Use of in vitro dose response effects to select antineoplastics for high-dose or regional administration regimens.

Several preclinical and clinical studies have documented that dose or dose intensity of chemotherapeutic agents are important factors for response of patients' tumors. This finding has prompted empiric trials of certain chemotherapeutic agents in high-dose or regional administration treatment regimens. The present study was performed to identify agents that would be particularly good candidates for high-dose or regional administration regimens against particular types of tumors. Using a human tumor cloning technique, we constructed dose in vitro response lines for ten different chemotherapeutic agents against seven different histologic types of malignancies. Slopes of the lines indicated the agents with the greatest increases of in vitro response per increment in dose of the agent. Tumors against which the agents gave the steepest dose response lines included lymphoma, head and neck cancer, ovarian cancer, and small-cell lung cancer, while the dose response lines for non-small-cell lung cancer, breast cancer, and colon cancer were quite flat. Suggestions for clinical trials based on these findings include the use of high-dose melphalan for patients with lymphoma, head and neck, and ovarian cancer; the use of mitoxantrone in high-dose regimens for patients with breast cancer; high-dose cisplatin regimens for patients with small-cell lung cancer; high-dose bleomycin regimens for patients with non-small-cell lung and head and neck cancer; and regional perfusion of liver metastases from colorectal cancer with cisplatin. Prospective testing of high-dose or regional administration regimens suggested by this new model should indicate its use for prediction of the best agent to use in high-dose regimens against a particular tumor type.