PURPOSE: The purpose of this study was to examine whether recombinant human endostatin (rhEndostatin), an antiangiogenic agent, is effective against a human neuroblastoma cell line (designated TNB9). We employed a human neuroblastoma xenograft model, and we investigated whether continuous infusion is more effective than an intermittent administration. METHODS: In the first experiment, when the tumors on the backs of nude mice reached a weight of 90 mg, rhEndostatin was administered subcutaneously to the mice (n=5) every day for 10 consecutive days. In the second experiment, the same daily dose of rhEndostatin was administered continuously to the TNB9- bearing mice (n=6) via subcutaneous infusion pumps for 3 consecutive days with the total dose being 30% of the dose given in the first experiment. Nestin and factor VIII expression levels were assessed immunohistochemically to elucidate whether the effects of rhEndostatin was present according to the histologic evidence at day 4 in the second experiment. RESULTS: In the first experiment, the relative tumor weight in treated mice (n=5) was significantly less than that in the controls (n=12) on day 2 after treatment initiation only (P<.05). The maximum inhibition rate (MIR) of TNB9 xenograft growth by rhEndostatin was 46.4%, indicating the lack of efficacy. In the second experiment, the effects of rhEndostatin were much more marked than those noted in the first experiment, with the MIR being 60.7%. The mean relative tumor weight in the treated group (n=6) in the second experiment was significantly less than that in the controls (n=10) on days 2, 4 and 6 (P<.01), as well as on days 8 and 10 (P<.05). The nestin staining in the endothelium of the control tumors (n=2) was remarkable, whereas the nestin staining showed as a loss of fibrillar structure in the rhEndostatin-treated tumors (n=2). The number of vessels immunostained with antifactor VIII antibody was markedly reduced in the tumors (n=2) from the rhEndostatin-treated mice compared with that from the control mice (n=2). CONCLUSION: Continuous administration of rhEndostatin resulted in more significant tumor regression than an intermittent administration of the agent. This result suggests that the continuous infusion of rhEndostatin is an effective agent and administration method for treating patients with neuroblastoma in the future.