Breast Cancer Research and Treatment

, Volume 121, Issue 2, pp 261–271

Efficacy of ixabepilone in ER/PR/HER2-negative (triple-negative) breast cancer


    • Division of Hematology and Oncology, College of MedicienMayo Clinic
  • Tejal Patel
    • Division of Hematology and Oncology, College of MedicienMayo Clinic
  • Alvaro Moreno-Aspitia
    • Division of Hematology and Oncology, College of MedicienMayo Clinic

DOI: 10.1007/s10549-010-0824-0

Cite this article as:
Perez, E.A., Patel, T. & Moreno-Aspitia, A. Breast Cancer Res Treat (2010) 121: 261. doi:10.1007/s10549-010-0824-0


Patients with ER/PR/HER2-negative (triple negative) breast cancer are not candidates for hormonal therapy or HER2-targeted agents. Ongoing research is aimed at identifying and understanding the benefit of established and emerging therapies in this disease setting. Triple-negative patients may achieve early responses to anthracyclines and taxanes, but novel strategies are also eagerly sought. The epothilone B analog ixabepilone acts to stabilize microtubules and demonstrates antitumor activity in recent breast cancer studies. Herein, we have analyzed efficacy and safety data of ixabepilone specifically for the treatment of women with triple-negative disease. A retrospective analysis was completed using activity and toxicity data in the triple-negative subsets from 5 phase II studies. In addition, a prospective pooled analysis of triple-negative patients from 2 phase III trials is also reviewed. Of 2,261 patients evaluated in these trials, 556 (24.5%) had triple-negative tumors. In the neoadjuvant setting, ixabepilone produced a pathologic complete response rate in the breast of 26% in triple-negative patients (vs. 15% in the non-triple-negative population). In patients with metastatic breast cancer whose pretreatment status ranged from no prior therapy to progression on several classes of agents, overall response rates (ORR) in the phase II ixabepilone monotherapy trials ranged from 6 to 55%, comparable to rates seen in patients with non-triple-negative tumors. The combination of ixabepilone and capecitabine in the phase II study resulted in an ORR of 23% in triple-negative patients. A similar ORR (31%) was observed for a preplanned pooled analysis of triple-negative patients in the phase III trials of ixabepilone plus capecitabine. The median progression-free survival (PFS) was significantly longer for triple-negative patients treated with ixabepilone plus capecitabine (4.2 months) compared with treatment with capecitabine alone (1.7 months). No increase in toxicity was noted in the triple-negative subgroup compared with other patients. Ixabepilone shows notable antitumor activity in patients with triple-negative breast cancer when used in a variety of settings. The addition of ixabepilone to capecitabine results in an approximately twofold increase in median PFS for triple-negative patients versus capecitabine alone and responses to ixabepilone in triple-negative disease are comparable to those seen in patients with non-triple-negative tumors.


Breast cancerER/PR/HER2-negativeTriple negativeBasal-likeIxabepiloneEpothilone



Breast-conserving surgery


Epidermal growth factor receptor


Estrogen receptor


Human epidermal growth factor receptor 2


Hazard ratio




Metastatic breast cancer


Overall response rate


Overall survival


Pathologic complete response


Progression-free survival


Progesterone receptor


Invasive breast cancer represents a heterogeneous group of diseases, with significant variability in morphological and pathological features such as size, histologic grade, invasiveness, ploidy and proliferation index, hormone receptor expression, and human epidermal growth factor receptor 2 (HER2) status. Molecular characterization is supplanting older classification methods and providing a better idea of how elements of this heterogeneity influence clinical response. Such methodologies are yielding prognostic and predictive information that may improve treatment outcomes [1, 2]. Some of the most well-established of the molecular approaches have revealed a distinct subtype of breast cancer that does not express receptors commonly found on many breast cancers, including those for the estrogen receptor (ER) and progesterone receptor (PR) and HER2 (ER/PR/HER2-negative, or “triple-negative” breast cancer).

Neither hormonal therapy nor HER2-targeted therapy is indicated in triple-negative disease [3], but chemotherapy has proven efficacious against this type of breast cancer. Unfortunately, as awareness of the triple-negative subtype postdates many of the major studies of today’s standard chemotherapeutic regimens, clinicians have very little data that have specifically addressed clinical benefit in this tumor type. As such, no standard therapeutic regimen for triple-negative breast cancer has emerged, and these patients have a relatively poor prognosis.

In attempts to improve these clinical outcomes, numerous classes of agents with novel mechanisms of action are being evaluated for the management of triple-negative disease. Among these agents are a new class of antineoplastic compounds known as the epothilones; like taxanes, they stabilize microtubules and cause cell cycle arrest and apoptosis [4]. Unlike taxanes, however, many of the epothilones have been found to maintain activity in breast cancer cells with the multidrug resistance phenotype [5, 6].

