Combination Versus Sequential Single-Agent Therapy in Metastatic Breast Cancer

Abstract

Learning Objectives

After completing this course, the reader will be able to:

1. Describe the current and evolving role of combination therapy in the management of the breast cancer patient with advanced (metastatic) disease.

2. Appreciate the principles of combination treatment design for breast cancer management.

3. Describe those combination treatment regimens, which are designed rationally.

4. Explain the current status of combination or sequential treatment in the management of advanced breast cancer and how clinicians might select options based on patient characteristics.

Access and take the CME test online and receive one hour of AMA PRA category 1 credit at CME.TheOncologist.com

The optimal treatment approach in metastatic breast cancer is controversial. Until recently, the arguments for administering antitumor agents sequentially in the metastatic setting have traditionally outweighed those for administering agents in combination. Older, often empirically designed drug combinations were no more effective and often more toxic than monotherapy, compromising quality of life for little or no clinical benefit. However, more recent studies show that combination and sequential therapy both have their place in the treatment of metastatic breast cancer, with the heterogeneity of breast cancer demanding flexibility in approach. Newer drug combinations, such as paclitaxel/trastuzumab or capecitabine/docetaxel, show survival advantages over single-agent therapy and have manageable safety profiles. Such combination treatments may be preferable to sequential therapy for patients requiring urgent reduction in their tumor burden. Sequential therapy allows the optimal delivery of single-drug therapy and potentially reduces the risk of toxicity, which may improve quality of life. Sequential therapy may be especially appropriate in frail or elderly patients, who may be unable to tolerate the toxicity of combination therapy, or in patients with slowly growing tumors. This article expands on these issues by reviewing trials comparing combination regimens with sequential approaches.

Capecitabine, Taxanes, Breast cancer, Docetaxel, Sequential, Combination

Introduction

The optimal treatment approach in metastatic breast cancer, with antitumor agents given sequentially or in combination, remains controversial. It is doubtful that either strategy is appropriate for all patients. Breast cancer patients are most likely treated on an individual basis according to each patient's demographic and disease characteristics, based on clinical trial data and influenced by the personal experience of the physician. In the future, prospective evaluation of specific subpopulations may help identify optimal treatment approaches for patient groups. This article discusses the advantages and disadvantages of sequential and combination therapies in metastatic breast cancer, given the evidence currently available.

The Current Status of Combination Therapies in Breast Cancer

In the adjuvant setting, combination chemotherapy is used routinely due to widespread evidence that polychemotherapy offers a survival advantage compared with single-agent therapy [1]. A retrospective analysis of trials performed in 1992 by the Early Breast Cancer Trialists' Collaborative Group showed that adjuvant treatment with combination chemotherapy regimens was associated with a 12% reduction in the risk of death compared with no adjuvant treatment, translating into a 4.4% reduction at 5 years (unpublished data). The same group showed that anthracycline-containing regimens provided advantages in overall and recurrence-free survival rates over non-anthracycline-containing regimens [2]. Consequently, anthracycline-containing polychemotherapy became a standard therapy in the adjuvant setting.

In contrast, the use of combination treatments in the metastatic/advanced cancer setting is more controversial. A meta-analysis of trials in the metastatic setting published by Fossati et al. concluded that there was a survival benefit for polychemotherapy over single-agent therapy, the hazard ratio of 0.82 (0.75-0.90) translating into an 18% reduction in the risk of death [3]. However, most of the trials included were small (median number of patients <150), were sometimes poorly designed, and often featured old regimens. For example, the overall hazard ratio was biased by a 30% reduction in the risk of death in a subset of trials in which the combination treatment was based on cyclophosphamide and 5-fluorouracil (5-FU) [3]. The entire number of patients in the meta-analysis (n <1,000) only modestly exceeds that in a single, recent, important randomized trial of sequential single-agent versus combination chemotherapy conducted by the Eastern Cooperative Oncology Group (n = 739) [4], as discussed later in this review.

