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

The HER2-Fab-PAS₂₀₀ [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.

Radiolabeling of Df-HER2-Fab-PAS₂₀₀ was performed according to a published procedure [15] using ⁸⁹Zr as supplied by Perkin Elmer (Boston, MA). Briefly, 93 MBq of ⁸⁹Zr in oxalic acid were neutralized with Na₂CO₃ and incubated with 260 μg of the purified Df-HER2-Fab-PAS₂₀₀ 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 ⁸⁹Zr∙Df-Her2-Fab-PAS₂₀₀ 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-PAS₂₀₀ 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-PAS₂₀₀ was charged with non-radioactive zirconium (^natZr) using the same protocol as for the radioisotope. The product was analyzed by ESI-TOF mass spectrometry, revealing successful complexation of 1–3 ^natZr ions per protein molecule.

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 ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀, 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-PAS₂₀₀ charged with ^natZr. The first group was sacrificed 24 h, and the second group was sacrificed 14 days after treatment with ^natZr∙Df-Her2-Fab-PAS₂₀₀. 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.

⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀ imaging was offered to support individual therapy planning and to identify the primary tumor under the German Pharmaceuticals Act (“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 ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀ PET/CT scans.

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 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.

Non-toxicity of Df-HER2-Fab-PAS₂₀₀ charged with ^natZr 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 ^natZr∙Df-HER2-Fab-PAS₂₀₀ was well tolerated.

The mBCa patient underwent whole-body PET/CT imaging 6 and 24 h post injection (p.i.) with ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀ (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 ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀ uptake in other non-tumor tissues (i.e., lung, muscle, bone) was low.

Lesions were detectable as early as 24 h p.i. of ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀ (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 ⁸⁹Zr∙Df-HER2-Fab-PAS₂₀₀.