In a landmark article published in the May 1, 2001, issue of Clinical Cancer Research, Lee and colleagues reported the original preclinical studies demonstrating anticancer activity of BMS-247550 (ixabepilone) against taxane-sensitive and taxane-resistant cancers. Subsequent clinical trials established the clinical efficacy of ixabepilone, leading to its regulatory approval for the treatment of drug-resistant metastatic or locally advanced breast cancers. Clin Cancer Res; 21(6); 1237–9. ©2015 AACR.
See related article by Lee et al., Clin Cancer Res 2001;7(5) May 2001;1429–37
Microtubules are an ideal target for anticancer therapy due to their essential role in mitosis. The Vinca alkaloids and taxanes are the two major classes of microtubule-targeted anticancer drugs that have significantly improved the treatment of wide-ranging malignancies. The introduction of paclitaxel in the early 1990s and demonstration of its impressive clinical activity against solid tumors was followed by the realization for a further need of improvements in both safety and efficacy. This recognition was primarily due to two factors: resistance and toxicity, which limited the clinical utility of existing antimicrotubule agents such as paclitaxel. Indeed, a majority of initially responsive patients developed resistance, and some solid tumors, such as colorectal cancer and melanoma, were inherently resistant to treatment. Moreover, although side effects such as neutropenia were clinically manageable, the occurrence of severe, potentially debilitating peripheral neurotoxicity was a major concern. As a result, a significant effort was made to develop next-generation microtubule-targeted anticancer drugs that could be active against taxane-resistant cancers, have enhanced oral absorption properties, and show improved toxicity profiles.
Multiple compounds with microtubule-targeting activity have been actively pursued as alternatives to taxanes, and the class of epothilones was among those that showed particularly potent microtubule-targeting activity. In 1992, Höfle and colleagues (1) first isolated epothilones from the myxobacterium Sorangium cellulosum. Epothilones are 16-membered ring macrolides that were originally shown to have both antifungal and cytotoxic activities. Significant interest in epothilones was generated after the report by Bollag and colleagues (2) that epothilones have a paclitaxel-like mechanism of action and had significant cytotoxicity in multiple drug-resistant cancer cell lines. In the late 1990s, in collaboration with Hans Reichenbach and Gerhard Höfle (Gesellschaft für Biotechnologische Forschung), Bristol-Myers Squibb (BMS) further developed epothilones for clinical use as anticancer drugs (3). During an initial lead optimization, two major limitations of the original epothilones had to be overcome, namely metabolic instability and chemical synthesis of this complex natural compound.
Epothilones were subsequently shown to possess impressive in vitro activity against various paclitaxel-resistant cancer cell lines; however, they had very modest in vivo antitumor activity. This was found to be due to their poor metabolic stability because the lactone ring of epothilones is particularly prone to cleavage by esterases. As a result of metabolic instability, these compounds had poor pharmacokinetic properties and limited anticancer efficacy in vivo. Eventually, hundreds of semisynthetic analogues of epothilones were developed and tested for anticancer efficacy in vitro and in vivo. These efforts led to the identification of BMS-247550 (ixabepilone), a lactam analogue of epithilone B.
Francis Lee and colleagues reported on their preclinical study of ixabepilone as an anticancer drug in the May 2001 issue of Clinical Cancer Research (4). In this work, it was first established that ixabepilone had cytotoxicity against both sensitive and resistant cancer cell lines in vitro. The potency was comparable with the parental epothilones and significantly higher than that observed for paclitaxel. Importantly, well-characterized cell lines with either ABCB1 (P-glycoprotein) overexpression or β-tubulin mutations that contributed to taxane-resistant phenotypes were found to retain sensitivity to ixabepilone. Moreover, the agent was found to affect both tubulin polymerization and cell-cycle progression. These in vitro studies established that ixabepilone retained the microtubule-targeting activity of parental epothilones, and that it had significant activity against taxane-resistant cell lines.
The authors subsequently used a wide panel of in vivo xenograft mouse models to examine the anticancer efficacy of ixabepilone. In these studies, they worked with paclitaxel-sensitive xenograft models (A2780, HCT116, and LS174T cell lines) and showed significant activity with ixabepilone that was comparable with the effects of paclitaxel. As compared with other epothilones, which did not have any in vivo activity, these results clearly showed that ixabepilone was active in vivo. More importantly, the authors used three resistant, patient-derived xenograft models (ovarian, breast, and pancreatic cancer) and showed very impressive anticancer activity of ixabepilone with each model. Similar results were obtained in cell lines (A2780Tax and HCT116/VM46), with β-tubulin mutations and ABCB1 overexpression contributing to paclitaxel resistance. In an interesting conclusion, the authors also described the anticancer effects of orally administered ixabepilone in a paclitaxel-resistant and -sensitive xenograft model. Overall, these studies suggested that, as compared with taxanes, ixabepilone had activity against resistant tumors and could possibly be delivered by oral administration, two potentially significant improvements over other microtubule-targeting chemotherapeutics.
