Dual‐Target Nectin‐4/FAP Heterobivalent Probe ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1 for Triple‐Negative Breast Cancer: Preclinical Theranostic Evaluation and First‐in‐Human Translation

ABSTRACT

Triple‐negative breast cancer (TNBC) features pronounced molecular heterogeneity and poor prognosis, restricting the efficacy of single‐targeted strategies. Leveraging the complementary theranostic targets fibroblast activation protein (FAP) and Nectin‐4, we designed the novel hetero‐bivalent agent FZ‐NF‐1 for dual targeting. Preclinically, both ⁶⁸Ga‐labeled and ¹⁷⁷Lu‐labeled FZ‐NF‐1 exhibited high radiochemical purity, good stability, and high affinity for both Nectin‐4 and FAP. ⁶⁸Ga‐FZ‐NF‐1 revealed better tumor uptake and prolonged retention time compared to monomeric FAP‐ or Nectin‐4 targeted tracers. Moreover, ¹⁷⁷Lu‐FZ‐NF‐1 exhibited sustained tumor accumulation for upto 120 h post‐injection, leading to significant, dose‐dependent tumor regression in TNBC mouse models, with no obvious systemic toxicity observed. In a preliminary first‐in‐human translational study involving 13 patients with TNBC, ⁶⁸Ga‐FZ‐NF‐1 demonstrated comparable tumor uptake to that of ¹⁸F‐fluorodeoxyglucose (¹⁸F‐FDG), while successfully identifying additional lesions missed by ¹⁸F‐FDG. Complementary tissue microarray analysis of 60 TNBC cases confirmed the heterogeneous expression of Nectin‐4 and FAP in TNBC. By concurrently targeting the tumor cells and stroma, ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1 effectively mitigates the limitations associated with single‐target imaging and therapy. These results indicate that the heterodimeric theranostic pair offers enhanced imaging performance and significant therapeutic efficacy, providing a promising strategy for the personalized management of refractory TNBC.

A novel hetero‐bivalent agent, ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1, concurrently targets Nectin‐4 and FAP in triple‐negative breast cancer. It yields significant preclinical tumor regression and clinically identifies additional metastatic lesions missed by ¹⁸F‐FDG PET, demonstrating promising theranostic potential for personalized management.

1. Introduction

Triple‐negative breast cancer (TNBC) is characterized by high aggressiveness and complex heterogeneity. Due to the long‐standing lack of effective targeted therapeutic strategies, it remains the breast cancer subtype with the poorest prognosis [1, 2]. The TNBC‐FUSCC subtyping system, developed by Chinese researchers, stratifies TNBC into four molecular subtypes categorized as luminal androgen receptor, immunomodulatory, mesenchymal‐like, and basal‐like immune‐suppressed [3]. The TNBC‐FUSCC subtyping not only elucidates the intricate molecular landscape of TNBC but also explains the limited clinical efficacy of traditional single‐target diagnostic and therapeutic approaches [4]. Recently, the theranostic strategy has demonstrated outstanding clinical efficacy, as exemplified by the remarkable success of ¹⁷⁷Lu‐PSMA [5] in prostate cancer and ¹⁷⁷Lu‐DOTATATE [6] in neuroendocrine tumors, offering renewed hope for patients with refractory advanced malignancies. However, while theranostics has achieved a major milestone in multiple solid tumors, its clinical potential in TNBC remains largely unrealized [7].

Nectin‐4 is a cell adhesion molecule that is frequently overexpressed in TNBC and correlates with adverse clinical outcomes [8]. Nectin‐4 has emerged as a promising target for various radiopharmaceuticals, with preliminary clinical translation highlighting its potential for the precise imaging and staging of TNBC [9, 10]. However, strategies targeting Nectin‐4 alone are limited by the profound molecular heterogeneity of TNBC [11]. The inconsistent expression of Nectin‐4 across lesions may lead to false negatives in single‐target imaging, hindering the ability to map the tumor landscape. Furthermore, current monomeric tracers often exhibit suboptimal pharmacokinetics characterized by limited tumor uptake and short retention time, which currently limits their translational potential for therapy [12].

Fibroblast activation protein (FAP) is specifically and highly expressed in cancer‐associated fibroblasts [13]. In TNBC, elevated stromal FAP expression is strongly correlated with aggressive disease progression and poor survival [14, 15]. While FAP inhibitors (FAPIs) demonstrate excellent tumor‐to‐background contrast in imaging, their translation into radioligand therapy faces significant challenges. Most current monomeric FAPIs suffer from rapid washout and insufficient retention [16, 17]. This limitation often necessitates elevated administered doses to maintain therapeutic concentrations, thereby increasing the risk of radiation exposure to healthy tissues while the ultimate therapeutic benefit remains uncertain [18]. Previous studies on hetero‐bivalent agents, such as FAPI‐RGD and FAPI‐SSTR, have demonstrated that dual‐targeting FAP alongside another tumor target can markedly optimize pharmacokinetics [12, 19]. These hetero‐bivalent probes improve diagnostic sensitivity while providing the prolonged tumor retention necessary for effective radionuclide delivery and subsequent cytotoxic effects [20].

We hypothesized that dual‐targeting of Nectin‐4 and FAP could achieve precise diagnosis and intense accumulation with prolonged retention for theranostics in TNBC. In this study, we designed and synthesized ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1, the first hetero‐bivalent radiopharmaceutical concurrently targeting Nectin‐4 and FAP. We systematically evaluated its pharmacokinetic properties in preclinical models and conducted a first‐in‐human translational study.

2. Results

2.1. Synthesis and Characterization of ⁶⁸Ga‐FZ‐NF‐1

The FZ‐NF‐1 precursor, incorporating a DOTAGA chelator and a hetero‐bivalent moiety comprising the FZ‐NR‐1 [9] and FUSCC‐FAPI‐II [16], was successfully synthesized (Figure 1a). Mass spectrometry (Figure S1) and high‐performance liquid chromatography (HPLC) (Figure S2a) analyses confirmed the expected molecular weight and the high chemical purity of the FZ‐NF‐1 precursor, respectively. Radiolabeling of ⁶⁸Ga‐FZ‐NF‐1 was achieved at an activity concentration of approximately 150 MBq/mL. Following purification, the radiochemical purity exceeded 99%, with a distinct radioactivity peak observed at 12.22 min (Figure 1b). HPLC analysis demonstrated that ⁶⁸Ga‐FZ‐NF‐1 exhibited high stability for up to 2 h, with no significant demetallation observed in either saline or fetal bovine serum (FBS) (Figure 1b) and (Figure S2b,c). The distribution coefficient (Log P) of ⁶⁸Ga‐FZ‐NF‐1 was calculated as −2.49 ± 0.29, suggesting a hydrophilic profile favorable for renal excretion. Additionally, ⁶⁸Ga‐FZ‐NF‐1 exhibited a high plasma protein binding rate over a 120‐min incubation period (Figure S2d). Pharmacokinetics analysis in normal mice revealed rapid blood clearance, with a half‐life of 1.14 min for the distribution phase and 18.86 min for the elimination phase (Figure S3A). Biodistribution profile of normal organs demonstrated a gradual, time‐dependent decline in radioactivity across most major organs, including the heart, liver, spleen, lung, and kidneys, confirming predominant clearance via the urinary system (Figure S3b).

FIGURE 1.

(A) Structure of ⁶⁸Ga‐FZ‐NF‐1. (B) Radiochemical purity and stability of ⁶⁸Ga‐FZ‐NF‐1 via radio‐high‐performance liquid chromatography analysis. FBS, fetal bovine serum.

