Tailoring Fibroblast-Activation Protein Targeting for Theranostics: A Comparative Preclinical Evaluation of the 68Ga- and 177Lu-Labeled Monomeric and Dimeric Fibroblast-Activation Protein Inhibitors DOTA.SA.FAPi and DOTAGA.(SA.FAPi)2

Background: FAP radiopharmaceuticals show promise for cancer diagnosis; however, their limited tumor residency hinders treatment. This study compared two FAPi derivatives, DOTA.SA.FAPi and DOTAGA.(SA.FAPi)2, labeled with gallium-68 and lutetium-177, aiming to determine an optimum combination for creating theranostic pairs. Methods: The radiotracers were studied for lipophilicity, binding to human serum proteins, and binding to human cancer-associated fibroblasts (CAFs) in vitro, including saturation and internalization/externalization studies. PET/SPECT/CT and biodistribution studies were conducted in PC3 and U87MG xenografts for [68Ga]Ga-DOTA.SA.FAPi and [68Ga]Ga-DOTAGA.(SA.FAPi)2. [177Lu]Lu-DOTA.SA.FAPi and [177Lu]Lu-DOTAGA.(SA.FAPi)2, were evaluated in PC3 xenografts. Biodistribution studies of [68Ga]Ga-DOTA.SA.FAPi were performed in healthy male and female mice. Results: All radiotracers exhibited strong binding to FAP. Their internalization rate was fast while only [177Lu]Lu-DOTAGA.(SA.FAPi)2 was retained longer in CAFs. [68Ga]Ga-DOTAGA.(SA.FAPi)2 and [177Lu]Lu-DOTAGA.(SA.FAPi)2 displayed elevated lipophilicity and affinity for human serum proteins compared to [68Ga]Ga-DOTA.SA.FAPi and [177Lu]Lu-DOTA.SA.FAPi. In vivo studies revealed slower washout of [68Ga]Ga-DOTAGA.(SA.FAPi)2 within 3 h compared to [68Ga]Ga-DOTA.SA.FAPi. The tumor-to-tissue ratios of [68Ga]Ga-DOTAGA.(SA.FAPi)2 versus [68Ga]Ga-DOTA.SA.FAPi did not exhibit any significant differences. [177Lu]Lu-DOTAGA.(SA.FAPi)2 maintained a significant tumor uptake even after 96 h p.i. compared to [177Lu]Lu-DOTA.SA.FAPi. Conclusions: Dimeric compounds hold promise for therapy, while monomers are better suited for diagnostics. Finding the right combination is essential for effective disease management.

