First In-Human Medical Imaging with a PASylated 89Zr-Labeled Anti-HER2 Fab-Fragment in a Patient with Metastatic Breast Cancer

Purpose PASylation® offers the ability to systematically tune and optimize the pharmacokinetics of protein tracers for molecular imaging. Here we report the first clinical translation of a PASylated Fab fragment (89Zr∙Df-HER2-Fab-PAS200) for the molecular imaging of tumor-related HER2 expression. Methods A patient with HER2-positive metastatic breast cancer received 37 MBq of 89Zr∙Df-HER2-Fab-PAS200 at a total mass dose of 70 μg. PET/CT was carried out 6, 24, and 45 h after injection, followed by image analysis of biodistribution, normal organ uptake, and lesion targeting. Results Images show a biodistribution typical for protein tracers, characterized by a prominent blood pool 6 h p.i., which decreased over time. Lesions were detectable as early as 24 h p.i. 89Zr∙Df-HER2-Fab-PAS200 was tolerated well. Conclusion This study demonstrates that a PASylated Fab tracer shows appropriate blood clearance to allow sensitive visualization of small tumor lesions in a clinical setting.


Introduction
Human epidermal growth factor receptor 2 (HER2) is a cell membrane receptor tyrosine kinase that plays a key role in cell development, proliferation, and differentiation [1]. Moreover, HER2 is overexpressed in a variety of cancers, including bladder; lung; gastric; ovarian; prostate; and, in particular, breast cancer (BCa) [2]. Overexpression of HER2 on tumor cells is associated with a high rate of proliferation and aggressive disease, poor prognosis, and short overall survival [3]. Notably, HER2-targeted therapies with monoclonal antibodies (mAbs) such as trastuzumab (Herceptin; Genentech, South San Francisco, CA) have significantly improved survival for up to 20% of patients suffering from BCa [4].
Antonia Richter and Karina Knorr contributed equally to this work.
* Antonia Richter antonia.richter@tum.de * Arne Skerra skerra@tum.de Precise determination of HER2 expression is the basis for success of HER2-targeted therapy. Routinely, HER2 status in BCa is determined on tissue biopsies either via immunohistochemistry (IHC) or by fluorescence in situ hybridization (FISH). However, both methods may be inaccurate in up to 20% of cases [5]. Also, due to the small size and limited number of tissue samples, tumor heterogeneity poses a challenge [2].
Molecular imaging using specific radiopharmaceuticals that target HER2 could offer a non-invasive option for better quantification and localization of HER2 overexpression and, thus, identify patients who may benefit from HER2-directed therapy. Several imaging tracers targeting HER2 have been reported, including radiolabeled mAbs, antibody fragments (Fab or F(ab) 2 ), nanobodies, or affibodies [6]. One of the most well studied radiotracers for HER2 imaging is 89 Zr-labeled trastuzumab, which allowed visualization and quantification of uptake in HER2-positive lesions for patients with metastatic BCa and other cancers in several clinical trials [7][8][9][10]. However, due to the large molecular size (150 kDa) and the intrinsically slow blood clearance of the full-length antibody, optimal detection of lesions is seen only 4-5 days after injection and accompanied by a considerable dose exposure to healthy organs, which is several-fold higher than for PET scans with 18 F-fluorodeoxyglucose [11]. Furthermore, uptake of 89 Zr-labeled trastuzumab has been found to be false-positive, i.e., to occur in the absence of clinically relevant HER2 expression, in a significant fraction of patients [8].
Generally, Fab fragments (48 kDa) offer rapid clearance and thereby better tumor contrast at early imaging time points. However, the fast elimination from the blood stream can also limit tumor uptake of such antibody fragments. In a preclinical study, we systematically examined the impact of plasma half-life of modified Fab fragments in a HER2-positive breast cancer model [12]. Tailoring of plasma half-life was conveniently achieved using the PASylation technology [12][13][14]. To this end, conformationally disordered 100-600 residue chains consisting of Pro, Ala, and Ser (PAS 100-600 ) were genetically fused to the C-terminus of the light chain of the trastuzumab Fab ( Fig. 1) and compared with the unmodified Fab in positron emission tomography (PET) and biodistribution experiments. In this preceding study, the radiotracer 89 Zr•Df-HER2-Fab-PAS 200 revealed an optimal PET imaging contrast 24 h post injection (p.i.) in mice and, thus, appeared promising for clinical translation [15]. Here we report the first in-human PET imaging with 89 Zr•Df-HER2-Fab-PAS 200 of a HER2-positive patient suffering from metastatic BCa.