Among the epothilones, ixabepilone has undergone the most extensive clinical development and was the first in the class to receive approval from the United States Food and Drug Administration. Its current indications are in metastatic breast cancer (MBC) that has relapsed following treatment with other chemotherapeutic agents: in combination with capecitabine (after taxane and anthracycline failure) or as monotherapy (after failure of a taxane, anthracycline, and capecitabine).

An ever-increasing understanding of triple-negative breast disease has spurred extensive evaluations of this subpopulation in more recent breast cancer research, such as that associated with the clinical development of ixabepilone. As a result, although relatively little data support the efficacy of most chemotherapeutics in this subgroup, a large number of patients who received ixabepilone therapy in clinical trials were classified as triple-negative, permitting retrospective subgroup analysis. Several of the phase II and III ixabepilone trials in various breast cancer settings reported encouraging outcomes for triple-negative tumors. Herein, we review these outcomes for the purposes of evaluating the efficacy and tolerability of ixabepilone for the treatment of triple-negative breast disease.

Basal-like breast cancer subtype

The triple-negative subtype can best be understood in context of its related subtype, basal-like breast cancer. Gene array profiling studies have identified five distinct subtypes of breast cancer: 2 ER-related subtypes (luminal A and B), the HER2 positive subtype, the hormone receptor-negative subtype (basal-like), and a normal breast-like subtype [7, 8]. Hierarchical clustering of expression patterns among breast cancer genes demonstrate that basal-like breast cancers are distinct from the luminal A and B subtypes and the HER2-positive group, having an expression profile closer to that of normal breast basal epithelium [7, 8]. Moreover, most (although not all) basal-like tumors lack expression of ER, PR, and HER2. The basal-like subtype may represent as many as 15% of all breast cancers [3], with a higher incidence in African American, premenopausal women, and female carriers of the mutated BRCA1 gene [9, 10].

Although basal-like tumors typically lack hormone receptor and HER2 expression, their expression of the proliferative cluster genes encoding Ki-67, proliferating cell nuclear antigen, and topoisomerase II is increased, as are genes encoding cytokeratins (CK 5/6 and 17), laminin, and integrin [11]. The epidermal growth factor receptor (EGFR) is expressed in more than 50% of cases and c-Kit in 31% [12, 13]. Morphologically, this type of breast cancer is associated with high nuclear and histologic grade, ductal histology, and an elevated mitotic index [9]. Comparative genomic hybridization studies have indicated that basal-like tumors tend to have more DNA copy number alterations (including specific gene amplifications or deletions) and known cytogenetic alterations, both of which indicate genomic instability [3, 14, 15]. In support of this idea, more single nucleotide polymorphisms are observed in basal-like breast cancer; this may contribute to the observed high rate of loss of heterozygosity and also affect tumorigenesis or progression [16]. Many basal-like tumors (40–80%) also harbor p53 mutations [11, 17]. Given the role of this tumor suppressor gene in repair of DNA damage and apoptosis, such mutations may affect prognosis or response to therapy.

Basal-like breast cancer is associated with an aggressive phenotype and a poor prognosis. The Carolina Breast Cancer Study evaluated 496 incident cases of invasive breast cancer and assessed outcomes for patients with basal-like tumors. Patients with basal-like breast cancer had a poorer disease-specific survival than those with the luminal A tumor phenotype (hazard ratio [HR], 1.8; 95% CI, 1.1–2.9; P = 0.03). Disease-specific survival in women with ER- and HER2-negative breast cancer was poorer still (HR, 3.5; 95% CI, 1.9–6.2; P < 0.001) [9]. A Cancer and Leukemia Group B study (CALGB 9342) found a similar reduction in overall survival (OS) for patients with a basal-like subtype versus those without (8.7 months vs. 12.8 months; P = 0.008), although response rate and time to treatment failure did not significantly differ [18]. This aggressive nature and poor outcome support the hypothesis that this tumor type represents a distinct biological subtype of breast cancer that may require different therapy.