Various categories of theranostic FAP radiotracers can be designed by considering their structural characteristics and target specificity.When categorizing them, they can be classified as small-molecule FAP inhibitors (FAPi), peptides, and antibodies [1,2].Although the diagnostic FAPi-based radiotracers offer numerous advantages, such as their clinically proven ability to detect a variety of cancer entities, their therapeutic counterparts present a set of challenges, making them a fascinating field of research.One of the main drawbacks of therapeutic FAPi-based radiotracers is their relatively short tumor residence time with biological half-lives much shorter than the physical half-lives of the therapeutic radionuclides lutetium-177 or actinium-225 [27].This limits the exposure time of the tumor to radiation, compromising the effectiveness of the cancer treatment [3,8,20,28,29].Several strategies have been pursued to increase the tumor residence time of therapeutic FAPibased radiotracers, mainly focusing on improving their affinity and/or pharmacokinetics, by creating dimeric analogs or introducing albumin-binding moieties to the monomeric vectors [4][5][6][10][11][12]15,17].
In the current study, we set out to close this gap by conducting a comprehensive preclinical assessment of the 68 Ga-and 177 Lu-labeled DOTA.SA.FAPi and DOTAGA.(SA.FAPi) 2 .The resulting four radiotracers were characterized in terms of radiolabeling, in vitro stability, lipophilicity, saturation binding (Kd and Bmax), internalization and externalization in cells, ex vivo organ distribution and in vivo PET/SPECT/CT imaging in prostate and glioblastoma tumor-bearing mice.
The labeling yields for the 177 Lu-labeled conjugates were >95% with apparent molar activities of 7-20 GBq/µmol, depending on the study.
The formulated 68 Ga-labeled and 177 Lu-labeled radiotracers proved highly stable, as neither radiolysis nor chemical decomposition was observed for a period of 4 and 48 h post labeling, respectively.
All of the radioligands were found to be well associated with CAFs within the incubation time (Figure 1).Continued exposure of CAFs to the radioligands resulted in a gradual increase of the total cell associated uptake.A side-by-side comparison of [ 177 Lu]Lu-DOTA.SA.FAPi and [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 regarding their externalization rate from CAFs showed that [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 exhibited higher retention in CAFs compared to [ 177 Lu]Lu-DOTA.SA.FAPi (Figure 2).After 24 h at 37 • C, 48% of the total internalized activity had remained in the CAFs for [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 , while at the same time point more than 93% of the total internalized activity had been released from the CAFs into the medium when [ 177 Lu]Lu-DOTA.SA.FAPi was evaluated.
face bound fraction plus specific internalized fraction.Total specific cell uptake is expressed as the percentage of the total applied radioactivity.Nonspecific binding was determined in the presence of 1 µM UAMC1110.
Biodistribution studies of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice at 1 h p.i. also indicated significantly higher background uptake in female compared to male mice (p = 0.0226, 0.0268, 0.0351 and 0.028 for muscle, lung, bone and blood).Blocking experiments performed with excess UAMC1110, showed negligible nonspecific binding on the CAFs's surface, demonstrating high specificity of the radioligands toward FAP-positive CAFs (Figure S8).Ex vivo biodistribution data and tumor-to-tissue ratios of [ 68 Ga]Ga-DOTA.SA.FAPi and [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 in PC3 and U87MG tumor-bearing mice as well as ex vivo biodistribution data of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice are depicted on Figures 3-5, respectively.Detailed tables with the ex vivo biodistribution values and tumor-to-tissue ratios are also given in the Supplementary Data (Tables S1-S3).
Biodistribution studies of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice at 1 h p.i. also indicated significantly higher background uptake in female compared to male mice (p = 0.0226, 0.0268, 0.0351 and 0.028 for muscle, lung, bone and blood).S4).Both are taken up by the PC3 tumors at early time points, exhibiting similar uptake (8.9 ± 0.    S4).Both are taken up by the PC3 tumors at early time points, exhibiting similar uptake (8.9 ± 0.

Small-Animal PET/SPECT/CT Studies
Representative PET/SPECT/CT images of PC3 and U87MG tumor-bearing mice are depicted in Figures 7 and 8.The U87MG and PC3 tumors are well delineated and the background, which is dominated by high blood pool, salivary glands and joints uptake, is higher for [ 68 Ga]Ga-DOTAGA.(SA.FAPi)2 compared to [ 68 Ga]Ga-DOTA.SA.FAPi.A quantitative analysis of the PET/CT images in terms of time-activity curves for the key organs and the corresponding tumor-to-tissue ratios is presented in Figures 9 and 10

Small-Animal PET/SPECT/CT Studies
Representative PET/SPECT/CT images of PC3 and U87MG tumor-bearing mice are depicted in Figures 7 and 8.The U87MG and PC3 tumors are well delineated and the background, which is dominated by high blood pool, salivary glands and joints uptake, is higher for [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 compared to [ 68 Ga]Ga-DOTA.SA.FAPi.A quantitative analysis of the PET/CT images in terms of time-activity curves for the key organs and the corresponding tumor-to-tissue ratios is presented in Figures 9 and 10, respectively.The PET-based quantification of the organ uptake aligns well with the ex vivo biodistribution data, revealing a higher blood pool (heart uptake in the PET images) for [ 68