2.2. Selective Binding of ⁶⁸Ga‐FZ‐NF‐1 to Human FAP and Nectin‐4

Surface plasmon resonance experiments confirmed the strong binding affinity of FZ‐NF‐1 for both human FAP and Nectin‐4 proteins, yielding K_D values of 60.2 pm and 32.55 nm, respectively (Figure S4). Flow cytometry was subsequently employed to profile target expression across various cell lines. MDA‐MB‐231‐hNectin‐4‐hFAP and MDA‐MB‐468‐hFAP were confirmed as dual‐target‐positive. HT1080‐hFAP was characterized as an FAP‐positive cell line, whereas MDA‐MB‐468 and HT1376 were Nectin‐4‐positive. The wild‐type MDA‐MB‐231 served as a dual‐negative cell line (Figure 2a and Figures S5 and S6a). In vitro cellular uptake assays following a 60‐min incubation revealed that dual‐positive cell lines exhibited significantly higher ⁶⁸Ga‐FZ‐NF‐1 uptake compared to single‐positive cells. These single‐positive cells, in turn, demonstrated greater uptake than the dual‐negative cell line, consistent with the target expression profiles established by flow cytometry (Figure S6b). Moreover, in dual‐positive, FAP‐positive, and Nectin‐4‐positive cells, the uptake of ⁶⁸Ga‐FZ‐NF‐1 was significantly attenuated by the co‐incubation with the corresponding unlabeled precursors. This competitive blockade confirmed the dual‐targeting specificity of the probe for both human FAP and Nectin‐4 (Figure 2b).

FIGURE 2.

Nectin‐4 and FAP expression, cell uptake assay of ⁶⁸Ga‐FZ‐NF‐1 and blocking experiments in MDA‐MB‐231‐hNectin‐4‐hFAP, MDA‐MB‐468‐hFAP, HT1080‐hFAP, MDA‐MB‐468, HT1376, and MDA‐MB‐231 cells. (A) Representative flow cytometry plots of FAP‐positive or Nectin‐4‐positive cells. (B) Cell uptake assay of ⁶⁸Ga‐FZ‐NF‐1 and blocking experiments on different cell lines (n = 6). *** p < 0.001.

2.3. ⁶⁸Ga‐FZ‐NF‐1 Outperforms Corresponding Monomers in Tumor Uptake and Retention

Small‐animal positron emission tomography/computed tomography (PET/CT) was performed on MDA‐MB‐468‐hFAP tumor xenografts, which were dual‐positive. Representative whole‐body scan images at different time points after injection of ⁶⁸Ga‐FZ‐NF‐1, ⁶⁸Ga‐FAPI‐FUSCC‐II, or ⁶⁸Ga‐FZ‐NR‐1 are shown in Figure 3. ⁶⁸Ga‐FZ‐NF‐1 demonstrated rapid tumor accumulation within 0.5 h post‐injection (13.77 ± 1.19%ID/g), and maintained sustained uptake up to 4 h (1 h, 13.90% ± 1.65%ID/g; 2 h, 14.17% ± 1.67%ID/g; and 4 h, 15.50% ± 1.97%ID/g) (Figure 3a). PET imaging in another dual‐positive model, MDA‐MB‐231‐hNectin‐4‐hFAP, showed similar characteristics (Figure S7). Concurrently, tracer uptake in background organs such as the heart, liver, and muscle remained relatively low and decreased over time, resulting in progressively improved target‐to‐background ratios. Furthermore, the high accumulation in the kidneys and bladder indicated that ⁶⁸Ga‐FZ‐NF‐1 was predominantly excreted via the renal‐urinary pathway. In contrast, tumor uptake of ⁶⁸Ga‐FAPI‐FUSCC‐II remained stable only until 1 h (0.5 h, 9.80% ± 0.40%ID/g and 1 h, 9.77% ± 0.81%ID/g) post‐injection before declining at 2 h (2 h, 7.83% ± 0.76%ID/g; 4 h, 7.20% ± 0.10%ID/g) (Figure 3b). Similarly, ⁶⁸Ga‐FZ‐NR‐1 showed a gradual washout from 1–4 h post‐injection (0.5 h, 4.00% ± 0.26%ID/g; 1 h, 3.57% ± 0.15%ID/g; 2 h, 2.50% ± 0.30%ID/g; 4 h, 1.77% ± 0.15%ID/g) (Figure 3c). Collectively, the hetero‐bivalent tracer ⁶⁸Ga‐FZ‐NF‐1 demonstrated superior tumor uptake and prolonged retention compared to the corresponding monomeric tracers.

FIGURE 3.

Representative PET/CT imaging and quantification results in MDA‐MB‐468‐hFAP tumor‐bearing mice with (A) ⁶⁸Ga‐FZ‐NF‐1 (n = 3), (B) ⁶⁸Ga‐FAPI‐FUSCC‐II (n = 3), (C) ⁶⁸Ga‐FZ‐NR‐1 (n = 3). (D) Representative PET imaging and quantification results of ⁶⁸Ga‐FZ‐NF‐1 and blocking with FAPI‐04, unlabeled Nectin‐4, or FAPI‐04 plus unlabeled Nectin‐4 in the MDA‐MB‐468‐hFAP tumor model (n = 3). * p < 0.05, *** p < 0.001.

To further validate targeting specificity, blocking studies were conducted using the MDA‐MB‐468‐hFAP model. At 1 h post‐injection, tumor uptake (14.03% ± 1.57%ID/g) was partially inhibited by the administration of unlabeled Nectin −4 (4.37% ± 0.59%ID/g, 67% blockade) or FAPI‐04 (2.20% ± 0.10%ID/g, 82% blockade), and was substantially abrogated (1.21% ± 0.25%ID/g, 91% blockade) by the co‐administration of both blocking agents (Figure 3d). The biodistribution results were highly consistent with the imaging data (without block, 27.63% ± 2.39%ID/g; Nectin‐4 block, 7.36% ± 1.61%ID/g; FAPI‐04 block, 5.63% ± 1.34%ID/g; dual block, and 1.87% ± 0.05%ID/g) (Figure S8).

Specificity was further confirmed in additional single‐target‐positive models, where ⁶⁸Ga‐FZ‐NF‐1 showed favorable tumor uptake and retention that was significantly blocked by the co‐injection of corresponding blocking agents (Figures S9 and S10a and Table S1). Finally, FAP and Nectin‐4 immunohistochemistry ((IHC)) staining of the corresponding tumor xenografts further validated the correlation between radiotracer uptake and antigen expression (Figure S10).

2.4. Biosafety of ⁶⁸Ga‐FZ‐NF‐1

Toxicology assessment was performed 14 days after intravenous injection of ⁶⁸Ga‐FZ‐NF‐1 in healthy mice. No significant differences were observed in complete blood counts (Figure S11a), serum biochemistry parameters (Figure S11b), or body weights (Figure S11c). Furthermore, hematoxylin and eosin ((H&E)) staining of the heart, liver, spleen, lung, and kidneys revealed no discernible pathological abnormalities (Figure S11d). These results indicated that ⁶⁸Ga‐FZ‐NF‐1 is well‐tolerated in mice, supporting its safety for further clinical translation.

2.5. Preclinical SPECT/CT Imaging and the Ex Vivo Biodistribution of ¹⁷⁷Lu‐FZ‐NF‐1 in TNBC

Given the excellent tumor uptake and retention exhibited by ⁶⁸Ga‐FZ‐NF‐1 in PET imaging, its potential as a theranostic agent was further explored by radiolabeling the FZ‐NF‐1 precursor with the therapeutic radionuclide ¹⁷⁷Lu. The stability of ¹⁷⁷Lu‐FZ‐NF‐1 was first evaluated, demonstrating that it remained stable in both saline and FBS for up to 48 h (Figure S12).