Women harboring BRCA1 mutations may experience a higher incidence of basal-like tumors, and interesting phenotypic and genotypic similarities are observed between basal-like breast cancers and those arising in carriers of BRCA1 germline mutations. Mutations in BRCA1 are thought to be responsible for 21–40% of all inherited cases of breast cancer, but are rare in sporadic tumors [19]. Like basal-like tumors, BRCA1-associated breast cancers are predominantly negative for ER and HER2. Both BRCA1-associated and triple-negative tumors tend to have high nuclear grades and express EGFR, CK 5/6, and high levels of Ki-67. Both have a high degree of cytogenetic abnormalities and share certain chromosomal changes. These similarities were noted in patients with early-stage breast cancer who were followed for locoregional relapse and metastasis subsequent to surgery and radiation therapy [10]. Eight of 10 patients with BRCA1 mutations were also classified as triple-negative; in contrast, only one of seven BRCA2-mutated carriers had the triple-negative phenotype [10]. Gene expression array studies provide further support for the relationship between these 2 subtypes, in that BRCA1 mutations segregate with the basal-like subtype [8]. The defects in DNA repair and p53 mutations observed in both tumor types may contribute to other shared characteristics between these tumor types, such as enhanced chemosensitivity to selected agents and distinct drug resistance profiles [3, 17, 20].

Triple-negative breast cancer

Although the terms “triple-negative” and “basal-like” are often used interchangeably, these groups may not be identical since a small proportion of basal-like cancers appear to express ER, PR, or HER2. In one study, of the 18 tumors that were classified ER- and HER2-negative by immunohistochemistry, only 15 were dubbed “basal-like” by DNA microarray analysis. Other studies indicate that 80 to 90% of triple-negative tumors are basal-like, implying that up to 20% may differ from basal-like breast cancers in phenotype (and presumably clinical response).

Therefore, in all probability, triple-negative breast cancers comprise a subset of basal-like tumors. Pathologically, triple-negative tumors tend to have a high mitotic index, central necrosis, a large percentage of apoptotic cells, a pushing margin of invasion, and a stromal lymphocytic response. Other common features are shown in Table 1 [3, 21, 22].
Table 1

Characteristics of ER/PR/HER2-negative breast cancer [3, 21, 22]

• Lack of detectable expression of ER, PR, and HER2

• Basal-like phenotype

  – Characteristic molecular profile on gene expression array

  – Expression of cytokeratin 5/6 and vimentin

  – Increased expression of EGFR and c-Kit

  – Increased frequency of p53 mutations

• Histopathologic findings

  – High nuclear grade, high histologic grade, and increased mitotic index

  – Central necrosis, pushing margin of invasion, and stromal lymphocytic response

• Subtype more commonly seen in

  – Carriers of BRCA1 mutations

  – African American premenopausal women

  – Very young women (<45 years)

• Clinical differences compared with non-basal subtypes

  – Increased rate of local relapse

  – Overall poorer prognosis, with shorter distant metastasis-free survival and OS

  – Increased frequency of visceral and brain metastases compared with bony metastases

EGFR epidermal growth factor receptor, ER estrogen receptor, HER2 human epidermal growth factor receptor 2, OS overall survival, PR progesterone receptor

Women with triple-negative tumors tend to experience disease onset at an earlier age [23], more frequent local relapse [24], a higher proportion of visceral compared with bony metastases [23, 24], and an increased frequency of brain metastases [25] compared with women with other subtypes. Recent data also suggest survival may be shorter. Haffty and colleagues reviewed data from 482 women with early-stage breast cancer who were conservatively managed (treated with breast conservation surgery followed by radiation, with or without systemic therapy), 117 (24.2%) of whom had triple-negative disease. Despite use of adjuvant therapy, distant metastasis-free survival at 5 years was significantly shorter in the triple-negative group than with other subtypes (68% vs. 83%; P = 0.002) [10]. Despite identical local relapse rates for triple-negative tumors and other subtypes (83% for both groups), triple-negative status was an independent predictor of distant metastasis (HR, 2.15; 95% CI, 1.31–3.53; P = 0.002) and cause-specific survival (HR, 1.79; 95% CI, 1.03–3.22; P = 0.047). Similarly, a review of nearly 2,000 cases of invasive breast cancer, 16% of which were triple-negative, associated this phenotype with a poorer OS (P < 0.0001) and disease-free interval (P = 0.0002) [22]. The worse prognosis may reflect underlying differences in the biology and progression of triple-negative disease compared with other subtypes, as suggested by the increased proportion of visceral and cerebral metastases.

