Treatment of Metastatic Breast Cancer: Chemotherapy
Sing-Huang Tan
Antonio C. Wolff
INTRODUCTION
In 2013, the United States had an estimated 232,340 new cases of breast cancer and 39,620 breast cancer-related deaths occurred in women (1). Survival has continually improved over the last six decades (1). In many developing countries, the incidence of breast cancer is rising sharply due to changes in lifestyle, reproductive factors and increased life expectancy (2). More than half of incident cases occur in the developing world, with the percentage of deaths in these countries double that in high-income countries (2). The treatment of advanced breast cancer is often more resource intensive and associated with worse outcomes, further taxing patient populations that may have fewer resources.
Access to screening programs has resulted in a higher proportion of women being diagnosed with earlier stages of disease which are imminently curable. However, approximately 6% to 10% of breast cancers are metastatic at presentation and systemic recurrence occurs in about 30% of early breast cancer cases, many beyond the first 5 years (3, 4). Metastatic breast cancer (MBC) patients have a median survival of 2 to 3 years. Patients have a diverse clinical behavior driven by histological and molecular subtype (and increasingly by therapy) though historically fewer than 5% survive beyond 10 to 15 years (5, 6 and 7). Patients with very limited disease, particularly if limited to soft tissue and/or bone, were the ones most likely to have long periods of progression-free survival (6). Newer therapies, particularly those targeting the estrogen receptor (ER) and the human epidermal growth factor receptor 2 (HER2) may have favorably altered the natural history of these breast cancer phenotypes (8).
CARE OF PATIENTS WITH METASTATIC DISEASE
MBC is by definition an incurable disease and most patients with metastatic disease will likely die from their disease. At the same time, the growing number of therapeutic options has changed the outcome for many patients, particularly for those with ER-positive and/or HER2-positive disease whose treatment backbone will center around the use of anti-estrogens and HER2-targeted therapies, respectively. Despite the lack of therapies targeting specific pathways, it is now well accepted that not all triple-negative breast cancers (TNBC) are the same, and there is a growing understanding that this nomenclature encompasses a spectrum of tumors with diverse clinical behavior (9). Patients with TNBC often will have durable and clinically meaningful
responses to conventional chemotherapy regimens, often given as single agents, and many are able to have prolonged progression-free survival (PFS) intervals.
responses to conventional chemotherapy regimens, often given as single agents, and many are able to have prolonged progression-free survival (PFS) intervals.
Goals of Therapy
It is critical for patients and their doctors to establish early on the overall goals of therapy, which will primarily center around symptom control, prevention of complications, and in many cases prolongation of overall survival (OS). It is strongly recommended that members of the health team (especially physicians) discuss early on with patients and their family members and caregivers, including preliminarily when to discontinue chemotherapy and focus primarily on symptom management. Multidisciplinary care is key, and should include pain management, nutritional and psychosocial support, family therapy, and other medical specialists as needed. As discussed in greater detail in other chapters of this book, patients should be referred early on for a palliative care evaluation.
Disease Monitoring
A growing body of evidence suggests a role for maintenance chemotherapy in patients with MBC (10). It is often simple to assess if patients are benefiting from systemic therapy or not. At the same time, there is great interest in identifying early on if the treatment started will help them or not. Tumor markers like CA27.29/CA15-3 and CEA are often used to monitor disease status, but oncologists are strongly discouraged from making therapy changes exclusively based on a test result, especially if patients are clinically stable without worsening of symptoms or evident disease progression by imaging studies. Enumeration of circulating tumor cells (CTCs) has been a topic of great interest, and a commercial assay has been available since the mid-2000s based on evidence suggesting an early prognostic role for progressionfree and OS in patients starting a new systemic regimen (11). Unfortunately, the enumeration assay detects circulating tumor cells in only 30% of patients with MBC. A recently completed trial (SWOG S0500, NCT00382018) is now examining whether treatment decision-making to continue or change therapy based on blood levels of CTCs in women with MBC lead to an improved clinical outcome (clinical utility).
In the meantime, more specific measures of circulating tumor DNA are now being tested with potential implications for surveillance in the adjuvant setting, monitoring of disease in the metastatic setting, and characterization of disease biology. Proof of concept studies have now shown that these assays are feasible, albeit costly if dependent on whole-genome sequencing approaches, and must now be prospectively tested (12, 13). Other groups have tested more selective approaches that targeted the detection of circulating tumor DNA of more commonly observed somatic genomic abnormalities like mutations in PIK3CA (14).
Patients with Small Volume of Metastatic Disease
Improvements in staging imaging studies have also caused stage migration and a growing number of patients are now diagnosed with limited or oligometastatic disease (15). These patients, specially those presenting with locally advanced disease and small volume isolated systemic metastases may often benefit from combine modality therapy, including surgery and radiation therapy for local control and/or for resection of focal sites of systemic disease. The role of systemic therapy for patients who become free of macroscopic disease with no evidence of disease (NED) remains controversial.
In late 2012, investigators reported on the role of chemotherapy as adjuvant therapy for patients with locally recurrent breast cancer (the CALOR trial, NCT00074152). Unfortunately, the study was closed due to slow accrual after 162 eligible patients were randomized to ER and/or HER2-targeted therapy, with or without chemotherapy, but after 8 years of follow-up, a 5-year improvement in OS (88% vs. 76%) favored those also given chemotherapy. There was also a striking difference in disease-free survival (DFS) rate between the ER-negative population who received chemotherapy and those who did not (67% vs. 35%), but this was less pronounced in the ER-positive population (70% vs. 69%), suggesting that the majority of benefit was in ER-negative tumors. In view of the challenge of mounting a definitive study, the data suggest that combined multimodality therapy is a reasonable approach for a select group of patients.
PRINCIPLES OF CHEMOTHERAPY
Breast cancer is generally considered to be a chemosensitive disease. The short term aims of chemotherapy in MBC are to increase response rates and maximize symptom control; the medium term end points are to anticipate complications and extend PFS; and the long term goals are to attempt to improve OS while minimizing therapy toxicity or disruption of quality of life (QOL). Chemotherapy treatment is individualized based on disease and patient-related factors such as endocrine responsiveness and HER2 expression, tumorrelated symptoms, disease-free interval, extent and sites of metastatic disease, organ function, comorbidities, age, and performance status. Treatment-related issues such as drug efficacy, side effect profile, previous systemic therapy, quality of life considerations, costs, and patient preferences have also to be taken into account. Tailoring chemotherapy options is made complex by a myriad of tumor and patientrelated factors that must be considered, the large number of chemotherapy regimens available without specific predictive markers of benefit, and the likelihood of disease progression that will force patients to go through a sequence of treatment regimens. Therefore, early on it is critical to discuss with patients and their caregivers the overall goals of therapy, the limitations of existing regimens, the concept of palliative care, and limitations of active therapy.
Prospective trials comparing chemotherapy to best supportive care have not been pursued to a greater extent due to the number of therapy options with clinical efficacy in the adjuvant and advanced settings. Although there have been numerous trials in MBC, the higher response rates and longer PFS seen in certain regimens over others have not translated into survival benefits in most trials in part due to tumor biology, small sample sizes, study designs allowing crossover to the investigational agent, and the increasing availability of newer more efficacious therapeutic regimens.
Endocrine therapy must be the initial consideration in ER-positive MBC with a less aggressive phenotype, such as those with a long disease-free interval and predominantly bone and soft-tissue disease. Chemotherapy is meant for ER-negative disease or ER-positive disease that has become endocrine-resistant or that displays higher risk features such as the often discussed but infrequently observed visceral crisis and/or a short disease-free interval. Visceral crises are uncommon but could be defined as presence of symptomatic lymphangitic lung metastases, bone marrow replacement, carcinomatous meningitis, or symptomatic liver metastases (16). Small volume lung or liver metastases (especially if minimally symptomatic) must not be considered a definitive indication for chemotherapy if endocrine therapy is suitable.
