In North America, approximately 182,000 women are diagnosed with breast cancer each year, and 46,000 women die of the disease . As many as 10% of these women present with metastatic disease at the time of diagnosis, and a substantial proportion of patients initially diagnosed with localized disease will develop metastases. The vast majority of women who develop metastatic disease will ultimately die of it, with a median survival of 2 to 3 years [2,3].
The prognosis of patients with advanced breast cancer is highly variable; some patients die within a few months of developing metastatic disease and others survive for 10 years or longer. More treatment options exist for patients with metastatic breast cancer than for most patients with cancer. A wide range of both hormonal therapies and cytotoxic treatments have been in use over the past 10 to 20 years, and several new agents have recently become available. Although systemic therapy probably prolongs survival [4,5], definitive evidence that treatment prolongs life is difficult to find. The primary goal of treatment is to improve quality of life by palliating the symptoms related to cancer. The heterogeneity among patients, the wide range of treatment options, and the need to weigh the risks versus the benefits of available therapies pose a challenge for medical oncologists caring for women with advanced breast cancer.
For patients with estrogen and/or progesterone(Drug information on progesterone) receptor-positive tumors, hormonal therapy is often the first-line treatment of metastatic disease because of its high degree of efficacy and minimal toxicity. Chemotherapy is usually recommended for patients with hormone receptor-negative tumors or for patients who have become refractory to hormonal therapy.
When chemotherapy is indicated, combination therapy is often administered. Some of the most common first-line regimens are listed in Table 1. Response rates for these regimens range from approximately 35% to 75%.6 Doxorubicin(Drug information on doxorubicin)-containing regimens are usually associated with higher response rates than regimens that do not contain this agent. For this reason, many oncologists opt for doxorubicin-based regimens as initial chemotherapy for patients with metastatic breast cancer, particularly for patients with troublesome cancer symptoms. Less toxic regimens (eg, cyclophosphamide(Drug information on cyclophosphamide) [Cytoxan, Neosar], methotrexate(Drug information on methotrexate), and fluorouracil(Drug information on fluorouracil) [CMF] or mitoxantrone(Drug information on mitoxantrone) [Novantrone]-containing regimens) are often administered to older patients, individuals with comorbidities, or patients who have particular difficulty tolerating the side effects of chemotherapy. Data from the M. D. Anderson Cancer Center, however, suggest that age alone is not a reason to avoid an anthracycline-based regimen .
For patients with metastatic disease who have received adjuvant treatment with doxorubicin, a taxane-based regimen, usually single-agent paclitaxel(Drug information on paclitaxel) (Taxol), often is used as first-line treatment in the metastatic setting [8,9]. Gianni et al  have reported a response rate in excess of 90% with a combination of paclitaxel and doxorubicin as first-line therapy for metastatic disease. Unfortunately, this regimen was also associated with higher cardiotoxicity than would have been expected given the total doxorubicin dose. Additional clinical trials evaluating doxorubicin/paclitaxel regimens must be completed before this option can be considered a standard first-line regimen.
The use of high-dose chemotherapy with autologous bone marrow or peripheral stem cell support has gained great attention in recent years, but this treatment is usually reserved for younger patients who have chemotherapy-responsive disease . Of the total population of patients with metastatic breast cancer in North America, a relatively small percentage receive this intensive treatment during the course of their illness.
Regardless of the choice of first-line chemotherapy, virtually all patients will develop disease progression, often after an initial response to treatment or a period of disease stabilization. Most patients with metastatic breast cancer are potential candidates for second-line chemotherapy. Although large population- based studies of treatment of advanced breast cancer have not been performed, the majority of patients with breast cancer in the United States probably receive two or more chemotherapeutic regimens in the metastatic setting. In the remainder of this article, the term "refractory breast cancer" will be used to describe patients with hormone-refractory metastatic breast cancer whohave developed disease progression following an initial chemotherapeutic regimen.