Single‐photon emission computed tomography/computed tomography (SPECT/CT) was performed on dual‐positive MDA‐MB‐231‐hNectin‐4‐hFAP tumor xenografts. Representative images at different time points (4, 10, 24, 48, 72, 96, and 120 h) after injection of ¹⁷⁷Lu‐FZ‐NF‐1, ¹⁷⁷Lu‐FAPI‐FUSCC‐II, or ¹⁷⁷Lu‐FZ‐NR‐1 are shown in Figure 4a. The hetero‐bivalent radioligand ¹⁷⁷Lu‐FZ‐NF‐1 exhibited sustained tumor retention throughout the observation period. Intense radiotracer accumulation was continuously maintained at the tumor, with prominent radioactive signals clearly visible even at 120 h post‐injection. In contrast, ¹⁷⁷Lu‐FAPI‐FUSCC‐II, demonstrated rapid tumor washout, with the radioactive signal at the tumor becoming nearly undetectable as early as 24 h post‐injection. Similarly, the monomeric Nectin‐4‐targeted probe, ¹⁷⁷Lu‐FZ‐NR‐1, showed moderate initial accumulation, but the tumor uptake decreased substantially over time, leaving only faint residual signals by 72 h. Concurrently, a pronounced accumulation of ¹⁷⁷Lu‐FZ‐NF‐1 was observed in the kidneys and bladder, indicating that this radioligand is mainly cleared via the renal excretion pathway.

FIGURE 4.

In vivo SPECT/CT imaging, ex vivo biodistribution, and therapeutic efficacy of 177Lu‐FZ‐NF‐1 and corresponding monomers in a TNBC mouse model. (A) Representative SPECT/CT images and (B) Ex vivo biodistribution results of ¹⁷⁷Lu‐FZ‐NF‐1 in MDA‐MB‐231‐hNectin‐4‐hFAP tumor‐bearing mice (n = 3). (C) Changes in tumor volume and body weight in MDA‐MB‐231‐hNectin‐4‐hFAP models with an average tumor volume of 50 mm³ at the onset of the treatments (n = 8). (D) Changes in tumor volume and body weight in MDA‐MB‐231‐hNectin‐4‐hFAP models with an average tumor volume of 150 mm³ at the onset of the treatments (n = 6).

Furthermore, negligible radiotracer uptake in other normal organs provided a clean background contrast. These findings indicate that the hetero‐bivalent ¹⁷⁷Lu‐FZ‐NF‐1 exhibits enhanced tumor accumulation and longer retention compared to the corresponding monomeric agents.

These imaging results were confirmed by ex vivo biodistribution studies (Figure 4b). The tumor uptake of ¹⁷⁷Lu‐FZ‐NF‐1 was measured at 11.05% ± 1.01%ID/g at 4 h, peaking at 12.03% ± 1.19%ID/g after 24 h, while maintaining a substantial retention of 3.45% ± 1.29%ID/g at 120 h post‐injection. Collectively, these data confirm the enhanced uptake and retention time of the dual‐targeting agent, demonstrating the potential of ¹⁷⁷Lu‐FZ‐NF‐1 for radioligand therapy (RLT).

2.6. Radiopharmaceutical Therapy of 177Lu‐FZ‐NF‐1 in TNBC Mouse Model

To evaluate the in vivo therapeutic efficacy, MDA‐MB‐231‐hNectin‐4‐hFAP tumor‐bearing mice were randomly divided into four groups when the tumor volume reached approximately 50 mm³. The groups included a saline control group and three treatment groups receiving a low dose (7.4 MBq) of either ¹⁷⁷Lu‐FZ‐NF‐1, ¹⁷⁷Lu‐FAPI‐FUSCC‐II, or ¹⁷⁷Lu‐FZ‐NR‐1. Tumor volumes and body weights were monitored every 2 days. As shown in Figure 4c and Figure S13, tumors in the saline control group exhibited rapid growth, and leading to 100% mortality by day 24 post‐treatment. Although both the 7.4 MBq ¹⁷⁷Lu‐FAPI‐FUSCC‐II and ¹⁷⁷Lu‐FZ‐NR‐1 treatment groups exhibited relatively slower tumor growth rates than those in the control group, all mice died by day 28. The low‐dose ¹⁷⁷Lu‐FZ‐NF‐1 treatment group demonstrated the most remarkable anti‐tumor efficacy, effectively suppressing tumor proliferation compared to both the control and single‐target radioligands, with a 100% survival rate maintained over the 28‐day observation period. Furthermore, no significant body weight loss was observed across any of the treatment groups (Figure 4c), indicating the favorable biosafety and general tolerability of this low‐dose regimen.

Subsequently, the dose‐dependent therapeutic potential of ¹⁷⁷Lu‐FZ‐NF‐1 was evaluated in a more challenging, established tumor model. MDA‐MB‐231‐hNectin‐4‐hFAP tumor‐bearing mice with larger tumors (∼150 mm³) were randomly divided into five groups: a saline control group, a medium‐dose ¹⁷⁷Lu‐FZ‐NF‐1 group (14.8 MBq), a high‐dose ¹⁷⁷Lu‐FZ‐NF‐1 group (29.6 MBq), and two high‐dose monomeric groups (29.6 MBq ¹⁷⁷Lu‐FAPI‐FUSCC‐II and 29.6 MBq ¹⁷⁷Lu‐FZ‐NR‐1). As exhibited in Figure 4d and Figure S14, the tumor volume in the control group steadily increased, and leading to total mortality by day 14 post‐treatment. Treatment with 29.6 MBq ¹⁷⁷Lu‐FAPI‐FUSCC‐II and ¹⁷⁷Lu‐FZ‐NR‐1 provided moderate tumor growth delay. While all mice in the ¹⁷⁷Lu‐FAPI‐FUSCC‐II group died by day 26 post‐treatment, only a single mortality was recorded in the ¹⁷⁷Lu‐FZ‐NR‐1 group during the monitoring period. Conversely, tumor volumes in both ¹⁷⁷Lu‐FZ‐NF‐1 treatment groups began to regress from day 2. Tumors in the 14.8 MBq ¹⁷⁷Lu‐FZ‐NF‐1 group resumed growth on day 14, whereas tumors in the 29.6 MBq ¹⁷⁷Lu‐FZ‐NF‐1 group maintained suppression significantly longer, with regrowth observed from day 20 onward. No mortality was observed in the two ¹⁷⁷Lu‐FZ‐NF‐1 groups during the monitoring window. The mice in 29.6 MBq ¹⁷⁷Lu‐FZ‐NF‐1 treatment group exhibited slight weight loss (< 15%). However, H&E staining of the heart, liver, spleen, lung, and kidneys following the completion of treatment revealed no discernible pathological abnormalities in low‐, medium‐, and high‐dose ¹⁷⁷Lu‐FZ‐NF‐1 treatment groups (Figure S15). Additionally, there were no significant differences in hepatic and renal functions between the high‐dose ¹⁷⁷Lu‐FZ‐NF‐1 treatment group and the saline group (Figure S16). This finding provides a promising indication for the long‐term safety profile of ¹⁷⁷Lu‐FZ‐NF‐1 treatment.

