Formulation of a kit under Good Manufacturing Practices (GMP) for preparing [111In]In-BnDTPA-trastuzumab-NLS injection: a theranostic agent for imaging and Meitner-Auger Electron (MAE) radioimmunotherapy of HER2-positive breast cancer

Background 111In[In]-BnDTPA-trastuzumab-NLS is a radiopharmaceutical with theranostic applications for imaging and Meitner-Auger electron (MAE) radioimmunotherapy (RIT) of HER2-positive breast cancer (BC). Nuclear localization sequence (NLS) peptides route the radiopharmaceutical to the nucleus of HER2-positive BC cells following receptor-mediated internalization for RIT with subcellular range MAEs. The γ-photons emitted by 111In permit tumour imaging by SPECT. Our aim was to formulate a kit under Good Manufacturing Practices conditions to prepare 111In[In]-BnDTPA-trastuzumab-NLS injection for a first-in-human clinical trial. Results Trastuzumab was derivatized with p-SCN-BnDTPA to introduce Bn-DTPA for complexing 111In, then modified with maleimide groups for conjugation to the thiol on cysteine in NLS peptides [CGYGPKKKRKVGG]. BnDTPA-trastuzumab-NLS (5 mg in 1.0 mL of 0.05 M ammonium acetate buffer, pH 5.5) was dispensed into unit dose sterile glass vials to produce kits for labeling with 100–165 MBq of 111In[In]Cl3. The kits met specifications for protein concentration (4.5–5.5 mg/mL), volume (0.95–1.05 mL), pH (5.5–6.0), appearance (clear, pale-yellow, particulate-free), BnDTPA substitution level (2.0–7.0 BnDTPA/trastuzumab), purity and homogeneity (SDS-PAGE and SE-HPLC), 111In labeling efficiency (> 90%), binding to HER2-positive SK-BR-3 human breast cancer cells (Ka = 1–8 × 108 L/mmol; Bmax = 0.5–2 × 106 sites/cell), NLS peptide conjugation (upward band shift on SDS-PAGE), sterility (USP Sterility Test) and endotoxins (USP Bacterial Endotoxins Test). 111In-BnDTPA-trastuzumab-NLS injection met specifications for pH (5.5–6.5), radiochemical purity (≥ 90%), radionuclide purity (≥ 99%), appearance (clear, colourless, particle-free) and sterility (retrospective USP Sterility Test). Kits were stable stored at 2–8 °C for up to 661 days (d) meeting all key specifications. Protein concentration remained within or just slightly greater than the specification for up to 139 d. 111In[In]-BnDTPA-trastuzumab-NLS injection was stable for up to 24 h. An expiry of 180 d was assigned for the kits and 8 h for the final radiopharmaceutical. Conclusion A kit was formulated under GMP conditions for preparing 111In[In]-BnDTPA-trastuzumab-NLS injection. This radiopharmaceutical was safely administered to 4 patients with HER2-positive BC to trace the uptake of trastuzumab into brain metastases before and after MRI-guided focused ultrasound (MRIg-FUS) by SPECT imaging.

