First-in-human Study of a 99mTc-Labeled Single-Domain Antibody for SPECT/CT Assessment of HER2 Expression in Breast Cancer

Background: Accurate determination of human epidermal growth factor receptor 2 (HER2) expression is essential for HER2-targeted therapy. HER2 expression in a complex environment, such as in a heterogenous tumor, makes precise assessment dicult using current methods. Therefore, we developed a novel 99m Tc-labeled anti-HER2-single domain antibody ( 99m Tc-NM-02) as a molecular imaging tracer for non-invasive detection of HER2 expression and investigated its safety, radiation dosimetry, biodistribution, and tumor-targeting potential in breast cancer patients. Methods: A lead compound (NM-02) was screened from a library of hexahistidine-tagged anti-HER2-single domain antibodies and labeled with 99m Tc for the preparation of 99m Tc-NM-02 tracer. Ten women with breast cancer were administered 99m Tc-NM-02 at a mean dose of 458 ± 37 MBq (406−510 MBq), corresponding to 100 μg of NM-02. Whole-body and local SPECT/CT images were acquired at 1 and 2 h post-administration to investigate the tumor-targeting potential in primary and metastatic lesions. Additional images were acquired at 10 min, 3 h, and 24 h in three patients to calculate radiation dosimetry. Physical evaluation and blood analysis were performed for safety assessment. Results: No drug-related adverse reactions occurred. The tracer mainly accumulated in the kidneys and liver with mild uptake in the spleen, intestines, and thyroid, but only background levels were observed in other organs where primary tumors and metastases typically occurred. The mean effective dose was 6.56 × 10 −3 mSv/MBq, and tracer uptake was visually observed in primary tumors and metastases. Owing to the fast clearance of the tracer, we were able to suciently discern uptake over normal background in both primary lesions and metastases within 2 h after injection. A maximal standard uptake value of 1.5 could be a reasonable cutoff for determining HER2 positivity using SPECT/CT imaging. Conclusions: Our 99m Tc-NM-02 tracer can be safely used for imaging in breast cancer patients with reasonable radiation doses, favorable biodistribution and imaging characteristics. 99m Tc-NM-02 SPECT imaging may provide an accurate and non-invasive method to detect HER2 status in breast cancer patients.


Background
Breast cancer has become the most common cancer in women worldwide [1][2][3]. One in eight women will be diagnosed with breast cancer in her life. Breast cancers can be classi ed into cancer subtypes based on the expression of molecular biomarkers, which allows physicians to adjust treatment plans and response assessments for improved survival and morbidity [4][5][6]. This development in breast cancer diagnosis and treatment has led to an evolution from standard strategies to precision medicine [7]. In addition to hormone receptors, human epidermal growth factor receptor 2 (HER2), which is highly expressed in about 20% of breast cancer patients, is a crucial biomarker for breast cancer classi cation [8,9]. These patients usually respond poorly to endocrine therapy and chemotherapy but obtain a marked survival bene t from HER2-targeted therapy [9][10][11]. Therefore, accurate determination of HER2 expression is a primary concern when selecting therapy. Immunohistochemistry (IHC) and uorescence in situ hybridization (FISH) of tumor tissues are routinely performed in the clinic to determine HER2 expression. Based on pathology results, breast cancer patients who have strong overexpression of HER2 (HER2 3+) or ampli cation of HER2/neu (FISH +) are suitable for HER2-targeted therapy, because they will gain the most clinical bene ts from this treatment [12]. The guidelines for HER2 testing have been revised several times to improve the accuracy and reproducibility of the interpretation of HER2 status [13]. However, despite these revisions, multiple studies have shown inaccurate HER2 testing results for breast cancer patients. The discordance rate was reported to be as high as 20% in an early study [14], although it had decreased to less than 10% in a more recent study [15]. Nevertheless, incorrect classi cation of HER2 status may lead clinicians to not administer anti-HER2 therapy when it might be effective and, conversely, increase needless cost and prolonged treatment for patients with false-positive results. With the increasing administration of HER2-directed therapies, the accurate determination of HER2 expression is critical.
Single-domain antibodies (sdAbs) are small antigen-binding fragments that are naturally derived from heavy-chain-only antibodies and are mainly produced in camelids [16]. Due to the outstanding features, such as fast blood clearance, deep tissue penetration, and high a nity, sdAbs are identi ed as ideal candidates for nuclear medicine applications [16,17]. Several studies have demonstrated the safety and feasibility of radiolabeled sdAbs for cancer diagnosis and treatment [18][19][20][21][22][23][24][25][26]. In our previous study, we reported a 99m Tc-labeled antiprogrammed death ligand-1 (PD-L1) sdAb as a SPECT/CT tracer for PD-L1 expression in non-small cell lung cancer (NSCLC) [26]. The tracer had an excellent safety pro le, favorable imaging characteristics, and a signi cant correlation with PD-L1 IHC results in NSCLC patients. Here, we used another novel sdAb (NM-02) to develop a 99m Tc-labeled anti-HER2 sdAb ( 99m Tc-NM-02) and examined its use as a novel tracer for SPECT/CT assessment of HER2 expression in breast cancer patients.