2.7. Pharmacokinetics and Dosimetry of ⁶⁸Ga‐FZ‐NF‐1 in Humans

The patient enrollment and clinical trial design flow diagram are illustrated in Figure S17. From October 2025 to June 2026, 13 patients with TNBC were prospectively enrolled. The patient clinical characteristics are shown in Table S2.

For semiquantitative analysis, SUV_mean for heart, liver, spleen, lung, and kidney at 3 h after injection were 1.59 ± 0.40, 0.96 ± 0.48, 0.89 ± 0.43, 0.50 ± 0.15, and 10.47 ± 1.76, respectively (Table S3). Elevated uptake was observed in certain normal glands, with SUV_mean of 8.54 ± 2.20 for the pancreas and 7.06 ± 1.23, and 7.70 ± 2.25 for the salivary gland and thyroid at 3 h after injection, respectively. The favorable pharmacokinetics demonstrated a gradual decline over time in most major organs, including the heart, liver, spleen, lung, and kidneys, and the tracer was mainly excreted from the urinary system (Figure 5a and Figure S18). The radiation dose in the organs is detailed in Table 1, with an effective dose of 0.0148 mSv/MBq. No vital sign changes or adverse events were observed, confirming the safety of ⁶⁸Ga‐FZ‐NF‐1 for human PET imaging.

FIGURE 5.

(A) Time‐activity curves depict accumulation and clearance profiles of ⁶⁸Ga‐FZ‐NF‐1 in normal organs (n = 7). (B) Quantification results of lesions in TNBC with ⁶⁸Ga‐FZ‐NF‐1 at different time points. (C) Representative PET/CT images of TNBC patient (patient 7) at 10 min, 1 h, and 3 h after injection of ⁶⁸Ga‐FZ‐NF‐1. (D) MIP images of ¹⁸F‐FDG or ¹⁸F‐FAPI, and ⁶⁸Ga‐FZ‐NF‐1 PET/CT at 1 h timepoint in TNBC patients. (E) Representative PET images of patients 2 and 4 with additional lesions detected by ⁶⁸Ga‐FZ‐NF‐1 that appeared as false‐negatives on ¹⁸F‐FDG PET/CT. Also shown is Patient 7 with a lesion exhibiting intense ¹⁸F‐FDG but negative ⁶⁸Ga‐FZ‐NF‐1 uptake, which was subsequently confirmed as benign on follow‐up. (F) Patient‐level SUV_max comparison between ⁶⁸Ga‐FZ‐NF‐1 and ¹⁸F‐FDG. ** p < 0.01; ns, not significant. MIP, Maximum‐intensity‐projection.

TABLE 1.

⁶⁸Ga‐FZ‐NF‐1 Dosimetry Summary of Effective Dose.

Target Organs	Mean (mSv/MBq)	SD (mSv/MBq)
Adrenal gland	1.72E‐02	6.96E‐03
Brain	3.02E‐03	4.76E‐04
Breast	1.35E‐02	4.82E‐03
Colon wall	1.46E‐02	2.66E‐03
Gallbladder wall	1.39E‐02	4.43E‐03
Heart wall	4.77E‐02	7.99E‐03
Kidney	5.08E‐02	4.35E‐02
Liver	1.38E‐02	3.62E‐03
Lung	1.96E‐02	4.77E‐03
Ovaries	1.48E‐02	4.51E‐03
Pancreas	1.68E‐02	1.12E‐02
Salivary glands	2.17E‐02	2.78E‐03
Spleen	1.19E‐02	5.87E‐03
Stomach wall	1.15E‐02	1.08E‐03
Thymus	8.55E‐02	5.03E‐02
Thyroid	1.27E‐02	2.45E‐03
Urinary bladder wall	1.18E‐02	3.26E‐03
Uterus	1.47E‐02	4.60E‐03
Effective dose	1.48E‐02	2.94E‐03

Abbreviation: SD, standard deviation.

All observed vital signs (including blood pressure, heart rate, and body temperature) remained normal during the injection and at the 3‐h follow‐up. No individuals reported any adverse events.

Regarding the tumor uptake of ⁶⁸Ga‐FZ‐NF‐1, TNBC could be visualized within 10 min after the tracer was administered. Delayed imaging revealed a gradual increase in the tumor accumulation of ⁶⁸Ga‐FZ‐NF‐1 (median SUV_max: 10 min, 7.30 [IQR, 5.30–9.70]; 1 h, 7.90 [IQR, 5.50–10.70]; 3 h, 9.30 [IQR, 6.30–12.70]) (Figure 5b and Table S4). Representative PET/CT images of TNBC at different timepoints after injection of ⁶⁸Ga‐FZ‐NF‐1 were shown in Figure 5C.

2.8. ⁶⁸Ga‐FZ‐NF‐1 PET/CT in Patients With TNBC

Paired ⁶⁸Ga‐FZ‐NF‐1 and ¹⁸F‐fluorodeoxyglucose (¹⁸F‐FDG) PET/CT was conducted in 12 patients (patient 1–12) with TNBC history. Three patients were treatment‐naïve, while nine presented with recurrent or metastatic disease (Table S2). However, patient 4 was pathologically confirmed to have an incidental primary hepatocellular carcinoma (HCC) and was therefore excluded from the comparative analysis (Figure S17). A total of 100 lesions were evaluated, including 4 breast lesions, 5 chest wall metastases, 38 metastatic lymph nodes, 49 bone metastases, 2 lung metastases, and single metastasis in the muscle and adrenal gland, all of which were validated by contrast‐enhanced CT and follow‐up imaging. ⁶⁸Ga‐FZ‐NF‐1 identified 100% (100/100) of the lesions, whereas ¹⁸F‐FDG detected 98% (98/100) (Table 2). The smallest metastasis successfully identified by ⁶⁸Ga‐FZ‐NF‐1 was a lymph node metastasis measuring 6.5 mm in diameter, which was completely negative on the paired ¹⁸F‐FDG PET/CT. In patient 2, who presented with widespread systemic metastases, and ⁶⁸Ga‐FZ‐NF‐1 complemented ¹⁸F‐FDG by identifying additional skull metastatic lesions that appeared as false‐negatives on ¹⁸F‐FDG PET/CT (Figure 5e, green arrow). Patient 4 was a 67‐year‐old woman who underwent PET/CT to detect tumor recurrence. ⁶⁸Ga‐FZ‐NF‐1 PET/CT revealed intense uptake in the right posterior superior segment of the liver (Figure 5e, yellow arrow), but ¹⁸F‐FDG PET/CT demonstrated a false‐negative finding. In patient 7, a left axillary lymph node lesion demonstrated intense ¹⁸F‐FDG uptake (Figure 5e, red arrow) but negative ⁶⁸Ga‐FZ‐NF‐1 uptake, which was subsequently confirmed as benign during follow‐up. The SUV_max of all lesions was 10.00 (Range, 2.00–23.50) for ⁶⁸Ga‐FZ‐NF‐1 and 7.55 (Range, 1.90–33.80) for ¹⁸F‐FDG (Table 2). To evaluate the quantitative uptake while accounting for intra‐patient clustering, we performed a patient‐level comparative analysis. The maximum SUV_max of the target lesions for each patient was selected as the representative value. As a result, the patient‐level SUV_max and tumor background ratio (TBR) of ⁶⁸Ga‐FZ‐NF‐1 was comparable to those of ¹⁸F‐FDG, with no statistically significant difference (Figure 5f and Figure S19).

TABLE 2.

Comparison of SUV_max on ⁶⁸Ga‐FZ‐NF‐1 PET/CT and ¹⁸F‐FDG PET/CT Images in Primary and Metastatic Tumors.