Radioimmunotherapeutic agents (RIT) are analogous to ADCs and link monoclonal antibodies (mAbs) to radionuclides that emit β-particles (e.g. 131 I, 177 Lu, 90 Y), α-particles (e.g. 225 Ac, 213 Bi, 211 At) or Meitner-Auger electrons (MAEs; e.g. 111 In, 125 I, 67 Ga) for radiation treatment of tumours (Aghevlian et al. 2017;Ku et al. 2019). Trastuzumab labeled with the β-particle emitter, 177 Lu was cytotoxic in vitro to HER2-positive human SK-BR-3 and MDA-MB-453 human BC cells and SK-OV-3 ovarian cancer cells (Sharma et al. 2020). Treatment of athymic mice with intraperitoneal (i.p.) HER2positive LS-174 T human colon cancer xenografts with i.p. administered 177 Lu-labeled trastuzumab prolonged survival more than 12-fold compared to untreated mice (Ray et al. 2011). 177 Lu-labeled trastuzumab administered to patients with HER2-positive BC localized in primary and metastatic tumors and these were detected by single photon emission computed tomography/computed tomography (SPECT/CT) illustrating its potential for imaging and RIT of HER2-positive BC (Bhusari et al. 2017). Our group reported that a bispecific RIT agent composed of trastuzumab Fab linked through a polyethylene glycol (PEG 24 ) spacer to epidermal growth factor (EGF) labeled with 177 Lu was cytotoxic in vitro to human BC cells expressing HER2 or epidermal growth factor receptors (EGFR) or both receptors (Razumienko et al. 2016). This agent was effective for RIT of HER2 and EGFR-positive MDA-MB-231/H2N tumours in athymic mice but hematopoietic toxicity limited the dose that could be safely administered. RIT of solid tumours in humans with β-particle emitters has similarly proven dose-limited by off-target hematopoietic toxicity (Larson et al. 2015) due to the long (several millimeters) path length of β-particles (cross-fire effect) (Richman et al. 2005). In addition, β-particles have very low linear energy transfer (LET = 0.1-1 keV/μm) that makes them theoretically less potent than shorter range (28-100 μm) and higher LET α-particles (LET = 50-230 keV/μm) or nanometer-micrometer range MAEs (LET = 4-26 keV/μm) for killing cancer cells (Aghevlian et al. 2017).
Trastuzumab labeled with α-particle emitting, 225 Ac administered by intraperitoneal (i.p.) injection was effective for RIT of HER2-positive SK-OV-3 human ovarian cancer tumours in athymic mice (Borchardt et al. 2003). 225 Ac-labeled trastuzumab was also effective for treatment of SUM225 ductal carcinoma in situ of the breast tumours in NCG mice after intraductal injection (Yoshida et al. 2016). However, intravenous (i.v.) injection of 225 Ac-labeled trastuzumab in these studies caused hematopoietic system toxicity, possibly due to irradiation of bone marrow stem cells by the 28-100 μm range α-particles emitted by circulating 225 Ac-labeled trastuzumab perfusing the bone marrow (cross-fire effect). In addition, the 213 Bi decay product of 225 Ac poses a risk for renal toxicity (Yoshida et al. 2016).
MAE-emitting radionuclides are an alternative to α-particle emitters for RIT since they similarly exhibit high LET, but their cytotoxic effects are restricted to cells that bind and internalize the radioimmunoconjugates (RICs) due to the subcellular range of the electrons. There is no cross-fire effect from MAE-emitting radionuclides, which greatly reduces off-target toxicity (e.g. hematopoietic toxicity) compared to longer range β-emitters such as 177 Lu (Ku et al. 2019;Aghevlian et al. 2017). Moreover, MAE-emitters decay to stable elements which avoids toxicity from radioactive daughter products, a challenge for α-particle emitters such as 225 Ac. MAEs kill cancer cells by causing oxidative damage to the cell membrane (Paillas et al. 2016) or by inflicting lethal DNA doublestrand breaks (Ku et al. 2019). DNA damage from MAEs is the most potent and may be maximized by conjugation of the mAbs to peptides that incorporate nuclear translocation sequences (NLS) or by targeting a receptor that harbours an endogenous NLS. NLS are short sequences of 4 or more cationic amino acids [arginine (R) or lysine (K)] that bind to importin-α which forms a complex with importin-β and mediates active transport of proteins across the nuclear pore complex (NPC) (Costantini et al. 2008b). Only proteins with molecular weight (MW) < 40-45 kDa are able to diffuse across the NPC due to the size of the pores (25-30 nm). A classical NLS is the Simian Virus-40 (SV-40) large T-antigen NLS [ 126 PKKKRKV 132 ] but many peptide growth factors and their receptors harbour endogenous NLS (Costantini et al. 2008b).