Response of triple-negative disease to traditional chemotherapy

Despite the current lack of options for triple-negative breast cancer, the data clearly demonstrate that some of these patients may respond well to he treatment with chemotherapy. As shown in Table 2, regimens incorporating anthracyclines, taxanes, or a combination of these agents were as effective if not more effective in triple-negative tumors as they were with non-basal-like breast cancer subtypes. In some studies, patients with basal-like tumors had a significantly greater clinical response rate than did non-basal-like cancers [26, 27]. Many BRCA1-mutated breast cancers (which, as noted above, are often triple-negative) also exhibit high in vitro chemosensitivity to agents that induce DNA cross-links and double-strand breaks [28, 29].
Table 2

Neoadjuvant and metastatic chemotherapy for basal-like or ER/PR/HER2-negative breast cancer

Neoadjuvant regimen



Basal-like tumorsa

Non-basal-like tumors

Neoadjuvant therapy

Anthracycline- and paclitaxel-based combination chemotherapy [26]


pCR: 45%


pCR: 6%b


Anthracycline-based chemotherapy [9]


ORR: 71%

OS: 75%

ORR: 51–65%c

OS: 52–87%

Taxane-based chemotherapy with or without anthracycline [27]


pCR: 25%

5-y OS: 66%

pCR: 11%

5-y OS: 83%

Metastatic regimen



Basal-like or ER/PR/HER2-negative tumors

Other subtypes

Metastatic therapy

Nab-paclitaxel plus gemcitabine [58]


ORR: 77%


Gemcitabine plus carboplatin [59]


ORR: 20%

N = 15d

ORR: 34%

N = 47f


ORR: 66%

N = 27 g

Paclitaxel with or without bevacizumab, respectively (ECOG 2100) [60]


PFS: 8.8 vs. 4.6 mo [HR, .53]; N = 233d

PFS: 14.4 vs. 8.0 mo [HR, .54]; N = 289e

ER estrogen receptor, HER2 human epidermal growth factor receptor 2, HR hazard ratio, mo months, NR not reported, ORR overall response rate, OS overall survival, pCR pathologic complete response, PFS progression-free survival, PR progesterone receptor

aBasal-like or HER2 subtype

bLuminal subtype

cRange for 4 other breast cancer subtypes

dER-, PR-, and HER2-negative

eER- and PR-positive

fAll HER2-negative, including the 15 ER/PR/HER2 patients

gHER2-positive; these patients also received trastuzumab

Paradoxically, increased response to chemotherapy does not always translate into improved survival. The explanation for this contradiction is likely multifactorial and includes presence of residual disease, biology of the disease, and limited availability of other therapies. Among women with breast tumors in general, those who receive neoadjuvant anthracycline and/or taxane-based chemotherapy and achieve a pathologic complete response (pCR) have a better long-term outcome. A study from Peking University People’s Hospital reported that patients with triple-negative breast cancer had a higher pathological response to neoadjuvant anthracycline-based treatment (38 vs. 12% for those with non-triple negative disease; P = 0.002) [30]. If pCR was achieved, patients with triple-negative cancer and patients with non-triple-negative breast cancer had similar survival (P = 0.497). However, patients with triple-negative tumors who did not achieve pCR had significantly worse survival compared with non-triple negative breast cancer (P < 0.05). The triple-negative cohort of patients had an overall decreased disease-free survival rate (P = 0.004). Similar results have been observed in other trials.

Novel systemic therapies for triple-negative breast cancer

Since triple-negative tumors are thought to be especially sensitive to DNA cross-linking agents, patients with triple-negative disease have been undergoing therapy with platinum-based chemotherapy in both the neo/adjuvant and advanced breast cancer settings. To date, clinical trials that have examined this combination have enrolled small numbers of triple-negative patients; thus, it is difficult to ascertain the regimen’s efficacy in this population. Further investigation is needed to determine whether this type of therapy will be particularly effective for this subgroup of patients.

In addition, some reports suggest that triple-negative tumors commonly express high levels of poly (ADP-ribose) polymerase 1 (PARP-1), a DNA repair enzyme that is the molecular target of the PARP inhibitors [31]. The investigational PARP inhibitor BSI-201 significantly improved overall response rate (ORR) (48 vs. 16%, P = 0.002), progression-free survival (PFS) (6.9 vs. 3.3 months, HR, 0.342, P < 0.0001), and median OS (9.2 vs. 5.7 months, HR, 0.348, P = 0.0005) when added to carboplatin and gemcitabine in a randomized phase II study in patients with metastatic triple-negative disease [31]. A follow-up phase III trial comparing these regimens has recently completed accrual (NCT00938652), and clinical trials with other PARP1 inhibitors, such as AGO14699, AZD-2281 and ABT-888, are ongoing or in development.

A high EGFR expression and high rate of proliferation may mean that patients with basal-like breast carcinomas will benefit from treatment with EGFR inhibitors like cetuximab [13]. Although triple-negative tumor cell lines appear less sensitive to EGFR inhibitors than HER2-positive lines, EGFR inhibitors may enhance chemosensitivity [32]. The utility of this approach is being evaluated in an ongoing clinical trial of the EGFR tyrosine kinase inhibitor erlotinib in patients with EGFR-positive, triple-negative breast disease (NCT00491816).