An obvious important difference between the adjuvant and the metastatic disease settings is life expectancy. In the meta-analysis by Fossati et al., anthracycline-containing combination regimens reduced the risk of death by 13% when compared with monotherapy [3], a reduction similar to that discussed above for polychemotherapy versus single-agent therapy in the adjuvant setting. However, the life expectancy for patients receiving treatment in the metastatic setting is much shorter, typically months rather than years, than for those treated in the adjuvant setting. Furthermore, patients treated for metastatic disease typically receive several chemotherapy regimens, and there is general concern that any survival advantage may be counterbalanced by additive toxicity. Consequently, as treatment in the metastatic setting is palliative, tolerability and patients' quality of life are major factors guiding treatment decisions.

Principles of Combination Therapy Design

Much of the concern about the toxicity and effectiveness of combination therapy has arisen from empirically chosen combination treatments. Ideally, a combination therapy should meet three criteria:

• Each component should have single-agent activity with no cross-resistance

• There should be preclinical evidence of synergy between the components

• The components should have nonoverlapping safety profiles

However, all three criteria are rarely met. Consequently, many combination therapies have failed to significantly improve outcomes compared with sequential administration of single agents because their components are administered at suboptimal doses due to dose-limiting toxicities.

Combination Therapies that are not Optimally Designed

An example of a combination therapy that does not meet all of the criteria described above, but which is frequently used in breast cancer, is the combination of a taxane and an anthracycline. There is little preclinical evidence of synergy, and myelosuppression is a major toxicity with both classes of agents [5]. The primary rationale for combining these agents is that each has high single-agent activity [6, 7]. Therefore, it is not surprising that the clinical impact of taxane/anthracycline combination regimens in metastatic breast cancer has been disappointing. Taxane/ anthracycline combinations have been proven to be only marginally more effective and often more toxic treatments for metastatic breast cancer than anthracycline/cyclophosphamide combinations (Table 1) [8–13]. A difference in survival was reported for only one of five studies [9]. A superior response rate and superior time to disease progression were seen in two trials [9, 10], although the clinical relevance of a 5-week difference in time to disease progression at the median may be questionable [11]. In terms of safety, three trials reported markedly higher incidences of febrile neutropenia in the taxane/anthracycline arms than in the anthracycline/cyclophosphamide arms [8, 9, 11].

Table 1

Review of trials comparing anthracycline/taxane combinations with anthracycline/cyclophosphamide combinations

Reference	Regimen	Response rate (%)	Median time to disease progression (months)	Median overall survival (months)	Febrile neutropenia (%)
Biganzoli et al., 2002 [8]	AP versus AC	58/54 NS	6.0/6.0 NS	20.6/20.5 NS	32/9 (p <0.001)
Jassem et al., 2001 [9]	AD versus FAC	68/55 (p = 0.032)	8.3/6.2 (p = 0.034)	23.3/18.3 (p <0.013)	68/55 (p <0.001)
Nabholtz et al., 1999 [10]	AD versus AC	60/47	7.3/8.5	NR	33/10
Nabholtz et al., 2001 [ 11]		(p = 0.012)	(p = 0.01)		(p <0.001)
Lück et al., 2000 [12]	EP versus EC	46/40 NS	9.0/7.4 NS	16.8/20.3 NS	2/2 NS
Carmichael et al., 2001 [13]	EP versus EC	67/56 NS	6.5/6.7 (p = 0.072)	13.7/13.8 (p <0.092)	NR

Abbreviations: A = doxorubicin; C = cyclophosphamide; D = docetaxel; E = epirubicin; F = 5-FU; P = paclitaxel; NR = not reported;
 NS = not significant.

Another problematic drug combination according to the criteria above is doxorubicin/vinorelbine. Several trials have investigated anthracyclines with or without vinorelbine. In one of these, 303 patients with metastatic breast cancer were randomized to either doxorubicin, 50 mg/m² on day 1, plus vinorelbine, 25 mg/m² on days 1 and 8, or doxorubicin, 70 mg/m² alone on day 1 of a 21-day cycle [14]. As both components are associated with myelosuppression, toxicity was significant in the combination arm. After 16 of the first 59 patients developed febrile neutropenia, the regimens were changed to doxorubicin, 40 mg/m², plus vinorelbine, 20 mg/m², versus doxorubicin, 60 mg/m² alone. At the end of the trial, there was no significant difference between treatments for any of the efficacy parameters (Table 2). Moreover, toxicity was worse in the combination arm than in the monotherapy arm. There were significantly higher incidences of neurotoxicity/neuropathy, constipation, and venous reactions in the combination arm than in the monotherapy arm [14]. Doxorubicin/ vinorelbine, therefore, provides no advantage over doxorubicin monotherapy.