Based on these encouraging preclinical results, our aim in the current study was to formulate a kit for preparing 111 In[In]-BnDTPA-trastuzumab-NLS injection under Good Manufacturing Practices (GMP) conditions to enable advancement to a first-in-human clinical trial. The p-isothiocyanate ester of DTPA (p-SCN-BnDTPA) was used rather than DTPA dianhydride employed prevously to modify trastuzumab with DTPA (Costantini et al. 2007), since this bifunctional chelator provides more stable complexes with 111 In and minimizes cross-linking of trastuzumab molecules, which may occur due to the two reactive groups present in DTPA dianhydride (Brechbiel et al. 1986). In addition to the MAE emissions, 111 In emits γ-photons [Eγ = 171 keV (90.7%) and 245 keV (94.1%)] that permit imaging by single photon emission computed tomography (SPECT). Thus, for this first-in-human clinical trial, we formulated a kit for labeling 5 mg of DTPA-trastuzumab-NLS with 111-165 MBq of 111 In to assess the tumour and normal tissue uptake of this radiopharmaceutical in humans by SPECT. These mass and activity amounts were selected based on those previously reported for SPECT imaging of HER2-positive BC in humans using 111 In[In]DTPA-trastuzumab (100-150 MBq; 5 mg) (Perik et al. 2006). 111 In[In]-BnDTPA-trastuzumab-NLS injection prepared from the kit described here was recently administered safely for the first time to 4 patients with HER2-positive BC to trace the uptake of trastuzumab into brain metastases, prior to and after application of MRI guided focused ultrasound (MRIg-FUS) (Meng et al. 2021). MRIg-FUS is an interventional technique that transiently and spatially disrupts the blood-brain-barrier (BBB), facilitating penetration of trastuzumab into metastatic lesions in the brain (Park et al. 2012). There were no adverse reactions associated with administration of the radiopharmaceutical. This study revealed that MRIg-FUS significantly improved the delivery of trastuzumab into brain metastases by as much as 4.5-fold. These results further suggest that 111 In[In]-BnDTPA-trastuzumab may be a promising MAE-emitting RIT agent for treating brain metastases in patients with HER2-positive BC by employing MRIg-FUS-enhanced delivery into the brain. Thus, 111 In-BnDTPA-trastuzumab-NLS is a theranostic agent with promising application for imaging and MAE RIT of HER2-positive BC.

Trastuzumab, NLS peptides and p-SCN-BnDTPA
Trastuzumab (Herceptin ® ; Hoffman La Roche, Mississauga, ON, Canada) was purchased from the Princess Margaret Cancer Centre (Toronto, ON, Canada) hospital pharmacy and reconstituted with the supplied Bacteriostatic Water for Injection, USP to 21 mg/mL following the manufacturer's directions. Trastuzumab identity and purity were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), Western blot and size-exclusion high performance liquid chromatography (SE-HPLC). We previously used these analytical methods to study Page 5 of 21 Chan et al. EJNMMI Radiopharmacy and Chemistry (2022)

Pharmaceutical quality buffers
Sterile 0.1 M NaHCO 3 buffer, pH 8.2 in Sodium Chloride for Injection, USP and 0.05 M NH 4 CO 2 CH 3 buffer, pH 5.5 and 0.1 M Na 2 HPO 4 buffer in Sterile Water for Injection USP, pH 7.3 were prepared as previously reported (Lam et al. 2015). Trace metals were removed from all buffers by passing through a 10 mL column filled with Chelex-100 cation exchange resin (BioRad) followed by re-adjustment of the pH with sterile 1 N HCl, 1 N NaOH or glacial acetic acid USP. Buffers were sterilized by filtration through a 0.22 μm Millex GV filter (Sigma-Aldrich) into 30 mL glass vials and stored at 2-8 °C.
All buffers were tested for sterility by the USP Sterility Test. The concentration of NaHCO 3 was assayed by titration with 0.1 N sulfuric acid according to the USP method. The concentration of Na 2 HPO 4 was determined by a colorimetric assay using ammonium molybdate and stannous chloride (Truong and Meyer 1929). The concentration of NH 4 CO 2 CH 3 was not assayed due to unavailability of a method but instead was calculated by dividing the weight of NH 4 CO 2 CH 3 used by the volume of the solution. Clarity and colour of the buffers were assessed against a light and dark background.