Discussion
FAP imagining is a new approach in pan-cancer theranostics based on highly potent targeting vectors.In particular, the FAP inhibitor UAMC 1110 has been translated into gallium-68 and fluorine-18 labeled radiopharmaceuticals, which have been investigated to be equivalent or even superior to [ 18 F]F-FDG [21,24,[31][32][33][34][35][36].The design of those compounds consists of a FAP inhibitor (FAPi) coupled to a spacer, a linker, a chelator and either the radiolabeled unit gallium-68 or fluorine-18.These structures are referred to as monomeric FAPi-based radiopharmaceuticals (Figure 12A).The initially great enthusiasm in the community to easily transform these diagnostic FAPi monomers into therapeutic analogs by simply exchanging the positron emitting gallium-68 with beta electron emitters

Discussion
FAP imagining is a new approach in pan-cancer theranostics based on highly potent targeting vectors.In particular, the FAP inhibitor UAMC 1110 has been translated into gallium-68 and fluorine-18 labeled radiopharmaceuticals, which have been investigated to be equivalent or even superior to [ 18 F]F-FDG [21,24,[31][32][33][34][35][36].The design of those com-pounds consists of a FAP inhibitor (FAPi) coupled to a spacer, a linker, a chelator and either the radiolabeled unit gallium-68 or fluorine-18.These structures are referred to as monomeric FAPi-based radiopharmaceuticals (Figure 12A).The initially great enthusiasm in the community to easily transform these diagnostic FAPi monomers into therapeutic analogs by simply exchanging the positron emitting gallium-68 with beta electron emitters such as yttrium-90 and lutetium-177 or the alpha emitting actinium-225 were somewhat tempered when a first in human study with a 177 Lu-labeled FAP inhibitor did not meet the high expectations [20,[37][38][39][40][41][42][43][44][45][46].A major root cause for the disappointing therapeutic performance appeared to be the short residence times of the monomeric FAPi-based radiopharmaceuticals in the tumor microenvironment (TME), which was in the order of hours only, much shorter than the physical half-lives of the therapeutic radiometals used [27].Fortunately, this problem was successfully addressed by developing homodimeric FAPI structures (Figure 12B) [4,6,[10][11][12][13][14]17].Compounds like DOTAGA.(SA.FAPI) 2 or DOTAGA.Glu.(FAPI) 2 remain in the lesions of various tumors over days and were found to substantially prolong overall and progression-free survival [20,22,23].