Table 2

Doxorubicin plus vinorelbine combination therapy versus doxorubicin alone [14]

	Doxorubicin plus vinorelbine (n  = 145)	Doxorubicin (n  = 144)	p value
Overall response rate (%)	38.0	30.0	NS
Median duration of response (months)	7.2	6.8	NS
Median time to disease progression (months)	6.2	6.1	NS
Median overall survival (months)	13.8	14.4	NS

Abbreviation: NS = not significant.

It has also been shown that CEF (cyclophosphamide/ epirubicin/5-FU) does not offer superior efficacy to single-agent mitoxantrone [15]. In that study, 412 high-risk patients with metastatic breast cancer were randomized to either 500 mg/m² cyclophosphamide, 50 mg/m² epirubicin, and 500 mg/m² 5-FU, or to 12 mg/m² single-agent mitoxantrone, given every 3 weeks. There was no difference between the CEF arm and the mitoxantrone arm in terms of disease control (complete or partial response or stable disease: 66% versus 57% of patients, respectively, p = 0.182), time to disease progression (median 4.4 versus 6.2 months, respectively, p = 0.18), and overall survival (median 15.8 versus 14.1 months, respectively, p = 0.66).

In that study, quality-of-life-adjusted efficacy (modified Brunner score) was also calculated by adding scores assigned for time to disease progression, change in performance status, patient's subjective judgment of treatment utility, and toxicity. There were no significant differences between treatments for any of the parameters except toxicity (p = 0.0001). The overall quality-of-life-adjusted efficacy scores were -2.07 for the combination arm and 3.92 for the monotherapy arm, indicating that monotherapy was preferable to the combination therapy.

Trials Comparing Combination with Sequential Therapy Regimens

The trials described above and the meta-analysis by Fossati et al. have compared combination therapies with either other combination therapies or monotherapies [3]. However, perhaps a more pertinent question is how combination therapy compares with sequential chemotherapy. This question has largely gone unanswered, at least partly because the term “sequential chemotherapy” can be interpreted in several ways. It can mean sequential administration of chemotherapy, each new regimen introduced following disease progression, or it can mean a planned multicourse sequence of chemotherapy without a break between each different treatment. Furthermore, chemotherapy may be either monotherapy or polychemotherapy. Consequently, the design of trials comparing combination therapy with sequential therapy has been difficult, and only three trials have been reported examining this question to date.

The first trial to compare combination therapy with sequential therapy was a three-arm, phase III trial by Sledge et al. [4, 16]. A total of 739 chemotherapy-naïve patients were randomized to one of three treatment regimens: doxorubicin, 60 mg/m², followed at progression by paclitaxel, 175 mg/m²/day; paclitaxel, 175 mg/m²/day, followed at progression by doxorubicin, 60 mg/m²; or a combination regimen of doxorubicin, 50 mg/m², plus paclitaxel, 150 mg/m²/day. The overlapping toxicity profiles of taxanes and anthracyclines and resulting risk of toxicity made it necessary to use lower doses of doxorubicin and paclitaxel in the combination arm than in the single-agent arms as well as G-CSF to ameliorate anticipated neutropenia.

The combination regimen resulted in a superior response rate and time to treatment failure compared with either of the sequential therapies (Table 3). However, combination therapy did not result in a survival benefit. There was no difference between the treatment arms in terms of safety or quality of life (functional assessment of cancer therapy-B scale).

Table 3

Doxorubicin and paclitaxel: combination therapy versus sequential therapy [16]

Regimen	A then P	P then A	A + P
Response rate (%)	34.0	33.0	46.0*
Median time to disease progression (months)	6.2	5.9	8.0*
Median overall survival (months)	20.1	22.2	22.4
Grade 3/4 adverse events (n of patients)	109.0	169.0	135.0
Treatment-related deaths (n of patients)	6.0	2.0	4.0

*p <0.01, A + P versus A then P and P then A

Abbreviations: A = doxorubicin, P = paclitaxel.