Kit formulation
Kits for the preparation of 111 In[In]-BnDTPA-trastuzumab-NLS injection were formulated under GMP conditions in a Class II Type A2 biosafety cabinet (Model 425-400; NuAire, Plymouth, MN, USA). Trastuzumab (Herceptin, Roche; 3.6 mL; 75.6 mg) reconstituted in Bacteriostatic Water for Injection, USP was buffer exchanged into 0.1 M NaHCO 3 buffer, pH 8.2 using a 15 mL Amicon Ultra-15 centrifugal filter device [molecular weight cut-off (MWCO) = 30 kDa] (MilliporeSigma, Burlington, MA, USA). Briefly, trastuzumab was dispensed into the device and diluted to 12.0 mL with 0.1 M NaHCO 3 buffer, pH 8.2. The device was centrifuged for 10 min at 5,000 × g at 18 °C, until ~ 2 mL of solution was retained. This solution was re-diluted to 12 mL with 0.1 M NaHCO 3 buffer, pH 8.2 and centrifuged again. This was repeated a total of 4 times. The retentate was recovered and a sample of 2 μL was diluted 1:40 to determine the concentration of IgG by measuring the absorbance at 280 nm using the extinction coefficient 1.5 mL mg −1 cm −1 for IgG. The trastuzumab concentration was adjusted to 16 mg/mL with 0.1 M NaHCO 3 buffer, pH 8.2 and a volume of ~ 5.0 mL. Trastuzumab was then reacted with a 15-fold molar excess of p-SCN-BnDTPA (10 mg/mL freshly prepared in 0.1 M NaHCO 3 buffer, pH 8.2) in a sterilized 10 mL glass Reacti-Vial ® (ThermoFisher Scientific, Waltham, MA) at RT for 1 h. Duplicate samples (12 μL) of the conjugation reaction were removed for subsequent determination of conjugation efficiency (CE) in order to estimate the number of BnDTPA conjugated to trastuzumab (see Quality Control Testing of Kits). The reaction mixture was transferred to an Amicon Ultra-15 centrifugal filter device (MWCO = 30 kDa) and re-diluted to 12 mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3. The device was centrifuged at 5,000 × g at 18 °C for 10 min until ~ 2 mL was retained. This solution was re-diluted to 12 mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3 and centrifuged again. This was repeated a total of 16 times to completely remove unconjugated p-SCN-BnDTPA. Purified BnDTPA-trastuzumab was recovered in ~ 8 mL of 0.1 M Na 2 HPO 4 buffer, pH 7.3 and transferred to a pre-weighed 15 mL sterile polypropylene tube (Sarstedt, Nümbrecht, Germany). The tube was re-weighed to calculate the volume of recovered solution by difference, assuming 1 g = 1 mL. A sample (2 μL) Page 7 of 21 Chan et al. EJNMMI Radiopharmacy and Chemistry (2022) 7:33 was removed and diluted 40-fold with 0.1 M Na 2 HPO 4 buffer, pH 7.3 and the absorbance was measured at 280 nm to calculate the concentration of BnDTPA-trastuzumab. Finally, BnDTPA-trastuzumab was diluted to 6.0 mg/mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3. BnDTPA-trastuzumab was then reacted with a fivefold molar excess of Sulfo-SMCC (10 mg/mL freshly dissolved in 0.1 M Na 2 HPO 4 buffer, pH 7.3) in a 10 mL sterilized glass Reacti-vial at RT for 1 h to introduce maleimide functional groups for reaction with the thiol on cysteine in NLS peptides. Excess unconjugated Sulfo-SMCC was removed by transferring the reaction mixture to an Amicon Ultra-15 centrifugal filter device (MWCO = 30 kDa), diluting to 12 mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3 and centrifuging the device at 5000 × g at 18 °C for 10 min. This retained solution was diluted again to 12 mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3 and the device centrifuged again. This was repeated a total of 10 times. Purified maleimide-functionalized BnDTPA-trastuzumab solution was recovered into a pre-weighed 15 mL sterile polypropylene tube and the volume of the solution was determined by difference, assuming 1 g = 1 mL. A sample (2 μL) was diluted 40-fold with 0.1 M Na 2 HPO 4 buffer, pH 7.3 and the absorbance measured at 280 nm to determine the concentration of maleimide-functionalized BnDTPA-trastuzumab. The concentration was adjusted to 5.0 mg/mL with 0.1 M Na 2 HPO 4 buffer, pH 7.3.