Tumor location	Total lesion (n)	68Ga‐FZ‐NF‐1		18F‐FDG
		Positive lesions (n)	Median SUVmax (range)	Positive lesions (n)	Median SUVmax (range)
Breast	4	4	10.65 (9.90–12.10)	4	11.00 (8.40–26.70)
Chest wall	5	5	9.90 (5.40–15.10)	5	6.80 (5.80–14.00)
LN mets	38	38	7.90 (3.30–19.30)	37	8.80 (1.90–33.80)
Bone mets	49	49	12.60 (3.50–23.50)	48	7.00 (1.90–13.20)
Muscle mets	1	1	3.60	1	4.50
Lung mets	2	2	6.75 (2.00–11.50)	2	4.60 (2.90–6.30)
Adrenal gland mets	1	1	9.70	1	11.20
Total	100	100	10.00 (2.00–23.50)	98	7.55 (1.90–33.80)

Abbreviations: SUV_max , maximum of standardized uptake value; LN, lymph node.

In a head‐to‐head comparison in patient 13 involving ¹⁸F‐FAPI positron emission tomography/Magnetic resonance imaging (PET/MRI) and ⁶⁸Ga‐FZ‐NF‐1 PET/CT, both radiotracers showed diffusely increased hepatic uptake while the ⁶⁸Ga‐FZ‐NF‐1 PET/CT revealed more extensive disease involvement with higher uptake intensity (SUV_max, 15.3 vs. 10.1) (Figure 5d).

Representative PET/CT images with H&E staining (Figure S20), IHC staining for FAP and Nectin‐4 are presented in Figure 6. IHC analysis revealed strong dual‐positivity in patient 1 (Figure 6a), patient 3 (Figure 6c), and patient 13 (Figure 6d). In contrast, Patient 2 (Figure 6b) exhibited weak FAP positivity but strong Nectin‐4 positivity, underscoring the value of the dual‐targeting strategy.

FIGURE 6.

Representative ¹⁸F‐FDG and ⁶⁸Ga‐FZ‐NF‐1 PET/CT images alongside IHC staining of the corresponding lesions. Biopsy site is indicated by arrow. Scale bar = 50 µm.

2.9. FAP and Nectin‐4 Expression in Tissue Microarrays of TNBC

Tissue microarray analysis (TMA) of 60 TNBC samples showed that the majority (60.0%) exhibited co‐expression of FAP and Nectin‐4. Discordant expression of FAP and Nectin‐4 was observed in 35.0% of the cases. This included 23.3% of cases that were FAP positive but Nectin‐4 negative and 11.7% of cases that were FAP negative but Nectin‐4 positive. Only 5% were negative for both markers. (Figure 7). Notably, the independent H‐score evaluation demonstrated excellent inter‐observer agreement between the two pathologists (Kappa = 0.82).

FIGURE 7.

FAP and Nectin‐4 expression in 60 TNBC patients. (A) Representative imaging of paired IHC staining of 60 human TNBC specimens using anti‐Nectin‐4 and anti‐FAP antibodies. (B) H‐score of FAP and Nectin‐4. (C) Proportion of different for FAP and Nectin‐4 expression level. TNBC, triple negative breast cancer.

3. Discussion

Nectin‐4 is highly expressed in multiple malignancies [8, 21, 22], including urothelial carcinoma and TNBC, and has recently emerged as a promising target for cancer imaging [9, 10] and antibody‐drug conjugate therapy [23, 24]. FAP, a marker of CAFs within the tumor microenvironment [25, 26], is one of the most extensively investigated theranostic targets in recent years [27, 28]. Both proteins represent highly promising targets for theranostic applications with significant translational potential. In the present study, we report the synthesis, preclinical evaluation, and preliminary clinical translation of the ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1 theranostic pair, a first‐in‐class heterodimeric radiotracer targeting both Nectin‐4 and FAP in TNBC.

Structurally, we utilized an L‐glutamate (Glu) moiety as a central trifunctional scaffold to bridge the Nectin‐4‐targeting ligand and the FAP‐targeting ligand with the DOTAGA chelator. The asymmetric configuration of two distinct carboxyl groups and one amino group allows for the precise, site‐specific conjugation of distinct pharmacophores while ensuring high chemical purity and structural reproducibility during the synthesis of this hetero‐bivalent probe [12, 19]. Beyond synthetic precision, the Glu‐based framework offers exceptional metabolic stability. The amide bonds within this framework are more resistant to enzymatic cleavage than ester‐linked bridges, which are often compromised by circulating esterases [29]. Unlike long‐chain aliphatic linkers that often increase lipophilicity and lead to undesirable hepatobiliary accumulation, the Glu scaffold promotes rapid blood clearance via the kidneys, thereby ensuring a clean imaging background [29, 30]. The favorable pharmacokinetics minimize non‐specific background and enhance imaging contrast, allowing for the precise detection of small metastatic lesions [19, 31]. This linker strategy shares structural similarities with other well‐established dual‐targeted tracers, such as FAPI‐RGD [32] and FAPI‐LM3 [12], further confirming the stability and reliability of this scaffold.

Compared to the monomeric tracers ⁶⁸Ga‐FAPI‐FUSCC‐II and ⁶⁸Ga‐FZ‐NR‐1, ⁶⁸Ga‐FZ‐NF‐1 exhibited better imaging performance in microPET imaging. In dual‐positive MDA‐MB‐468‐hFAP tumor models, the tumor uptake of ⁶⁸Ga‐FZ‐NF‐1 was significantly higher than that of monomeric probes. Furthermore, while the monomeric tracers showed a gradual decline in tumor uptake starting from 1 h post‐injection as reported [9, 16, 33], ⁶⁸Ga‐FZ‐NF‐1 demonstrated a divergent pharmacokinetic pattern with tumor uptake rising steadily from 1 to 4 h. The enhanced tumor uptake and prolonged retention may be attributed to the simultaneous engagement of Nectin‐4 and FAP. Even if one targeting moiety detaches from its receptor, the secondary motif can anchor the probe to the alternative target [12, 20]. We further evaluated the performance of ⁶⁸Ga‐FZ‐NF‐1 in various dual‐target or single‐target positive xenograft models including MDA‐MB‐231‐hNectin‐4‐hFAP, HT1080‐hFAP, MDA‐MB‐468, and HT1376 tumor models. Consistent with the reported performance of other dual‐targeting probes [32, 34], ⁶⁸Ga‐FZ‐NF‐1 maintained substantial tumor uptake and high TBR in these models despite the absence of one target. These findings demonstrate that ⁶⁸Ga‐FZ‐NF‐1 can be used for visualization of a wide spectrum of tumors, including those of FAP+/Nectin‐4+, FAP+/Nectin‐4‐, and FAP‐/Nectin‐4+, thereby ensuring maximal coverage of tumor lesions within complex heterogeneous tumor microenvironment.

