Preclinical evaluation of 5-methyltetrahydrofolate-based radioconjugates—new perspectives for folate receptor–targeted radionuclide therapy

Purpose The folate receptor (FR) is frequently overexpressed in a variety of tumor types and, hence, an interesting target for radionuclide therapy. The aim of this study was to evaluate a new class of albumin-binding radioconjugates comprising 5-methyltetrahydrofolate (5-MTHF) as a targeting agent and to compare their properties with those of the previously established folic acid-based [177Lu]Lu-OxFol-1. Methods [177Lu]Lu-6R-RedFol-1 and [177Lu]Lu-6S-RedFol-1 were investigated in vitro using FR-positive KB tumor cells. Biodistribution studies were performed in KB tumor-bearing mice, and the areas under the curve (AUC0 → 120h) were determined for the uptake in tumors and kidneys. [177Lu]Lu-6R-RedFol-1 was compared with [177Lu]Lu-OxFol-1 in a therapy study over 8 weeks using KB tumor-bearing mice. Results Both radioconjugates demonstrated similar in vitro properties as [177Lu]Lu-OxFol-1; however, the tumor uptake of [177Lu]Lu-6R-RedFol-1 and [177Lu]Lu-6S-RedFol-1 was significantly increased in comparison with [177Lu]Lu-OxFol-1. In the case of [177Lu]Lu-6S-RedFol-1, also the kidney uptake was increased; however, renal retention of [177Lu]Lu-6R-RedFol-1 was similar to that of [177Lu]Lu-OxFol-1. This led to an almost 4-fold increased tumor-to-kidney AUC0 → 120h ratio of [177Lu]Lu-6R-RedFol-1 as compared with [177Lu]Lu-6S-RedFol-1 and [177Lu]Lu-OxFol-1. At equal activity, the therapeutic effect of [177Lu]Lu-6R-RedFol-1 was better than that of [177Lu]Lu-OxFol-1, reflected by a slower tumor growth and, consequently, an increased median survival time (49 days vs. 34 days). Conclusion This study demonstrated the promising potential of 5-MTHF-based radioconjugates for FR-targeting. Application of [177Lu]Lu-6R-RedFol-1 resulted in unprecedentedly high tumor-to-kidney ratios and, as a consequence, a superior therapeutic effect as compared with [177Lu]Lu-OxFol-1. These findings, together with the absence of early side effects, make [177Lu]Lu-6R-RedFol-1 attractive in view of a future clinical translation. Electronic supplementary material The online version of this article (10.1007/s00259-020-04980-y) contains supplementary material, which is available to authorized users.

The folate receptor (FR) is a membrane-anchored glycoprotein, which is overexpressed in gynecological and other tumor types, including lung, breast, and colon cancers [7][8][9]. Folic acid radioconjugates have been translated to clinics for nuclear imaging of FR-positive tumors [10,11]; however, their therapeutic exploitation remains challenging. The relatively low tumor-to-kidney ratio of accumulated folate radioconjugates would limit the applicable quantity of activity in order to prevent the risk of damage to the kidneys [12].
The high renal uptake of folate-based radiopharmaceuticals has been addressed with various strategies [13], including pharmacological interactions [14][15][16]. A major step forward was achieved by introducing an albuminbinding entity into the structure of radiofolates to prolong their b lood circul ation [ 1 7, 18 ] . T h e r e s u l t i n g radioconjugates revealed significantly increased tumor uptake and improved tumor-to-kidney ratios, which enabled their use in preclinical therapy studies in mice [17,19]. The therapeutic effects of this approach were promising; however, the kidneys remained the dose-limiting organ.
Recently, Boss et al. reported on results obtained with a novel class of fluorine-18-based radiotracers, in which folic acid (oxidized version of folate) was exchanged with the two s t e r e o i s o m e r s ( 6 R a n d 6 S , r e s p e c t i v e l y ) o f 5methyltetrahydrofolate   [20,21]. It was found that 6R-aza-[ 18 F]fluoro-5-MTHF as well as 6S-aza-[ 18 F]fluoro-5-MTHF accumulated to a significantly higher extent in the tumor tissue than aza-[ 18 F]fluorofolic acid ( 18 F-AzaFol), in which folic acid was employed as a targeting agent [21]. Importantly, the 6R-aza-[ 18 F]fluoro-5-MTHF isomer was cleared much more effectively through the kidneys as compared with 18 F-AzaFol.
Thus, the aim of this study was to translate the concept of using 5-MTHF as a targeting agent to albumin-binding DOTA conjugates in order to increase the tumor uptake and possibly reduce renal retention of activity. 6R-RedFol-1 and 6S-RedFol-1, based on 6R-5-MTHF and 6S-5-MTHF, respectively, were designed as structural equivalents to the previously developed albumin-binding DOTA-folate conjugate (cm10, herein referred to as OxFol-1 [18]) (Fig. 1). 6R-RedFol-1 and 6S-RedFol-1 were labeled with lutetium-177 and evaluated in vitro and in vivo for comparison of their characteristics with those of [ 177 Lu]Lu-OxFol-1 [18]. Therapy studies with KB tumor-bearing mice were performed in order to compare the therapeutic effect of the more promising [ 177 Lu]Lu-RedFol-1 isomer with [ 177 Lu]Lu-OxFol-1.