A meta-analysis of published and unpublished trials comparing endocrine therapy or chemotherapy treatment in advanced breast cancer revealed no significant difference in OS (hazard ratio [HR] 0.94, 95% confidence interval [CI], 0.79-1.12, p = .5) (17). Notably, over 50% of women in these trials had visceral disease, which is often seen as an indicator for chemotherapy use. A pooled estimate of reported response rates (8 trials; n = 817) showed a significant advantage for chemotherapy over endocrine therapy (relative risk [RR] 1.25, 95% CI, 1.01-1.54, p = .04), but the two largest trials showed results in opposite directions with a significant test for heterogeneity (p = .0018), thus questioning this observation (18, 19). Six of the seven fully published trials reported increased toxicity with chemotherapy, in particular nausea, vomiting, and alopecia. The shortcomings of these analyses were that the trials were from 1963 to 1995, and contained ER-negative or unknown tumors. These trials used endocrine agents which are not commonly used as first-line choices today and also had outdated chemotherapy regimens. Hence, the role of more contemporary chemotherapy agents in comparison to endocrine therapy remains uncertain. Clinical trials have failed to show a survival benefit of combined chemo endocrine therapy over either modality used separately (20).
Biopsy of metastatic sites of disease is always encouraged as this may impact subsequent treatment decisions, first to confirm the development of metastatic disease and second to retest ER and HER2, especially for tumors that may have tested negative when first diagnosed but the clinical behavior (e.g., long DFS) might suggest otherwise. Discordance in receptor results may occur due to tumor heterogeneity missed by the limited tissue sampling of a needle biopsy, limited reproducibility and concordance of assays especially when testing of the primary tumor and metastatic sites done years apart, or true biologic change over time (21, 22).
Single versus Combination Chemotherapy
The active classes of chemotherapy drugs are the anthracyclines, taxanes, vinca alkaloids, antimetabolites, alkylating agents, epothilones, and other antimicrotubule agents such as eribulin. Randomized trials comparing combination versus sequential single agent chemotherapy have shown combination regimens give better response rates and tumor time to progression (TTP) or PFS, but survival benefit tend to be observed in studies that did not allow crossover from the single-agent arm to the new investigational drug upon progression (see Table 71-1). Well-defined comparisons of combination versus the same agents used in sequence (a registration strategy less favored by the pharmaceutical industry) are unavailable, hence the true impact on survival outcomes is not known.
In an early systemic review of 15 trials (n = 2,442) comparing poly chemotherapy to single agents in MBC treatment, the complete and partial responses for polychemotherapy were significantly better than that associated with single agents (20). Survival data from 12 trials (n = 1,986) also favored polychemotherapy regimens (HR 0.82, 95% CI, 0.75-0.90), translating into an 18% reduction in risk of death (20). The limitations of this meta-analysis were its usage of published material instead of individualized patient data, trials from the pretaxane era consisting of outdated regimens, small sample sizes, poorly designed studies, heterogeneity of patients and their previous treatments, and a lack of data on exposure to adjuvant therapy, and prior treatment for metastatic disease. Its modest total number of patients (n = 996) only slightly exceeded that of an important randomized ECOG study comparing combination versus sequential single-agent therapy which is discussed later (23).
A Cochrane meta-analysis of 43 trials consisting of 9,742 women of whom 55% were receiving first-line chemotherapy for metastatic disease showed a statistically significant advantage for the combination regimens in terms of OS (HR 0.88; p < .00001), TTP (HR 0.78; p < .00001), and response rates (RR 1.29; p < .0001) (24). They were however associated with more leukopenia, nausea, vomiting, and alopecia.
Clinical trial designs have incorporated comparisons of (i) a particular agent versus combination regimens consisting of completely different agents, or (ii) a particular drug versus a regimen containing that same drug in addition to other agents. With regards to the first trial design, taxanes as single agents have shown superiority in terms of survival over older regimens such as cyclophosphamide/methotrexate/5-fluorouracil/prednisone (CMFP) or mitomycin/vinblastine (25, 26). Capecitabine was reported to have comparable TTP and OS compared with intravenous (IV) cyclophosphamide/methotrexate/5-fluorouracil (CMF) (27). As for study design (2), the seminal Eastern Cooperative Oncology Group (ECOG) trial E1193 assigned 739 patients with MBC to doxorubicin alone, paclitaxel alone, or the combination, with crossover allowed for the single agent therapy arms (23). Combination therapy demonstrated significantly higher complete and partial responses compared to the single-agent doxorubicin or paclitaxel arms (47% vs. 36% vs. 34%), and median time-to-treatment failure (TTF) (8 months vs. 5.8 months vs. 6 months), although median survivals were similar (22 months vs. 18.9 months vs. 22.2 months). Responses were seen in 20% of patients crossing from doxorubicin to paclitaxel and 22% of patients crossing from paclitaxel to doxorubicin (p = not significant). Global QOL measurements from onstudy to week 16 were similar in all three groups. While this is the largest randomized trial to address this issue, it remains a relatively small study with limited statistical power.
Other trials of similar design comparing concomitant epirubicin and paclitaxel versus sequential therapy in MBC (28), or capecitabine and taxanes in sequence or combination (29), did not show a survival benefit. Other combinations such as vinorelbine/doxorubicin and gemcitabine/vinorelbine have found no difference in OS between the combinations versus single agents doxorubicin or vinorelbine respectively (30, 31). Two important trials have demonstrated a survival benefit of taxane-containing combinations over the taxane itself and will be discussed in detail later in the chapter (32, 33). One of them showed that the docetaxel/capecitabine combination had significantly superior RRs, TTP, and OS over single-agent docetaxel, while the other showed that the gemcitabine/paclitaxel regimen had superior RRs, TTP, and OS compared to paclitaxel, although there was a lack of a planned crossover design in both studies.
In general, sequential single agents have a more favorable toxicity profile and a better QOL without compromising crucial end points such as OS and TTP. Hence, sequential therapy is useful in the metastatic setting where efficacy should be balanced with a good QOL. This sequential singleagent strategy is also useful in patients with less aggressive disease, those who are older, or with a poorer performance status. However, in cases where rapid tumor shrinkage is needed due to symptomatic disease, combination therapy is preferred.
Intermittent versus Continuous Chemotherapy
The issue of chemotherapy duration in the metastatic setting remains an unresolved issue. Several randomized trials have attempted to address the potential benefits of continuous chemotherapy versus chemotherapy for a fixed
number of cycles, and then resumption only upon disease progression. Trial designs have varied with regards to the maintenance treatment with some continuing the same chemotherapeutic agents while others have utilized different regimens (see Table 71-2). However, earlier studies in the pretaxane era comparing shorter versus longer chemotherapy durations were hampered by insufficient sample sizes, chemotherapy drugs considered obsolete, nonstandard chemotherapy schedules, and limited durations in the control arms. More recent trials with newer agents have been carried out. A recent meta-analysis of 11 randomized controlled trials (RCTs) (n = 2,269) demonstrated that longer first-line chemotherapy had a marginally improved OS (HR 0.91, 95% CI, 0.84-0.99; p = .046) and substantially longer PFS (HR 0.64, 95% CI, 0.55-0.76; p < .001) (34). No statistically significant variation in effects on OS and PFS were seen when trials were stratified according to timing of randomization, study design, number of cycles in the control arm, and concomitant endocrine therapy. Lengthening PFS is considered clinically beneficial as this may improve QOL by delaying symptoms of progressive disease that may be perceived as valuable by the patient. Unfortunately, only the study by Coates et al. (35) in this meta-analysis evaluated this issue, reporting that QOL was indeed better in the extended chemotherapy arm. Shortcomings of this analysis were lack of individualized patient data, no quality control on original records and analyses, limitation of subgroup analyses to those only on trial, moderate number of trials and sample numbers, outdated chemotherapeutic agents, and heterogeneity of study designs, chemotherapy regimens and publication status. There were three studies in the meta-analysis which included more recent agents like paclitaxel and liposomal doxorubicin, none of which demonstrated a survival benefit with maintenance therapy as well (36, 37 and 38). In the Spanish Breast Cancer Research Group (GEICAM) 2001-01 study, patients without disease progression after three cycles of doxorubicin followed by three cycles of docetaxel were randomized to pegylated liposomal doxorubicin (PLD) for six cycles or to observation (38). PLD significantly prolonged the primary end point of TTP by 3.3 months compared to observation although OS was not significantly prolonged. PLD toxicities were manageable with up to 5% experiencing fatigue, mucositis, and palmar-plantar erythrodysesthesia.
