It has been estimated that close to 213,000 persons will be diagnosed with breast cancer in the United States in 2005.[1] Most of these patients will present with locoregional breast cancer, which relapses despite appropriate initial treatment. Mortality from breast cancer has decreased substantially, especially in patients with localized disease.[2] Both endocrine therapy and chemotherapy have led to substantial increases in survival among women with early-stage breast cancer.[3] Adjuvant endocrine therapy, which remains an essential component of treatment for patients with hormone receptor-positive disease, is not discussed in this article.[4,5] Adjuvant chemotherapy is a work in progress: It started as single-agent treatment,[6] was followed in the 1970s by combination chemotherapy with CMF (cyclophosphamide, methotrexate(Drug information on methotrexate), and fluorouracil(Drug information on fluorouracil) [5-FU]),[7] and now entails the use of various effective agents. Among the chemotherapy agents that are associated with greater survival are older drugs such as the alkylating agents (mainly cyclophosphamide(Drug information on cyclophosphamide)), the antimetabolites (methotrexate and 5-FU), the anthracyclines (epirubicin [Ellence] and doxorubicin(Drug information on doxorubicin)), and more recently, the taxanes (paclitaxel and docetaxel(Drug information on docetaxel) [Taxotere]). The 15-year update of findings from the Early Breast Cancer Trialists' Collaborative Group meta-analysis has shown that the risk of recurrence is 23.5% lower and the risk of mortality is 15.3% lower with several months of adjuvant combination chemotherapy than with no chemotherapy (both 2P < .001).[8] This article discusses the adjuvant regimens that are associated with the best outcomes, the selection of regimens on the basis of patient characteristics, and trials that have been completed or are in progress and are expected to substantially increase our knowledge in the next several years. Strategies for the Use of Combination Chemotherapy Choice of Agents

• Anthracycline and Nonanthracycline Regimens-The 2001 update of the Early Breast Cancer Trialists' Collaborative Group meta-analysis reported that mortality was lower with anthracycline-containing regimens than with nonanthracycline regimens- a further proportional reduction of 11% in the annual rate and an absolute reduction at 5 years of about 3%.[8] These reductions were greater in younger women and were not affected by lymph node status. This meta-analysis includes data on anthracyclines from several large trials of epirubicin(Drug information on epirubicin)-containing regimens. Doxorubicin- and epirubicin-containing regimens appear similar in efficacy, but epirubicin-containing regimens may be less cardiotoxic.
The optimal duration of adjuvant therapy with anthracycline-containing regimens is still being investigated. The efficacy of 2 months (four cycles) of AC (doxorubicin [Adriamycin] and cyclophosphamide) and of 6 months (six cycles) of classic CMF using oral cyclophosphamide are similar.[ 9] Six months (six cycles) of an anthracycline-containing regimen (CEF [cyclophosphamide, epirubicin, and 5-FU) does, however, appear to be superior to six cycles of CMF in terms of disease-free and overall survival.[ 10] Furthermore, preliminary results from the National Epirubicin Adjuvant Trial (NEAT), a phase III trial that compared six cycles of classic CMF with four cycles of epirubicin followed by four cycles of CMF in 2,021 patients, showed significantly greater rates of recurrence-free survival (P < .001) and overall survival (P < .001) in the epirubicin-containing arm.[11] These two studies suggest that treat ment with more than four cycles of chemotherapy (particularly anthracycline- containing regimens) is associated with the best outcomes. The ongoing CALGB 40101 clinical trial, which is comparing four and six cycles of AC in women with node-negative breast cancer, should help establish the optimal duration of anthracycline-based adjuvant therapy (Table 1).
• Taxanes-There remains substantial room for improvement, however, and treatments with nonoverlapping toxicity are needed. The taxanes paclitaxel(Drug information on paclitaxel) and docetaxel have been shown to be effective in both the metastatic and adjuvant settings (Figure 1).[12-19] Tumor cells have little cross-resistance to taxanes and anthracyclines, and the taxanes can be used with almost all other drugs in a variety of combinations and schedules. Single-agent taxanes are effective in metastatic disease, and combinations of taxanes and other agents have produced higher response rates and longer times to progression than older combination regimens.[20-22]
Paclitaxel and docetaxel have similar antitumor activity, owing to their high-affinity binding to beta-tubulin and subsequent blocking of cell division.[ 23] Paclitaxel has pharmacokinetic interactions with anthracyclines, resulting in greater cardiotoxicity when it is used concurrently with doxorubicin (if the cumulative dose of doxorubicin exceeds 300 mg/m²)[24,25] and greater myelotoxicity when it is used concurrently with epirubicin.[26] To minimize toxicity, paclitaxel should be administered after the anthracyclines. Unlike paclitaxel, docetaxel has no clinically significant effects on the pharmacokinetics of anthracyclines.[23] Ongoing trials that are expected to help define the optimal use of these agents are summarized in Table 1.
