Feasibility study of intraarterial vs intravenous cisplatin, BCNU, and teniposide combined with systemic cisplatin, teniposide, cytosine arabinoside, glycerol and mannitol in the treatment of primary and metastatic brain tumors.

Sixteen patients with intracerebral tumors received intraarterial cisplatin, teniposide, and BCNU combined with intravenous cisplatin, teniposide, and cytosine arabinoside. Oral glycerol and intravenous mannitol were given along with the intravenous chemotherapy in an attempt to increase drug delivery to tumor by augmenting tumor blood flow. Thirteen additional patients were treated with the same regimen, but received all the chemotherapy intravenously. Of the 16 patients receiving intraarterial chemotherapy (median survival, 14 weeks), none responded, 5 (31%) were stable for > 8 weeks, 8 (50%) failed, and 3 (19%) were unevaluable due to early death. Of the 13 patients receiving all their treatment intravenously (median survival, 13 weeks), 3 (23%) responded, 1 (8%) was stable, 7 (54%) failed, and 2 (15%) were unevaluable due to early death. In the patients receiving intraarterial chemotherapy, toxicity included ipsilateral retinal toxicity (2 patients), ocular pain or headache (10), periorbital swelling and flushing (6), increased brain edema with focal neurological deficits and drowsiness (5), and catheter-related carotid artery thrombosis followed by fatal herniation (1). Myelosuppression was worse in patients who received all their treatment intravenously than in those receiving intraarterial chemotherapy (p < 0.05). Neutropenic sepsis developed in 4 patients on the intraarterial arm (1 fatal) and in 5 patients on the intravenous arm (2 fatal). Other toxic effects were similar whether or not patients received intraarterial treatment or only intravenous treatment. Overall, toxicity of this regimen was excessive, and response rates were lower than would have been expected with single agent therapy.