number of cycles, and then resumption only upon disease progression. Trial designs have varied with regards to the maintenance treatment with some continuing the same chemotherapeutic agents while others have utilized different regimens (see Table 71-2). However, earlier studies in the pretaxane era comparing shorter versus longer chemotherapy durations were hampered by insufficient sample sizes, chemotherapy drugs considered obsolete, nonstandard chemotherapy schedules, and limited durations in the control arms. More recent trials with newer agents have been carried out. A recent meta-analysis of 11 randomized controlled trials (RCTs) (n = 2,269) demonstrated that longer first-line chemotherapy had a marginally improved OS (HR 0.91, 95% CI, 0.84-0.99; p = .046) and substantially longer PFS (HR 0.64, 95% CI, 0.55-0.76; p < .001) (34). No statistically significant variation in effects on OS and PFS were seen when trials were stratified according to timing of randomization, study design, number of cycles in the control arm, and concomitant endocrine therapy. Lengthening PFS is considered clinically beneficial as this may improve QOL by delaying symptoms of progressive disease that may be perceived as valuable by the patient. Unfortunately, only the study by Coates et al. (35) in this meta-analysis evaluated this issue, reporting that QOL was indeed better in the extended chemotherapy arm. Shortcomings of this analysis were lack of individualized patient data, no quality control on original records and analyses, limitation of subgroup analyses to those only on trial, moderate number of trials and sample numbers, outdated chemotherapeutic agents, and heterogeneity of study designs, chemotherapy regimens and publication status. There were three studies in the meta-analysis which included more recent agents like paclitaxel and liposomal doxorubicin, none of which demonstrated a survival benefit with maintenance therapy as well (36, 37 and 38). In the Spanish Breast Cancer Research Group (GEICAM) 2001-01 study, patients without disease progression after three cycles of doxorubicin followed by three cycles of docetaxel were randomized to pegylated liposomal doxorubicin (PLD) for six cycles or to observation (38). PLD significantly prolonged the primary end point of TTP by 3.3 months compared to observation although OS was not significantly prolonged. PLD toxicities were manageable with up to 5% experiencing fatigue, mucositis, and palmar-plantar erythrodysesthesia.
TABLE 71-1 Selected Clinical Trials of Single versus Combination Chemotherapy | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
OS seems not to be influenced by continuing chemotherapy indefinitely, with the benefit primarily in PFS. Thus, patients who need to stop treatment due to drug-related toxicities can be reassured that this is not detrimental to their survival. In those who have symptomatic disease and
remain responsive to chemotherapy, continuing therapy can be a favorable option to prolong time to disease progression. In clinical practice, no predefined number of courses of chemotherapy must be delivered, and factors such as treatment tolerability and disease response in terms of disease stabilization as opposed to tumor shrinkage must be taken into account. If the patient shows improvement after two to three cycles, then the same regimen is continued for another two to three cycles before further reassessment. The duration of treatment in patients who have stable disease or who continue to respond is controversial. Although there is limited evidence for maintenance chemotherapy long-term, chemotherapy can be continued beyond six to eight cycles for PFS benefit, and only ceased upon disease progression or intolerable toxicities. Endocrine therapy could be used as maintenance therapy in ER-positive disease, or the patient closely monitored for recurrence without any systemic therapy if ER-negative disease and a therapy holiday is being considered. However, the targeted therapy such as trastuzumab should be continued if HER2-positive disease, either alone or with endocrine therapy in endocrine responsive disease. For those who progress while on one line of treatment or during the chemotherapy-free period, they are generally switched to an alternative agent or regimen based upon their performance status, previous chemotherapy exposure, and potential for further treatment response.
remain responsive to chemotherapy, continuing therapy can be a favorable option to prolong time to disease progression. In clinical practice, no predefined number of courses of chemotherapy must be delivered, and factors such as treatment tolerability and disease response in terms of disease stabilization as opposed to tumor shrinkage must be taken into account. If the patient shows improvement after two to three cycles, then the same regimen is continued for another two to three cycles before further reassessment. The duration of treatment in patients who have stable disease or who continue to respond is controversial. Although there is limited evidence for maintenance chemotherapy long-term, chemotherapy can be continued beyond six to eight cycles for PFS benefit, and only ceased upon disease progression or intolerable toxicities. Endocrine therapy could be used as maintenance therapy in ER-positive disease, or the patient closely monitored for recurrence without any systemic therapy if ER-negative disease and a therapy holiday is being considered. However, the targeted therapy such as trastuzumab should be continued if HER2-positive disease, either alone or with endocrine therapy in endocrine responsive disease. For those who progress while on one line of treatment or during the chemotherapy-free period, they are generally switched to an alternative agent or regimen based upon their performance status, previous chemotherapy exposure, and potential for further treatment response.
TABLE 71-2 Selected Clinical Trials Comparing Maintenance versus Intermittent Chemotherapy | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Chemotherapy Scheduling
The impact of chemotherapy scheduling has been less well studied. The Norton-Simon hypothesis derived from clinical and laboratory observations states that “therapy results in a rate of regression in tumor volume that is proportional to the rate of growth that would be expected for an unperturbed tumor of that size” (39). This hypothesis has led to the dose-dense approach to breast cancer chemotherapy, thus short circuiting the Gompertzian growth curve before tumor regrowth achieves its mathematically greatest gains. This has been illustrated by the Cancer and Leukemia Group-B (CALGB) 9741 adjuvant trial, which confirmed the clinical efficacy in terms of DFS and OS of a dose-dense 2-weekly schedule of sequential doxorubicin/cyclophosphamide and paclitaxel for four cycles each with colony stimulating support over an identical regimen but administered in a 3-weekly fashion (40). The use of dose-density as strictly defined has not been shown to be effective in the metastatic setting.
Weekly scheduling attempts to maximize frequency and cumulative doses with a more favorable toxicity profile. A weekly scheduling at a lower chemotherapy dose with regards to taxanes has been shown to be superior to the 3-weekly regimen. In the CALGB 9840 trial, a higher response rate and TTP favoring weekly paclitaxel over the 3-weekly schedule has been demonstrated (41).
Oral CMF (cyclophosphamide for 14 days, methotrexate and 5-FU on days 1 and 8) every 28 days has been shown to be better than IV CMF every 3 weeks with respect to response rates (48% vs. 29%; p = .003) and OS (17 months vs. 12 months; p = .016) possibly due to scheduling leading to a higher dose intensity achieved (42).
Trials exploring other scheduling approaches such as theoretically non-cross-resistant agents utilized in a sequential fashion, or adding on additional different agents have not shown any remarkable clinical significance (43, 44 and 45). In summary, the trials in general do not demonstrate any major survival benefit from using dose-dense chemotherapy, sequential non-cross resistant regimens or “intensified” regimens whereby new agents are added on, and appear to support simpler single-agent regimens to minimize therapy-related toxicities. At the same time, weekly regimens (e.g., paclitaxel) or even daily regimens (e.g., oral etoposide) appear to be more efficacious with less side effects.
CHEMOTHERAPY OPTIONS (SINGLE AGENT AND COMBINATIONS)
In view of the lack of specific biomarkers that predict differential responsiveness to conventional chemotherapy regimens according to the established phenotypes, the regimens discussed below in principle equally apply to patients with TNBC and those with ER-positive/HER2-negative disease that has become endocrine-resistant. Later in this chapter we discuss strategies that may potentially apply more specifically to TNBC, especially in tumors that might have basal-like features.
Meaningful improvements in survival have been seen with the advent of newer therapeutic options and better supportive medical care. Median OS is about 18 to 24 months, with a range of a few months to many years, according to tumor subtype as well as sites and burden of metastatic disease. Several favorable prognostic factors include ER-positivity, a longer relapse-free interval of more than 2 years and metastases involving the chest wall, bones, or lymph nodes. Weight loss, poor performance status, elevated serum lactate dehydrogenase, and less than 35 years old are poor prognostic factors.
The goals of treatment are alleviation of symptoms, prolongation of survival, and improvement in QOL. OS in MBC trials is the gold standard although this end point requires prolonged follow-up and is now frequently diluted by increasingly more effective subsequent treatment options.
There is no standardized treatment regimen sequence to be utilized in MBC although anthracyclines and taxanes are the mainstay of initial treatment. Eighty reports of 63 trials were identified from the Cochrane Breast Cancer Group (CBCG) and abstracts from the American Society of Clinical Oncology (ASCO) annual scientific meeting (2000-2007) (46). There was little evidence from trials reported from 2000 to 2007 that major survival differences existed between many commonly employed chemotherapy regimens.
