Tissue distribution and pharmacokinetics (PK) for KOS-1584: A rationally designed microtubule polymerization stabilizer

3098

KOS-1584 (9,10-didehydroepothilone D) was designed to optimize tumor tissue penetration while limiting the exposure to selected organ tissues (including the central nervous system). PK studies in the rat and dog had shown that the compound had a large volume of distribution (Vss > 10 L/kg in all species). Whole body autoradiography studies in the rat demonstrated wide distribution of the compound into highly perfused tissues (highest concentrations were seen in the liver, small bowel, lung, kidney, and bone marrow), with clearance from peripheral tissues 96 hr following the injection. The predominant route of elimination of drug was via the feces (90.1%). Prior to clinical testing, experiments were conducted to determine whether time of administration would result in differential distribution of the compound. Subsequent distribution studies used tumor-bearing nude mice (MX-1 breast cancer cell line) with direct quantification of KOS-1584 in various organs following a dose of 2.5 mg/kg administered over 5-min, 1-hr or 3-hr (comparing end-of-infusion, 4, 6 and 24 hours following the dose). Based on previous experiments, exposure over these 3 times of administration were predicted to be essentially equivalent; the primary PK difference would be in plasma Cmax. Treatment of mice with these different times of administration demonstrated superior penetration of KOS-1584 to tumor tissue for the 1-hr and 3-hr infusions (maximum concentrations at the end-of-infusion of 131, 1677, and 963.3 ng/g in tumors comparing the 5-min, 1-hr, and 3-hr infusions). Both 1-hr and 3-hr infusion regimens provided a favorable tumor-to-brain distribution ratio comparing either the end-of-infusion values or values observed 24-hours following exposure to the drug. At the end of the 1-hr and 3-hr infusions, KOS-1584 levels were 5 times higher in the tumor than in the brain (in contrast to the 5-min infusion in which the concentration of KOS-1584 in tumor was only 28% of that of brain). Overall penetration to the brain was lowest with the 3-hr infusion. With the different infusion regimens, highly perfused organs, such as kidney, lung, heart, and spleen, had highest concentrations of KOS-1584 with relatively rapid clearance of the compound.Predictions from nonclinical species to humans estimated the volume of distribution to be large (Vd values for mouse, rat, monkey and dog were 51-83, 30-50, 27-40 and 12-18 L/kg in these species). The compound began phase 1 clinical trials in patients with advanced malignancies in late 2004. Data from the first in-human studies of KOS-1584 confirmed the predicted value, with Vz of ∼800 liters following a 3-hr infusion (n=19 patients at doses between 0.8 - 6.5 mg/m²). The 1-hr infusion is also being tested in humans. These first-in-human results are consistent with the observation of wide distribution of KOS-1584 in the rodent model.