Maleimide-functionalized BnDTPA-trastuzumab was reacted with a 60-fold molar excess of NLS peptides (20 mg/mL in 0.1 M Na 2 HPO 4 buffer, pH 7.3) in a 10 mL sterilized glass Reacti-vial at 4 °C overnight. Unconjugated NLS peptides were removed by transferring the reaction mixture to an Amicon Ultra-15 centrifugal filter device and diluting with 0.05 M NH 4 CO 2 CH 3 buffer, pH 5.5. The solution was centrifuged at 5000 × g for 10 min, the filtrate discarded and the retained solution diluted again with 0.05 M NH 4 CO 2 CH 3 buffer, pH 5.5. This was repeated a total of 15 times. The concentration of recovered purified BnDTPA-trastuzumab-NLS was determined by measuring the absorbance at 280 nm and the solution was diluted to a final concentration of 5.0 mg/ mL with 0.05 M NH 4 CO 2 CH 3 buffer, pH 5.5. Tween 20 ® surfactant (Sigma-Aldrich) was added into the BnDTPA-trastuzumab-NLS solution to a final concentration of 0.1% to prevent protein aggregation (Strickley and Lambert 2021). The BnDTPA-trastuzumab-NLS solution was drawn up in a 5 mL sterile syringe with an 18G × 1½" needle (Becton-Dickenson, Franklin Lakes, NJ, USA) and sterilized by filtration through a 0.22 μm Millex GV low protein-binding filter (Sigma-Aldrich) into a 30 mL sterile glass vial with grey butyl rubber septum and aluminum seal. The integrity of the sterilizing filter was checked by the bubble test. Then 1.0 mL aliquots (5.0 mg of BnDTPA-trastuzumab-NLS) were drawn up using a 1 cc U-1000 insulin syringe with 28G × ½" gauge needle (Becton-Dickenson) and aseptically dispensed into sterile 5-mL glass vials to produce unit-dose kits. Kits were stored in the refrigerator at 2-8 °C.

Quality control testing of kits
Kits were tested against specifications for protein concentration (4.5-5.5 mg/mL), volume (0.95-1.05 mL), pH (5.5-6.0), appearance (clear, pale-yellow, particulate-free), BnDTPA substitution level (2.0-7.0 BnDTPA/trastuzumab), purity and homogeneity (SDS-PAGE and SE-HPLC), 111 In labeling efficiency (> 90%), binding to HER2-positive Page 8 of 21 Chan et al. EJNMMI Radiopharmacy and Chemistry (2022) 7:33 SK-BR-3 human breast cancer cells (K a = 1-8 × 10 8 L/mmol; B max = 0.5-2 × 10 6 sites/ cell), NLS peptide conjugation (upward shift in the protein band on SDS-PAGE compared to BnDTPA-trastuzumab), sterility (USP Sterility Test) and endotoxins (USP Bacterial Endotoxins Test). Protein concentration was determined by measuring the absorbance at 280 nm. The volume in each kit vial was measured by the difference in weight of the vial before and after dispensing 1.0 mL of BnDTPA-trastuzumab-NLS solution into the vial, assuming 1 g = 1 mL. The pH was measured by spotting a sample on pH 4.5-7.5 range pH paper (pHydrion ® , Micro Essential Laboratory, Brooklyn, NY, USA). Appearance was inspected by examining the colour, clarity and presence of any particles against a light or dark background. SDS-PAGE was performed by electrophoresing ~ 2 μg on a 4-20% Tris-HCl gradient mini-gel (BioRad) under reducing (DTT) and non-reducing conditions. Protein bands were stained with Coomassie R-250 Brilliant Blue. The gel was calibrated by electrophoresing broad range MW standards. The stained SDS-PAGE gel was imaged on a BioRad ChemiDoc Imaging System (Mississauga, ON, Canada) and the relative density of each band determined using BioRad Image Lab (ver 6.0) software. The number of NLS peptides conjugated to trastuzumab was estimated by comparing the approximate MW of electrophoresed BnDTPA-trastuzumab-NLS versus BnDTPAtrastuzumab, assuming that each NLS peptide has a MW = 1,419 Da. SE-HPLC was performed on a BioSep SEC-s4000 column (Phenomenex) eluted with 100 mM NaH 2 PO 4 buffer, pH 7.0 at a flow rate of 0.8 mL/min with UV detection at 280 nm (Agilent Technologies). To avoid sticking of the immunoconjugates to the matrix of the column, the cationic charges on NLS peptides were masked prior to analysis by reaction with a 50-fold molar excess of citraconic