CONTROL OF VASOGENIC EDEMA IN A BRAIN-TUMOR MODEL - COMPARISON BETWEEN DEXAMETHASONE AND SUPEROXIDE-DISMUTASE

OBJECTIVE: The production of prostaglandin (PC) within brain tumors pr obably generates excessive amounts of oxygen free radicals that may di srupt microvessel permeability within the tumor and in the adjacent br ain. We evaluated the effect of systemic therapy with recombinant huma n manganese-superoxide dismutase (r-hMnSOD) and with dexamethasone on the vascular permeability (VP) of a brain tumor and the adjacent brain . Treatment effect was also evaluated in control animals subjected to mild penetration injury. METHODS: Fischer rats were injected stereotac tically with either 10(5) cells of malignant sarcoma or with vehicle i nto the right parietal hemisphere. Nine days later, the animals were t reated with r-hMnSOD (50 mg/kg of body weight every 12 h [one intraven ous, then two intraperitoneal injections]; serum levels, 1100-1800 mu g/ml), dexamethasone (2 mg/kg every 12 h [one intravenous, then two in traperitoneal injections]), or vehicle and were killed after 30 hours for evaluation of VP and PC production. RESULTS: The VP was markedly i ncreased within the tumor (P < 0.001), in the brain adjacent to it, an d in the vehicle injection site. The VP of the normal brain was unaffe cted by r-hMnSOD or dexamethasone treatment, unlike the VP in the tumo r, the adjacent brain, and the injection sites of control animals, whe re it was reduced by 50, 54, and 23%, respectively (P < 0.04), for r-h MnSOD and 50, 41, and 71%, respectively (P < 0.05), for dexamethasone. A one- to threefold increase in synthesis of thromboxane and PGE(2) w as measured within the tumor, the adjacent brain, and the injection si tes of control animals (P < 0.0001). Treatment with r-hMnSOD had no ef fect on tumor PG production, but it reduced the synthesis in the brain tissue adjacent to the tumor and in traumatized control animals (P < 0.04). Immunohistochemical evaluation revealed vascular proliferation with abnormal basal membrane, atypical astrocytes, and large numbers o f reactive macrophages present in the adjacent brain and at the inject ion sites of control animals but not within the tumor mass. CONCLUSION : Oxygen free radicals probably enhance vasogenic brain edema resultin g from tumor and penetration injury. The edema can be attenuated by sy stemic r-hMnSOD therapy, which has been proven to be as effective as s teroid treatment. An inflammatory response may account for oxygen free radical production in brain tissue adjacent to the tumor and at the i njection site of vehicle solution, but other mechanisms probably gener ate oxygen free radicals within the tumor mass.