• Paclitaxel-Cancer and Leukemia Group B (CALGB) trial 9344, the first large randomized trial to evaluate paclitaxel in the adjuvant setting, randomized 3,121 patients with nodepositive primary breast cancer to four cycles of AC, using three different doses of doxorubicin (60, 75, and 90 mg/m²), followed by either four cycles of paclitaxel (175 mg/m² every 3 weeks) or no further treatment.[14] The risk of recurrence was 17% lower (P = .002) and risk of death was 18% lower (P = .006) with the addition of paclitaxel, with a moderate increase in toxicity. The survival benefit with paclitaxel was similar to that with an anthracycline-containing regimen over CMF.[3] Higher doses of doxorubicin, however, provided no improvement.
National Surgical Adjuvant Breast and Bowel Project (NSABP) trial B-28 also compared adjuvant AC alone with AC followed by paclitaxel, in 3,060 patients with node-positive operable breast cancer, using a higher dose of paclitaxel than that in CALGB 9344 (225 rather than 175 mg/m²). The addition of paclitaxel reduced the risk of recurrence by 17% (P = .008) but did not have an effect on overall survival (P = 0.48).[15] Further followup is needed. Factors that may have contributed to the lack of an overall survival benefit in NSABP B-28 are that the patients had a better prognosis than those in CALGB 9344 and that a high proportion (25%) of the patients in the paclitaxel arm could not be treated with all four cycles of paclitaxel because of toxicity.[15] The results of NSABP B-28 may have been influenced by the concurrent use of tamoxifen(Drug information on tamoxifen) with the chemotherapy, which has been associated with lower survival than that with the use of tamoxifen after the chemotherapy.[27] It is also possible that the treatment effect that was seen in both trials may be due to the duration of treatment rather than to the addition of a sequential, non-cross-resistant agent. A recent randomized trial compared two paclitaxel regimens in 1,830 patients with stage I, II, or III breast cancer: four cycles of AC (60/600 mg/m²) followed by paclitaxel (Taxol) at 175 mg/m² every 3 weeks (AC→T) and four cycles of doxorubicin and paclitaxel (50/200 mg/m²) followed by paclitaxel at 80 mg/m²/wk for 12 weeks (AT→T).[28] At 3 years, there were significant differences favoring AT→T in both disease-free survival (90% vs 86%; P = .02) and overall survival (95% vs 92%; P = .02). The incidence of neuropathy was greater in the AT→T arm, but there was no difference in cardiotoxicity. The results of this trial suggest that substituting paclitaxel for cyclophosphamide in the AC regimen, using weekly paclitaxel, or both, further improves on the AC→T regimen that proved superior to AC alone in CALGB 9344. Additonal follow-up of this important trial is needed.
• Docetaxel-The combination of docetaxel, doxorubicin, and cyclophosphamide (TAC) has produced significantly higher complete and partial response rates than the combination of 5-FU, doxorubicin, and cyclophosphamide (FAC) in patients with metastatic breast cancer.[29] This regimen was tested in the adjuvant setting in the Breast Cancer International Research Group (BCIRG) 001 trial, which randomized 1,491 patients with node-positive breast cancer to six cycles of TAC (75/50/500 mg/m²) or FAC (500/50/500 mg/m²) every 3 weeks (Table 2).[16]
Interim analysis at a median follow- up of 55 months showed that 5-year disease-free survival was 7 percentage points greater and 5-year overall survival was 6 percentage points greater with TAC. The benefits of TAC on disease-free survival were not affected by hormone receptor status or HER2 status. Febrile neutropenia occurred in 24% of patients treated with TAC and in 2% of those treated with FAC.[30] Granulocyte colonystimulating factor (G-CSF, Neupogen) was used in all cycles after an occurrence of febrile neutropenia (secondary prophylaxis). An ongoing trial of TAC vs FAC in patients with nodenegative disease has confirmed the benefits of initiating G-CSF in the first cycle for minimizing neutropenic complications in patients treated with TAC.[31] NSABP B-27, a randomized trial of neoadjuvant chemotherapy in 2,411 women with operable breast cancer, compared three treatments: preoperative AC alone, preoperative AC followed by preoperative docetaxel, and preoperative AC followed by postoperative docetaxel.