Production of HER2-Fab-PAS 200
The HER2-Fab-PAS 200 [12] was produced by bench top fermentation in E. coli as previously described [13]. Bacterial harvest and preparation of the periplasmic extract were performed under sterile conditions by crossflow filtration in a closed system, followed by chromatographic purification to homogeneity. Conjugation of p-SCN-Bn-Deferoxamine (Df; Macrocyclics, Plano, TX) was carried out according to a published protocol [15] following good manufacturing practice guidelines. On average, 2-3 Df chelates were coupled per Fab molecule as assessed by ESI-TOF mass spectrometry.

Zr-Labeling, Formulation, and Quality Control
Radiolabeling of Df-HER2-Fab-PAS 200 was performed according to a published procedure [15] using 89 Zr as supplied by Perkin Elmer (Boston, MA). Briefly, 93 MBq of 89 Zr in oxalic acid were neutralized with Na 2 CO 3 and incubated with 260 μg of the purified Df-HER2-Fab-PAS 200 in HEPES/ NaOH buffer (pH 7.0) for 60 min at room temperature, followed by gel filtration on a PD-10 column (GE Healthcare, Munich, Germany). Radiolabeling efficiency was 92.4%, and the radiochemical purity was > 95%, as determined by instant thin-layer chromatography (TLC). Two milliliters of the isolated 89 Zr•Df-Her2-Fab-PAS 200 was diluted with 9 ml sterile 0.9% saline and sterilized by filtration through a 0.2-μm Millex LG syringe filter (Merck Millipore, Darmstadt, Germany) under aseptic conditions (with only negligible amounts of radioactivity accumulating in the filter). The amount of protein was quantified by Bradford assay (Bio-Rad Laboratories, CA) using a dilution series of the unlabeled Df-HER2-Fab-PAS 200 preparation as reference. As a further quality control, a radio-HPLC of a sample was performed, which revealed a single peak at the expected retention time. The final product (9.6 μg/ml) was documented to be sterile and free of particles at pH 7.0, and the bacterial endotoxin content was < 0.5 EU/ml.
For the toxicity study, Df-HER2-Fab-PAS 200 was charged with non-radioactive zirconium ( nat Zr) using the same protocol as for the radioisotope. The product was analyzed by ESI-TOF mass spectrometry, revealing successful complexation of 1-3 nat Zr ions per protein molecule.

Single-Dose Toxicity Study
To obtain information on the general toxicity of the PASylated Fab fragment, we performed a single-dose toxicity study in female CD1-Foxn1 nu mice (7 months age, average weight 38.9 ± 5 g). Based on our preclinical findings [12], a maximum dose of 100 μg injected protein ("microdose") was assessed as a starting point for the first clinical application of 89 Zr•Df-HER2-Fab-PAS 200 , corresponding to 1.4 μg/kg body weight for a 70-kg patient. Application of the same total protein amount to these mice was equal to a > 1000-fold dose, in line with the "ICH guideline M3(R2) on non-clinical safety studies for the conduct of human clinical trials and marketing authorization for pharmaceuticals." Therefore, two groups of mice (n = 11) were injected once intravenously with 100 μg of Df-HER2-Fab-PAS 200 charged with nat Zr. The first group was sacrificed 24 h, and the second group was sacrificed 14 days after treatment with nat Zr•Df-Her2-Fab-PAS 200 . Six mice (three per group) treated with saline served as reference. All tissues and organs were examined histologically by two veterinary pathologists, and findings were reported according to the INHAND criteria of the Society of Toxicologic Pathology (STP) in line with the most recent recommendations. Hematology, clinical chemistry, and urinalysis as well as analyses of organs and blood samples were performed as described elsewhere [16]. The animal experiments were approved by local authorities (General Administration of Upper Bavaria; license 55.2-1-54-2532-46-12) and in compliance with regulatory and institutional guidelines.

