Evaluation of recombinant human tumor necrosis factor by scheduled intratumoral administration in mice bearing transplantable tumors.

The antitumor effect of recombinant human tumor necrosis factor (rTNF) was examined against Meth A fibrosarcoma in BALB/c mice and Sarcoma-180 in ddY mice. Significant hemorrhagic necrosis in tumor tissues occurred within 24 hr when optimal rTNF (1,000 to 5,000 units per mouse) was injected intratumorally on day 5 after intradermal inoculation of 5 x 10(5) tumor cells. Complete tumor regression resulted when two repeated courses of administration a week, each for 3 consecutive days, were given. For this marked effect to occur, however, initial tumor weight should not be greater than 1 g. When the initial tumor was greater than 1 g the surgical removal of tumor tissues was conducted and followed by rTNF administration. This caused hemorrhagic necrosis and the regression was the case with smaller tumors. When the cured mice were rechallenged with same tumors, more than 60% of mice rejected the tumors in a specific manner. In spite of such demonstration of specific immunity, well-known immunological effector mechanisms such as augmentation of natural killer cell activity, activation of antibody dependent cellular cytotoxicity or induction of interferon activity by rTNF were not detected in normal and tumor-bearing mice, suggesting that the activation of immunoregulatory cells by TNF itself may not involve at least in an early stage of TNF treatment. These results suggest that rTNF is a potent therapeutic agent for a certain solid tumor when the protocol of administration is optimized.