Discussion
FAP imagining is a new approach in pan-cancer theranostics based on highly potent targeting vectors.In particular, the FAP inhibitor UAMC 1110 has been translated into gallium-68 and fluorine-18 labeled radiopharmaceuticals, which have been investigated to be equivalent or even superior to [ 18 F]F-FDG [21,24,[31][32][33][34][35][36].The design of those compounds consists of a FAP inhibitor (FAPi) coupled to a spacer, a linker, a chelator and either the radiolabeled unit gallium-68 or fluorine-18.These structures are referred to as monomeric FAPi-based radiopharmaceuticals (Figure 12A).The initially great enthusiasm in the community to easily transform these diagnostic FAPi monomers into therapeutic analogs by simply exchanging the positron emitting gallium-68 with beta electron emitters such as yttrium-90 and lutetium-177 or the alpha emitting actinium-225 were somewhat tempered when a first in human study with a 177 Lu-labeled FAP inhibitor did not meet the high expectations [20,[37][38][39][40][41][42][43][44][45][46].A major root cause for the disappointing therapeutic performance appeared to be the short residence times of the monomeric FAPi-based radiopharmaceuticals in the tumor microenvironment (TME), which was in the order of hours only, much shorter than the physical half-lives of the therapeutic radiometals used [27].Fortunately, this problem was successfully addressed by developing homodimeric FAPI structures (Figure 12B) [4,6,[10][11][12][13][14]17].Compounds like DOTAGA.(SA.FAPI)2 or DOT-AGA.Glu.(FAPI)2remain in the lesions of various tumors over days and were found to substantially prolong overall and progression-free survival [20,22,23].Although the strategy of increasing the residence time in the tumor by employing homodimeric instead of monomeric FAP inhibitors has proven successful, the underlying molecular basis and mechanism still remain elusive.In our current work, we aimed to gain a better understanding of the mechanisms governing the increase in the residence time at the oncological targets FAP or CAF when employing homodimeric instead of monomeric FAP inhibitors.We are confident that this exploration will not only enhance our understanding but also catalyze future advancements in the field.There are several concepts that may be relevant: (a) avidity versus affinity, (b) dual-binding mechanisms, (c) differences in membrane passage and internalization, (d) differences in externalization rates, and (e) active intracellular reactions by enzymes, i.e., trapping.
The present study first focused on the cellular aspects and investigated differences between monomeric and homodimeric FAP inhibitors labeled with trivalent radiometals, addressing binding affinity in terms of Kd values, membrane passage, and internalization and externalization rates.The systematic cell studies were then compared with the in vivo and ex vivo studies on tumor-bearing mice.
The chemical modification of the UAMC1110 lead structure with the SA linker and the chelator was well tolerated, as both DOTA.SA.FAPi and DOTAGA.(SA.FAPi) 2 , showed K d values in the sub-nanomolar range.This compared well with data published by Moon et al. [12][13][14].
177/nat Lu-DOTA.SA.FAPi and 177/nat Lu-DOTAGA.(SA.FAPi) 2 exhibited a somewhat reduced affinity when compared to 68/nat Ga-DOTA.SA.FAPi and 68/nat Ga-DOTAGA.(SA.FAPi) 2 , with this effect appearing to be more prominent for 177/nat Lu-DOTA.SA.FAPi.This may be explained by differences in terms of the charge and charge distribution arising after the coordination of gallium and lutetium in the chelator cavity [47].Throughout all of the evaluated time points, internalization accounted for more than 95% of the total cellular activity.This phenomenon is influenced by several factors such as FAP expression, cell line characteristics, radioligand-FAP internalization mechanisms, affinity, and specificity.Furthermore, functional activation of FAP requires both dimerization and glycosylation, and given FAP's short cytoplasmic domain, other receptors, such as integrins, may also act as intermediaries for FAP's impact on the intracellular signaling and internalization rate [11,48].Interestingly, in our study, the total internalized activity for the dimer was approximately 1.5 times lower than that of the monomer, indicating that different internalization mechanisms may be triggered upon the binding of monomers and dimers to FAP.
