First-in-human phase I trial of a novel epothilone, KOS-1584

Abstract

2003 Background: KOS-1584 (9,10-didehydroepothilone D) was discovered as part of a screening program to develop a new generation of epothilones with higher potency and an improved pharmacologic/pharmacokinetic (PK) profile. Epothilones stabilize microtubule polymerization, inducing rapid G2/M arrest and apoptosis. Antitumor activity of KOS-1584 (Chou et al 2003) is approximately 3–12 fold more potent when compared to the structurally related Epothilone D. KOS-1584 demonstrates enhanced tumor tissue penetration and reduced exposure to selected tissues (including CNS). We report the results of the initial dose-escalation trial in which KOS-1584 was administered to pts with advanced solid malignancies. Methods: Define the MTD, toxicity profile and PK of KOS-1584 when administered via 3-hour infusion every 3 weeks. PK was determined after the 1st and 2nd infusion. Pharmacodynamics were assessed by serial sampling of PBMCs for microtubule bundle formation. Results: 27 pts (17 F; median age 60; median ECOG PS 1; median prior regimens 3, range 0–7) enrolled in 8 dose levels (between 0.8 - 11.3 mg/m2). To date, no Cycle 1 DLT has been seen. Toxicities (n=24) did not show obvious dose dependency; common toxicity (Grade 1–2) included gastrointestinal (diarrhea, constipation, nausea), fatigue, and ↑AST. Drug-related Grade 3 toxicity: constipation, fatigue and ↑AST (1 each). Drug-related neurotoxicity was not notable. PK/parent (n=25): t½ 17.7 ± 4.6h, Vz 741±330 L and CL 30.2± 16.5 L/h. At 8.5 mg/m2 Cmax 78 ± 29 ng/mL; AUCtot 631 ± 337 ng*h/mL. Cmax and AUCtot increased linearly with dose over the range tested. Vz is ∼5-fold and t½ 2-fold higher than that of Epothilone D. Dose dependent increases in microtubule bundle formation were observed (8.5 mg/m2: 40–50% at end of infusion, compared to 60–65% for ixabepilone and 50–60% for Epothilone D using the same assay at their phase 2 dose). A sigmoidal Emax model described the relationship between plasma concentration and microtubule bundle formation. Activity consisted of 4 pts with extended stable disease (6 cycles leiomyosarcoma and ovarian cancer; 5 cycles colon cancer; a 2nd patient with ovarian cancer is active at 5 cycles). Of these, 3 had document progressive disease prior to study. Conclusions: Accrual is continuing in order to define the optimal dose on the 3-week regimen. [Table: see text]