Combination chemotherapy versus sequential single agent use depends on disease characteristics and patientrelated factors. For those with rapidly progressive symptomatic disease or a visceral crisis, combination therapy would be the preferred choice particularly in the first-line setting. Once disease stability is achieved (usually six to eight cycles), a switch to maintenance therapy should be considered. In the absence of rapid clinical progression or life-threatening visceral metastases, sequential single agent chemotherapy is recommended for most patients with low risk disease such as no multi-organ involvement and a longer disease-free interval, with consideration also given to endocrine therapy to this group if ER-positive. Single-agent therapy given in weekly schedules is also preferable for those with visceral impairment or bone marrow suppression due to metastatic disease where dose-reduced single agents may attempt to control the disease at first. For those with asymptomatic or minimally symptomatic disease and/or without visceral involvement, endocrine therapy can be used upfront. If PFS is short or the disease becomes more symptomatic, then switch of treatment to chemotherapy may be considered before another line of endocrine therapy.
Single-Agent Chemotherapy
Single agents include anthracyclines, taxanes, antimetabolites and other microtubule inhibitors. Response rates range
from 25% to 45% and PFS 5 months to 8 as first-line therapy; 15% to 30% and 2 months to 5 respectively as second-line therapy; and 0% to 2% and 1 to 4 months respectively as third-line therapy (47). From fourth-line therapy and beyond, data are scant, although chemotherapy is often continued with further lines of treatment. While worth considering in patients with good performance status and whose disease might have responded to earlier lines of chemotherapy, there is little or no evidence to support pursuing this in patients with poor performance status or whose disease is refractory to earlier lines of therapy. Rather, palliative care measures aiming at reducing symptoms and improving QOL should take primacy over a simple switch to another chemotherapy drug.
from 25% to 45% and PFS 5 months to 8 as first-line therapy; 15% to 30% and 2 months to 5 respectively as second-line therapy; and 0% to 2% and 1 to 4 months respectively as third-line therapy (47). From fourth-line therapy and beyond, data are scant, although chemotherapy is often continued with further lines of treatment. While worth considering in patients with good performance status and whose disease might have responded to earlier lines of chemotherapy, there is little or no evidence to support pursuing this in patients with poor performance status or whose disease is refractory to earlier lines of therapy. Rather, palliative care measures aiming at reducing symptoms and improving QOL should take primacy over a simple switch to another chemotherapy drug.
Lack of response to first-line chemotherapy treatment or a short progression-free period portends a poorer response to subsequent lines of treatment. The drug of choice for first-line therapy would depend on the disease-free interval since the end of adjuvant chemotherapy. Those whose disease recurs in <12 months would reflect a degree of resistance to the previous regimen. Sequential single agents may have lower response rates but also lower toxicities and no compromise in survival.
Anthracyclines
Since their introduction in the 1980s, anthracyclines have remained one of the most active agents for breast cancer. In an overview of randomized trials, inclusion of anthracyclines—such as doxorubicin—were found to confer a benefit in RR, TTF, and OS over nonanthracycline containing regimens (48). Anthracyclines have response rates of 35% to 50% for those who are anthracycline-naïve, or those who develop metastases after 12 months after anthracycline-based adjuvant therapy (23, 30, 49), but their efficacy in those with an anthracycline-free interval of <12 months is uncertain. Common dosing schedules are doxorubicin 60-75 mg/m2 3-weekly or 20 mg/m2 weekly; or epirubicin 75-100 mg/m2 3-weekly or 20-30 mg/m2 weekly. The risk of congestive heart failure is dose-related and rises from about 5% at a cumulative doxorubicin dose of 400 mg/m2 to 16% at cumulative doses of more than 500 mg/m2 (50). The use of anthracyclines in the metastatic setting is limited by acute toxicities such as nausea, vomiting, myelotoxicity, alopecia, and long-term issues such as leukemogenic risks and cardiotoxicity.
Liposome-encapsulated and PLD have comparable efficacy to doxorubicin but with reduced cardiotoxicity. Women with MBC who were randomized to receive either first-line PLD 50 mg/m2 every 4 weeks or doxorubicin 60 mg/m2 every 3 weeks showed comparable objective RR (33% vs. 38%), PFS (6.9 months vs. 7.8 months) and OS (21 months vs. 22 months), but there was significantly higher risk of cardiotoxicity with doxorubicin (RR = 3.16, 95% CI, 1.58-6.31; p < .001) (51). PLD caused less alopecia (20% vs. 66%), nausea (37% vs. 53%), vomiting (19% vs. 31%), and neutropenia (4% vs. 10%), but had more palmar-plantar erythrodysesthesia (48% vs. 2%), stomatitis (22% vs. 15%), mucositis (23% vs. 13%), and infusion reactions (13% vs. 3%).
Using PLD as salvage therapy has been shown to have comparable efficacy to that of common salvage regimens such as vinorelbine in taxane-refractory MBC (52). In women whose disease progressed after first or second-line taxanecontaining chemotherapy for MBC, PLD 50 mg/m2 every 28 days compared to vinorelbine 30 mg/m2 weekly or mitomycin 10 mg/m2 on day 1 every 28 days plus vinblastine 5 mg/m2 on days 1,14, 28, 42 every 6 to 8 weeks demonstrated similar PFS (HR 1.26; p = .11) and OS (HR 1.05; p = .71) (52).
In anthracycline-naive patients, PLD had a significantly longer PFS compared to the comparator arms (HR 2.4; p = .01), although those patients on PLD had more palmarplantar erythrodysesthesia (37% vs. <1%) and stomatitis. These results suggest that liposomal doxorubicin may offer an alternative therapeutic option for those who are at increased cardiac risk such as elderly patients, those who have cardiac risk factors, and those previously exposed to anthracyclines.
In patients previously treated with adjuvant anthracyclines, the combination of liposomal doxorubicin with docetaxel compared to docetaxel alone showed an improved RR and PFS but not OS (53). However, this combination was not granted regulatory approval because of excessive palmar-plantar erythrodysesthesia. Nonetheless, rechallenging patients with liposomal doxorubicin either singly or in combination with other agents remains an option for those previously treated with adjuvant anthracyclines if more than 12 months have elapsed since their completion (54).
Anthracycline Combinations
Data have shown that anthracycline-based regimens have improved response rates and PFS compared to regimens that do not contain doxorubicin (20, 48, 49), with some showing a survival benefit (48). Anthracycline-based combination regimens like doxorubicin/cyclophosphamide (AC), epirubicin/cyclophosphamide (EC), cyclophosphamide/doxorubicin/5-FU (FAC or CAF), or 5-FU/epirubicin/cyclophosphamide (FEC) are more active than single-agent anthracyclines but also have more toxicities, mainly myelosuppression, gastrointestinal toxicity, cardiotoxicity, and alopecia. They have not demonstrated a survival benefit over monotherapy. For example, the French Epirubicin Study Group comparing firstline epirubicin 75 mg/m2 alone with fluorouracil 500 mg/m2, epirubicin 50 mg/m2, cyclophosphamide 500 mg/m2 (FEC 50), and 5-FU 500 mg/m2, epirubicin 75 mg/m2, cyclophosphamide 500 mg/m2 (FEC 75) as first-line treatment for advanced breast cancer patients found superior response rates for the combination regimen compared to single-agent therapy (FEC 50 [44.6%] and FEC 75 [44.7%] vs. epirubicin [30.6%] [p = .04 and p = .0006]) (55). The epirubicin alone group showed better tolerability than the two combination groups, which did not differ significantly. PFS and OS were not significantly different among the three groups but more early relapses occurred in the epirubicin and FEC 50 groups.
Another randomized prospective study comparing sequential monotherapy versus sequential combination regimens as first and second-line therapies: weekly epirubicin (E) 20 mg/m2 until progression or until cumulative dose of 1,000 mg/m2 followed by mitomycin (M) 8 mg/m2 every 4 weeks (n = 153) versus combination cyclophosphamide 500 mg/m2, epirubicin 60 mg/m2, and fluorouracil 500 mg/m2 every three weeks (CEF) followed by mitomycin 8 mg/m2 plus vinblastine (V) 6 mg/m2 every 4 weeks (n = 150) found higher objective response rates for the anthracycline regimens (CEF 55%, E 48%, M 16%, MV 7%) (56). Between the two anthracycline regimens, the combination gave longer response durations (CEF 12 months vs. E 10.5 months) but there was no difference in PFS or OS between the two treatment arms. Toxicities were less and QOL was better in the single-agent arm.
