Radioimmunotherapy of human pancreatic cancer xenografts in NOD-scid mice with [Cu]Cu-NOTA-panitumumab F(ab') alone or combined with radiosensitizing gemcitabine and the PARP inhibitor, rucaparib.

INTRODUCTION

Our objective was to determine the feasibility of extending our previously reported PET imaging study of pancreatic cancer (PnCa) with [Cu]Cu-NOTA-panitumumab F(ab') to radioimmunotherapy (RIT) by exploiting the β-particle and Auger electron emissions of Cu (PET theranostic concept). To enhance the effectiveness of [Cu]Cu-NOTA-panitumumab F(ab'), we further combined RIT with radiosensitizing gemcitabine (GEM) and the poly(ADP)ribose polymerase inhibitor (PARPi), rucaparib.

METHODS

Normal tissue toxicity was assessed in non-tumor-bearing NOD-scid mice injected i.v. with [Cu]Cu-NOTA-panitumumab F(ab') (1.85-9.25 MBq; 10 μg) or [Cu]Cu-NOTA-anti-mouse EGFR Ab30 F(ab') (12.95 MBq). Body weight was monitored, and hematopoietic (CBC), liver (ALT) and kidney [creatinine (SCr)] toxicity were assessed. RIT studies were performed in NOD-scid mice with s.c. OCIP23 human PnCa patient-derived xenografts (PDX) administered [Cu]Cu-NOTA-panitumumab F(ab') (3.7 MBq; 10 μg), unlabeled panitumumab F(ab') (10 μg) or normal saline every two weeks. Subsequent studies evaluated RIT with [Cu]Cu-NOTA-panitumumab F(ab') (12.95 MBq; 10 μg) administered alone or combined with GEM and the PARPi, rucaparib administered on a 14-day treatment cycle for up to 6 cycles in NOD-scid mice with s.c. PANC-1 human PnCa xenografts. The radiation absorbed dose in PANC-1 tumors and normal organs in mice after a single i.v. injection of [Cu]Cu-NOTA-panitumumab F(ab') (12.95 MBq; 10 μg) was estimated based on previously reported biodistribution studies of [Cu]Cu-NOTA-panitumumab F(ab').

RESULTS

No normal tissue toxicity was observed in non-tumor-bearing NOD-scid mice administered up to 3.7 MBq (10 μg) of [Cu]Cu-NOTA-panitumumab F(ab') but slightly increased ALT was noted at 9.25 MBq. Administration of [Cu]Cu-NOTA-anti-mouse EGFR Ab30 F(ab') (12.95 MBq; 10 μg) caused some hematopoietic toxicity but no increase in ALT or SCr or decreased body weight. A slight tumor growth delay and increased survival was noted in NOD-scid mice with s.c. OCIP23 PDX treated with [Cu]Cu-NOTA-panitumumab F(ab') (3.7 MBq; 10 μg) or unlabeled panitumumab F(ab') (10 μg) compared to normal saline treated mice. RIT with [Cu]Cu-NOTA-panitumumab F(ab') (12.95 MBq; 10 μg) combined with GEM + PARPi for up to 6 cycles was most effective for the treatment of PANC-1 tumors. Tumor doubling time increased to 13.3 ± 0.9 days vs. 7.8 ± 3.7 days for RIT alone and 9.3 ± 2.2 days for normal saline treatment. Median survival was significantly longer (P < 0.05) than in mice treated with normal saline (35 days) for RIT + GEM + PARPi (71 days), GEM + PARPi (44 days) and RIT + GEM (43 days) but not for RIT alone (25 days). RIT + GEM + PARPi provided a longer median survival than RIT (P < 0.01), GEM + PARPi (P = 0.01) but not RIT + GEM (P = 0.23). Nonetheless, PANC-1 tumors grew exponentially in all treatment groups. The absorbed dose in PANC-1 tumors after a single i.v. injection of [Cu]Cu-NOTA-panitumumab F(ab') (12.85 MBq; 10 μg) was 0.8 Gy, while the dose in normal organs ranged from 0.6-1.2 Gy.

CONCLUSIONS

We conclude that RIT with [Cu]Cu-NOTA-panitumumab F(ab') did not cause significant normal tissue toxicity but was not effective when administered alone for treatment of PnCa xenografts in NOD-scid mice. Combining RIT with GEM and the PARPi, rucaparib enhanced its effectiveness but tumors continued to grow exponentially. Our results suggest that Cu is not feasible for RIT of PnCa due to low tumor absorbed doses. Lu which has a higher abundance of moderate energy β-particle emissions may be more effective than Cu. The hematopoietic toxicity of [Cu]Cu-NOTA-anti-mouse EGFR Ab30 F(ab') may be mediated by binding to mouse EGFR expressed on some hematopoietic stem cells.

ADVANCES IN KNOWLEDGE AND IMPLICATIONS FOR PATIENT CARE

Direct extension of PET with Cu(Cu)-NOTA-panitumumab F(ab') to RIT exploiting the β-particle and Auger electron emissions of Cu is not feasible. Theranostic approaches that combine PET with RIT employing Lu may be more promising and should be explored.