Determination of logD values
Distribution coefficients (logD values) of the folate radioconjugates were determined by a shake-flask method using n-octanol and PBS pH 7.4 followed by phase separation, as previously reported (Supplementary Material) [18].

Binding affinity to mouse and human plasma proteins
To compare the plasma protein-binding properties of the folate radioconjugates, the percentage of the fraction bound to mouse and human plasma proteins was determined at variable plasma dilutions calculated as [albumin]-to-[radioconjugate] molar ratios. Determination of the binding affinity was performed by measuring the free fraction of the radioconjugate separated from the albumin-bound fraction using an ultrafiltration method as previously reported (Supplementary Material) [23].
Tumor cell culture and cell uptake studies KB tumor cells (cervical carcinoma cell line, subclone of HeLa cells, ACC-136) were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ, Germany). Cells were cultured in folate-deficient RPMI medium (FFRPMI, Cell Culture Technologies GmbH, Gravesano, Switzerland) supplemented with 10% fetal calf serum, L-glutamine, and antibiotics.
Cellular uptake and internalization studies were performed with all folate radioconjugates according to a previously published procedure (Supplementary Material) [18]. The results were expressed as percentage of total added activity and presented as average ± SD of 3-6 independent experiments performed in triplicate.

In vivo studies
All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. In particular, all animal experiments were carried out according to the guidelines of Swiss Regulations for Animal Welfare. The preclinical studies have been ethically approved by the C antonal C ommittee of A nimal Experimentation and permitted by the responsible cantonal authorities.
Five-to 6-week-old female, athymic nude mice (CD-1 Foxn-1/nu) were purchased from Charles River Laboratories (Sulzfeld, Germany) and fed with a folatedeficient rodent diet (ssniff Spezialdiäten GmbH; Soest, Germany). Mice were subcutaneously inoculated with KB tumor cells (5 × 10 6 cells in 100 μL PBS) on both shoulders for biodistribution and imaging studies or with KB tumor cells (4.5 × 10 6 cells in 100 μL PBS) on the right shoulder for the therapy study.

In vivo stability of folate radioconjugates
In vivo stability studies were performed in mice without tumors (n = 2), injected with the folate radioconjugates (25 MBq, 0.5 nmol, 100 μL). After precipitation of proteins in plasma of blood samples collected at 4 h p.i., the samples were analyzed using HPLC (Supplementary Material).