Rechallenging with Anthracyclines/Liposomal Doxorubicin
Rechallenging with anthracycline-based chemotherapy up to cumulative doses of doxorubicin 450-550 mg/m2 and epirubicin 800-900 mg/m2 is acceptable. Retrospective studies
have examined the feasibility of rechallenging with anthracyclines in the first-line setting for patients who have been exposed to anthracyclines in the adjuvant setting (57, 58, 59 and 60). All studies had a similar schema comprising a chemo-naive group, others who had received CMF or CMF-like regimens, or anthracyclines. There was a trend towards a worse clinical outcome (response rates and survival) for those with prior adjuvant chemotherapy, and this was statistically significant in two studies (58, 59). However, there was no difference between CMF and anthracycline-based adjuvant regimens with regards to impact on first-line anthracycline therapy outcomes.
have examined the feasibility of rechallenging with anthracyclines in the first-line setting for patients who have been exposed to anthracyclines in the adjuvant setting (57, 58, 59 and 60). All studies had a similar schema comprising a chemo-naive group, others who had received CMF or CMF-like regimens, or anthracyclines. There was a trend towards a worse clinical outcome (response rates and survival) for those with prior adjuvant chemotherapy, and this was statistically significant in two studies (58, 59). However, there was no difference between CMF and anthracycline-based adjuvant regimens with regards to impact on first-line anthracycline therapy outcomes.
A small randomized phase III study compared rechallenging with epirubicin/docetaxel versus docetaxel alone as first-line chemotherapy in patients exposed to adjuvant or neoadjuvant epirubicin, and found similar antitumor efficacy in the two arms with more leukopenia, nausea and stomatitis in the epirubicin/docetaxel arm (61). Hence, the use if anthracyclines as first-line treatment for those already exposed to adjuvant anthracyclines is generally not recommended. In this case, taxane-based therapy is usually considered instead.
Liposomal doxorubicin is active in MBC patients who have been previously treated with conventional anthracyclines, with an overall clinical benefit rate of 24% (62). In a study conducted on behalf of the Spanish Breast Cancer Research Group, patients who did not have progression after three cycles of doxorubicin 75 mg/m2 followed by three cycles of docetaxel 100 mg/m2 administered 3-weekly were then randomized to liposomal doxorubicin 40 mg/m2 once every 28 days for six cycles or observation (38). At a median follow-up of 20 months, liposomal doxorubicin significantly improved TTP (8.4 months vs. 5.1 months; HR = 0.54; p = .0002) compared to the control arm, although OS was not significantly prolonged (24.8 months vs. 22 months, HR = 0.86; p = .44). Liposomal doxorubicin was well tolerated with only 5% experiencing grade 3 or 4 fatigue, mucositis or palmar-plantar erythrodysesthesia. Hematologic toxicities were slightly greater with 12% experiencing neutropenia but only two with febrile neutropenia. However, non-crossresistance between conventional and liposomal doxorubicin and the appropriateness of treating patients who have progressed on conventional doxorubicin with liposomal doxorubicin cannot be assessed from this study.
Data pooled from two prospective randomized phase III clinical trials comparing non-PLD or conventional doxorubicin combined with cyclophosphamide or liposomal doxorubicin versus conventional doxorubicin as monotherapy in patients previously treated with anthracyclines revealed significant differences for overall RR and median TTP in favor of liposomal doxorubicin, although there was no difference in OS (63, 64 and 65). Moreover, there was less cardiotoxicity for the liposomal doxorubicin formulation.
In summary, data suggest that a greater benefit is achieved when rechallenging with non-PLD than conventional doxorubicin, either as monotherapy or in combination with other agents. Hence, patients who were exposed to adjuvant anthracyclines may remain responsive to liposomal formulations of anthracyclines.
Epirubicin versus Doxorubicin
In order to compare the efficacy of epirubicin to doxorubicin in the treatment of MBC, 13 randomized controlled trials (11 published reports and 2 reports in abstract form) were reviewed (66). No significant differences in response rates or median survival were observed at equal doses of epirubicin and doxorubicin. Although higher RRs were observed for higher epirubicin doses, this did not translate into a survival advantage. However, epirubicin was associated with less nausea, vomiting, neutropenia, and cardiotoxicity. A recent meta-analysis however failed to demonstrate a significant difference in congestive heart failure (CHF) risk between epirubicin and doxorubicin, although there was a suggestion of a lower rate of clinical heart failure for patients treated with epirubicin (67).
Taxanes
Until the development of taxanes in the 1990s, treatment options were much more limited. In a systematic review of taxane-containing chemotherapy compared with nontaxane regimens, taxane-based regimens showed a higher RR (odds ratio [OR] 1.34; p < .001), TTP (HR 0.92; p = .02), and OS (HR 0.93; p = .05), although there was significant heterogeneity in the trials, partially due to varying efficacy of the comparator regimens (68). The conclusion was that taxanecontaining regimens were more effective than some, but not all nontaxane regimens. Taxanes have been studied in two main groups of patients; those who are anthracycline-naive and those who have been anthracycline pretreated.
Anthracycline-Naive Patients
In the TAX 303 phase III study evaluating docetaxel 100 mg/m2 versus doxorubicin 75 mg/m2 every 3 weeks for a maximum of seven cycles in MBC patients who had previous alkylating agent chemotherapy, docetaxel was significantly better than doxorubicin in terms of RR (48% vs. 33%; p = .008), even in those with poor prognostic factors such as visceral metastases and resistance to prior chemotherapy (69). Median TTP was longer in the docetaxel (26 weeks vs. 21 weeks) although the difference was not significant, and median OS was similar in the two groups (15 months vs. 14 months). Febrile neutropenia was more prevalent in the doxorubicin group, including cardiotoxicity, nausea, vomiting and stomatitis, whereas there was more diarrhea, neuropathy, fluid retention, skin and nail changes with docetaxel.
In an European Organisation for Research and Treatment of Cancer (EORTC) study (n = 331) for MBC comparing first-line doxorubicin 75 mg/m2 with paclitaxel 200 mg/m2 both once every 3 weeks, with a crossover design incorporated, doxorubicin was significantly better than paclitaxel for objective response rates (41% vs. 25%; p = .003), median PFS (7.5 months vs. 3.9 months; p < .001), but not in OS (18.3 months vs. 15.6 months; p = .38) (49). At crossover to doxorubicin or paclitaxel during second-line therapy, response rates were 30% and 16%, respectively. The doxorubicin arm was more toxic than paclitaxel in terms of hematologic, gastrointestinal, and cardiac side effects, but counterbalanced by better symptom control. There was no difference in QOL between the two treatment groups.
In the ECOG trial described previously, first-line paclitaxel 175 mg/m2 versus doxorubicin 60 mg/m2 produced equivalent outcomes in terms of RRs, TTF, and median survival, although this may be attributable to the lower dose of doxorubicin at 60 mg/m2 used (23).
Paclitaxel 200 mg/m2 every 3 weeks when compared to oral CMF plus prednisone (CMFP) every 28 days in untreated patients with MBC, had a significantly longer OS after adjustment for prognostic factors without a difference in overall response rate (ORR) and TTP (25). CMFP had more leukopenia, thrombocytopenia, nausea, and vomiting, but overall QOL assessments were similar in the two treatment arms. However, the dose of paclitaxel used in practice is usually 175 mg/m2 as higher doses have greater toxicities but have not demonstrated a better efficacy (70).
In conclusion, the evidence does not indicate a clear superiority of an anthracycline or taxane in anthracyclinenaive patients, and either agent can be used as first-line therapy after taking into account previous adjuvant therapy exposure, tolerability, and side effect profile.