Studies in triple-negative patients are also exploring other targeted therapies, such as the Src family kinase inhibitor dasatinib [33], the angiogenesis inhibitor bevacizumab [34], and MEK inhibitors [35]. Although these compounds are in earlier stages of clinical evaluation for triple-negative breast cancer, they may be useful in combination regimens if they are found to be active in this setting.


The backbone of therapy for triple-negative disease remains chemotherapy, particularly in the metastatic setting. Because continued therapy with taxanes and anthracyclines is shown to encourage tumor resistance to these agents, cytotoxic approaches that are fundamentally different from these classes of agents are eagerly being explored in ongoing breast cancer research. A product of such research, the epothilone ixabepilone, has efficacy in triple-negative breast cancer and is hence the subject of the remainder of this review.

Epothilones are microtubule-stabilizing agents that are thought to bind to the paclitaxel binding site on the β-tubulin side of the microtubule interior [36]. The epothilones, however, bind to β-tubulin in a different manner from the taxanes, and the two classes are structurally distinct [37]. Notably, epothilone B has demonstrated a potential for treatment of taxane-resistant tumors, including those that are refractory to multiple classes of agents.

Ixabepilone, a semisynthetic analog of epothilone B with improved solubility and stability over the natural compound, was 2.5- and 10-fold more potent than paclitaxel at inducing tubulin polymerization and promoting cell cycle arrest, respectively, in cell lines from various tumor types [4, 5]. In vitro, ixabepilone is cytotoxic in cell lines that express high amounts of the drug efflux pump P-glycoprotein or the taxane-refractory tubulin isotype βIII-tubulin [46, 38].

Efficacy of ixabepilone in triple-negative breast cancer

In light of its encouraging preclinical antitumor activity in drug-resistant breast cancer models, ixabepilone was subsequently evaluated for the treatment of women with breast cancer at various stages. Many of these patients had heavily pretreated metastatic disease, and a portion of these tumors were triple-negative. The present review discusses 5 phase II and 2 phase III breast cancer studies. The phase II studies include a neoadjuvant trial of ixabepilone monotherapy, three monotherapy trials in MBC patients with differing pretreatment histories, and 1 trial of ixabepilone and capecitabine in women who had received prior anthracycline and taxane therapy. The 2 phase III studies further evaluated this combination in patients pretreated with an anthracycline and a taxane (Tables 3, 4, 5).
Table 3

Phase II ixabepilone neoadjuvant monotherapya


ER/PR/HER2-negative tumors (N = 42)

Non-ER/PR/HER2-negative tumors (N = 119)

Breast cancer ≥3 cm not amenable to primary BCS

ORRb: 64%

pCRB: 26%

pCRBL: 19%

ORRb: 60%

pCRB: 15%

pCRBL: 8%

BCS breast-conserving surgery, ER estrogen receptor, HER2 human epidermal growth factor receptor 2, PR progesterone receptor, ORR overall response rate, pCRB pathologic complete response in the breast, pCRBL pathologic complete response in the breast and lymph nodes

aIxabepilone 40 mg/m2 3-h infusion on day 1 every 3 weeks

bInvestigator-assessed ORR

Table 4

Phase II Ixabepilone Monotherapy in MBCa


ER/PR/HER2-negative tumors

Non-ER/PR/HER2-negative tumors

Resistant to anthracycline, taxane, and capecitabine

N = 42

ORRb: 17%

PFS: 2.7 (1.5–5.9 mo)

N = 119

ORRb: 19%

PFS: 3.6 (2.9–4.4 mo)

Resistant to taxane

N = 18

ORR: 6%

PFS: 1.6 (1.3–2.3 mo)

N = 31

ORR: 16%

PFS: 2.5 (1.7–4.0 mo)

Prior therapy with an anthracycline

N = 11

ORR: 55%

PFS: 4.6 (2.8–9.3 mo)

N = 54

ORR: 39%

PFS: 5.7 (4.1–7.7 mo)

aIxabepilone 40 mg/m2 3-h infusion on day 1 every 3 weeks

bInvestigator-assessed ORRs

ER estrogen receptor, HER2 human epidermal growth factor receptor 2, MBC metastatic breast cancer, ORR overall response rate, PFS progression-free survival, PR progesterone receptor

Table 5

Phase II and III ixabepilone and capecitabine combination therapy

Phase II population

Ixabepilone plus capecitabinea

ER/PR/HER2-negative tumors

Non–ER/PR/HER2-negative tumors

Prior therapy with anthracycline and taxane

N = 22 ORRb: 23%

N = 28 ORRb: 32%

Phase III Population

ER/PR/HER2-Negative Tumors

Ixabepilone plus capecitabinea

Capecitabine monotherapyc

Resistant to anthracycline and taxane (BMS 046)

