The Epothilone Dilemma

Abstract

Because of their potent antitumor activity, drugs targeting microtubules are among the most commonly prescribed anticancer agents. These agents fall into the following two broad categories: agents that inhibit tubulin/microtubule polymerization and agents that stabilize preformed microtubules under depolymerizing conditions (the taxanes). The broad clinical antitumor activity of taxanes has stimulated the search for mechanistically similar agents that overcome the difficulties related to formulation and administration and yet subvert the typical mechanism of cellular resistance conferred by the drug efflux by P-glycoprotein and show preserved activity in paclitaxel-resistant cells with mutations in beta-tubulin. The epothilones (a name derived from its molecular features; epoxide, thiazole, and ketone) have emerged as a new class of microtubule-targeting agents and were originally isolated from the broth of a fermenting soil bacteria, Sorangium cellulosum. The epothilones and taxanes have overlapping binding sites to the surface of beta-tubulin, and the microtubule polymerization–inducing mechanism of epothilones seems to be similar to that of taxanes. Preclinical studies indicate a relatively broad spectrum of activity for epothilones. They are generally 5 to 25 times more potent than paclitaxel in inhibiting cell growth in cultures. Epothilones and taxanes seem to differ distinctly in their drug-resistance mechanisms because the cytotoxic effects of the epothilones are preserved in multi-drug resistance– expressing cell lines and in cells harboring tubulin mutations, which demonstrate resistance to paclitaxel. Five epothilone analogs are currently undergoing clinical assessment; these are ixabepilone (aza-epothilone, BMS-247550), BMS-310705 (a water-soluble analog of epothilone B), patupilone (epothilone B, EPO906), epothilone D (KOS-862), and ZK-EPO. Patupilone was one of the earliest epothilones entering clinical development. It has been evaluated in phase I studies using a 5-minute bolus administration weekly, administered 6 of every 9 weeks or 3 of every 4 weeks, once every 3 weeks, or daily times five as a prolonged infusion. In this issue, Rubin et al describe a dose-finding study of the weekly administration of patupilone in patients with advanced solid tumors. A total of 91 patients were treated at dose levels ranging from 0.3 to 3.6 mg/m/wk in a classic, phase I, 3 3 dose-escalation design. Dose-limiting toxicity consisting of grade 3 diarrhea was observed at both of the dose levels of 3.6 and 3.0 mg/m in the 6 weeks on and 3 weeks off dosing schedule. Therefore, the recommended dose for this schedule was set at 2.5 mg/m/wk. Because the drug-related diarrhea often peaked in severity during the fourth week of treatment, an alternative schedule (3 weeks on and 1 week off) was also studied, resulting in the same recommended dose per administration (2.5 mg/m/wk). The most common non–dose-limiting toxicities included nausea, vomiting, and fatigue. Importantly and in contrast to the results with taxanes, only one patient developed alopecia, and there was no neurologic toxicity. As with weekly taxane schedules, hematologic toxicity was limited. The pharmacokinetic data revealed that drug clearance was nonrenal and not related to body-surface area (BSA). Over the dose range studied, systemic drug exposure was dose proportional. There were three partial responses, two of which occurred in patients previously treated with taxanes, and 42% of the patients experienced disease stabilization for a median duration of 16.3 weeks. Given the absence of data on tumor growth rate before study entry, the latter data are obviously difficult to interpret. As stated, the side effects observed with weekly administration of patupilone are strikingly different from those associated with weekly administration of taxanes. Yet, it is important to note there are also major differences in adverse effect profiles between the various epothilones presently in clinical development. Ixabepilone (BMS-247550) is formulated in polyoxyethylated castor oil. Not unexpectedly, during phase I studies, hypersensitivity reactions were seen necessitating prophylactic administration of oral histamine-1 and histamine-2 blockers. Both the once every 21 days schedule and the weekly schedule resulted in doselimiting neutropenia and sensory neuropathy. Early sensory neuropathic changes were already detectable after two cycles of treatment. During the early phase II studies using the schedule of every 3 weeks, grade 3 or 4 neuropathy JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 23 NUMBER 36 DECEMBER 2