Anthracycline Pretreated Patients
Taxanes have been compared to older nonanthracycline regimens in MBC patients with prior anthracycline exposure and found to be superior in terms of ORR, TTP, and OS in some instances. Docetaxel 100 mg/m2 every 3 weeks compared to methotrexate/fluorouracil after anthracycline failure in advanced breast cancer had a significantly higher ORR (42% vs. 21%) and median TTP (6.3 months vs. 3 months), but no significant difference in OS although notably crossover was allowed upon progression (71). The same dose of docetaxel has also been shown to be significantly better to mitomycin plus vinblastine in MBC progressing after previous anthracycline therapy in terms of RR (30% vs. 11.6%; p < .0001), median TTP (19 weeks vs. 11 week; p = .001) and OS (11.4 months vs. 8.7 months; p = .0097) (26). Docetaxel 100 mg/m2 every 3 weeks was found to be equivalent in terms of TTP and OS when compared to a regimen of vinorelbine and continuous infusional fluorouracil in MBC patients who had been exposed to anthracyclines in the adjuvant, neoadjuvant, or palliative setting (72).
Paclitaxel monotherapy is also active in those who have been exposed to anthracyclines. Paclitaxel 175 mg/m2 every 3 weeks was found to be inferior to 3-weekly cisplatin/oral etoposide in patients with advanced breast cancer pretreated with anthracyclines (73). The cisplatin/etoposide arm was superior to paclitaxel with respect to RR (36.3% vs. 22.2%; p = .038), TTP (5.5 months vs. 3.9 months; p = .003), and median OS (14 months vs. 9.5 months; p = .039).
In the TAX-311 multicenter open-label phase III study (n = 449) comparing docetaxel 100 mg/m2 versus paclitaxel 175 mg/m2 in patients with advanced breast cancer that had progressed after an anthracycline-containing chemotherapy regimen, docetaxel demonstrated significantly superior median OS (15.4 months vs. 12.7 months, HR 1.41; p = .03) and TTP (5.7 months vs. 3.6 months, HR 1.64; p < .0001), with a higher ORR which did not reach statistical significance (32% vs. 25%; p = .10) (74). Both hematologic and nonhematologic toxicities were greater for docetaxel but QOL scores were not significantly different between the groups over time. Another trial directly comparing second-line docetaxel 100 mg/m2 or paclitaxel 175 mg/m2 after failure of anthracyclines, confirmed that TTP and OS were significantly better for docetaxel (75).
Inference from the available data suggests that docetaxel may be superior to 3-weekly paclitaxel. However, docetaxel maintenance therapy is often limited by hematologic toxicities, peripheral neuropathy, fatigue, nail changes, and fluid retention. Notably, docetaxel has not been compared to the more commonly used weekly paclitaxel schedule, which has demonstrated a survival advantage over the 3-weekly regimen.
In the CALGB 9840 trial for MBC patients who had received up to one line of prior chemotherapy in the metastatic setting, weekly paclitaxel was compared to 3-weekly paclitaxel 175 mg/m2 (41). Owing to a 30% incidence of grade 3 sensory neuropathy, the starting dose of weekly paclitaxel was amended from 100 mg/m2 to 80 mg/m2. The weekly regimen was superior to the 3-weekly regimen in terms of RR (42% vs. 29%; p = .0004), TTP (9 months vs. 5 months; p < .0001), and OS (24 months vs. 12 months; p = .0092). The statistical validity may have been reduced by inclusion of 158 patients who received paclitaxel 175 mg/m2 from the CALGB 9342 study evaluating three doses of single-agent paclitaxel (76).
Although grade 3 or more neutropenia was more frequent with the 3-weekly compared to weekly regimen (15% vs. 9%), febrile neutropenia requiring hospitalization remained infrequent in both arms (4% vs. 3%). Grade 3 neuropathy was a treatment-limiting toxicity more common with the weekly regimen (24% vs. 12%; p = .0003).
These results were confirmed by the Anglo-Celtic study, which showed a better response rate for the weekly paclitaxel 90 mg/m2 for 12 cycles compared to 3-weekly paclitaxel 175 mg/m2 for four cycles (42% vs. 27%; p = .002) (77). Although the TTP was not significantly different, it was thought that the mismatch in treatment duration may have accounted for this.
A meta-analysis of randomized controlled trials comparing weekly and 3-weekly taxanes in advanced breast cancer reported that weekly paclitaxel 80-100 mg/m2 had an OS survival benefit over 3-weekly paclitaxel 175 mg/m2 therapy (5 studies, 1,471 patients, HR 0.78, 95% CI, 0.67-0.89; p = .001), but with worse sensory neuropathy (78). In contrast, no difference between weekly docetaxel 35-40 mg/m2 and 3-weekly docetaxel 100 mg/m2 was reported for ORR, PFS, and OS, the only advantage being significantly less neutropenia or neutropenic fever in the weekly docetaxel schedule. On the contrary, nail changes and epiphora were significantly lower in the 3-weekly docetaxel schedule. Limitations of this meta-analysis were small sample sizes with none of the trials designed to measure OS as a primary end point, lack of individualized patient data with only published trials used, and considerable heterogeneity in design, modes of treatment, and response rates.
Docetaxel 100 mg/m2 as approved for MBC treatment in the U.S. and Europe is not a tolerable dose in Asian patients due to increased toxicities. In a study examining three different doses of second-line docetaxel at 60 mg/m2, 75 mg/m2 and 100 mg/m2 for at least six cycles in pretreated MBC patients, significantly higher ORRs (22.1% vs. 23.3% vs. 36%; p = .007) and TTP (13.7, 13.9, 18.6 weeks; p = .014) were obtained with higher doses in the assessable population, but there was no difference in OS in the intent-to-treat population at a median follow-up of 30 months (79). About 80% of patients were exposed to prior anthracyclines in each arm. Most hematologic and nonhematologic toxicities were related to increasing doses, including those of febrile neutropenia rates (4.7% vs. 7.4% vs. 14.1%). Hence, lower doses of docetaxel must be considered for those who are more frail or who have tolerability issues. As a consequence, in clinical practice we usually use weekly paclitaxel 80 mg/m2 or 3-weekly docetaxel 60-100 mg/m2.
Cross-resistance between the Taxanes
There is evidence of an incomplete cross-resistance between paclitaxel and docetaxel, since modest responses are still seen in those exposed to the alternate taxane (80, 81). However, using a taxane after progression on the other may be best reserved for patients who relapse more than 12 months after adjuvant taxane-containing therapy or who had previous clinical response to taxanes with a reasonable time lapse of at least a year. In a small retrospective study (n = 44) of patients with docetaxel-resistant MBC, paclitaxel 80 mg/m2 weekly obtained objective responses in 14 of 44 women (31.8%, 95% CI, 17.5-46.1), seven of which had primary resistance to docetaxel (82). The median duration of response and time to progression were 6.1 months and 5 months respectively (82). In another larger retrospective study of weekly paclitaxel 80 mg/m2 in 82 patients with docetaxel-resistant MBC, the patients were classified into those with primary or secondary resistance (short interval ≤120 days, long interval >120 days) (83). The response rate
to paclitaxel for those with primary docetaxel resistance (n = 24) was 8.3%, and those with secondary resistance (n = 58) was 24.1% (short interval [n = 39] 17.9%, long interval [n = 19] 36.8%), the differences in response rates being statistically significant (p = .0247). Conversely, docetaxel every 3 weeks was reported to have a response rate of about 18% to 25% in paclitaxel-refractory MBC (81, 84).
to paclitaxel for those with primary docetaxel resistance (n = 24) was 8.3%, and those with secondary resistance (n = 58) was 24.1% (short interval [n = 39] 17.9%, long interval [n = 19] 36.8%), the differences in response rates being statistically significant (p = .0247). Conversely, docetaxel every 3 weeks was reported to have a response rate of about 18% to 25% in paclitaxel-refractory MBC (81, 84).
The data suggest that treatment with an alternative taxane can result in objective responses. Studies support the notion that there is only partial cross-resistance between paclitaxel and docetaxel. However, it should be noted that there was wide variation in extent of prior anthracycline and/or taxane exposure in these studies, as well as the dose and schedule of taxanes used.
Nab-paclitaxel
Nab-paclitaxel is a Cremophor-free, albumin-bound formulation designed to distribute into tumor tissue more rapidly and at higher concentrations than conventional paclitaxel, thus possibly improving drug delivery and reducing toxicity. It is FDA-approved for MBC as monotherapy after failure of anthracycline-based combination chemotherapy for MBC, or after relapse within 6 months of adjuvant anthracyclines. It allows higher doses of paclitaxel infusion, over a shorter duration of 30 minutes, with no need for antihistamine or corticosteroid premedications.
