The Supplementary Information covers cell-line authentication and culture conditions, serum-stability, intraperitoneal implantation, immunohistochemistry, cell and bead-binding assays, autoradiography, pathology analysis, and the synthesis and characterization of mcp-PEG4-TFP.

RadiochemistryZirconium-89 bioconjugation and radiolabeling for PET

SG (Gilead) was reconstituted (10 mg/mL, sterile saline), and conjugated to p-NCS-Bn-Deferoxamine (DFO) (Macrocyclics, B-705); in chelex-treated PBS (pH 8.7) using a 6-fold molar excess of deferoxamine (10 mg/mL in DMSO), incubated at 37 °C for 1 h, purified via a PD-10 column, and concentrated via a 30 kDa Amicon filter. Radiolabeling of SG-DFO with [89Zr]Zr-oxalate (1 M oxalic acid) was achieved. By adjusting the pH of solution to ~ 7.4 with HEPES buffer (1 M), followed by addition of antibody, and incubation in thermomixer at 37 °C for 1 h. Radiochemical conversion was assessed via radio-iTLC with EDTA buffer (50 mM, pH 5.5). Serum stability was assessed over 7 days at 37 °C in human serum (100167-WEB).

Lutetium-177 and Actinium- 225 bioconjugations and radiolabeling for radiotherapies

SG was buffer exchanged with chelex-treated PBS (pH 8.7) and conjugated to p-NCS-CHX-A”-DTPA (DTPA) (Macrocyclics, B-355), which was in a 6-fold excess, to yield DTPA-SG. Lutetium-177 radiolabeling was performed in 250 mM ammonium acetate (pH 5.5), incubated at 37°C for 1 h, and analyzed via radio-iTLC. Doses of 9.25 MBq (20 µg) and 18.5 MBq (40 µg) were prepared of [177Lu]Lu-DTPA-SG.

For Actinium-225 studies, SG was conjugated to macropa (mcp) via two-strategies: (i) click chemistry using a trans-cyclooctene (TCO)-modified SG and mcp-PEG7-Tz (17), or (ii) direct conjugation with tetrafluorophenyl ester (mcp-PEG4-TFP). Conjugations to SG were performed in chelex-treated PBS (pH 8.9) for 4 h at a 6-fold molar excess of mcp-PEG4-TFP, or a 30-molar excess of NHS-TCO (axial). Products were purified by PD-10 and Amicon 30 kDa concentrated. Actinium-225 radiolabeling was performed at 37 °C: mcp-Direct-SG in HEPES (37 kBq, 20 µg and 25.9 kBq, 14 µg), and mcp-Click-SG in 0.25 M ammonium acetate (pH 5.5) for 15 min, then neutralized and clicked to TCO-SG (37 kBq, 20 µg). Radiochemical conversion via radio-iTLC (0.4 M citrate buffer, pH 4) at 1 and 24 h post-iTLC with greater than 99% actinium-225 at the origin.

Animal studiesXenograft OVCAR3 model

All IACUC-approved procedures (MSKCC #08-07-013) used female nude mice (6–8 weeks, Charles River) implanted on the right flank with 10 million OVCAR3 cells in 150 µL 1:1 medium/BD Matrigel. Initial xenografts were given an additional 5 million OVCAR3 cells implanted as published previously [16]. Tumors reaching 100–350 mm3 were assigned to imaging (n = 5) or therapy (n = 10) cohorts. At start of therapy, cohort averages were between 100 and 350 mm3, though some individual tumors were larger than 350 mm with only one in the entire study over 550 mm3.

PET and Cerenkov imaging

Female nude mice (RA00077606) bearing subcutaneous tumors received 9.2 MBq, 20 µg of [89Zr]Zr-DFO-SG, with imaging at 24–144 h post-injection on a Siemens Inveon PET/CT capturing 40 million events per mouse per timepoint. Reconstructed images (2DOSEM 512 binning) were then calibrated to percent injected activity per cubic centimeter. For the Lutetium-177 study, mice were imaged 24 h after administration of [89Zr]Zr-DFO-SG to confirm positivity. After second dose of [177Lu]Lu-DTPA-SG, mice were imaged 24 h post-injection via Cerenkov luminescence (IVIS, open filter, 300 s). Tumor radiance was reported in p/s/cm2/sr using Living Image V 4.2.

Radioimmunotherapy (RIT)

Mice bearing OVCAR3 tumors (100–350 mm3) were enrolled into five therapy studies (n = 10):

Control – one dose of 20 µg of SG.

Standard of Care – 180 µg of SG on days 1, 7, and 14; cycle repeated monthly after 1-week break.

SG 177Lu Therapy – 9.25 MBq (20 µg) followed by 18.5 MBq (40 µg) of [177Lu]Lu-DTPA-SG.

Click-SG 225Ac Therapy – 37 kBq, 20 µg of [225Ac]Ac-mcp-Click-SG.

Direct-SG 225Ac Therapy – 37 kBq (20 µg) followed by 25.9 kBq (10 µg) of [225Ac]Ac-mcp-Direct-SG.

Complete blood count (CBC) parameters (white blood cells (WBC), red blood cells (RBC), and platelets) were monitored weekly in n = 3 mice/group. Tumor sizes and body weight were recorded twice weekly. Mice whose weight decreased received supplemental DietGel. Therapy endpoints were 2,000 mm3, petechia, and > 20% body weight loss. Tumor dimensions were measured with vernier caliper, and TV calculated by:

$$\:TV=\left(\frac3\right)\left(\frac\alpha2\right)^2\left(\frac b2\right)$$

TV = tumor volume (mm)3

α = longest axis of the tumor (mm)

b = axis perpendicular to the longest axis (mm), α (mm)

Ex vivo biodistribution

Ex vivo biodistribution analyses were performed on separate cohorts of xenografted mice that were administered [89Zr]Zr-DFO-SG, [225Ac]Ac-mcp-Direct-SG, or [225Ac]Ac-mcp-Click-SG. For serial biodistribution studies with [89Zr]Zr-DFO-SG, mice were injected with 9.25MBq (20 µg). Mice (n = 3) were euthanized at 4, 24, and 144 post-injection. For the [225Ac]Ac-mcp-Direct-SG and [225Ac]Ac-mcp-Click-SG cohort, mice were injected with 37 kBq (20 µg). Mice (n = 3) were euthanized at 4, 24, 72 and 168 post-injection. Organs harvested include blood, heart, lung, liver, spleen, pancreas, small intestine, large intestine, kidneys, muscle, bone, and tumor. Organs were weighted, and activity was measured using a Perkin Elmer Wizard 3 series gamma counter with the respective radionuclide protocol. Values were reported as % injected activity per gram of tissue.

Statistical analysis

Data were analyzed using GraphPad Prism v10. Unpaired, two-tailed t tests assessed bead and cell-binding; one-way ANOVA with Tukey-Kramer correction was used for organ uptake in biodistribution experiments. Survival was analyzed via a log rank (Mantel–Cox) test. A P value less than 0.05 was considered statistically significant.