anhydride (Sigma-Aldrich, 1.25 g/ mL in H 2 O) for 2 h at RT. Excess citraconic acid was removed by ultrafiltration on an Amicon Ultra-0.5 mL centrifugal filter device, repeated 8 times and a 20 μL sample (  -BnDTPA-trastuzumab-NLS to HER2-positive SK-BR-3 cells was determined by a direct (saturation) binding assay as previously reported (Lam et al. 2015). Briefly, total binding (TB) was determined by incubating increasing concentrations [0.07-300 nmoles/L) of 111 In[In]-BnDTPA-trastuzumab-NLS in phosphate-buffered saline (PBS) with 1 × 10 6 SK-BR-3 cells in 1.5 mL Eppendorf tubes at 4 °C for 3.5 h with gentle shaking. The tubes were centrifuged at 2,000 rpm for 5 min to recover the cell-bound (pellet) and free (supernatant) radioactivity which was then measured in a γ-counter (Model Wallac Wizard 1480, PerkinElmer, Waltham, MA, USA). The assay was repeated in the presence of an excess (15.2 μmoles/L) of unlabeled trastuzumab to determine non-specific binding (NSB). Specific binding (SB) was calculated by subtraction of NSB from TB and plotted vs. the concentration of free 111 In[In]-BnDTPA-trastuzumab-NLS (nmoles/L). The curve was fitted to a one-site receptor-binding model by Prism Ver. 4.0 software (GraphPad, San Diego, CA, USA) and the affinity constant (K a ) and maximum number of binding sites per cell (B max ) were calculated.
Kits were tested by the USP Sterility Test at the Microbiology Laboratory at Mount Sinai Hospital (Toronto, ON, Canada). Kits were tested for endotoxins by the USP Bacterial Endotoxins Test using the QCL-1000 Endpoint Chromogenic LAL Assay (Lonza, Walkersville, MD, USA). The stability of kits stored at 2-8 °C was assessed at selected intervals up to 661 d to establish an expiry by re-testing against specifications for protein concentration, pH, appearance, purity and homogeneity, labeling efficiency with 111 In and HER2-binding properties. Sterility and endotoxins were not re-tested as it was expected that these would be maintained in storage since the kit solution was maintained sterile and endotoxin-free by packaging in sealed sterile, apyrogenic 5 mL glass vials.

111
In[In]-BnDTPA-trastuzumab-NLS injection was prepared by aseptically decapping a single kit vial in a Class II Type A2 biosafety cabinet (Model 425-400; NuAire) and adding 100-165 MBq (~ 19 μL) of 111 In[In]Cl 3 (BWXT Medical Ltd.) mixed with 38 μL of 0.05 M NH 4 CO 2 CH 3 buffer, pH 5.5 using a digital pipette and sterile tip. The vial was incubated at RT for 1-2 h, then diluted to a final volume of 2.0 mL by addition of Sodium Chloride Injection, USP. The radiopharmaceutical solution was then drawn up in a 3 mL syringe with an 18G x 1½" needle and the solution sterilized by passing through a 0.22 μm Millex GV filter into a 5 mL sterile glass vial with grey butyl rubber septum and aluminum seal. The integrity of the sterilizing filter was checked by the bubble test. 111 In[In]-BnDTPA-trastuzumab-NLS injection was assayed in a radioisotope dose calibrator (Capintec Model CRC25R, Ramsey, NJ, USA) and the specific activity calculated by dividing the radioactivity (MBq) by the mass of BnDTPA-trastuzumab-NLS in the kit (5 mg). The specification was 20-33 MBq/mg. The radiopharmaceutical was tested against specifications for pH (5.5-6.5), radiochemical purity (RCP ≥ 90%), radionuclide purity (≥ 99%), appearance (clear, colourless, particle-free) and sterility (retrospective USP Sterility Test). The pH was measured using pH 4.5-7.5 range pH paper (pHydrion ® ). The RCP was measured by ITLC-SG developed in 0.1 M sodium citrate, Page 10 of 21 Chan et al. EJNMMI Radiopharmacy and Chemistry (2022)  In-BnDTPA-trastuzumab-NLS injection were tested retrospectively for sterility by the USP Sterility Test after decay for at least 60 d. The stability of 111 In[In]-BnDTPAtrastuzumab-NLS injection stored at 2-8 °C was assessed over 24 h by re-testing against specifications for clarity and RCP (≥ 90%) to establish an expiry.