It has been compared to docetaxel and paclitaxel, both as 3-weekly regimens. In a phase III trial comparing 3-weekly nab-paclitaxel 260 mg/m2 without premedication versus paclitaxel 175 mg/m2 with premedication in women with MBC (majority of whom had ≤1 prior MBC chemotherapy regimen not including a taxane) (85), nab-paclitaxel had significantly superior RRs (33% vs. 19%; p = .001) and TTP (23.0 weeks vs. 16.9 weeks, HR = 0.75; p = .006) compared with standard paclitaxel, but no difference in OS. OS was significantly longer in the subgroup who received nab-paclitaxel compared with paclitaxel as second-line or greater therapy (56.4 weeks vs. 46.7 weeks, HR 0.73; p = .024). Grade 4 neutropenia was significantly lower for nab-paclitaxel compared with standard paclitaxel (9% vs. 22%, respectively; p < .001) despite a 49% higher paclitaxel dose. Grade 3 sensory neuropathy was more common with nab-paclitaxel (10% vs. 2%, respectively; p < .001), lasting for a median of 22 days. No hypersensitivity reactions occurred with nab-paclitaxel despite the absence of premedication.
In another phase II trial (n = 302) comparing first-line nabpaclitaxel 300 mg/m2 3-weekly, 100 mg/m2 weekly or 150 mg/m2 weekly, versus docetaxel 100 mg/m2 3-weekly, there was a significant prolongation of PFS (>5 months) in patients receiving nab-paclitaxel 150 mg/m2 weekly compared with docetaxel 100 mg/m2 q3w (86). Nab-paclitaxel 150 mg/m2 weekly showed a significantly longer PFS than docetaxel by independent radiologist assessment (12.9 months vs. 7.5 months, respectively; p = .0065). Both 150 mg/m2 (49%) and 100 mg/m2 (45%) weekly of nab-paclitaxel demonstrated a higher ORR than docetaxel (35%), but this did not reach statistical significance. Nab-paclitaxel 3-weekly versus docetaxel was not different for PFS or ORR. Disease control rate (stable disease ≥16 weeks or confirmed overall complete or partial response) was significantly higher for patients receiving either dose of weekly nab-paclitaxel compared with docetaxel, but survival data were not mature at the point of this publication. Grade 3 or 4 fatigue, neutropenia, and febrile neutropenia were less frequent in the nab-paclitaxel arms, whereas the frequency and grade of peripheral neuropathy were similar in all arms.
Nab-paclitaxel has also shown activity, albeit limited, in taxane-resistant MBC patients (87), with a tolerable toxicity profile of only about 17% grade 1 or 2 sensory neuropathy in taxane-pretreated MBC patients (88). In general, nabpaclitaxel demonstrated better response rates and PFS compared to 3-weekly paclitaxel or docetaxel. However, nab-paclitaxel (and ixabepilone) failed to demonstrate superior efficacy compared to standard weekly paclitaxel in the three-arm phase III open-label randomized trial CALGB 40502/NCCTG N063H study (NCT00785291) comparing the three therapies given in a weekly fashion with bevacizumab (which became optional subsequently) as first-line for metastatic breast cancer patients (89). The PFS for ixabepilone was found to be significantly inferior to paclitaxel, while nab-paclitaxel was not superior to paclitaxel. There was no difference in OS for all treatment arms and unfortunately the investigational arms were more toxic (e.g., peripheral neuropathy) than the conventional weekly paclitaxel arm.
The higher cost of nab-paclitaxel may compare favorably to the cost of docetaxel (90). However, the lack of meaningful clinical efficacy when compared to conventional paclitaxel suggest that the extra cost associated with the use of nab-paclitaxel can only be justified in patients who cannot tolerate use of steroids needed in most patients treated with paclitaxel.
Anthracyclines versus Taxanes
In a meta-analysis of individualized patient data on three single-agent trials comparing taxanes with anthracyclines (n = 919), taxanes fared significantly worse for PFS (5.1 months vs. 7.2 months, HR1.19; p = .011), although response rates for taxanes versus anthracyclines (38% vs. 33%; p = .08), and survival (19.5 months vs. 18.6 months, HR 1.01; p = .9) were similar (91).
Single-agent anthracyclines may be used for firstline therapy of MBC if the patient is anthracycline-naive. Re-using anthracyclines in the metastatic setting after adjuvant exposure is not usually preferred due to the presence of dose-limiting cardiotoxicity and the availability of multiple other drug options. A taxane may also be used as firstline treatment for those who are taxane-naive or who have an adjuvant taxane-free interval of more than 12 months. In the latter scenario, an alternative taxane (docetaxel or paclitaxel) to that used in the adjuvant setting may be preferred. Both are reasonable options in the first-line setting with neither being definitively superior to the other. Resistance to anthracyclines and taxanes has been defined as disease recurrence occurring within 6 to 12 months of an adjuvant, neoadjuvant or first-line metastatic regimen or while on active treatment.
Anthracycline Combined with Taxanes
It is not clear if the combination of the two most active agents anthracyclines and taxanes are more beneficial compared to sequential use of these agents in the treatment of MBC. This has been compared in several phase III studies evaluating a paclitaxel-based or docetaxel-based combination (23, 92).
ECOG 1193 trial compared doxorubicin 60 mg/m2, paclitaxel 175 mg/m2 over 24 hours, and the combination of doxorubicin 50 mg/m2 followed 3 hours later by paclitaxel 150 mg/m2 over 24 hours, the latter plus granulocyte colonystimulating factor, as first-line therapy in 739 women with MBC (23). Although complete and partial responses, and PFS were significantly higher in the combination arm compared to either the doxorubicin or paclitaxel arms, there was no significant difference in median OS as well as QOL measurements. Cardiac toxicity was equivalent in patients receiving single-agent doxorubicin and combination therapy perhaps due to the dose and administration schedule of the combination arm.
A Spanish Breast Cancer Research Group (GEICAM-9903) phase III study evaluated three cycles of doxorubicin 75 mg/m2 followed by three cycles of docetaxel 100 mg/m2, both every 21 days or six cycles of the combination doxorubicin 50 mg/m2 and docetaxel 75 mg/m2 every 21 days, to determine if sequential therapy could reduce the incidence of hematological toxicity especially febrile neutropenia (primary end point) while maintaining antitumoral activity (secondary end point) (92). Febrile neutropenia was significantly less common in the sequential compared to the combination arm (29.3% vs. 47.8%), and so were other toxicities like asthenia, diarrhea, and fever. There were no significant differences between the sequential versus the combination arms in terms of ORRs (61% vs. 51%), median duration of response (8.7 months vs. 7.6 months), median TTP (10.5 months vs. 9.2 months), and median OS (22.3 months vs. 21.8 months).
The ERASME 3 study compared docetaxel/doxorubicin or paclitaxel/doxorubicin every 3 weeks for four cycles followed by docetaxel or paclitaxel respectively for four cycles as first-line therapy in MBC (93). At a median follow-up of
50.2 months, there was no significant difference in QOL scores (measured after the first four cycles), ORR, PFS, and OS between the two treatment arms. However, hematologic toxicity and asthenia were significantly increased in the docetaxel arm and neuropathy in the paclitaxel arm.
The therapeutic benefit of anthracycline/taxane combinations compared to nontaxane anthracycline combinations has been studied as first-line therapy in several studies (94, 95, 96, 97, 98 and 99). Only a few studies have demonstrated an OS advantage for the anthracycline/taxane combination over anthracycline-based regimens (97, 99). An Eastern European phase III trial of doxorubicin 50 mg/m2 followed 24 hours later by paclitaxel 220 mg/m2 over 3 hours versus 5-FU 500 mg/m2 IV, doxorubicin 50 mg/m2 IV, and cyclophosphamide 500 mg/m2 (FAC) showed an overall RR (68% vs. 55%; p = .032), median TTP ( 8.3 vs. 6.2 months; p = .034), and OS (23.3 vs. 18.3 months; p = .013) significantly in favor of the doxorubicin/paclitaxel arm (97). The percentages of second-line therapy were similar in both arms except that taxane use was more prevalent in the FAC arm (24% vs. 2%). The grade 4 neutropenia rate was significantly higher in the doxorubicin/paclitaxel arm (89% vs. 65%; p < .001), although the incidences of fever, infection, and cardiotoxicity were low. QOL measurements were similar in the two arms.