[12] The doses were the same in each arm-four cycles of AC (60/600 mg/m²) every 3 weeks and four cycles of docetaxel (100 mg/m²) every 3 weeks. The pathologic complete response rate with preoperative AC and docetaxel was twice that with preoperative AC alone (26.1% vs 13.7%; P < .001), and relapse-free survival at 5 years was also significantly higher (74% vs 69%; P = .03). Overall survival at 5 years was the same in all three arms. The concurrent use of tamoxifen and chemotherapy may have affected the outcomes in this trial. The PACS 01 trial compared three cycles of FEC (5-FU, epirubicin, and cyclophosphamide at 500/100/500 mg/m²) followed by three cycles of docetaxel (100 mg/m²) with six cycles of adjuvant FEC (500/100/500 mg/m²) in 1,999 patients with node-positive operable breast cancer. At 5 years, both disease-free survival (78.3% vs 73.2%; P = .04) and overall survival (90.7% vs 86.7%; P = .05) were significantly higher with FEC followed by docetaxel than with FEC alone.[13] Since both regimens were used with the same total number of cycles (avoiding a potentially confounding factor seen in the earlier randomized trials of paclitaxel, CALGB 9344, and NSABP B-28, in which paclitaxel recipients were given more cycles of chemotherapy), the results in this trial suggest that docetaxel plays a major role in improving treatment outcomes. Moreover, cardiotoxicity was lower in the FEC/docetaxel arm. Two other large randomized trials are further evaluating the dose and sequence of anthracycline-docetaxel combinations (see Table 1). In NSABP B-30, patients with node-positive disease have been randomized to four cycles of concurrent doxorubicin and docetaxel (50/75 mg/m²), four cycles of TAC (75/50/500 mg/m²), or four cycles of AC (60/600 mg/m²) followed by four cycles of docetaxel (100 mg/m²). All cycles are repeated at 3-week intervals. In BCIRG 005, patients are randomized to six cycles of TAC (75/50/500 mg/m²) or four cycles of AC (60/600 mg/m²) followed by four cycles of docetaxel (100 mg/m²). The results of these trials should shed light on the optimal dosing and scheduling of anthracycline-docetaxel combinations in patients with early-stage breast cancer.

Increasing Dose Intensity
Strong evidence supports the concept that the dose intensity of the chemotherapy, or the amount of drug delivered per unit of time, correlates with outcomes in the adjuvant setting.[ 32-34] A 30-year follow-up of patients in an early trial of adjuvant CMF shows that the survival advantage was greatest in those who were treated with more than 85% of the planned dose intensity (overall survival = 40% vs 21%).[35] At least three randomized trials have investigated the relationship between dose intensity and outcomes with anthracycline-based regimens. CALGB 8541 compared three regimens of adjuvant CAF (cyclophosphamide, doxorubicin, and 5-FU) in 1,550 patients with node-positive breast cancer: four cycles of 600/60/600 mg/m² every 4 weeks (high dose intensity), six cycles of 400/40/400 mg/m² every 4 weeks (moderate dose intensity), and four cycles of 300/30/300 mg/m² every 4 weeks (low dose intensity).[ 33] The rates of both disease-free survival (P < .001) and overall survival (P = .004) were greater with the high- and moderate-dose-intensity regimens than with the low-doseintensity regimen. In a later trial in patients with nodepositive disease (CALGB 9344), the outcomes with the dose of doxorubicin increased to 75 or 90 mg/m² were no better than those with the 60-mg/m² dose.[14] Bonneterre and colleagues showed that an FEC regimen with epirubicin, 100 mg/m², was superior to FEC with epirubicin, 50 mg/m².[34] The results of these trials suggest that the effect of doxorubicin is maximal at a dose of about 50 to 60 mg/m² and that the effect of epirubicin is maximal at a dose of about 100 mg/m². More-intensive schedules of AC, tested in several large NSABP randomized trials, have proved to be no more efficacious as a standard schedule but much more toxic.[36,37] Extremely high doses of chemotherapy with stem cell rescue have not been found to be superior to standard regimens in randomized trials. Increasing Dose Density