Patient Characteristics
Between September 2019 and August 2020, 10 female breast cancer patients at rst diagnosis, at relapse, or under treatment, participated in this study. All tissue samples used for pathologic HER2 testing were obtained less than 15 days prior to imaging. Five patients had no HER2 expression, and 5 patients had moderate or high expression of HER2 on IHC (2 + or 3+). Three of 10 patients had only primary tumors, and the other 7 patients had both primary tumors and metastases. The patient characteristics are summarized in Table 1.

Safety Assessment
No signs or symptoms of drug-related adverse effects were reported during the 1-week post-injection follow-up period. Clinical blood tests showed no signi cant changes that could be related to the study drug.
Biodistribution Figure 1 shows the whole-body SPECT images of a representative patient at different time points. 99m Tc-NM-02 was mainly observed in the liver and kidneys, with mild uptake in the spleen, intestines, and glandular tissues such as the thyroid, submandibular, and parotid glands.
This biodistribution pattern was already present at 10 min post-injection and decreased over time. Notably, fast clearance was seen in the lung and blood, with only a very low level of radioactivity remaining at 1 h post-injection, allowing improved tumor-to-background signal within 2 h after injection. Although there was radiotracer uptake in the liver and intestines, there were no signs of hepatobiliary excretion, such as an accumulation in the gallbladder or duodenum. In contrast, high uptake in the kidneys and continuous excretion of activity into the bladder could be clearly seen during the studied time period.
Radiation Dosimetry Table 2 summarizes the individual organ doses and individual effective doses for three patients involved in the dosimetry study. The kidneys showed the highest organ dose (0.031 ± 0.0021 mSv/MBq), followed by the thyroid (0.019 ± 0.0050 mSv/MBq) and liver (0.014 ± 0.0019 mSv/MBq). Time activity curves for the six organs with the highest radiotracer uptake are shown in Fig. 2. The mean effective dose was 6.56 × 10 − 3 mSv/MBq. The SPECT/CT images of all patients were carefully evaluated by visual interpretation (Table 3). Other than one HER2-negaitive patient (Patient BR007) who showed no distinct uptake in the tumor, nine patients had varying degrees of tracer uptake in their primary tumors, and the tumors were classi ed as positive based on imaging results. The obvious tumor uptake in the primary tumors of the ve HER2-positive patients was consistent with their IHC results (HER2 2 + or 3+). Figure 3 shows the representative images of Patient BR007 (HER2 0) who had no distinct tumor uptake, and Patient BR001 (HER2 3+) who had high tracer accumulation in the primary lesions. Moreover, the tracer accumulation observed in Patient BR001 matched well with the uptake patterns of a recent 18 F-FDG scan (Fig. S1). The remaining four HER2-negative patients (HER2 0) had visually evident uptake in the primary tumors, but varying levels of intensity were observed in different tumor areas, suggesting intratumoral heterogeneity in HER2 expression. Fig. S2 shows a representative case of tumor heterogeneity (Patient BR005). Tracer accumulation was also observed in the metastases with varying degrees of uptake. Figure 4 shows the images of Patient 6 (HER2 3+) and Patient 10 (HER2 3+), both of whom had clear tracer accumulation in their primary lesions but different uptake levels in their metastases, which may re ect intertumoral heterogeneity in HER2 expression [27,28]. Discussion HER2-positive breast cancer patients have poor prognosis due to the aggressive nature of HER2-positive disease [6]. Many studies have reported that HER2-targeted treatments, including anti-HER2 antibodies (such as trastuzumab and pertuzumab) and small-molecule tyrosine kinase inhibitors (such as lapatinib and neratinib), have excellent overall survival bene ts in this patient cohort [1]. Therefore, determining the HER2 expression level in tumors is vital when selecting targeted therapy. Currently, this means analysis of HER2 expression by IHC and/or HER2/neu gene ampli cation by FISH. However, despite standardization of testing methods and interpretation recommendations, assessment of HER2 status can still be inaccurate. Several factors contribute to this imprecise HER2 interpretation, such as HER2 heterogeneity, gain in chromosome enumeration probe 17 copy number, and HER2 status alteration after neoadjuvant chemotherapy or during metastatic progression [13]. As more therapeutics are approved for HER2-postive tumors, there is a growing need to accurately determine the HER2 status of both primary and metastatic tumors.
In this study, we utilized an accurate, whole-body and non-invasive method for HER2 status determination to help clinicians to identify breast cancer patients who will bene t from HER2-targeted therapy. This rst-in-human study of 99m Tc-NM-02 as a radiotracer for SPECT/CT assessment of HER2 expression was conducted in 10 female breast cancer patients. The safety of this tracer was excellent, with no reported adverse reactions and acceptable radiation dosimetry. The SPECT/CT data demonstrated favorable biodistribution and satisfactory imaging characteristics of 99m Tc-NM-02. As with other radiolabeled sdAbs [23][24][25], the urinary system received the highest radiation dose, but it was tolerable. The radiation dosimetry of 99m Tc-NM-02 was similar to that of the 99m Tc-labeled anti-PD-L1 sdAb reported in our previous study [26]. Although the individual organ doses are acceptable for a diagnostic procedure and in line with other 99m Tc-labeled radiotracers in clinical use [29], they could potentially be further minimized by faster excretion using methods such as coinjection of diuretic agents.