With these results, Lee and colleagues (4) provided the incentive for a large-scale clinical development program of ixabepilone involving a variety of schedules, infusion durations, and combinations. In phase I trials, responses were observed in a wide range of tumors, and the main toxicities reported were neutropenia, peripheral sensory neuropathy, and hypersensitivity reactions (5). The main clinical benefits and subsequent development of ixabepilone have been in metastatic breast cancer. In several phase II trials, single-agent ixabepilone showed activity as a first-line treatment in patients with metastatic breast cancer and in those resistant to an anthracycline and taxane (6). The combination of ixabepilone and capecitabine was also extensively explored in phase I and II trials and found to be both safe and active in women with metastatic breast cancer previously treated with an anthracycline and a taxane. This combination was subsequently compared with single-agent capecitabine in the same disease setting in two large randomized phase III trials. Based on these studies, the FDA approved the use of ixabepilone in combination with capecitabine in metastatic or locally advanced breast cancer resistant to treatment with an anthracycline and a taxane, and as monotherapy in tumors that are also resistant to capecitabine (7). Pooled subset analyses of the two phase III studies have suggested that ixabepilone plus capecitabine shows superior overall survival compared with capecitabine alone in patients with Karnofsky performance status scores of 70 to 80 (8). It should be noted, however, that the European Medicines Agency did not approve ixabepilone based on the consideration that its therapeutic benefits do not outweigh the risk of significant toxicities, in particular neuropathy, despite post hoc support for a positive benefit–risk ratio based on Q-TWiST analysis (9).
In the context of observed side-effect profiles, it is of interest that the poor aqueous solubility of ixabepilone resulted in a formulation strategy similar to that used earlier for paclitaxel, as a vehicle containing large amounts of polyoxyethylated castor oil (Kolliphor EL, previously Cremophor EL) and ethanol. This formulation requires the use of relatively prolonged infusion times and has been associated with acute hypersensitivity reactions that require premedication with antihistamines. In addition, this formulation may limit drug penetration in tumor cells, can act as a perpetrator in pharmacokinetic drug interactions, and could directly or indirectly contribute to treatment-related peripheral neuropathy (10). This latter possibility would be consistent with clinical data on the comparative side-effect profiles of the two FDA-approved paclitaxel formulations, which suggest that, at equimolar doses, the formulation containing Cremophor EL is substantially more neurotoxic. In light of this prior knowledge, it is unfortunate that the development by BMS of a Cremophor EL-free formulation of ixabepilone that was found in a phase I trial to cause no hypersensitivity reactions or grade ≥3 peripheral sensory neuropathy was discontinued (11).
Since the original FDA approval of ixabepilone, significant effort has been focused on the identification of potentially useful pharmacodynamic markers to monitor drug effects on tumors and normal tissue, as well as on the development of predictive markers to allow more tailored therapy for patients (12). Unfortunately, efforts to validate the predictive value of such putative biomarkers for ixabepilone have not yet been successful. An unexplored area of research for ixabepilone relates to the study of population pharmacokinetics as a means to individualize dose and schedule of administration. This discipline seeks to identify the measurable (patho)physiologic factors that cause changes in the dose–concentration relationship and the extent of these alterations so that, if these factors are associated with clinically significant shifts in the therapeutic index, the dosage can be appropriately modified in an individual patient. It is obvious that a careful collection of data during the development of drugs and subsequent analyses could be helpful. Unfortunately, important information is often lost by failing to analyze these data or because the relevant samples or data were never collected, as in the case of ixabepilone. Another currently unexplored opportunity for therapy refinement with ixabepilone is to evaluate the contribution of germline variants in genes with a confirmed or suspected role in the highly unpredictable pharmacokinetic profile of ixabepilone. The identification of genetic and environmental factors associated with interindividual variability in the disposition of highly metabolized drugs, such as ixabepilone, is potentially vital to predicting or eventually adapting appropriate individualized doses.
The broad clinical development program of ixabepilone has also involved exploration of single-agent use across the spectrum of breast cancer stages as well as combinations with other cytotoxic regimens, including carboplatin, liposomal doxorubicin, and epirubicin, and with molecularly targeted agents such as bevacizumab, cetuximab, sorafenib, dasatinib, and trastuzumab. These studies have demonstrated that the potential benefits of ixabepilone in patients with breast cancer are maintained across the conventional molecular subgroups, including in women with triple-negative disease. Moreover, other structurally distinct epothilones have been developed in recent years and are being evaluated in a variety of other advanced solid tumors in addition to breast cancer as monotherapies and in combination with other agents. These studies indicate that, despite their relative minor structural differences, the individual epothilones show distinct safety and efficacy profiles across a range of malignant diseases and may ultimately have nononcologic indications as well (12).
The original report from Lee and colleagues (4) in Clinical Cancer Research serves as an example in which preclinical efficacy data were predictive of future clinical success, and it has provided the basis for the subsequent development of multiple cytotoxic agents targeting microtubules. Indeed, the FDA approval of ixabepilone for treatment of breast cancer shows that an important role remains for the development of cytotoxic drugs in cancer therapy, especially for patients with advanced disease (5).
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Conception and design: N. Pabla, A. Sparreboom
Writing, review, and/or revision of the manuscript: N. Pabla, A. Sparreboom
A. Sparreboom was supported by the NIH under award number R01CA151633.
- Received January 12, 2015.
- Accepted January 13, 2015.
- ©2015 American Association for Cancer Research.