Preclinical characterization of pretargeted radioimmunotherapy (PRIT) with GD2-SADA, a self-assembling and disassembling bispecific fusion protein.

e15101 Background: GD2-SADA is a bispecific fusion protein with binding domains for the tumor-associated antigen GD2 and Lutetium 177 (Lu-177)-DOTA complex, and a p53-derived domain that mediates tetramer self-assembly and disassembly (SADA). In Step 1 of PRIT, SADA tetramers bind to GD2+ tumor cells, while unbound protein disassembles into renally cleared monomers. In Step 2, the radioactive payload Lu-177-DOTA is delivered and binds to tumor-bound GD2-SADA. Here, we characterize the binding activities of GD2-SADA to a range of GD2+ tumors and Lanthanide (Ln) metal-DOTA complexes and report the dose responsive, anti-tumor efficacy of GD2-SADA with Lu-177-DOTA in mice. Methods: GD2-SADA binding to GD2+ cell lines (small-cell lung cancer [SCLC], melanoma, osteosarcoma) was evaluated by flow cytometry (FC). A colorimetric competition ELISA was used to investigate binding to Ln-DOTA complexes (terbium [Tb], europium [Eu], and lanthanum [La]), trace metal TM-DOTA complexes (iron [Fe], copper [Cu], or zinc [Zn]), and empty DOTA (used in excess during radiolabeling to ensure complete Lu-177 chelation). GD2 binding of tetrameric GD2-SADA and monomeric p53 mutant SADA was compared using ELISA and FC assays. Tumor response to GD2-SADA with Lu-177-DOTA was evaluated in mice with SC SCLC xenografts (NCI-H524). Mice received 3 weekly treatments with a fixed GD2-SADA dose (10 mg/kg) followed by a Lu-177-DOTA dose (32 mCi or 47mCi/kg; n=6-10 per group). Tumor volumes were monitored weekly for ≥60 days after treatment. Results: GD2+ cell lines showed dose-dependent binding of GD2-SADA, with half-maximal effective concentration (EC50) values between 1.6 and 5 μg/mL. In ELISA and FC assays, the monomeric p53 mutant SADA demonstrated 85-93% reductions in binding compared with GD2-SADA. Competition ELISA demonstrated comparable binding by all Ln-DOTA complexes (50% inhibition concentration, IC50 [nM]: Lu, 2.3; Tb, 1.0; Eu, 1.2; La, 1.9). TM-DOTA and empty DOTA showed negligible binding to GD2-SADA. In mouse SCLC models, median time to tumor volume for 10 mg/kg GD2-SADA with Lu-177-DOTA dosing at 32 mCi and 47 mCi/kg exceeded control groups by 30+ and 40+ days, respectively. Conclusions: The anti-GD2 domain of GD2-SADA mediated binding to GD2+ tumor cells, a highly avid interaction that required a functional p53 tetramerization domain. Selective binding of GD2-SADA to multiple Ln-DOTA complexes further supports the SADA platform’s theranostic applications. Negligible binding to TM-DOTA complexes or empty DOTA strongly suggests the absence of competitive antagonism, an important consideration for targeted delivery of the radioactive payload in patients. GD2-SADA Lu-177-DOTA showed potent and dose-responsive anti-tumor efficacy in mice. Clinical development of GD2-SADA with Lu-177-DOTA is underway in adults and adolescents with GD2+ tumors (Trial 1001).