Biodistribution studies
Biodistribution studies were performed 10-14 days after tumor cell inoculation when the tumor size reached a volume of 300 mm 3 . Mice (n = 4) were injected into a lateral tail vein with the respective folate radioconjugate (3 MBq, 0.5 nmol, 100 μL) diluted in PBS containing 0.05% bovine serum albumin (BSA). The animals were sacrificed at various timepoints up to 120 h after administration of the radioconjugates. Additional mice (n = 3) were injected with excess folic acid (100 μg in PBS pH 7.4),~5 min prior to the folate radioconjugates and sacrificed 1 h later (Supplementary Material). Selected tissues and organs were collected, weighed, and counted using a γ-counter (PerkinElmer, Wallac Wizard 1480). The results were listed as a percentage of the injected activity per gram (% IA/g) of tissue mass, using counts of a defined volume of the original injection solution measured at the same time, resulting in decay-corrected values.

Determination of areas under the curve
Biodistribution data were converted to non-decay-corrected values to obtain the time-dependent curves of accumulated activity in the tumor xenografts, blood, kidneys, and liver. The data points were used to calculate the areas under the curve (AUC) using GraphPad Prism (version 7) as previously reported [18]. The AUC 0 → 120h values of the [ 177 Lu]Lu-OxFol-1 were set as 1.0 to determine the relative values of

SPECT/CT imaging studies
The acquisition and analysis of images were performed with a dedicated small-animal SPECT/CT scanner (NanoSPECT/ CT™, Mediso Medical Imaging Systems, Budapest, Hungary) as previously reported (Supplementary Material) [18]. Mice were injected with the folate radioconjugates (25 MBq, 0.5 nmol) and scanned at 4 h and 24 h post injection (p.i.) Images were prepared using VivoQuant post-processing software (version 3.5, inviCRO Imaging Services and Software, Boston, USA). A Gauss post-reconstruction filter (FWHM = 1 mm) was applied, and the scale of activity was set as indicated on the images.

Therapy study
Mice were randomly assigned to five groups consisting of 6-9 animals 4 days after tumor cell inoculation when tumors reached an average size of 60-100 mm 3 Table 1). The relative body weight (RBW) was defined as [BW x /BW 0 ], where BW x is the body weight in gram at a given day x and BW 0 is the body weight in gram at day 0. The tumor dimensions were determined by measuring the longest tumor axis (L) and its perpendicular axis (W) with a digital caliper. The tumor volume (TV) was calculated according to the equation [TV = 0.5 × (L × W 2 )]. The relative tumor volume (RTV) was defined as [TV x / TV 0 ], where TV x is the tumor volume in cubic millimeters at a given day x and TV 0 is the tumor volume in cubic millimeters at day 0. Endpoint criteria were defined as (i) a tumor volume of ≥ 1000 mm 3 , (ii) loss of ≥ 15% of initial body weight, (iii) a combination of a tumor size of ≥ 800 mm 3 and body weight loss of ≥ 10% and/or (iv) ulceration of the tumor, and/or (v) abnormal behavior, indicating pain or unease. Mice were removed from the study and euthanized when an endpoint was reached.

Assessment of the therapy study
The efficacy of the radionuclide therapy was assessed by determination of the tumor growth delay (TGD x ), which was calculated as the time required for the tumor volume to increase x-fold over the initial volume at day 0. The tumor growth delay indices [TGDI x = TGD x (T)/TGD x (C)] were calculated as the TGD x ratio of treated mice (T) over control mice (C) for a 2-fold (x = 2, TGD 2 ) and 5-fold (x = 5, TGD 5 ) increase of the initial tumor volume. The percentage of tumor growth inhibition (TGI) was calculated as [100 − (RTV T / RTV C × 100)], where RTV T is a relative tumor volume of treated mice at day 14, when the first mouse of the control group (group A) reached the endpoint and the average relative tumor volume of control mice was RTV C .
The average of relative body weights of mice from each group was compared with that of control mice at day 14 and at the endpoint. Blood plasma parameters were determined once an endpoint was reached or at the end of the study (Supplementary Material). After euthanasia, the kidneys, liver, spleen, and brain were collected and weighed. The organ mass-to-brain ratios were calculated using the organ masses collected at the day of euthanasia (Supplementary Material).
A full macroscopic examination was performed in each animal, and the kidneys, bone marrow (sternum and femur), and spleen were sampled for histological assessment as previously reported (Supplementary Material) [24]. Histological lesions were semi-quantitatively scored by a veterinary pathologist in a blind manner using a severity grading scheme that ranged from 0 to 5.

