Determination of the therapeutic index of floxuridine by its circadian infusion pattern.

To test if circadian timing of a drug is important for its toxicity and antitumor activity, we compared the circadian patterns observed with seven equal doses of floxuridine (FUDR) infused either at a variable rate or at a constant rate in female F344 rats. For the variable-rate infusion, the daily dose of FUDR was divided into four 6-hour portions of 68%, 15%, 2%, and 15% to achieve a quasisinusoidal pattern. Peak drug delivery occurred during one of six different times of day. At a dose level resulting in 50% overall mortality, lethal toxicity differed significantly, depending on the circadian stage of maximum drug delivery. Depending on the circadian stage of maximum drug flow, variable-rate infusions were more toxic than or as toxic as constant-rate infusion. FUDR lethality was lowest when constant-rate infusion was used or when variable-rate infusion peaked during the late activity-early rest span of the recipients. The circadian pattern of variable-rate infusion also determined the antitumor activity in tumor-bearing rats. At a therapeutic dose level and at identical dose intensity, the variable-rate-infusion pattern, with peak drug flow during the late activity-early rest span, resulted in significantly greater delay in tumor growth than was observed with either the constant-rate infusion or other variable-rate patterns. We conclude that the toxicity and antitumor activity of FUDR depend on the circadian timing of the infusion peak when the drug is given by variable-rate infusion. Since some of the circadian-shaped infusions studied are toxicologically and therapeutically inferior to constant-rate infusion, the circadian pattern and not the quasi-intermittency of circadian FUDR administration is primarily responsible for these pharmacodynamic differences.