N = 191

PFS: 4.2 (3.6–4.4 mo)

ORR: 31%

N = 208

PFS: 1.7 (1.5–2.4 mo)

ORR: 15%

Pretreated with anthracycline and taxane (BMS 048)

aIxabepilone 40 mg/m2 as 3-h infusion on day 1 every 3 weeks; capecitabine 2000 mg/m2 on days 1–14 every 3 weeks

bInvestigator-assessed ORRs

cCapecitabine 2500 mg/m2 on days 1–14 every 3 weeks

ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2, ORR overall response rate, PFS progression-free survival

Ixabepilone as monotherapy

A multicenter phase II trial evaluated ixabepilone as neoadjuvant therapy for patients with invasive breast adenocarcinoma who were not eligible for breast-conserving surgery (BCS) [39]. Women enrolled in the trial had histologically confirmed T2–4, N0–3 tumors. Treatment consisted of single-agent ixabepilone 40 mg/m2 administered as a 3-h infusion on day 1, given every 3 weeks for four cycles. Ixabepilone demonstrated substantial tumor activity in this study, with a pCRB rate of 18% and pCR rate for breast and lymph nodes (pCRBL) of 11%, comparable to rates observed in other studies of neoadjuvant therapy with docetaxel (pCRB 10% and pCRBL 7%), paclitaxel (9%),or doxorubicin/cyclophosphamide (13%) [4045]. Specifically, ER/PR/HER2-negative tumors had pCRB rate of 26%, compared with 10.6% in ER-positive/HER2-negative patients [46]. Notably, 33% of patients in the neoadjuvant ixabepilone study were able to undergo BCS following treatment with four cycles of ixabepilone.

A retrospective analysis of this study revealed that basal-like and triple-negative tumors were more likely to express high baseline levels of βIII-tubulin, a β-tubulin isotype linked to reduced efficacy of taxanes and poor response to taxane-based therapy in breast and other cancers. Receiver operating characteristics analysis suggested that in the overall study population, high βIII-tubulin expression levels might be predictive of response to ixabepilone (area under the curve = 0.66) [47].

The activity of ixabepilone monotherapy in the metastatic setting was explored in 3 phase II trials, with patients in each study having different levels of prior therapy and resistance to chemotherapy. Patient populations ranged from taxane-naïve patients who had been treated with anthracyclines to heavily pretreated patients whose disease had progressed after treatment with an anthracycline, a taxane, and capecitabine. Of note, all patients who were dubbed as having “taxane resistance” progressed while receiving therapy, within 6 months of receiving adjuvant taxane therapy, or within 8 weeks or 4 months of receiving taxane therapy for metastatic disease (depending on the study).

Ixabepilone was administered at a dose of 40 mg/m2 as a 3-h infusion on day 1 every 3 weeks. The ORR in these three trials ranged from 12 to 42%, depending on the pretreatment status of the patient population (Table 4). As expected, women who had only prior anthracycline therapy had a higher response rate (42%) compared with those who had already received three other classes of agents (12–18%). PFS in these three studies ranged between 1.6 and 5.7 months, again depending on the extent of prior chemotherapy or resistance.

To compare the activity of ixabepilone monotherapy in triple-negative breast cancer and in unselected subtypes, a retrospective analysis was conducted using data from the four ixabepilone monotherapy studies [48]. This analysis included a total of 288 ixabepilone-treated patients, 113 (39%) of whom had triple-negative tumors. ORR with single-agent ixabepilone ranged from 6 to 55% in triple-negative patients. As seen in Tables 3 and 4, the response rates within any given study were comparable between patients with triple-negative disease and those with hormone receptor- or HER2-positive disease. In the neoadjuvant trial, ORR values were again similar between the triple-negative and non-triple-negative groups, and pCR rates were slightly higher for triple-negative patients. Although numbers of triple-negative patients in these trials were small, every study reported activity for ixabepilone in this patient population, even when patients had received prior therapy with as many as three separate classes of chemotherapy (Table 4).

Ixabepilone in combination with capecitabine

Since capecitabine is widely used as second-line therapy in women whose tumors are resistant to anthracyclines and taxanes, three trials have evaluated ixabepilone in combination with capecitabine. The phase I/II study employed 21-day schedules: ixabepilone 40 mg/m2 (administered as either a 3-h infusion on day 1) plus capecitabine 1,650 to 2,000 mg/m2 (daily, dose divided, on days 1–14) or ixabepilone 8 to 10 mg/m2 (1-h infusion on days 1–3) plus capecitabine 1,650 mg/m2 (daily, dose divided, on days 1–14) [49]. Patients in this trial had received prior therapy with an anthracycline and a taxane. The ORR for the 62 patients treated for the phase II tumor response portion was 30%, median duration of response was 6.9 months, and median PFS was 3.8 months.