Considering the potential of the 177 Lu-labeled radiotracers for therapeutic use, it is crucial to not only emphasize a strong binding affinity and high internalization rate but also to take into account the residence time of the radioactivity within cells.As part of our investigation, we conducted externalization studies, revealing notable distinctions between [ 177 Lu]Lu-DOTA.SA.FAPi and [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 .For [ 177 Lu]Lu-DOTA.SA.FAPi, after a 24 h period, only 7% of the total internalized radioactivity remained within the cells.In contrast, for [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 , a significant 48% of the total internalized radioactivity was retained in the cells.Undoubtedly, the 177 Lu-labeled dimer exhibits a more favorable profile in the context of potential therapeutic possibilities.This corresponds to the observations made in patient studies [22].
Given the encouraging in vitro results and our prior success with FAP targeting of PC3 and U87MG tumor-bearing mice, we proceeded to in vivo evaluation using the same tumor models.While both [ 68 Ga]Ga-DOTA.SA.FAPi and [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 showed consistent results in terms of tumor uptake in both models, the dimer displayed slightly higher and more sustained absolute uptake, with no decline observed over time.In all cases, tumor uptake was specific, as demonstrated by the control group treated with the blocking agent.A considerably higher background uptake of [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 in comparison to [ 68 Ga]Ga-DOTA.SA.FAPi in both PC3 and U87MG tumor-bearing mice was observed.One possible explanation for this might be the elevated lipophilicity and the almost 2-fold higher serum protein binding of [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 relative to [ 68 Ga]Ga-DOTA.SA.FAPi.This could potentially be a limiting factor when employing [ 68 Ga]Ga-DOTAGA.(SA.FAPi) 2 for diagnostic applications.
An additional observation that warrants attention is the higher background uptake in the female U87MG compared to the male PC3 tumor-bearing mice for both radiotracers, possibly due to sex differences.Biodistribution studies of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice supported the sex assumption.However, further research is required to validate these preliminary findings.
In view of their potential therapeutic application, a direct in vivo comparison of the 177 Lu-labeled radiotracers was conducted in PC3 tumor-bearing mice.Even though both radiotracers exhibited early and relatively high tumor uptake, they demonstrated distinct clearance rates in line with our externalization data.[ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 maintained a significant tumor uptake even after 96 h, whereas [ 177 Lu]Lu-DOTA.SA.FAPi was nearly completely washed out already at 48 h p.i.The elevated liver uptake observed for [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 across all examined time points could be attributed to its increased lipophilicity compared to [ 177 Lu]Lu-DOTA.SA.FAPi.To achieve the appropriate balance and consequently the desired in vivo behavior, additional chemical modifications on the dimeric precursor's structure may be required [10].
Both 68 Ga-and 177 Lu-labeled radiotracers demonstrated elevated uptake in non-target organs including muscles, lungs, pancreas, salivary glands, and joints, a phenomenon also noted in clinical settings [3,7,8,17,[19][20][21][22][23]26,29,49].The development of these organs is, in part, influenced by the activity of fibroblasts and this might provide an explanation for the off-target uptake.The non-target organs uptake was lower for the monomer compared to the dimer and also for the male compared to the female mice.Additionally, the monomer exhibited a higher rate of washout over time (Figure S3).
The utilization of diverse tumor models from individual groups precludes a straightforward comparison between [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 and the previously reported 177 Lu-labeled dimeric FAPi-based radiopharmaceuticals. Nevertheless, a consistent pattern was observed across all, indicating an enhanced accumulation and extended retention time of the dimers in tumors compared to monomers [6,11,15,17].This highlights the promising therapeutic potential of the dimeric concept.