Statistical analysis and figure preparation
Binding affinity to plasma proteins was statistically analyzed using one-way ANOVA with Dunnett's multiple comparisons post-test. Analyses of biodistribution data and the absolute AUC 0 → 120h values were performed with two-way or oneway ANOVA with Tukey's multiple comparisons post-test. The therapy study was analyzed for significance using a oneway ANOVA with Tukey's or Dunnett's test. Survival of mice was assessed using Kaplan-Meier curves to determine median survival of mice of each group. All analyses were performed using GraphPad Prism program (version 7.0). A p value of < 0.05 was considered statistically significant. Graphs of Figs. 2 and 4 were prepared using GraphPad Prism software (version 7).

Biodistribution studies
Biodistribution studies of the folate radioconjugates were performed in KB tumor-bearing mice over a period of 5 days (   Material, Fig. S5). The uptake of activity in KB tumors and kidneys at 1 h p.i. of all folate radioconjugates was reduced to~5-7% IA/g and 5-9% IA/g, respectively, when folic acid was pre-injected to block FRs in these tissues (Supplementary Material, Table S4).

SPECT/CT imaging studies
The SPECT/CT images of mice injected with [ 177 Lu]Lu-6R-RedFol-1 showed high accumulation of activity in the tumors and less in the kidneys as compared with [ 177 Lu]Lu-OxFol-1 (Fig. 3). The same high tumor uptake was observed after injection of [ 177 Lu]Lu-6S-RedFol-1; however, in this case also, the kidney uptake was increased.

Therapy study
The tumor size of untreated control mice (group A) was constantly increasing over time, whereas a considerable tumor growth delay was observed in treated mice of groups B-E. This was reflected by significantly increased tumor growth delay indices (TGDI) in treated groups as compared with control mice where the TGDIs were defined as 1.0 ( Fig. 4; Supplementary Material, Tables S6/S7).

Assessment of the therapy study
The average relative body weight (1.00-1.08) was comparable in all groups of mice at day 14, when the first mouse of the control group reached an endpoint (Supplementary Material, Table S8). Organ-to-body weight as well as organ-to-brain mass ratios may serve as indicators of the health status, since it is known that the brain of the mice does not increase in size after the age of 3 weeks (Supplementary Material, Tables S9/ S10) [25,26]. The calculated organ-to-brain mass ratios were in the same range for untreated mice and mice treated with 10 MBq of the folate radioconjugates (groups B/C). The organ-to-brain mass ratios of mice treated with 15 MBq of the folate radioconjugates (groups D/E) were decreased (p < 0.05).
Blood plasma parameters determined at the time of euthanasia did not differ among treated mice and untreated controls (Supplementary Material, Table S11). Histological investigations of the kidneys, spleen, and bone marrow did, however, not reveal any significant lesion attributed to the treatment. In particular, bone marrow of mice that received [ 177 Lu]Lu-6R-RedFol-1 showed an overall hematopoietic cellularity comparable to the control animals and mice treated with [ 177 Lu]Lu-OxFol-1 (Supplementary Material, Table S12).