Crucially, this dual‐targeting anchoring effect not only optimizes diagnostic contrast but also provides the prolonged tumor residence time essential for RLT. This extended pharmacokinetic advantage was directly visualized and confirmed by SPECT/CT imaging and biodistribution studies. Unlike the ¹⁷⁷Lu‐labeled monomeric agents that exhibited rapid washout, ¹⁷⁷Lu‐FZ‐NF‐1 maintained substantial tumor accumulation for up to 120 h post‐injection. This sustained radiation delivery resulted in significant, dose‐dependent tumor regression in TNBC mouse models. While single‐target agents provided only moderate and transient tumor inhibition, ¹⁷⁷Lu‐FZ‐NF‐1 significantly prolonged tumor suppression without obvious systemic toxicity. Ultimately, these data highlight the strong potential of ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1 for TNBC theranostics. Recently, dual‐targeting strategies have shown great potential in developing theranostic radiopharmaceuticals. However, most research remains limited to ⁶⁸Ga‐ or ¹⁸F‐labeled PET imaging [12, 19, 29, 30, 31]. Only a few studies have explored ¹⁷⁷Lu‐labeled SPECT/CT imaging and ex vivo biodistribution [35, 36, 37]. Although dual‐targeting RLT is still in its early stages, this strategy has been well‐validated in other targeted oncological therapies. For example, bispecific antibodies targeting EGFR/MET [38] or CTLA4/PD‐L1 [39], dual‐target ADCs directed against EGFR/HER3 [40] or PSMA/B7‐H3 [41], and dual‐target CAR‐T therapies engaging CD19/CD20 [42] or MSLN/B7H3 [43] have been proven to overcome single‐agent resistance and antigen‐loss relapse.

In our preliminary clinical translational study, ⁶⁸Ga‐FZ‐NF‐1 exhibited intense physiologic uptake in the salivary glands, thyroid, pancreas and kidneys. Physiologic uptake of salivary gland and thyroid was reported in Nectin‐4 monomer [9]. Similarly, high uptake in the thyroid and pancreas has been observed in ⁶⁸Ga‐labeled FAPI dimers [19, 44, 45]. The reasons for the physiological high uptake in these normal organs remain poorly understood at this stage. Delayed imaging demonstrated that ⁶⁸Ga‐FZ‐NF‐1 tumor uptake exhibited a continuous, gradual increase from 10 min to 3 h post‐injection. Meanwhile, the physiological uptake in most normal organs underwent gradual clearance. Consequently, the delayed imaging provided optimal lesion contrast. Similar favorable imaging characteristics and durable tumor retention have also been observed with other well‐established dual‐targeted tracers [12, 30].

In the head‐to‐head comparison with ¹⁸F‐FDG, ¹⁸F‐FDG missed several metastatic lesions, including lymph node, liver, and bone lesions. Owning to its low physiological brain uptake, ⁶⁸Ga‐FZ‐NF‐1 avoids the intense background signal typical of ¹⁸F‐FDG, allowing brain and skull lesions to be clearly detected. In the case of patient 1, the administration of granulocyte colony‐stimulating factor (G‐CSF) within 2weeks of the PET scan induced diffuse high bone uptake on¹⁸F‐FDG PET due to bone marrow activation [46], which hindered the accurate assessment of bone lesions. In contrast, the uptake of ⁶⁸Ga‐FZ‐NF‐1 was completely unaffected by bone marrow stimulation, allowing bone metastases to be clearly identified and ensuring the diagnostic reliability. More significantly, ⁶⁸Ga‐FZ‐NF‐1 has also demonstrated the potential to detect multiple primary tumors. In the case of patient 4, ¹⁸F‐FDG failed to detect the liver lesion that was clearly recognized by ⁶⁸Ga‐FZ‐NF‐1, and was ultimately pathologically confirmed as HCC. These findings suggest that ⁶⁸Ga‐FZ‐NF‐1 holds substantial clinical promise for identifying other FDG false‐negative malignancies, such as well‐differentiated HCC and gastric signet ring cell carcinoma [47, 48]. While our current cohort focuses exclusively on TNBC, the clinical applicability of dual‐targeting strategies is underscored by the successful translation of tracers, including ⁶⁸Ga‐FAPI‐RGD [19, 30, 32] for a broad spectrum of solid tumors, ⁶⁸Ga‐FAPI‐LM3 [12] for nasopharyngeal carcinoma, and ⁶⁸Ga‐FAPI‐PSMA [29] for prostate cancer. Given that FAP is widely overexpressed within the tumor microenvironment of most epithelial tumors and Nectin‐4 is highly upregulated across multiple malignancies, ⁶⁸Ga‐FZ‐NF‐1 holds immense pan‐cancer diagnostic potential beyond breast cancer.

In this study, IHC of the biopsy lesions and TMA analyses revealed marked heterogeneity in the expression of Nectin‐4 and FAP across different patients and lesions. TMA data showed that while dual‐target co‐expression was observed in 60.0% of cases, 35.0% remained positive for only one target. This pathological evidence clearly illustrates the risk of false‐negative results in single‐target imaging driven by the profound molecular heterogeneity of TNBC [3, 4, 49] and mechanistically underscores the necessity of a dual‐targeting strategy in clinical applications. In the preliminary clinical study, the performance of ⁶⁸Ga‐FZ‐NF‐1 was highly consistent with TMA results, achieving a lesion detection rate of 100%. The high SUV_max and high rate of lesion detection confirm that by concurrently targeting both tumor cells and the tumor microenvironment, ⁶⁸Ga‐FZ‐NF‐1 effectively mitigates the risk of false negatives caused by tumor heterogeneity. By maximizing tumor signal capture, this dual‐target strategy provides a more reliable imaging tool for the precise staging and personalized management of TNBC.

Despite these promising findings, certain limitations remain. First, while we have successfully demonstrated the preclinical therapeutic efficacy of ¹⁷⁷Lu‐FZ‐NF‐1, its dosimetry, safety profile, and anti‐tumor efficacy in humans have yet to be evaluated. Besides, labeling FZ‐NF‐1 with alpha‐emitters such as ²²⁵Ac to maximize targeted radiotoxicity warrants future exploration. Additionally, the current clinical translation study enrolled only 13 patients, representing a relatively small sample size and failing to encompass all molecular subtypes of TNBC [3]. In the future, we will conduct large sample, multicenter clinical studies to further validate the broad applicability of ⁶⁸Ga‐FZ‐NF‐1 across different TNBC subtypes and actively explore its potential as a theranostic agent in the treatment of refractory TNBC.

4. Methods

4.1. Radiolabeling and Stability Testing of ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1

For ⁶⁸Ga labeling, 20 µg of the precursor in 110 µL of sodium acetate buffer (1.5 m) was allowed to react with 1 mL of ⁶⁸Ga solution (150 MBq in 0.1 m HCl) at 85°C for 15 min. The radiolabeled product was incubated in PBS, saline and FBS at 37°C for 60 min and 120 min to test the in vitro stability by HPLC. 6‐week‐old BALB/c mice were administrated 7.4 MBq of radiotracer in 200 µL saline via tail‐vein injection, the urine was collected 60 and 120 min after injection to assess the in vivo stability by HPLC. For ¹⁷⁷Lu labeling, 0.37 GBq of ¹⁷⁷LuCl₃ (ITM, Germany) was diluted with 0.1 mL of NaOAc (0.5 m, pH 5.5) and added to the precursor (10 µg). The reaction mixture was incubated at 95°C for 30 min. The radiolabeled product was incubated in saline and FBS at 37°C for 24 and 48 h to test the in vitro stability by HPLC.

4.2. Cell and Animal Models

MDA‐MB‐231, MDA‐MB‐468, and HT1376 cell lines were obtained from Fuheng Biology. MDA‐MB‐231‐hNectin‐4‐hFAP cells were derived from MDA‐MB‐231 cells and stably transfected with human Nectin‐4 and FAP. MDA‐MB‐468‐hFAP cells were derived from MDA‐MB‐468 cells and stably transfected with human FAP. HT1080‐hFAP cells were derived from HT1080 cells and stably transfected with human FAP.

