Radiopharmaceutical therapies (RPT) activate a type I interferon (IFN1) response in tumor cells. We hypothesized that the timing and amplitude of this response varies by isotope. We compared equal doses delivered by 90Y, 177Lu, and 225Ac in vitro as unbound radionuclides and in vivo when chelated to NM600, a tumor-selective alkylphosphocholine. Response in murine MOC2 head and neck carcinoma and B78 melanoma was evaluated by qPCR and flow cytometry. Therapeutic response to 225Ac-NM600+anti-CTLA4+anti-PD-L1 immune checkpoint inhibition (ICI) was evaluated in wild-type and stimulator of interferon genes knockout (STING KO) B78. The timing and magnitude of the IFN1 response correlated with radionuclide half-life and linear energy transfer. The ratio of CD8+ T cells to regulatory T cells (Treg) increased in tumors 7 days after 90Y- and 177Lu-NM600 and day 21 after 225Ac-NM600. 225Ac-NM600+ICI improved survival in mice with wild-type but not with STING KO tumors, relative to monotherapies. Thus, we have found that the immunomodulatory effects of RPT vary with radioisotope and promote tumor cell STING-dependent enhanced response to ICIs in murine models. These findings have implications for the optimization of RPT–immunotherapy combinations and could guide the relative timing of therapies, the selection of isotope, and patient selection through tumor biomarkers.
Purpose: Glypican-3 (GPC3), which is highly expressed in the most common liver cancer, hepatocellular carcinoma, but absent from normal liver, represents an attractive tumor-selective target. We evaluated GPC3-targeted paired positron emission tomography (PET) diagnostic and α particle-emitting therapeutic (radiotheranostic) agents in orthotopic models of liver cancer. Experimental Design: Using the GPC3-specific antibody codrituzumab (GC33), we engineered an immunoPET agent, [89Zr]Zr-DFO-GC33 (89Zr-GC33), and a therapeutic agent, [225Ac]Ac-Macropa-GC33 (225Ac-GC33). Binding specificity and biodistribution were assessed in orthotopic luciferase–expressing liver cancer models (HepG2 and Hep3B). Microscale co-localization of the radiotheranostic pair was validated using quantitative particle identification spectral autoradiography. A preclinical trial of the therapeutic efficacy of a single 9.3 kBq administration of 225Ac-GC33 was conducted. Primary endpoints were local control, as assessed by longitudinal bioluminescence imaging (BLI) and immunoPET with 89Zr-GC33 with histologic confirmation, and overall survival (OS). Results: The biodistribution of 225Ac-GC33 closely mirrored that of 89Zr-GC33, with both agents showing high accumulation in HepG2 tumors (>200% injected activity per gram) and a high degree of overlap in organ-level macrodistribution. Animals treated with 225Ac-GC33 showed excellent local control, including several complete pathologic responses and improved OS compared with control cohorts. We also observed efficacy in the orthotopic Hep3B tumors, suggesting activity even in tumors with moderate GPC3 expression. In both models, longitudinal 89Zr-GC33 immunoPET correlated with BLI. Conclusions: We demonstrate that GPC3-specific PET imaging can predict and assess therapeutic response to GPC3-targeted α particle therapy in orthotopic models of liver cancer. Such an approach can begin to address the unmet diagnostic and therapeutic needs for patients with liver cancer and may extend to other GPC3-expressing malignancies.