Discussion
In this study, albumin-binding radioconjugates of a new class, based on 5-MTHF as a FR-binding entity, were evaluated and compared with the previously developed [ 177 Lu]Lu-OxFol-1 [18]. [ 177 Lu]Lu-6R-RedFol-1 and [ 177 Lu]Lu-6S-RedFol-1 showed high stability in PBS and human plasma in vitro. In all three cases, the binding to human plasma proteins was stronger than to mouse plasma proteins, which is in line with the reported affinity of the p-iodophenyl entity [27] and recently reported results obtained with albumin-binding PSMAtargeted radioligands [23]. In vitro studies revealed that the exchange of folic acid with 5-MTHF slightly increased the affinity of the respective radioconjugates to both mouse and human plasma proteins when compared with the affinity It remains, however, unclear whether this was the reason for the increased blood retention of 5-MTHF-based folate radioconjugates or if it was due to another, yet unknown, mechanism. The hypothesis that the observed phenomenon was due to radiometabolite formation was refuted by stability experiments that showed only intact folate radioconjugates in the blood plasma of mice 4 h after injection. The FR-specific in vitro uptake of the radioconjugates into KB tumor cells was demonstrated in vivo, since / in vivo, as pre-injection of excess folic acid reduced the ac-cumulation of the radioconjugates in tumors and kidneys of mice.The application of excess non-labeled albumin-binding folic acid conjugate (cm13) was, however, much more effective in this regard due to its enhanced blood circulation similar to the folate radioconjugates in question. The in vivo results were likely due to the enhanced blood retention of the 5-MTHF-based radioconjugates; however, it may also be due to the easier release of 5-MTHF from the FR upon internalization and, thus, more efficient accumulation as compared with folic acid [28]. The latter hypothesis is further supported by the results of Boss et al. who observed an increased tumor uptake of 18 Flabeled 5-MTHF radiotracers as compared with the folic acid analogue even though these radiotracers did not comprise an albumin-binding entity [21]. In parallel to the increased tumor uptake, [ 177 Lu]Lu-6S-RedFol-1 showed also higher retention in the kidneys resulting in similar tumor-to-kidney ratios as observed with [ 177 Lu]Lu-OxFol-1. In the case of [ 177 Lu]Lu-6R-RedFol-1, the retention in the kidneys remained relatively low resulting in substantially improved tumor-to-kidneys AUC 0 → 120h ratios to a value that has never been achieved before. This radioconjugate was, thus, selected for further investigations in a preclinical therapy study.
As expected, the treatment of KB tumor-bearing mice with 10 MBq or 15 MBq [ 177 Lu]Lu-6R-RedFol-1, showed an activity-dependent tumor growth inhibition and survival of mice. In line with the higher tumor uptake, the outcome of the [ 177 Lu]Lu-6R-RedFol-1 therapy was superior over that of [ 177 Lu]Lu-OxFol-1. Comparison of the TGDI and TGI results suggested that the application of 10 MBq [ 177 Lu]Lu-6R-RedFol-1 was equipotent to the application of 15 MBq [ 177 Lu]Lu-OxFol-1 which confirmed the enhanced therapeutic potential of this novel radioconjugate.
As the tumor-to-kidney AUC 0 → 120h ratio of [ 177 Lu]Lu-6R-RedFol-1 was almost 4-fold increased compared with [ 177 Lu]Lu-OxFol-1, the use of [ 177 Lu]Lu-6R-RedFol-1 would most probably allow delivering an effective tumor dose without the risk of long-term damage to the kidneys. In our study, no obvious early side effects were observed. Neither the body weights nor the blood plasma parameters of treated mice were significantly different from the control values. Moreover, no significant histopathological changes in kidneys and spleen were observed that would indicate radiation-induced damage of these tissues. Most importantly, the evaluation of the bone marrow of mice treated with [ 177 Lu]Lu-6R-RedFol-1 confirmed the absence of hematological side effects, in spite of the enhanced blood retention of this novel radioconjugate.

Conclusion
It was demonstrated in this study that 5-MTHF-based radioconjugates have the potential to be used for targeted radionuclide therapy. Due to the unprecedentedly high tumorto-kidney ratios of [ 1 7 7 Lu]Lu-6R-RedFol-1, this radioconjugate outperformed any other folate radioconjugate. It was, thus, shown to have enhanced therapeutic efficacy as compared with [ 177 Lu]Lu-OxFol-1, which makes [ 177 Lu]Lu-6R-RedFol-1 attractive for clinical translation. deviation; SPECT, single-photon emission computed tomography; TGD, tumor growth delay; TGI, tumor growth inhibition; TGDI, tumor growth delay index; TV, tumor volume Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.