Female BALB/c nude and NCG mice (6–8 weeks old) were obtained from GemPharmatech Co., Ltd., and injected subcutaneously with different cell lines to generate cell‐derived xenograft (CDX) models. When the tumor length reached 0.5–1 cm, the mice were ready for PET imaging or biodistribution assay.

4.3. Flow Cytometry

Conjugated antibodies and probes used for flow cytometry are listed above. For Nectin‐4 and FAP staining, cells were resuspended in 100 µL PBS and incubated with FcR Blocking Reagent (Miltenyi, dilution 1:20) on ice for 20 min from light. Then, the cells were centrifuged and resuspended in 100 mL PBS containing 1% FBS (Gibco). Fixable Viability Dye (eFluor 506, dilution 1:1000), antibodies against surface markers, including anti‐Nectin‐4 (Abclonal, A25193, 1/20), anti‐FAP (Abclonal, A24965, 1/20) antibodies, along with Brilliant Stain Buffer Plus, were added to the cell suspension. Following staining, the cells were washed twice and resuspended in PBS containing 1% FBS before analysis on a BD FACSymphony A3 Cell Analyzer.

4.4. Distribution Coefficient of ⁶⁸Ga‐FZ‐NF‐1

The octanol‐PBS distribution coefficients of the tracer were determined by mixing 74 kBq of purified ⁶⁸Ga‐FZ‐NF‐1 with 500 µL double‐distilled water and 500 µL octanol in a 1.5 mL tube. The mixture was vortexed for 15 min and centrifuged at 12 000 rpm for 10 min. The radioactivity of each sampling layer (100 µL) was measured using γ‐counter (SN‐697, Shanghai Nuclear Institute Rihuan Photoelectric Instrument Company Limited). LogP = Log (counts in octanol layer/counts in double‐distilled water layer).

4.5. Plasma Protein Binding Rate of ⁶⁸Ga‐FZ‐NF‐1

Mouse plasma (SBJ‐P‐M003) was obtained from Sbj Biology. ⁶⁸Ga‐FZ‐NF‐1(0.74 MBq) was incubated with 100 µL mouse plasma. The mixture was centrifuged at 15 000 rpm for 10 min in a 30 K ultrafiltration tube after 30, 60, or 90 min incubation. The radioactivity from the membrane and lower liquid parts was detected separately with γ counter. The radioactivity from the membrane represents the amount of the radiotracer bound to plasma proteins.

4.6. Pharmacokinetics of ⁶⁸Ga‐FZ‐NF‐1

The blood pharmacokinetics study was performed using 6–8 weeks old female BALB/c mice (n = 5). Each mouse was intravenously injected with 5.5 MBq of ⁶⁸Ga‐FZ‐NF‐1. A trace amount of blood was taken from the tail tip of the mice using microscopic capillaries at 1, 3, 5, 10, 15, 30, 60, 90, 120, and 240 min post‐injection. The radioactivity of the blood samples was detected with γ counter.

For the normal organ pharmacokinetics study, 5.5 MBq ⁶⁸Ga‐FZ‐NF‐1 was injected into the tail vein of 18 female BALB/c mice divided into six groups of three, among which six groups were sacrificed at 5, 10, 15, 30, 60, and 120 min after the injection, respectively. The tissue samples of heart, liver, spleen, lung, kidney, stomach, intestine, muscle, bone, and brain were collected, blotted, weighed, and the radioactivity was counted by a γ counter. The data were normalized to %ID/g.

4.7. Cell Uptake Assay

MDA‐MB‐231‐hNectin‐4‐hFAP, MDA‐MB‐468‐hFAP, HT1080‐hFAP, MDA‐MB‐468, HT‐1376, and MDA‐MB‐231 cell lines were seeded in 24‐well plates and cultured in a routine medium until they reached approximately 90% confluence. During the experiment, the medium was replaced with 300 µL FBS‐free medium containing 150 kBq ⁶⁸Ga‐FZ‐NF‐1 with or without unlabeled precursor (FAPI‐04, Nectin‐4, or FAPI‐04 plus Nectin‐4) and incubated for 1 h at 37°C. Afterward, the cells were washed twice with PBS and lysed with 200 µL of NaOH (1 m). The supernatant and cells were collected separately, and the radioactivity was measured with a γ counter.

4.8. Small Animal PET/CT Imaging of ⁶⁸Ga‐FZ‐NF‐1

The tumor‐bearing mice were intravenously injected with approximately 5.5 MBq ⁶⁸Ga‐FZ‐NF‐1, ⁶⁸Ga‐FZ‐NR‐1, and ⁶⁸Ga‐FAPI‐FUSCC‐II (n = 3 for each point). Subsequent static PET scans were obtained at intervals of 0.5, 1, 2, and 4 h after injection using an Inveon small‐animal PET scanner (Biograph 6, Siemens Medical Solutions). The blocking experiments were performed by coinjection with approximately 500 nmol unlabeled precursor (FAPI‐04, Nectin‐4, or FAPI‐04 plus Nectin‐4). Images were reconstructed using a three‐dimensional ordered‐subset expectation maximum algorithm based on the CT data for attenuation correction and converted to the percent injected dose per gram of tissue (% ID/g).

4.9. Ex vivo Biodistribution of ⁶⁸Ga/¹⁷⁷Lu‐FZ‐NF‐1

For ⁶⁸Ga‐FZ‐NF‐1 biodistribution studies, the tumor‐bearing mice were sacrificed and dissected immediately 0.5, 1, 2, or 4 h after intravenous injection of ⁶⁸Ga‐FZ‐NF‐1 (5.5 MBq per mouse, n = 3 for each point). Additionally, the mice that were co‐injected with approximately 500 nmol unlabeled precursor (FAPI‐04, Nectin‐4, or FAPI‐04 plus Nectin‐4) were sacrificed at 1 h for the blocking study. For ¹⁷⁷Lu‐FZ‐NF‐1 biodistribution studies, the tumor‐bearing mice were sacrificed and dissected immediately 4, 10, 24, 48, 72, 96, or 120 h after intravenous injection of ¹⁷⁷Lu‐FZ‐NF‐1 (3 MBq per mouse, n = 3 for each point). Heart, liver, spleen, lung, kidney, stomach, intestine, bladder, muscle, bone, brain, tumor, and blood were collected and weighed, and the radioactivity was measured with a γ counter. The data were normalized to %ID/g.

4.10. H&E and IHC in CDX Tumor Tissue and Patient Biopsy Samples

CDX tumor tissue dissected from the above biodistribution were sectioned for H&E immunohistochemical analysis. The patient biopsy specimens were obtained from the Department of Pathology of Fudan University Shanghai Medical Center.

Immunohistochemical staining for FAP and Nectin‐4 expression was performed using the anti‐FAP antibody (Abcam, #ab207178, 1/500) or anti‐Nectin‐4 antibody (Abcam, #ab192033, 1/300). The results were processed by ImageJ software.

4.11. Small Animal SPECT/CT Imaging of ¹⁷⁷Lu‐FZ‐NF‐1

The tumor‐bearing mice were intravenously injected with approximately 11.1 MBq ¹⁷⁷Lu‐FZ‐NF‐1, ¹⁷⁷Lu‐FZ‐NR‐1, and ¹⁷⁷Lu‐FAPI‐FUSCC‐II (n = 3 for each point). Subsequent static SPECT scans were obtained at intervals of 4, 10, 24, 48, 72, 96, and 120 h after injection using nanoScan SPECT/CT (Mediso).