Some studies have attempted to decrease nonspeci c radiotracer accumulation in the liver and spleen by increasing the dose of nonradiolabeled sdAb, but the effect was very limited [24]. In our previous study, we encountered similar results [26]. In that study, we observed no signi cant differences in image quality, biodistribution, and tumor-to-background ratios in patients receiving 100 and 400 µg of sdAb.
Therefore, we injected 100 µg of NM-02 sdAb in this study. As anticipated, the 99m Tc-NM-02 tracer mainly accumulated in the liver and kidneys, and increased activity in the bladder indicated that 99m Tc-NM-02 was eliminated through the urinary system, in line with the excretion patterns described in the literature [24].
Although trace uptake in the liver, spleen, and intestines was visible early after injection, we observed rapid clearance from these organs at 1 h after injection, with no signs of hepatobiliary excretion, such as an enhanced accumulation in the gallbladder or duodenum. This might be explained by the rapid clearance of 99m Tc-NM-02 from the blood pool. The rapid blood clearance not only allowed us to perform SPECT imaging at early time points (within 2 h after injection) but also might reduce the risk of false-positive signals due to high activity in the blood pool. In addition, there was rapid tracer reduction in normal lung tissue, which further improved the contrast of primary breast cancer and metastases in images. Because of this continuous and rapid decreasing activity in the blood pool and lung background over time, images acquired at 2 h after injection had higher signal-to-noise ratios than those acquired at 1 h.
Unlike the biodistribution data in our previous work [26], mild uptake of 99m Tc-NM-02 was observed in the thyroid, submandibular, and parotid glands. This pattern has also been reported in other studies of similar tracers, such as the 68 Ga-labeled anti-HER2-nanobody and the 111 In-labeled anti-HER2-A body [24,30,31]. The tracer uptake may re ect the low-level HER2 expression in these glandular tissues, or it may be associated with chelator-mediated trapping mechanisms. However, studies performed with 18 F-anti-HER2-nanobody and 111 Inlabeled anti-HER2 monoclonal antibody showed insigni cant uptake in these glands [32][33][34]. Notably, glandular uptake has also been reported in prostate-speci c membrane antigen and broblast-activating protein inhibitor tracers [35,36]. These results reveal the complex mechanisms of glandular uptake, and its cause remains unknown.
The SPECT/CT images of primary lesions and metastases were carefully evaluated in these breast cancer patients and compared with their HER2 IHC results. As expected, ve patients with positive HER2 IHC results had obvious tracer uptake in their primary lesions. Remarkably, four of ve HER2-negative patients in this study also showed obvious local tracer uptake in their primary tumors, likely due to the intratumoral heterogeneity of HER2 expression. This discordance in uptake might be due to the tumor specimens being obtained by core needle biopsy from a small area of the tumor, which fails to capture the heterogeneity within the tumor. Additionally, intertumoral heterogeneity was observed in some patients with metastases. Distinct uptake was seen in the metastases of seven patients, although the primary tumor of one patient (Patient BR003) was classi ed as HER2-negative based on pathology results. Conversely, one patient (Patient BR001) with inconspicuous tracer uptake in metastases had high HER2 expression (HER2 3+) in the primary lesion. The inconsistency of tracer uptake in metastases could be attributable to a discordance in HER2 expression between the primary tumor and the metastases, which has been previously reported [37,38]. This intra-and intertumoral heterogeneity in HER2 expression highlights the need for the development of non-invasive imaging methods to provide more holistic information on HER2 status when making HER2-targeted therapy decisions.
Tumor uptake in the primary lesions and metastases of these breast cancer patients showed a wide range of SUV max (0. 35-11.18).
Although it was not the primary purpose of this early phase I study, based on these SUV max , we attempted to identify a cutoff for determining HER2 positivity in SPECT/CT imaging. An SUV max of 1.5 appeared to be a reasonable point. Using this cutoff, the results of visual interpretation from nuclear medicine doctors and the tumor SUV max at both 1 h and 2 h post-injection were consistent. Interestingly, Patient BR008 was originally identi ed to have a HER2-negative tumor based on imaging by visual interpretation at 1 h post-injection, but the tumor was rede ned at 2 h post-injection (SUV max of 1.36 and 1.76, respectively). This suggested not only the utility of the identi ed cutoff value but also the importance of delayed imaging for the patients with equivocal uptakes.
Several researchers have discussed the advantages of sdAbs as novel candidates in the development of molecular imaging tracers when compared with full therapeutic antibodies [16]. Based on these distinctive properties, many researchers have labeled various anti-HER2 sdAbs and antibodies with different radionuclides for PET and SPECT imaging, including 89 Zr, 64 Cu, 68 Ga and 111 In [23,24,[39][40][41].
However, the availability and production costs of these radionuclides may hinder their rapid clinical transformation and popularization. In this situation, 99m Tc-labeled sdAbs can be translated into clinical application with more ease, thanking to the ideal properties and popularity of 99m Tc radionuclide [42].
Although the relatively small sample size of this study is a potential limitation, this study provided su cient pilot feasibility data on the optimal imaging time point, radiation dosimetry analysis, biodistribution pattern, and, most importantly, safety of this novel tracer.
Moreover, recent advances in targeted radionuclide therapy for HER2-overexpressing tumors using radiolabeled sdAbs displayed impressive results [43,44]. This sdAb is planned to be studied in larger clinical trials for non-invasive detection of HER2 expression and targeted radionuclide therapy in breast cancer [45,46].