89
Zr•Df-HER2-Fab-PAS 200 imaging was offered to support individual therapy planning and to identify the primary t u m o r u n d e r t h e G e r m a n P h a r m a c e u t i c a l s A c t ("Arzneimittelgesetz", AMG), Sect. 13.2b, and with notification of the General Administration of Upper Bavaria. The patient was a 67-year-old woman with newly diagnosed HER2-positive metastatic BCa. Metastatic BCa had been proven by biopsy of an enlarged axillary lymph node, but no definitive abnormal findings were seen in both breasts on mammography. On immunohistochemistry, the tumor tissue in the axillary lymph node was positive for HER2 (score 3 +). An MRI scan of the brain showed multiple enhancing lesions, consistent with metastases (Fig. 2). Thus, the tumor stage was cTx pN1 cM1. The patient was treated with whole brain radiation therapy in combination with dexamethasone (4 mg per day) prior to the 89 Zr•Df-HER2-Fab-PAS 200 PET/CT scans.

Zr•Df-HER2-Fab-PAS 200 Administration and PET Imaging
The radiolabeled Fab fragment (37 MBq, 70 μg) was administered intravenously over 1-2 min without premedication. The patient was monitored for at least 1 h after injection for any reactions or adverse events. During this period, the documentation of blood pressure and pulse did not show any relevant change.

Image Acquisition and Interpretation
Image acquisition was performed using a Biograph 128 PET/ CT scanner (Siemens Molecular Imaging, Knoxville, TN). Whole-body PET/CT was conducted 6 and 24 h p.i. with 13 bed positions (5 min upper body, 2.5 min lower extremities). Images of the thorax and epigastrium were acquired 45 h p.i. with 2 bed positions. For quantitative evaluation, mean standardized uptake values (SUV) were calculated from attenuation corrected image data using the average intensity of the respective tissue as listed in Table 1.

Single-Dose Toxicity Study
Non-toxicity of Df-HER2-Fab-PAS 200 charged with nat Zr was substantiated in female mice prior to human application at > 1000-fold the clinical dose (on a mg/kg basis). No relevant histopathological findings for the main organs were detected after 24 h up to 14 days and nat Zr•Df-HER2-Fab-PAS 200 was well tolerated.

Biodistribution and Normal Organ Uptake
The mBCa patient underwent whole-body PET/CT imaging 6 and 24 h post injection (p.i.) with 89 Zr•Df-HER2-Fab-PAS 200 (70 μg, 37 MBq). Additionally, PET images of the upper body were acquired 45 h p.i. Images showed a biodistribution typical for protein tracers, characterized by a prominent blood pool 6 h p.i., which decreased over time ( Fig. 2 and Table 1). Increased uptake in the liver and kidneys was observed 24 h p.i., which-in case of the liver-decreased thereafter. Activity in the gastrointestinal tract was also prominent 24 h and 45 h p.i., suggesting hepatobiliary clearance. No significant activity was seen in the bladder, and no significant urinary excretion was noted. The 89 Zr•Df-HER2-Fab-PAS 200 uptake in other non-tumor tissues (i.e., lung, muscle, bone) was low.

Lesion Targeting
Lesions were detectable as early as 24 h p.i. of 89 Zr•Df-HER2-Fab-PAS 200 (Fig. 2). The cranial lymph node metastasis in the left axillary, earlier diagnosed by CT scan, was clearly visualized by PET imaging with this novel tracer (maximum standardized uptake value (SUV max ) = 5.4). Notably, a single lesion was also detectable in an area of dense parenchyma in the left breast, with an SUV max of 4.2 at 24 h p.i., which was indicative of the putative primary tumor that had remained elusive at this time.
Of note, no accumulation of the protein tracer was visible in the known brain metastases. Furthermore, the second axillar lymph node revealed a prominent area of central necrosis with no uptake of 89 Zr•Df-HER2-Fab-PAS 200 .