A number of chemotherapeutic agents and regimens are available for patients with refractory breast cancer. As with first-line therapy, the primary goal of treatment is the palliation of cancer-related symptoms. If there is any survival advantage as a result of treatment of refractory disease, it is almost certainly of a very small magnitude. In this setting, the potential for treatment-induced toxicity must be considered carefully before any therapy is initiated. Unfortunately, only a limited number of patients who receive treatment of refractory breast cancer will have an objective (or even subjective) response to treatment, but all patients (whether or not they derive benefit from treatment) are at risk for developing minor side effects and/or debilitating complications. For patients with few cancer symptoms, the option to withhold further chemotherapy until symptoms worsen is always worthy of consideration. Many physicians have difficulty with this "watchful waiting" approach, particularly for younger patients, for whom they may hope that a response to treatment could offer some prolongation of life. Many patients are also more comfortable receiving active treatment. In these situations, especially when patients are entirely asymptomatic, treatment regimens without serious toxicity should be strongly considered. Although continuing active treatment may facilitate a sense of hopefulness, it is difficult to know how much this enhanced hopefulness is counterbalanced by the side effects, risks, and inconvenience that result from chemotherapy.
As we consider new treatments for refractory breast cancer, it is helpful to view these treatment programs from a historical perspective. Between 1975 and 1991, 1,756 patients underwent treatment of metastatic breast cancer at Guy's Hospital in London . A total of 758 (43%) received first-line chemotherapy for metastases. More than half of these patients received an anthracycline-containing regimen as initial treatment, with most of the remaining patients receiving CMF. The overall response rate in patients receiving first-line therapy was 35%, and the median time to progression in responding patients was 7.8 months. As expected, a longer disease-free interval and the absence of metastases were both predictive of a response to therapy.
Of patients who underwent first-line treatment, approximately one-third received a second-line chemotherapeutic regimen. The overall response rate to second-line therapy was only 16%; patients who had responded to first-line treatment had a significantly higher chance of responding to treatment in the refractory setting than did patients who had not responded to first-line treatment. Of interest, the response rate seen in 58 patients who received third-line therapy was similar to that seen in the second-line experience, including a small number of responses in patients who had failed to respond to either of the first two regimens.
Investigators from Finland  have described the outcome of treatments administered for refractory disease in a group of patients who had participated in a study evaluating two different first-line anthracycline-based chemotherapeutic schedules. Although response to therapy could only be determined in approximately half of the treatment courses, the results echoed those of the Guy's Hospital experience. Where data were available, the response rates to second- and third-line chemotherapy were 20% and 12%, respectively. It is important to note that these results, like those from Guy's Hospital, do not reflect the availability of some of the newer cytotoxic agents, such as paclitaxel and docetaxel(Drug information on docetaxel) (Taxotere). Nevertheless, such findings indicate the need to evaluate critically the risks and benefits of treatment programs in this population. Of interest, occasional responses were seen with third-line chemotherapy in both the Guy's Hospital and Finnish series. This finding suggests that the use of chemotherapy cannot be arbitrarily discontinued after a patient has received two regimens.
A wide range of combination regimens have been evaluated over the past 5 to 10 years; a partial list of these regimens appears in Table 2 [14-23]. Response rates vary but rarely exceed the 30% to 40% range and are often much lower. The median duration of response in responders is consistently in the range of 5 to 7 months. Because most regimens have been evaluated in the setting of single-institution phase II trials, it is very difficult to compare one regimen with another. It remains unclear whether there is any advantage to combination therapy compared with single agents. Clearly, there is no combination regimen that could be considered the gold standard for patients with refractory disease.
Although some combination regimens are well tolerated, the toxicity of most multiagent regimens tends to be higher than that of single agents. Mitomycin(Drug information on mitomycin) (Mutamycin), which is often combined with one of the vinca alkaloids or other agents, is active in refractory disease. Unfortunately, it is highly myelosuppressive and can limit options for subsequent treatment regimens because of its profound effect on bone marrow reserve. With the availability of new agents, the use of mitomycin has probably declined in recent years. Cisplatin(Drug information on cisplatin) (Platinol), another agent often found in combination therapy, is active in patients with no prior exposure to chemotherapy in the metastatic setting . As a single agent, it has limited efficacy in patients with refractory disease, and combinations containing cisplatin tend to be quite toxic. For these reasons, cisplatin has no established role in the treatment of refractory breast cancer. Another platinum analog, carboplatin(Drug information on carboplatin) (Paraplatin), has no definitive role in the treatment of refractory metastatic disease [17,18].