Statistical analysis
Statistical significance was determined using an unpaired Student's t test (P < 0.05) using GraphPad software Ver. 9.0 for Mac.

Raw materials and pharmaceutical buffers
Trastuzumab migrated as a single major band on SDS-PAGE under non-reducing conditions corresponding to a protein with the expected MW = 170 kDa and two bands under reducing conditions corresponding to proteins with MW = 50 kDa and 25 kDa, representing the heavy and light chains of trastuzumab (Fig. 1A). These bands were immunopositive on Western blot when probed with goat anti-human Fab-specific horseradish peroxidase (HRP) immunoconjugates (Fig. 1B). SE-HPLC analysis of trastuzumab showed one major peak with retention time (t R = 11.7 min) accounting for > 98% of all peak areas (Fig. 1C). Proton NMR analysis of p-SCN-BnDTPA agreed with the reported 1 H spectrum (Brechbiel et al. 1986). The COA from the supplier (Macrocyclics) confirmed 94.1% purity. Amino acid composition analysis confirmed the identity of NLS peptides and the COA from the supplier (Bio Basic, Inc.) showed 98.6% purity and a mass spectrum consistent with the expected MW = 1,419 Da. Three lots of 0.1 M

Stability testing of kits and expiry
The stability of kits stored at 2-8 °C was assessed up to 304 d for lot 18N008, 589 d for lot 17N026 and 661 d for lot 17N014 by re-testing against key quality specifications except sterility and endotoxins ( Table 2). All lots met specifications for pH, appearance and purity and homogeneity by SDS-PAGE and SE-HPLC (representative lot 17N014 shown in Fig. 5). Protein concentration was slightly higher (5.6-5.9 mg/ mL) than specification (4.5-5.5 mg/mL) at extended storage times (304-661 d). However, protein concentration was within specification for kit lot 17N014 up to 127 d (5.4 mg/mL) and kit lot 17N026 up to 139 d (5.5 mg/mL) and only slightly higher than specification for kit lot 18N008 at 129 d (5.6 mg/mL). The labeling efficiency with 111 In remained high: 97.4%, 97.6% and 98.1% for kit lots 18N008, 17N026 and 17N014, respectively, when stored up to 661 d, 589 d and 304 d, respectively (Table 2). 111 In[In]-BnDTPA-trastuzumab-NLS Injection prepared from the kits exhibited preserved high affinity specific binding to HER2-positive SK-BR-3 human BC cells (mean ± SD: K a = 3.6 ± 0.4 × 10 8 L/mol; B max = 1.1 ± 0.1 × 10 6 sites/cell). Based on these testing results, and to limit the protein concentration exceeding specification, an expiry of 180 d was assigned to the kits stored at 2-8 °C. The specific binding (SB) was calculated by subtracting NSB from TB. In this representative HER2 binding assay, the K a was 3.8 ± 0.6 × 10 8 L/mole and the B max was 1.0 ± 0.03 × 10 6 sites per cell

In[In]-BnDTPA-trastuzumab-NLS injection
Fifteen sequential lots of 111 In[In]-BnDTPA-trastuzumab-NLS injection were prepared (Table 3). All lots passed specifications for specific activity, pH, appearance, radiochemical purity (RCP) and sterility (tested retrospectively). Re-testing after storage at 2-8 °C showed that there was no significant difference between the mean ± SD RCP immediately after preparation (98.4 ± 0.01%) and at 24 h in storage (98.0 ± 0.6%). All lots of 111 In[In]-BnDTPA-trastuzumab injection remained clear, pale yellow and were particle-free. An expiry of 8 h was assigned to 111 In[In]-BnDTPA-trastuzumab-NLS injection stored at 2-8 °C, since the product was intended to be administered to a patient within this time period.

