for the care and use of test animals were followed.