Preclinical analysis of radiolabeled anti-GD2 immunoglobulin G.

BACKGROUND: Unlabeled murine monoclonal anti-GD2 immunoglobulin (Ig)G (14G2a) reactive with nervous system diganglioside and neuroblastoma, melanoma, and small cell lung carcinoma produces tumor regression. However, serious acute abdominal pain, paresthesia, hypotension and hypertension, syndrome of inappropriate secretion of antidiuretic hormone (SIADH), and occasional motor weakness occur. Studies in preclinical animal models can elucidate the mechanism of the observed neurotoxicity and lead to anti-GD2 antibody treatment with a higher therapeutic ratio. METHODS: One mg of 14G2a or control IgG was labeled with 1-2 mCi of indium-111 and administered intravenously to beagles (n = 8). In 2 dogs, additional high dose (200 mg) unlabeled 14G2a was given over 5 days. Whole body gamma camera images and SPECT scans were obtained repeatedly over 7 days. On Day 7, sciatic nerve conduction studies were performed, and after euthanasia radioactivity was determined in major organs. RESULTS: Unlabeled high dose 14G2a administered to mice, rats, or rabbits did not cause neurotoxicity within 3 weeks. GD2 antigens were shown by immunochemistry to be present in brain and peripheral nerve tissues of rodents and beagles. After in vivo administration of radiolabeled 14G2a, canine lymph nodes showed specific uptake, but only minimal radioactivity was found in the nervous system. Dogs that received additional high dose unlabeled 14G2a showed much higher lymph node uptake and follicular lymph node hyperplasia. Low motor response amplitudes on nerve conduction studies were noted. CONCLUSIONS: A radioisotope label on IgG and its visualization in a large series of animal models indicate that a low protein dose of anti-GD2 IgG will not cause neurologic side effects in patients. High protein dose anti-GD2 IgG may enhance antineoplastic effects and contribute to neurotoxicity through stimulation of normal lymphocytes with subsequent release of cytokines.