4.12. Therapeutic Efficacy Evaluation of ¹⁷⁷Lu‐FZ‐NF‐1

MDA‐MB‐231‐hNectin‐4‐hFAP tumor models with positive Nectin‐4 and FAP expression were utilized in this study. MDA‐MB‐231‐hNectin‐4‐hFAP tumor‐bearing mice with smaller tumors (∼50 mm³) were randomly divided into four groups: saline control group, 7.4 MBq ¹⁷⁷Lu‐FZ‐NF‐1 group, 7.4 MBq¹⁷⁷Lu‐FAPI‐FUSCC‐II group, and 7.4 MBq ¹⁷⁷Lu‐FZ‐NR‐1 group. MDA‐MB‐231‐hNectin‐4‐hFAP tumor‐bearing mice with larger tumors (∼150 mm³) were randomly divided into five groups: saline control group, 14.8 MBq ¹⁷⁷Lu‐FZ‐NF‐1 group, 29.6 MBq ¹⁷⁷Lu‐FZ‐NF‐1 group, 29.6 MBq ¹⁷⁷Lu‐FAPI‐FUSCC‐II group, and 29.6 MBq ¹⁷⁷Lu‐FZ‐NR‐1 group. Tumor volume and mice body weight were examined every 2days after a single‐dose injection. Tumor volume was measured using the formula V = width [2] × length × 0.5. Mice were euthanized when the tumor volume exceeded 1000 mm³, the tumor ruptured, or when it lost more than 20% of its body weight.

4.13. Human Participants and PET/CT Imaging

The human study was approved by the Ethics Committee of Fudan University Shanghai Cancer Center (number 2407299‐29) and registered at ClinicalTrials.gov (NCT07708922). Written informed consent was obtained from all patients. The participant inclusion criteria for patients were as follows: (1) age ≥ 18 years with a previous histologically confirmed diagnosis of breast cancer; (2) suspected metastasis shown at conventional imaging (CT, MRI, US); (3) underwent ⁶⁸Ga‐FZ‐NF‐1 PET/CT and ¹⁸F‐FDG PET/CT or ¹⁸F‐FAPI PET/MRI within 1 week; (4) signed informed consent. Exclusion criteria were as follows: (1) with other malignancies; (2) incomplete case data. From October 2025 to June 2026, 13 patients with TNBC were prospectively enrolled. Twelve patients (patients 1–12) underwent paired ¹⁸F‐FDG and ⁶⁸Ga‐FZ‐NF‐1 PET/CT. Following the pathological confirmation and subsequent exclusion of patient 4 for the incidental HCC, 11 patients were ultimately included in the TNBC lesion‐level comparative analysis. To assess the pharmacokinetics and dosimetry of ⁶⁸Ga‐FZ‐NF‐1, and 7 of these 12 patients (patients 7–12) underwent dynamic imaging at 10 min, 1 h, and 3 h post‐injection using Biograph mCT Flow scanner (Siemens Medical Solutions). The remaining 1 patient (patient 13) in the cohort was evaluated using paired ¹⁸F‐FAPI PET/MR and ⁶⁸Ga‐FZ‐NF‐1 PET/CT at the 1 h timepoint.

4.14. Imaging Interpretation

Two experienced nuclear medicine physicians analyzed and interpreted the PET/CT images independently, and they reached a consensus in cases of inconsistency. For quantitative analysis, SUV_max normalized according to body weight was measured using a multimodality computer platform (Syngo; Siemens Healthineers). TBR was calculated according to the following formulas: TBR = SUV_max of tumor lesions/SUV_mean of the liver. If the number of lesions in an organ or region exceeded five, the count was truncated to five to avoid bias. The size of primary and metastatic lesions was measured by CT.

4.15. Radiation Dosimetry of ⁶⁸Ga‐FZ‐NF‐1

Patient underwent three whole‐body PET/CT scans at 10, 60, and 180 min post‐injection of ⁶⁸Ga‐FZ‐NF‐1. Volumes of interest (VOIs) covering target organs and the whole body were semi‐automatically delineated using a hybrid region‐growing intelligent delineation (HRID) algorithm. The activity concentration (kBq/mL) for each VOI was converted to the percentage of injected dose (%ID). The analyzed organs included brain, parotid glands, salivary glands, thyroid, heart, liver, spleen, lung, kidneys, gallbladder, pancreas, stomach wall, colon wall, bone, and muscle. The effective dose of ⁶⁸Ga‐FZ‐NF‐1 was determined by IDAC‐Dose [50].

4.16. FAP and Nectin‐4 IHC in Breast Cancer Patients

Tissue microarrays of human TNBC (HBreD060Bc01) were purchased from Shanghai Outdo Biotech Co. The Ethics Committee of the Shanghai Outdo Biotech Co. approved the study. Serial sections of tumor tissue microarrays were used to ensure consistency in verifying different biomarkers in IHC experiments. For IHC analysis of paraffin‐embedded areas, the BenchMark ULTRA (Ventana Medical Systems) automated slide stainer was used to stain cells with the anti‐FAP antibody (Abcam, #ab207178, and 1/1000) or anti‐Nectin‐4 antibody (Abcam, #ab192033, and 1/500). FAP and Nectin‐4 expression was semi‐quantitatively evaluated with the H‐score method. The staining intensity was graded as 0 (negative), 1 (weak), 2 (moderate), or 3 (strong). The H‐score was calculated using the standard formula: H‐score = ∑(i × P_i), where i is the intensity score and P_i is the percentage of stained cells (0%–100%), yielding a total score ranging from 0 to 300. To ensure objectivity, all slides were independently assessed by two experienced pathologists in a blinded manner, and inter‐observer agreement was evaluated using Cohen's kappa.

4.17. Statistical Analysis

All statistical analyses were conducted using GraphPad Prism (Version 10.00). All animal experiments were repeated at least twice to ensure reproducibility. Quantitative data from in vitro and preclinical in vivo experiments, as well as the SUV_mean of normal human organs, are presented as the mean ± standard deviation (SD). Mean values between two independent groups were compared using Student's t‐test‐, while a one‐way ANOVA was applied for multi‐group comparisons. For human lesion analysis, SUV_max and TBR are presented as the median [with the interquartile range (IQR) or with the range]. Accordingly, the paired comparisons of SUV_max and TBR between ¹⁸F‐FDG and ⁶⁸Ga‐FZ‐NF‐1 PET/CT scans were evaluated using the non‐parametric Wilcoxon matched‐pairs signed‐rank test. For the longitudinal comparison of ⁶⁸Ga‐FZ‐NF‐1 imaging parameters across the three different time points, the non‐parametric Friedman test was utilized.

Author Contributions

S.S. and X.X. designed and led the project. Y.L., L.S., X.P., S.L., J.Z., F.Q., F.Z., and L.B. performed experiments. Y.L., B.G., X.P., and D.W. analyzed data. B.G. and B.Z. provided clinical samples and information. B.G. and Y.L. guided data analysis. Y.L., B.G., and S.L. wrote the manuscript. All authors contributed to data interpretation and discussion of results on the manuscript.

Funding

This work was supported by the Key Program of National Natural Science Foundation of China (No. U23A20465 to S.S.), the Science and Technology Commission of Shanghai Municipality (No. 23DZ2291400 to S.S.), the National Natural Science Foundation of China (No. 82272305 to S.S. and No. 12275057 to X.X.), the Nuclear Technology R&D Program (No. HNKFHBZ202319 to S.S.), the Shanghai Collaborative Innovation Project (HCXBCY‐2024‐001 to S.S.), and the Shanghai Oriental Talents Program Youth Project (No. SSF828085 to X.X.).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: advs77606‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.