On the basis of the encouraging activity reported in this phase II study, two subsequent randomized phase III trials (BMS 046 and BMS 048) evaluated the efficacy and safety of the ixabepilone/capecitabine combination versus capecitabine alone. In BMS 046, Thomas and colleagues evaluated 752 patients with metastatic disease pretreated with or resistant to an anthracycline and resistant to a taxane; in BMS 048, Hortobagyi and colleagues evaluated the same combination in 1,221 patients that had been pretreated with taxanes and anthracyclines. In both trials, ixabepilone and capecitabine were administered at the same dose and schedule as in the phase II study, with ixabepilone given as a 3-h infusion on day 1. The patient populations for these two trials differed only with respect to their pretreatment status: patients in BMS 046 all met strict resistance criteria; patients in BMS 048 had been pretreated with anthracyclines and taxanes, but only about half met the resistance criteria utilized in BMS 046 [50, 51].

Prospective pooled analyses of patient data from both trials were undertaken to evaluate ORR, PFS, and OS within several pre-defined subgroups, one of which was triple-negative disease [52]. The two trials investigated a total of 443 patients with triple-negative metastatic disease, one of the largest pools of such patients reported to date. All patients in BMS 046 and those with measurable disease in BMS 048 were used to calculate ORR and PFS data (n = 399), whereas all randomized patients in both trials were used to calculate OS [52].

Adding ixabepilone to capecitabine improved responses in triple-negative patients over those seen with single-agent capecitabine, with respect to ORR (31 vs. 15%) and PFS (4.2 vs. 1.7 months; HR, 0.63 [0.52–0.77]). Combination therapy was also associated with a trend toward OS (10.3 vs. 9.0 months, respectively; HR, 0.87 [0.71–1.07]) [52]. These data suggest a clinical benefit associated with addition of ixabepilone to capecitabine in women with triple-negative MBC that has progressed after prior treatment with an anthracycline and a taxane. Notably, the results are not very different than those noted in a more recent combination study of the PARP1 inhibitor BSI-201 plus gemcitabine/carboplatin [31].

Ixabepilone safety

Subset analyses of patients with triple-negative breast cancer indicate that the frequency and severity of ixabepilone-related toxicities were not significantly different from those seen in patients with non-triple-negative tumors. This was true for both hematologic and nonhematologic grade 3/4 adverse events [48, 53], and whether treatment was administered in the metastatic or neoadjuvant setting. Similarly, the incidence of treatment-related discontinuations was not higher for the triple-negative subset (15%) compared with the total population (27%) [53]. These findings suggest that no significant differences exist between the triple-negative subset and the overall breast cancer population with respect to ixabepilone toxicity or treatment discontinuations. Patients receiving ixabepilone do not require corticosteroid premedication except in cases where the patient has experienced a previous hypersensitivity reaction to ixabepilone [54].


The most frequently occurring grade 3/4 treatment-related adverse events in the monotherapy trials were neutropenia (51 to 59%) and leukopenia (37 to 49%). For all ixabepilone studies, severe thrombocytopenia (grade 3/4) occurred in <10% of patients.

Sensory neuropathy is a common adverse event with microtubule-stabilizing agents [55, 56], and the incidence of sensory neuropathy observed with ixabepilone was fairly similar to those reported for the taxanes. In breast cancer studies, peripheral neuropathy was mainly sensory and cumulative. Most of the phase II trials of ixabepilone monotherapy enrolled heavily pretreated patients with metastatic disease; treatment-related grade 1/2 and 3/4 sensory neuropathy developed in 48 and 13% of patients, respectively. Other notable non-hematologic grade 3/4 adverse events in the monotherapy trials included fatigue (6–27%) and myalgia (7–10%).

Combination therapy

The incidence of neutropenia and leukopenia were 68 and 57%, respectively, in the ixabepilone/capecitabine combination arm of the BMS 046 study. Grade 3 and 4 sensory neuropathy (21 vs. 0%), fatigue (9 vs. 3%), and neutropenia (68 vs. 11%) occurred more frequently with combination therapy than with single-agent capecitabine. In patients with grade 2 or higher liver dysfunction, there was also an increased rate of death as a result of toxicity (3% vs. 1%). Thus, the combination regimen should be avoided in patients with aspartate aminotransferase and alanine aminotransferase values higher than 2.5 upper limits of normal or bilirubin above upper limit of normal [54].