Dose intensification can be achieved not only by increasing the dose intensity (increasing the dose of the drug per cycle) but also by increasing the dose density (decreasing the interval between the cycles while keeping the dose of the drug the same).[38] Mathematic models of tumor growth predict that shortening the interval between chemotherapy treatments would minimize tumor regrowth between cycles[39] and also lessen the emergence of drug-resistant mutations.[38,40] Bonadonna and colleagues were among the first to test dose-dense adjuvant therapy. In a study in women with early-stage breast cancer involving four or more lymph nodes, they compared 4 cycles of doxorubicin followed by 8 cycles of CMF (sequential regimen) and 2 cycles of CMF alternated with 1 cycle of doxorubicin for a total of 12 cycles (alternating regimen).[41] Overall survival at 10 years was 58% in the sequential arm and 44% in the alternating arm (P = .002). Thus, a clear survival advantage was shown for the sequential arm, which used the same total doses, but each agent was given at a higher dose density than in the alternating arm. More recently, Hudis and colleagues showed the feasibility of sequential doxorubicin, paclitaxel, and cyclophosphamide (A→T→C) given every 2 weeks (dose-dense regimen) with G-CSF support in 42 women with node-positive breast cancer.[42] The majority of patients (69%) developed febrile neutropenia, and red blood cell transfusions were required in 67%, but a 78% disease-free survival at 48 months was promising. Designed to further explore dosedense regimens, CALGB 9741 was a 2 * 2 trial that compared concurrent AC followed by paclitaxel and sequential A→T→C, both given either every 2 weeks (dose dense) or every 3 weeks, in 2,005 women with nodepositive breast cancer (Table 3).[43] The doxorubicin, cyclophosphamide, and paclitaxel were given at the same doses (60/600/175 mg/m²) in both arms. Patients in the dose-dense arms were given G-CSF to allow neutrophil recovery before the next cycle. Sequence had no effect on survival, but rates of both diseasefree survival (P = .01) and overall survival (P = .01) were significantly higher at a median follow-up of 36 months in the dose-dense arms. Furthermore, dose-dense therapy was not associated with greater toxicity, and the incidence of grade 4 neutropenia was greater in the every- 3-week sequential arm than in the two dose-dense arms (24%-43% for the every-3-week regimen vs 3%-9% for dose-dense therapy).[43] Platelet transfusions were not required in any patients, but red blood cell transfusions were required in 13% of patients in the concurrent dose-dense arm and in 3% of those in the every-3-week arm. An interim analysis of 89 patients (135 enrolled) treated with dose-dense therapy found that darbepoetin alfa(Drug information on darbepoetin alfa) (Aranesp), given at 200 μg when the hemoglobin level dropped to less than 12 g/dL and administered thereafter according to a preplanned algorithm, eliminated the need for red blood cell transfusions.[44] The results of two European trials of dose-dense epirubicin-based regimens have been disappointing. Venturini and colleagues conducted a phase III trial in 1,214 patients with node-positive or high-risk node-negative operable breast cancer who were randomized to six cycles of FEC (600/60/600 mg/m²) at either 2- or 3-week intervals with G-CSF.[45] There were no significant differences in survival. In the neoadjuvant setting in patients with locally advanced breast cancer, survival with doseintense EC (120/830 mg/m²) given with G-CSF every 2 weeks for six cycles was no greater than that with FEC (500/60/75 mg/m²) given every 4 weeks for six cycles.[46] These data suggest that dose density may be more important with the taxanes than with other drugs. The benefits seen with dose density should not be extrapolated to untested regimens. Another strategy used with the taxanes is weekly regimens. A regimen of weekly paclitaxel was found to have greater efficacy than an every-3-week regimen, especially in metastatic disease.[ 47,48] The benefits of weekly paclitaxel in the neoadjuvant setting have also been shown: In a study in 258 patients with operable breast cancer, the response rate with preoperative weekly paclitaxel followed by FAC was double that with standard paclitaxel followed by FAC (29% vs 14%; P < .01).[47] The value of weekly paclitaxel as adjuvant therapy continues to be investigated and should be determined in ongoing trials. ECOG 1199, a randomized four-arm trial is comparing paclitaxel and docetaxel given either weekly or every 3 weeks after four cycles of AC. In another 2 * 2 trial design, SWOG S0221 will evaluate AC (weekly or every 2 weeks) followed by paclitaxel (weekly or every 2 weeks). The ongoing Intergroup S0221 trial is the only trial that is comparing weekly and every-2- week paclitaxel in the adjuvant setting. These data support the importance of delivering full chemotherapy doses on schedule. Dose reductions and delays, however, are common in clinical practice,[49,33,34] and when substantial, may affect outcomes. Clinicians should therefore avoid dose reductions whenever possible and should attempt to maintain the dose and schedule of the regimens that have been established in clinical trials.