Discussion
We report here the formulation of a kit under GMP conditions for preparing 111   recently safely administered to 4 patients with HER2-positive BC to trace the uptake of trastuzumab into brain metastases enhanced by MRIg-FUS (Meng et al. 2021). Our report is the first to describe a kit for preparing 111 In[In]-BnDTPA-trastuzumab-NLS injection under GMP conditions and to our knowledge, is the first to report formulation of a radiopharmaceutical modified with NLS peptides for human administration.
To formulate the kit, trastuzumab was first reacted with p-SCN-BnDTPA to introduce BnDTPA chelators to complex 111 In. There were 3.9, 3.8 and 5.1 BnDTPA/trastuzumab molecule for kit lots 17N014, 17N026 and 18N008, respectively (Table 1). BnDTPA-trastuzumab was then reacted with Sulfo-SMCC to introduce maleimide groups for reaction with the thiol on cysteine on 13-mer peptides [CGYG PKK-KRKVGG] that harbour the NLS of SV-40 large T-antigen (italics) (Costantini et al. 2008b). NLS conjugation was confirmed by an upward shift in the band for BnDTPAtrastuzumab-NLS compared to BnDTPA-trastuzumab on SDS-PAGE under nonreducing conditions (Fig. 2). Based on the band shift, we estimated that there were 3, 2 and 3 NLS peptides per trastuzumab molecule for kit lots 17N014, 17N026 and 18N008, respectively. This level of NLS peptides is within the range that we previously reported for 111 In[In]-DTPA-trastuzumab-NLS (2.5-8.4 NLS peptides per trastuzumab molecule) (Costantini et al. 2007). We noted a small percentage (< 8%) of higher MW (> 250 kDa) protein on SDS-PAGE analysis of BnDTPA-trastuzumab-NLS under non-reducing conditions, that was not present in BnDTPA-trastuzumab or trastuzumab. Similarly, a small percentage (< 10%) of a higher MW protein was detected by SE-HPLC analysis of BnDTPA-trastuzumab-NLS ( Fig. 3A; t R = 10.6 min) but not for trastuzumab (Fig. 3B). We believe that this represents cross-linked IgG due to reaction of BnDTPA-trastuzumab with Sulfo-SMCC. Sulfo-SMCC incorporates a sulfosuccinimidyl ester that reacts with primary amines and a maleimide functional group that reacts with thiols. Our intent was to react amine groups on BnDTPA-trastuzumab with Sulfo-SMCC and use the introduced maleimide functionality to conjugate BnDTPA-trastuzumab to the NLS peptides by reaction with the thiol on the cysteine in the peptides. However, recombinant monoclonal antibodies may contain a small number of free thiols that could react with the maleimide functional groups and this may be responsible for the small amount of cross-linking of BnDTPA-trastuzumab molecules observed (Metcalfe 2022). 111 In[In]-BnDTPA-trastuzumab-NLS injection prepared from the kits exhibited high affinity and specific binding to HER2 on SK-BR-3 human BC cells in a direct (saturation) binding assay (K a = 4.6-6.2 × 10 8 L/mole; B max = 0.9 × 10 6 binding sites/cell; Table 1 and Fig. 4). The HER2 binding affinity was similar to that previously reported for 111 In[In]-DTPA-trastuzumab-NLS, measured in a competition receptor-binding assay using SK-BR-3 cells (K a = 1.7-3.2 × 10 8 L/mole) (Costantini et al. 2007). We previously reported that 111 In[In]-BnDTPA-trastuzumab exhibited a K a = 3.2 × 10 8 L/mole (K d = 3.1 × 10 -9 mol/L) for binding to HER2 on SK-BR-3 cells and B max = 1 × 10 6 binding sites/cell. (Chan et al. 2020). The product monograph for trastuzumab (Herceptin, Roche) states that the K d for binding to HER2 = 5 × 10 -9 mol/L (K a = 2 × 10 8 L/mole) in a cell-based assay (Roche 2021). Thus, the HER2-binding properties of 111 In[In]-BnDTPAtrastuzumab-NLS were similar to those of 111 In[In]-BnDTPA-trastuzumab or unmodified trastuzumab.