Importantly, in the majority of patients, peripheral neuropathy was responsive to dose reduction. It was also observed to be reversible to baseline grade or below within a matter of weeks, regardless of whether ixabepilone was administered as monotherapy (median resolution time of 5.4 weeks) or in combination with capecitabine (median resolution time of 6.0 to 6.2 weeks) [57].


Utility in triple-negative disease is largely unproven for most commonly used chemotherapeutics. In contrast, a relatively large pool of data demonstrates that ixabepilone is effective in this population as monotherapy or in combination with capecitabine. The seven trials presented herein indicate activity in triple-negative breast cancer across multiple disease settings, from neoadjuvant to refractory MBC. Ixabepilone elicited similar or slightly better clinical outcomes in patients with triple-negative disease compared with those with receptor-positive breast cancer, with comparable toxicity. Regarding the ixabepilone plus capecitabine doublet, results from the prospective analysis of the pooled phase III trial data are among the few to demonstrate a significant improvement in PFS for patients with triple-negative MBC. Taken as a whole, these results indicate that ixabepilone appears to be a viable option for this hard-to-treat patient population.

Ixabepilone is now undergoing further evaluation in combination with various novel targeted agents in an attempt to augment its overall efficacy, and/or in head-to-head trials with taxanes. Some of these trials are being conducted exclusively in triple-negative patients. A planned single-arm phase II trial will investigate ixabepilone (40 mg/m2 every 3 weeks) plus sunitinib (37.5 mg daily) as first-line therapy in triple-negative patients (BMS 168), and an ongoing randomized phase II trial is evaluating ixabepilone (same dose and schedule as above) as monotherapy or in combination with cetuximab (400 mg/m2 initial dose followed by 250 mg/m2 weekly) as first-line therapy for metastatic triple-negative disease (BMS 139; NCT00633464). The combination of ixabepilone and cetuximab will also be studied in the neoadjuvant setting in an upcoming trial (BMS 157).

Two ixabepilone trials in triple-negative patients undergoing adjuvant therapy are also underway. The phase III PACS-08 trial is comparing disease-free survival in patients who receive standard FEC (fluorouracil, epirubicin, cyclophosphamide 100 mg/m2 every 3 weeks for three cycles) followed by ixabepilone (same dose and schedule as above, three cycles) compared with FEC followed by three cycles of single-agent docetaxel (BMS 087; NCT00630032). The randomized phase III TITAN trial is comparing ixabepilone (same dose and schedule as above) with weekly paclitaxel (80 mg/m2) following standard anthracycline/cyclophosphamide therapy (doxorubicin 60 mg/m2 plus cyclophosphamide 600 mg/m2; BMS 181; NCT00789581).

Compared with other subgroups of breast cancer, triple-negative breast cancers have a higher degree of homogeneity in their RNA expression profiles, suggesting that these tumors may share a common pathogenesis; as such, they might also have a predictable response to effective therapeutic agents [7, 8]. Indeed, a genomic analysis of the neoadjuvant ixabepilone study discussed herein suggested that βIII-tubulin, a molecular marker of taxane resistance, was more commonly overexpressed in triple-negative, basal-like, and HER2-positive tumors than in luminal tumors. This isotype of β-tubulin may also serve as a predictive marker for pCR following neoadjuvant ixabepilone therapy [47]. Such pharmacogenomic studies are becoming standard in oncology research, and in time the results may yield a wealth of information regarding the optimal treatment for various subtypes of breast cancer.

Concurrently with pharmacogenomic research to reveal more about the biology of triple-negative disease, recent drug studies employing molecular approaches to therapy with rationally designed agents have yielded encouraging results in patients with triple-negative tumors. Besides evaluations of novel cytotoxic agents like ixabepilone, trials of several promising targeted therapies have yielded promising data: inhibitors of PARP, various tyrosine kinases, and key players in the angiogenesis pathway. In addition, more and more breast cancer trials are being designed to enrich the patient pool for those with triple-negative disease. It is hoped that these focused research efforts will continue to improve our understanding of the molecular biology and appropriate treatment for this challenging form of breast cancer.


As the lead author, Dr. Perez takes full responsibility for the content of this publication and confirms that it reflects the viewpoint and medical expertise of herself and her colleagues. The authors also wish to acknowledge StemScientific, funded by Bristol-Myers Squibb, for providing writing and editing support. Bristol-Myers Squibb did not influence the content of the manuscript, nor did the authors receive financial compensation for authoring the manuscript.

Copyright information

© Springer Science+Business Media, LLC. 2010