Methods
Radiopharmaceutical preparation NM-02 selection and preparation was performed as previously described [26]. 99m Tc-NM-02 was synthesized following the methodology published in the literature [26,47]. Brie y, the [ 99m Tc(OH 2 ) 3 (CO) 3 ] + complex was rst manufactured and added to a sealed vial containing 200 µg of NM-02. After incubation at 50°C for 1 h, the 99m Tc-NM-02 was prepared, and a small amount of sample was withdrawn for quality control analysis. The RCP was analyzed using instant thin-layer chromatography, and endotoxin analysis was performed using Limulus Ameboctye Lysate testing. Additional File 1 includes additional details on the preparation of the radiopharmaceutical.

Patients
This was an open-label, non-randomized early phase I ( rst-in-human) study in 10 breast cancer patients. This study was registered at ClinicalTrials.gov (NCT04040686) and approved by the Shanghai General Hospital Ethics Committee. Informed consent was obtained from all patients enrolled in this study. Details of patient screening are included in Additional File 1.

Spect/CT Imaging
Ten patients were injected with 458 ± 37 MBq 99m Tc-NM-02, corresponding to 100 µg of NM-02 nanobody, as an intravenous bolus. After injection, all patients were asked to drink more than 500 mL of water and empty their bladders before imaging to facilitate tracer clearance in normal tissue. All images were obtained using a GE Discovery NM670 SPECT/CT system (Denver, CO, USA). Routine whole-body SPECT images were acquired at 1 and 2 h post-injection for all patients. Images were also acquired at 10 min, 3 h, and 24 h after injection in three patients to calculate radiation dosimetry. At 1 h and 2 h post-injection, local SPECT/CT images of primary lesions and regional metastases were acquired. Visual interpretation of the images was performed by at least two experienced nuclear medicine doctors. Positive imaging was de ned as focal 99m Tc-NM-02 uptake greater than background corresponding to a lesion on conventional imaging. Negative imaging was de ned as no distinct pathological tracer uptake. Tumor uptake was quantitatively analyzed using Q.Metrix software (Denver, CO, USA), and the SUV max of tumor lesions were calculated at 1 and 2 h post-injection. The parameters used for SPECT/CT image acquisition and analysis are speci ed in Additional File 1.

Radiation Dosimetry Calculations
Radiation dosimetry calculations were performed as previously describe [26]. Three patients underwent additional whole-body planar imaging at 10 min, 3 h, and 24 h post-injection. A calibration source of 37 MBq was placed above the head of each patient at the time of injection to provide quantitative calibration of counts to activity. Radiation dosimetry calculations were performed by an experienced operator using OLINDA/EXM dose calculation software (version 1.1) and well-known models [48,49].

Statistics
Data are reported as mean ± standard deviation and were analyzed using SPSS software (Version 24.0). One-way analysis of variance was performed to evaluate the signi cance of the data. A P value of less than 0.05 was considered statistically signi cant.    Time activity curve of organs with highest radiotracer uptake