Self-Assembling and DisAssembling Bispecific Antibodies for Curative 2-step Pretargeted Radioimmunotherapy

Introduction: Conventional pre-targeted radioimmunotherapy (PRIT) has relied on long incubations between targeting and payload delivery (2-step) or synthetic clearing agents (3-step) to reduce the otherwise toxic side-effects of radiation exposure to the blood, liver, marrow and kidneys, leading to reduced therapeutic index or increased regulatory complexity, respectively. We now report a novel 2-step PRIT platform using Self-Assembling and DisAssembling (SADA) domains that induce rapid elimination of unbound antibody without clearing agent (CA), allowing for tumor specific delivery of ¹⁷⁷Lu and ²²⁵Ac isotopes.

Methods: Tandem bispecific antibody (BsAb) fragments (anti-GD2 x anti-DOTA) were fused to SADA domains to create tetrameric BsAb of approximately 200 kDa. SADA domains were screened to keep each monomeric subunit between 50 and 70 kDa, to allow for renal filtration. SADA domains were profiled for stability, affinity and purity in vitro, in addition to tissue biodistribution, serum pharmacokinetics and anti-tumor function in vivo by treating xenograft mouse models of GD2(+) neuroblastoma and small cell-lung cancer with ¹⁷⁷Lu or ²²⁵Ac isotopes.

Results: Analysis of multiple candidate SADA domains revealed the TP53-derived tetramerization sequence to be of consistent stability and purity, which became our lead candidate. This SADA-BsAb expressed at high purity (>90%) and displayed improved tumor binding avidity over conventional bivalent molecules (IgG). Ex vivo biodistribution analysis in xenograft mouse models revealed SADA-BsAb completely cleared from the blood within 48 hours, while achieving high and specific tumor uptake (for e.g., estimated doses of 320 cGy/MBq to the tumor, 3 cGy/MBq to the blood (100:1), and 13 cGy/MBq to the kidneys (25:1) for ¹⁷⁷Lu), which was verified by PET/CT using ⁸⁶Y. Treatment of subcutaneous neuroblastoma tumors (both cell line and patient-derived xenografts, PDX) using SADA-PRIT and ¹⁷⁷Lu (up to 6,000 MBq/kg), completely eliminated established tumors (median survival not reached after >120 days, vs 25 days for control groups). Histological analysis over 100 days after treatment revealed no clinical, biochemical or histologic toxicities to the kidneys, liver, spleen, or hematopoietic system including bone marrow. In addition, using the previously developed Proteus-DOTA (J Nucl Med 2018 59:123), SADA-PRIT safely delivered 1.4 MBq/kg of ²²⁵Ac to shrink neuroblastoma PDX tumor without any toxicity to the kidneys. Lastly in an aggressive PDX model of small cell lung cancer, treatment with SADA-PRIT and ²²⁵Ac induced complete responses in all treated mice and significantly extended survival (median survival 90 days vs 10 days for control groups). Metastatic relapses (as gross palpable masses distal from site of tumor inoculation) remained GD2(+) and responded to further GD2-directed SADA-PRIT to obtain additional complete responses.

Conclusions: 2-step SADA-PRIT proved capable of safely delivering high doses of radiation to cure mice of disease without any observable treatment-related toxicity. Additionally, the flexibility and modularity of the SADA-design allows for simple and facile adaptation to most tumor targets and payloads.