Discussion
Molecular imaging should help to guide physicians to tailor individual treatment of patients and to monitor the therapeutic response. To this end, biodistribution and lesion targeting of 89 Zr•Df-HER2-Fab-PAS 200 was evaluated in a patient with HER2-positive mBCa. Biodistribution of 89 Zr•Df-HER2-Fab-PAS 200 was in accordance with expectation from preclinical studies in mice [12,15]. Due to the genetic fusion of the Fab with the conformationally disordered PAS 200 polypeptide and the resulting increased hydrodynamic volume, with an enlarged apparent MW ≈ 165 kDa, 89 Zr•Df-HER2-Fab-PAS 200 showed moderately prolonged blood circulation compared with the unmodified Fab [12]. Similarly, a delayed blood clearance of 89 Zr•Df-HER2-Fab-PAS 200 was observed in the patient ( Fig. 2; cf. time points 6 and 24 h p.i.), confirming that PASylation is also effective in humans.
In the preclinical evaluation of 89 Zr•Df-HER2-Fab-PAS 200 , high tumor-to-background ratios were observed 24 h p.i. (tumor-to-blood, 3.4; tumor-to-muscle, 20) with even higher values at 48 h p.i. [15]. In the patient, blood pool activity was prominent 6 h p.i. and still seen at 24 and 45 h p.i. On the other hand, signals for the lesions became more evident 24 h p.i. compared with 6 h p.i., indicating ongoing accumulation in the tumor tissue. This is in agreement with the finding from our preceding preclinical study that a moderately prolonged plasma halflife can improve tumor uptake of a Fab-size protein tracer [12]. Furthermore, our present observations indicate a longer plasma half-life of the PASylated Fab in humans than in mice. Indeed, slower clearance rates of therapeutic antibodies in humans compared with mice have been reported, in line with the rules of allometric scaling [17]. An 111 In-labeled Fab fragment of trastuzumab was previously investigated in a phase I trial of intraoperative detection of tumor margins in patients with HER2-positive carcinoma, revealing a terminal plasma half-life of 20.7 h [18]. However, it was not feasible to reliably detect the margins of disease in those patients due to the low uptake of the 111 In•DTPA-trastuzumab Fab in the tumor.
In contrast to the signal for the presumed primary tumor and the metastasis in the axillary lymph node seen with 89 Zr•Df-HER2-Fab-PAS 200 , no accumulation of radioactivity was visible in the diagnosed brain metastases. Of note, even large molecules (e.g., T-DM1, 89 Zr-trastuzumab) were reported to penetrate HER2positive breast cancer brain metastases [7,19] due to local disruption of the blood-brain barrier at these sites [20]. Possibly, in the present case, the blood-brain barrier was stabilized by the pretreatment with dexamethasone during whole brain irradiation prior to the application of 89 Zr•Df-HER2-Fab-PAS 200 [21].

Conclusions
Based on these first results, 89 Zr•Df-HER2-Fab-PAS 200 has the potential to serve as a novel imaging agent to support individual therapy planning of HER2-positive BCa. PET imaging with 89 Zr•Df-HER2-Fab-PAS 200 was feasible and tolerated well. This study indicates that PASylation technology is effective in a human patient and leads to delayed blood clearance of a radiolabeled anti-HER2 Fab fragment and sensitive tumor accumulation. However, the pharmacokinetics of 89 Zr•Df-HER2-Fab-PAS 200 was longer than expected and may be further optimized, e.g., by use of a shorter PAS polypeptide or, possibly, by co-injection of trastuzumab in order to scavenge circulating shedded HER2 ectodomain and to effect more rapid clearance of immune complexes [22].
Acknowledgments The authors wish to thank XL-protein GmbH for providing the PAS gene cassette and Uli Binder for critically reading the manuscript. Excellent technical assistance by Klaus Wachinger and Olga Seelbach, both at TUM, is gratefully acknowledged. We also thank Gabriel Buschner (MRI) for his helpful support.
Funding Information Open Access funding provided by Projekt DEAL. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.