All the agents used in the combinations previously discussed, as well as several other cytotoxic agents, have been evaluated as single agents for patients with refractory breast cancer. In particular, fluorouracil (5-FU) is quite active in patients with refractory disease, especially when it is administered as a continuous infusion or in combination with leucovorin [25-27]. Toxicity with continuous infusion 5-FU is usually quite modest; when administered as a low-dose infusion over a number of weeks, it causes minimal myelosuppression and few serious toxicities. Although 5-FU has been deleted from some adjuvant regimens in recent years, it remains an effective agent in patients with refractory disease.
Table 3 includes trials with single-agent paclitaxel, docetaxel, and vinorelbine (Navelbine) [9,28-33]. Each agent is active in patients with refractory breast cancer. In many centers, paclitaxel has become the standard second-line regimen, based on promising results from initial phase II trials [9,28] and subsequent approval by the Food and Drug Administration (FDA). A number of questions concerning the optimal use of paclitaxel in the treatment of breast cancer remain. Ongoing studies are addressing several of these questions, including the importance of dose intensity and the role of different infusion durations.
Docetaxel (Taxotere) has been found to have remarkable activity in patients with anthracycline-resistant refractory breast cancer [30,31]. Unfortunately, significant fluid retention with cumulative doses of docetaxel has slowed the development of this agent, but premedication with steroids has decreased the severity of this problem. It will be important to see whether the promising early results are sustained with additional clinical trials. [Docetaxel was recently appproved by the FDA for treatment of patients with locally advanced or metastatic breast cancer that has progressed during anthracycline-based therapy or has relapsed during anthracycline-based adjuvant therapy.--Ed.]
Finally, vinorelbine has also been demonstrated to be active in patients with refractory disease [32,33], including patients with anthracycline-refractory and both anthracycline- and taxane-refractory breast cancer . Although response rates with single-agent vinorelbine have been somewhat lower than those seen in some of the taxane trials, the toxicity of vinorelbine is quite mild. Vinorelbine has yet to gain FDA approval for use in patients with breast cancer, but this agent ultimately may play an important role in the treatment of patients who cannot tolerate side effects from therapy or patients who have had severe toxic effects with prior treatment regimens.
Is there an ideal single agent for patients with refractory breast cancer? The single agents previously discussed have yet to be compared with one another or with combination regimens in randomized clinical trials, although some studies are underway. Until further information is available, the decision to treat patients with either a single agent or a combination regimen must be based on individual symptoms, comorbid illnesses, response to prior chemotherapy, ability to tolerate different types of toxicities, and eagerness to proceed with treatment. Whenever possible, patients should be encouraged to participate in clinical trials that address unresolved questions about the use of current therapies or that seek to identify new treatments for refractory breast cancer.
Little research has addressed physicians' attitudes about the treatment of patients with refractory breast cancer. In a study conducted several years ago, Benner and colleagues  surveyed oncologists in Maryland to characterize their knowledge and attitudes about the use of chemotherapy for hormone-unresponsive metastatic breast cancer. A total of 71% of oncologists cited palliation as the primary goal of treatment. Although they were enthusiastic about recommending first-line chemotherapy to their patients, they were less likely to encourage patients to consider second-line treatments. Of interest, however, is the fact that 47% of physicians indicated that they would not reduce the dose of second-line therapy, even if a patient had experienced febrile neutropenia. Whether physicians' attitudes vary in different geographic regions is unknown. It is also unclear as to what extent the availability of new agents such as the taxanes has affected practice patterns. Further research in this area will provide a fuller understanding of physicians' attitudes and expectations with regard to the treatment of refractory breast cancer.
Although palliation is the primary goal of treatment for patients with refractory breast cancer, obtaining a clear assessment of palliation is often quite difficult. Quality-of-life issues are often discussed, but incorporating meaningful quality-of-life measurements into clinical trials (and into clinical practice) is an arduous task. A number of validated quality-of-life instruments have been developed and are available as a means to assess outcomes in patients with refractory breast cancer. To date, few studies have successfully incorporated quality-of-life end points. Coates and colleagues  documented improvements in quality of life as a result of chemotherapy despite the side effects of treatment. Their study, however, was limited to first-line chemotherapy. In the future, studies of patients with refractory breast cancer should include quality-of-life assessments that can complement the more traditional end points of response and survival. Quality-of-life assessments can capture the subjective improvements often seen in patients who respond to treatment. As the number of treatment options increases for patients with breast cancer, well-designed quality-of-life assessments can help physicians and their patients choose from among the various alternatives.