Agents of the same class were used in a second, more recent trial comparing combination therapy with sequential therapy [17]. In this phase II trial, 193 patients were randomized to docetaxel, 100 mg/m², followed by doxorubicin, 75 mg/m², at progression (or after four cycles without progression), or to docetaxel plus doxorubicin combination therapy at one of two reduced dose levels: docetaxel, 75 or 60 mg/m², plus doxorubicin, 50 or 60 mg/m². Although there was no formal statistical analysis, there was no clinical difference among arms for overall survival, time to disease progression, response rate, or toxicity (Table 4). This trial, although smaller and less powerful, supports the notion that taxane/anthracycline combination therapy provides no advantage over sequential administration of the same agents.

Table 4

Doxorubicin and docetaxel: combination therapy versus sequential therapy [17]

Regimen	T 75 mg + A 50 mg	T 60 mg + A 60 mg	T 100 mg then A 75 mg
Response rate (%)	49.0	12.0	21.0
Median time to disease progression (months)	6.7	8.3	6.9
Median overall survival (months)	11.9	14.5	13.8
Grade 3/4 neutropenia (%)	85.0	90.0	85.0

Abbreviations: A = doxorubicin; T = docetaxel.

The third trial to compare combination therapy with sequential therapy was conducted by Joensuu et al. [18]. This trial had a very relevant design in which 303 patients with metastatic breast cancer were initially randomized to either monotherapy or combination therapy as first- and second-line treatment. In the monotherapy arm, patients received epirubicin, 20 mg/m² every week until progression (or until the cumulative dose was 1,000 mg/m²), and then mitomycin C, 8 mg/m² every 28 days. In the combination therapy arm, patients received cyclophosphamide, 500 mg/m², epirubicin, 60 mg/m², and 5-FU, 500 mg/m² every 21 days until progression, followed by mitomycin C, 8 mg/m², plus vinblastine, 6 mg/m², every 28 days (n = 88). No significant difference in survival (p = 0.93) was found between the two arms, but quality of life and tolerability were more favorable in the sequential monotherapy arm.

The three trials described above comparing combination therapy with sequential monotherapy have not provided evidence of an advantage with the combination approach.

Rationally Designed Combination Therapies

In contrast to the combinations described above, trastuzumab (a recombinant, humanized anti-HER2 antibody) plus a taxane represents a rationally designed combination fulfilling the criteria described earlier. Each component has single-agent activity with distinct mechanisms of action and without cross-resistance. There is also preclinical evidence of synergy: trastuzumab plus a taxane showed clearly enhanced antitumor activity in xenograft models [19]. Finally, the two components have nonoverlapping toxicities. It has now been established that trastuzumab plus paclitaxel offers a survival advantage of 25% over paclitaxel alone in women with metastatic breast cancer overexpressing HER2 [20].

Another rationally designed combination therapy is capecitabine plus docetaxel. Both agents have considerable single-agent activity and distinct mechanisms of action. There is preclinical evidence of synergy, as indicated in xenograft models [21]. Potential mechanisms underlying the observed synergy include thymidine phosphorylase (TP) upregulation and Bcl-2 downregulation, as shown in Figure 1 [21, 22]. Finally, the two agents have nonoverlapping key toxicities: the predominant toxicity with docetaxel is myelosuppression, but capecitabine monotherapy is associated with a particularly low incidence of myelosuppression.

Figure 1

Mechanism of capecitabine/docetaxel synergy.

Capecitabine/docetaxel combination therapy, compared with docetaxel monotherapy, significantly improves overall survival, time to disease progression, and response rates in women whose breast cancer has progressed during or following anthracycline treatment [23] (as described by O'Shaughnessy et al. pp. 4-12 [24]). Despite the lower dose of docetaxel administered in the combination arm, the addition of capecitabine to docetaxel resulted in greater efficacy compared with the full dose of docetaxel monotherapy. The overall survival curves separated early and maintained a clear separation over time. This survival advantage, associated with the combination therapy, was maintained despite balanced numbers of patients receiving poststudy therapy [25]. In terms of safety, the incidence of grade 3/4 adverse events was higher in the combination arm than the docetaxel-only arm (71% versus 49%), due primarily to grade 3 hand-foot syndrome, which was managed by dose adjustment of capecitabine.