In another Dutch study (n = 216), first-line doxorubicin 50 mg/m2 followed one hour later by docetaxel 75 mg/m2 compared with FAC showed a significantly higher objective RR (58% vs. 37%; p = .003), TTP (8 vs. 6.6 mo; p = .004), and OS (22.6 vs. 16.2 mo; p = .019) (99). Although febrile neutropenia rates were significantly higher in the doxorubicin/docetaxel arm (33% vs. 9%; p < .001), with two toxic deaths, the congestive heart failure rate was similarly low in both arms (3% vs. 6%). Additional taxanes as second-line therapy was administered to 67% and 23% of patients in the FAC and doxorubicin/docetaxel arms respectively. It should be noted that OS was not a primary study end point, there was a small sample size and the survival on the anthracycline-based arm was particularly poor.
An EORTC study looked at combinations of doxorubicin/paclitaxel versus doxorubicin/cyclophosphamide did not reveal any benefit in terms of RR, PFS, and OS although there was a significantly increased febrile neutropenia rate in the doxorubicin/paclitaxel arm (32% vs. 9%) (95). The doxorubicin/docetaxel combination was also compared to AC in the TAX 306 multicenter, multinational randomized phase III trial (n = 429) and showed a significantly higher RR (59%, CR 10%, PR 49%) than for those taking AC (47%, CR 7%, PR 39% [p = .009]) and TTP (37.3 vs. 31.9 weeks; p = .014), but no OS benefit and similar QOL measurements (98). Some 29% in the AC group received additional treatment with docetaxel compared with only 6% in the doxorubicin/docetaxel group. There was also a higher febrile neutropenia rate in the doxorubicin/docetaxel arm (33% vs. 10%; p < .001), although cardiotoxicity was similarly low in both arms (CHF 3% in doxorubicin/docetaxel and 4% in the AC arms).
In a phase III study comparing 3-weekly docetaxel/doxorubicin/cyclophosphamide (TAC) (75/50/500 mg/m2) to FAC (500/50/500 mg/m2) as first-line in MBC, TAC demonstrated a significantly higher RR (55% vs. 44%; p = .02), but there were no improvements in TTP or OS compared with FAC (94). A significantly higher incidence of grade 3 and 4 hematologic and nonhematologic toxicities, including more cardiotoxicity, was found in the TAC arm. A higher percentage in the FAC group received crossover docetaxel treatment than the TAC group (38% vs. 11%).
In a U.K.-driven trial assessing the combination of epirubicin/paclitaxel versus epirubicin/cyclophosphamide, a significantly better response rate (65% vs. 55%; p = .015) for the epirubicin/paclitaxel arm was demonstrated, although there were no differences in TTP and OS (96). Comparing epirubicin/docetaxel versus epirubicin/cyclophosphamide also did not yield any differences in the efficacy end points of ORR, PFS, and OS in a German study (100).
Two pooled analyses have not found an OS advantage for the anthracycline/taxane combination although there was a better response rate and TTP (91, 101).
In a combined pooled analysis and literature-based metaanalysis of seven phase III prospective randomized trials (three published and found abstracts; n = 2,805), a significant difference was found in favor of anthracycline/taxane combinations over standard anthracycline regimens for ORR (RR 1.21; p < .001), a borderline significance for TTP (RR 1.10; p = .05), but no significant OS difference (101). The neutropenia and febrile neutropenia rates were significantly higher in the anthracycline/taxane arms. This analysis has been hampered by incomplete and non-definitive abstract data, heterogeneity in median follow-up which could have affected survival analysis, and a lack of individualized patient data.
In another analysis consisting of individualized patient data collected on eight randomized combination trials (n = 3,034), there was a significant benefit of anthracycline/taxane combinations over nontaxane anthracycline-based regimens, in terms of response rate and PFS, with no significant difference in OS (91).
On the basis of these results, anthracycline/taxane combinations should not routinely replace anthracycline-based regimens in clinical practice. There was limited statistical power in these trials to detect an OS benefit. Meta-analyses confirmed a better response rate and PFS, but not OS. The disappointing results of the anthracycline/taxane combinations are not completely unexpected as there is no preclinical evidence of synergy between them and both have overlapping and limiting hematological toxicities. These regimens should be reserved for only those patients with good performance status and life-threatening disease.
Other Taxane Combinations
After progression on anthracyclines, taxane combinations may sometimes be used if a higher response rate is required. Certain combinations have shown a survival benefit compared to single-agent taxanes such as docetaxel/capecitabine and paclitaxel/gemcitabine (32, 33). However, there was no planned crossover in the studies, a third arm with singleagent capecitabine or gemcitabine were missing, and these combinations have been associated with increased toxicities.
Based on data from these trials, the FDA approved both combination regimens for treatment of MBC pretreated with anthracyclines: the capecitabine/docetaxel combination indicated for anthracycline and/or taxane-failed disease and the gemcitabine/paclitaxel combination as first-line therapy after failure of prior anthracycline-based adjuvant chemotherapy.
The capecitabine/docetaxel combination capitalizes on the synergistic antitumor activity of these two drugs observed in xenograft models (102). Capecitabine generates 5-FU preferentially in tumor tissue mimicking continuous 5-FU infusion, this tumor selectivity achieved because of the higher activity of thymidine phosphorylase in human tumor tissue compared with healthy tissue (103, 104). Docetaxel also causes upregulation of thymidine phosphorylase and Bcl-2 downregulation (105). Both drugs also capitalize on their nonoverlapping toxicities as docetaxel is myelosuppressive, but capecitabine has a low incidence of myelosuppression. An international phase III trial of 511 patients with unresectable locally advanced or metastatic disease previously exposed to anthracyclines in the neoadjuvant, adjuvant or metastatic setting were randomized to capecitabine 1,250 mg/m2 twice daily for 14 out of 21 days and docetaxel 75 mg/m2 on day 1 every 21 days (n = 255) compared to docetaxel 100 mg/m2 every 21 days (n = 256) (32). In addition, more than 90% of patients had received previous alkylating agents and 5-FU had been administered to about three-quarters of patients. The capecitabine/docetaxel arm demonstrated superior TTP (HR 0.652; p = .0001, 6.1 months vs. 4.2 months), OS (HR 0.775; p = .0126, 14.5 months vs. 11.5 months), and objective tumor response (42% vs. 30%; p = .006). Of note, approximately 30% of patients in each arm were ER-negative which indicates activity of the treatment in this subgroup. More grade 3 adverse events (AEs) in the capecitabine/docetaxel arm necessitated treatment interruption or dose reduction (71% vs. 49%), due largely to handfoot syndrome. The frequency of grade 3/4 neutropenia and neutropenic fever was 24% versus 28% in the combination versus docetaxel arms. Approximately 65% patients in the combination arm required dose reduction of capecitabine alone (4%), docetaxel alone (10%), or both drugs (51%) for AEs, while only 36% required dose reduction for the docetaxel arm. Myalgia, arthralgia, neutropenic fever and sepsis were more common with docetaxel, while hand-foot syndrome, diarrhea and stomatitis were more frequent with docetaxel/capecitabine, although there was no significant difference in QOL scores. A high proportion of patients received poststudy treatment in both arms (70% and 63% in combination and single-agent docetaxel arms respectively), with similar proportions in each arm receiving post-study 5-FU, vinorelbine, anthracyclines, trastuzumab and paclitaxel. Post-study docetaxel was administered in 20% and 7% of the combination and single-agent docetaxel arm respectively; and the use of post-study capecitabine was more common in the monotherapy compared to the combination arm (27% vs. 4%). As no crossover was planned and only a small proportion of patients on docetaxel subsequently received capecitabine, no definitive conclusions can be made regarding the relative merits of combination over sequential single-agent therapy (106). Lower doses of capecitabine and docetaxel may retain the efficacy while reducing the concomitant toxic effects as has been suggested by a retrospective analysis of this trial (107), which is important to consider when dealing with otherwise incurable disease and a primary goal of palliation.
Related posts:
Breast Cancer Screening
Ductal Carcinoma In Situ and Other Intraductal Lesions: Pathology, Immunohistochemistry, and Molecular Alterations
Postmastectomy Radiation Therapy
Preoperative Chemotherapy for Operable Breast Cancer
Management Summary for the Care of Patients with Metastatic Breast Cancer
Nursing Care in Patient Management and Quality of Life
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