Radiolabeling/Quality Control of the Radiotracers/Stability
68 Ga-labelings were performed using the Modular-Lab PharmTracer module (Eckert & Ziegler Berlin, Germany), and 177 Lu-labelings were performed manually (Supplemental Data).Radiochemical purity and stability for 4 h for the 68 Ga-labelled and 48 h for the 177 Lulabelled radiotracers were determined by reversed-phase high performance liquid chromatography (RP-HPLC) and radio thin-layer chromatography (radio-TLC).

Lipophilicity/Protein Binding Studies
The lipophilicity (LogD octanol/PBS ) and protein binding of the four radiotracers in human serum were determined as described in the Supplemental Data.

Cell Lines
PC3 (Cell Lines Service GmbH (CLS, Eppelheim, Germany), U87MG (American Type Culture Collection (ATCC, Manassas, VA, USA) and CAF (American Type Culture Collection (ATCC, Manassas, VA, USA) cells have been used.Cultivation conditions, materials and further details are described in the Supplemental Data (Gibco BRL, Life Technologies (Grand Island, NY, USA).

Animal Models
U87MG and PC3 tumor-bearing mice: Female athymic BALB/c (6 weeks/16-20 g) and male athymic BALB/c nude mice (6 weeks/20-25 g) were implanted with U87MG or PC3 cells (5 × 10 6 /100 µL PBS) into their right shoulder, respectively.The animals were used for biodistribution and PET/SPECT/CT imaging studies (Animal License No: BE98/2021).Ten pmol (0.08-0.1 MBq) of the radiotracers in 100 µL of NaCl 0.9% were injected intravenously into the tail vein of U87MG and PC3 tumor-bearing mice and biodistribution studies were conducted at 1, 2, and 3 h after injection.To demonstrate the specificity of binding, the mice were co-injected with 10 pmol of each radiotracer along with 20 nmol of UAMC1110 (total injected volume: 100 µL) and the biodistribution studies were performed 2 h p.i.

Statistical Analysis
Data are expressed as mean ± standard deviation (mean ± SD).Prism 8 software (GraphPad Software 8) was used to determine the statistical significance at the 95% confidence level, with a p-value of less than 0.05 being considered significantly different.

Conclusions
[ 68 Ga]Ga-DOTA.SA.FAPi exhibited highly advantageous characteristics for imaging.These qualities encompass tumor-to-background ratios that highlight its excellent utility in diagnostic applications for a wide range of cancer types.The present comprehensive preclinical evaluation of [ 177 Lu]Lu-DOTAGA.(SA.FAPi) 2 revealed substantial potential for effective radionuclide-based targeted therapy against FAP-positive tumors, supporting the current clinical data.Although this is not a "true" radiotheranostic pair, the optimal synergy of monomers for imaging and dimers for therapy could be a pivotal factor in the successful management of FAP-positive tumors.Further clinical trials are required to evaluate the efficacy and safety of various radiotheranostic combinations, providing hope for improved cancer care.

Figure 2 .
Figure 2. Percentage of the activity externalized from the cells expressed in relation to the total internalized activity (100%).

Figure 2 .
Figure 2. Percentage of the activity externalized from the cells expressed in relation to the total internalized activity (100%).

Figure 5 .
Figure 5. Biodistribution data of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice at 1 h p.i. Data have been calculated as %I.A./g of tissue and are presented as mean ± SD (n = 4).

Figure 5 .
Figure 5. Biodistribution data of [ 68 Ga]Ga-DOTA.SA.FAPi in healthy male and female mice at 1 h p.i. Data have been calculated as %I.A./g of tissue and are presented as mean ± SD (n = 4).
, respectively.The PET-based quantification of the organ uptake aligns well with the ex vivo biodistribution data, revealing a higher blood pool (heart uptake in the PET images) for [ 68 Ga]Ga-DOTAGA.(SA.FAPi)2 in comparison to [ 68 Ga]Ga-DOTA.SA.FAPi.Furthermore, again in accordance with the biodistribution studies, PET imaging showed that the accumulated activity for [ 68 Ga]Ga-DOTAGA.(SA.FAPi)2 in both the PC3 and U87MG tumors revealed slower washout over the course of 3 h compared to [ 68 Ga]Ga-DOTA.SA.FAPi.Despite this, the tumor-to-tissue ratios of [ 68 Ga]Ga-DOTAGA.(SA.FAPi)2 versus [ 68 Ga]Ga-DOTA.SA.FAPi in both the PC3 and U87MG tumor models did not exhibit any significant differences.

Figure 10 .
Figure 10.Tumor-to-tissue ratios generated from the quantification of the PET images of PC3 and U87MG tumor-bearing mice after the injection of [ 68 Ga]Ga-DOTA.SA.FAPi and [ 68 Ga]Ga-DOT-AGA.(SA.FAPi)2, respectively.In line with the ex vivo biodistribution studies, the SPECT/CT images (Figure 11) clearly illustrate the superior tumor retention of [ 177 Lu]Lu-DOTAGA.(SA.FAPi)2 compared to [ 177 Lu]Lu-DOTA.SA.FAPi after their administration to PC3 tumor-bearing mice.The specificity of tumor uptake was confirmed through blocking experiments.