Of note, the survival benefit in the capecitabine/docetaxel combination arm was achieved without compromising quality of life. This is an important finding as the overall spectrum of toxicity was greater. Nevertheless, in that study, the impact of toxicity on perceived quality of life did not differ between treatment arms.

Figure 2 highlights an evolution of therapy in patients with anthracycline-pretreated breast cancer since Nabholtz et al. demonstrated that it was possible to extend the survival of such patients with docetaxel monotherapy, compared with mitomycin C plus vinblastine [26], to the current survival benefit offered by capecitabine plus docetaxel [23]. The combination of capecitabine plus docetaxel has not been evaluated against other combination regimens, but none of these have been shown to improve survival compared with docetaxel monotherapy.

Figure 2

The evolution of therapy for anthracycline-pretreated metastatic breast cancer: probability of survival in two randomized trials.

However, it should be emphasized that neither the capecitabine/docetaxel trial [23] nor the trastuzumab/paclitaxel trial [20] has answered the question of whether combination therapy is preferable to sequential therapy because these combinations were compared only with monotherapy and a formal crossover was not part of the trial design. Indeed in the capecitabine/docetaxel trial, among the 164 patients initially randomized to docetaxel alone who received any kind of poststudy chemotherapy, only one quarter (27%) subsequently received capecitabine monotherapy [23], and those patients appeared to have a survival advantage over those receiving other poststudy chemotherapies [25]. The survival advantage of combination over true sequential therapy (capecitabine then docetaxel or docetaxel then capecitabine) has not been demonstrated, and this question remains, as yet, unanswered.

The Choice Between Combination and Sequential Therapy

Combination regimens appear to offer no advantage over sequential monotherapy in the metastatic setting, where palliation is traditionally the primary goal of treatment. However, newer, rationally designed combinations may offer clinically meaningful survival advantages over single monotherapy. Some cytotoxic combinations achieve this clinical benefit with a manageable safety profile, making them more attractive options in the metastatic setting. This was demonstrated in the capecitabine/docetaxel trial [23]. Furthermore, the early separation of the survival curves in the capecitabine/docetaxel trial shows that a high proportion of patients may benefit from combination treatment, whereas 30%-40% of patients initially randomized to docetaxel alone did not receive further therapy after disease progression. Combination therapy may be particularly useful in patients needing a prompt reduction in tumor burden (e.g., patients with visceral disease and fast-growing tumors). Therefore, as in the adjuvant setting, survival data should play a more prominent role in motivating the choice of treatment.

However, as described above, the majority of combinations are more toxic than monotherapy alternatives, and the importance of maintaining quality of life cannot be underestimated in the metastatic setting. In the adjuvant setting, chemotherapy causes a temporary impairment in patients' quality of life. Nonetheless, most patients are willing to receive adjuvant chemotherapy with the potential for clinical benefit [27, 28]. In patients with metastatic disease, the survival benefit associated with treatment is generally modest at best, and therefore, maintenance of quality of life assumes greater importance, although patients may be unwilling to trade quantity for quality of life [29].

The newer drug combinations, however, do not entirely supplant sequential monotherapy in the metastatic setting. In multiple randomized trials designed to compare these strategies, no survival benefit has been seen with combination versus true sequential therapy. The combination of capecitabine/docetaxel has not yet been compared with true sequential therapy. Sequential monotherapy allows the optimal delivery of single-agent therapy, potentially reducing the risk of toxicity, and thereby improving quality of life. Sequential monotherapy may be especially appropriate in patients with slowly progressing tumors or in frail or elderly patients who may be unable to tolerate the toxicity of combination therapy.

It can be concluded that combination and sequential therapy may both have their place in the metastatic breast cancer setting, and the heterogeneity of breast cancer argues for flexibility in approach. For example, in patients for whom sequential monotherapy may not be optimal (e.g., those with more rapidly progressing disease and/or visceral metastases), capecitabine/docetaxel combination therapy could be considered optimal therapy. However, in some groups of patients, such as those with poor performance status or more indolent disease, sequential monotherapy may be more appropriate, and oral capecitabine, an active monotherapy for breast cancer, can be considered a valuable option.