Introduction of Anthracyclines

Anthracyclines were developed as an anticancer antibiotic derived from Streptomyces bacterium that have several mechanisms of action, including intercalation of DNA and RNA inhibiting synthesis, and inhibition of topoisomerase II, which interferes with DNA supercoiling and relaxation. Initially found to be active against pediatric malignancies, anthracyclines were first tested in breast cancer in the metastatic setting where they were found to have anticancer activity. The National Surgical Adjuvant Breast and Bowel Project (NSABP) B-15 and B-16 studies investigated the regimen of adriamycin and cyclophosphamide (AC) in node-positive breast cancer and found that it was equivalent to CMF but better tolerated. A large meta-analysis by the Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) confirmed that four doses of AC was at least equivalent to CMF (relative risk [RR)] 0.98, standard error [SE] 0.05, 2 p = .67).

In efforts to improve the CMF regimen, methotrexate was substituted for an anthracycline, leading to the development of the 5-fluorouracil, adriamycin, and cyclophosphamide (FAC) and 5-fluorouracil, epirubicin, cyclophosphamide (FEC) regimens. The Grupo Español de Investigación en Cancer de Mama (GEICAM; Spanish Breast Cancer Research Group) compared FAC with CMF and found that in node-negative disease 5-year disease-free survival (DFS) was 75% versus 67% favoring FAC ( p = .0378). The National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) similarly performed a comparison of FEC versus CMF and found that 5-year relapse-free survival (RFS) rates were 63% compared with 53%, favoring FEC ( p = .009) with an overall survival (OS) advantage as well (77% vs. 70%, p = .03, chemotherapy, respectively). The importance of anthracyclines in the adjuvant setting was also highlighted in the National Epirubicin Adjuvant Trial (NEAT) and the BR9601 trial by the Scottish Cancer Trials Breast Group (SCTBG). In these studies, patients with early-stage breast cancer were randomized to epirubicin or no-epirubicin before CMF, and both 5-year RFS and OS were improved in the epirubicin group (76% vs. 69%, and 82% vs. 75%, p < .001 for all comparisons, respectively). The aforementioned meta-analysis by the EBCTCG also found that patients receiving anthracycline doses greater than four cycles of AC (i.e., FEC or FAC) had breast cancer mortality rates superior to CMF (RR = 0.78, SE 0.06, 2 p = .0004).

Addition of Taxanes to Anthracycline-Based Chemotherapy

Derived from plants of the genus Taxus, taxanes are a class of anticancer drugs developed in the 1970s that work by disrupting microtubule function. Paclitaxel and docetaxel were initially evaluated in metastatic breast cancer and found to be an effective option for anthracycline-resistant patients. Based on their activity in the metastatic setting, numerous studies have been conducted to assess the efficacy of adding taxanes to adjuvant.

The addition of paclitaxel to AC was evaluated in the Cancer and Leukemia Group B (CALGB) 9344 trial. In this trial, 3121 patients with axillary node-positive breast cancer were initially randomized to compared three doses of adriamycin (60 mg/m ² , 75 mg/m ² , and 90 mg/m ² ) as part of the standard AC regimen and then randomized to four cycles of paclitaxel (175 mg/m ² every 3 weeks) versus no further therapy. There was no difference in outcome based on the different AC regimens, but there was a significant improvement both in RFS (hazard ratio [HR] 0.83, p = .0023) and in OS (HR 0.82, p = .0064) with the addition of paclitaxel.

A weekly schedule of paclitaxel was investigated in the GEICAM 9906 study, which included a total of 1248 patients with axillary node–positive early-stage breast cancer. Patients were randomized into one of two treatment arms: (1) FEC90 (5-fluorouracil 600 mg/m ² , epirubicin 90 mg/m ² , cyclophosphamide 600 mg/m ² ) every 3 weeks for a total of six cycles, (2) or FEC90 every 3 weeks for four cycles followed by paclitaxel 100 mg/m ² every week for eight cycles. After a median of 46 months, there was a significant difference in DFS (85% vs. 79%, HR 0.63, p = .0008) in favor of the taxane arm; however, there was no significant difference in OS (95% vs. 92%, HR 0.74, p = .137).

A similar benefit was seen when adding docetaxel to anthracycline-based therapy. In the Protocole Adjuvant dans le Cancer du Sein (PACS; French Adjuvant Study Group) 01 trial, 1999 women with node-positive breast cancer were randomized to receive six cycles of FEC100 (5-fluorouracil 500 mg/m ² , epirubicin 100 mg/m ² , cyclophosphamide 500 mg/m ² ) or three cycles of FEC100 followed by three cycles of docetaxel 100 mg/m ² . At 5 years, DFS was 78% in the taxane-containing arm compared with 73% in the nontaxane arm (HR 0.82, p = .012), and OS also significantly favored the taxane-containing arm (91% vs. 87%, HR 0.73, p = .017).

The Eastern Cooperative Oncology Group (ECOG) E1199 trial was a four-arm study designed to assess whether docetaxel was better than paclitaxel and whether weekly versus every-3-week administration was better. AC was administered initially for four cycles every 3 weeks. Subsequently patients were randomized to one of four arms: (1) paclitaxel 175 mg/m ² every 3 weeks for four cycles (control group); (2) paclitaxel 80 mg/m ² every week for 12 doses; (3) docetaxel 100 mg/m ² every 3 weeks for four cycles; and (4) docetaxel 35 mg/m ² every week for 12 doses. The odds ratio (OR) for DFS was 1.27 (95% confidence interval [CI] 1.03–1.57) in the weekly paclitaxel arm ( p = .006) and 1.23 (95% CI 1.00–1.52) in the every-3-week docetaxel arm ( p = .02) compared with the control group; however, only the weekly paclitaxel group had an OS benefit (OR = 1.32, 95% CI 1.02–1.72, p = .01). Exploratory analysis demonstrated that benefit was seen primarily in HER2-negative patients irrespective of ER status.

In contrast, retrospective studies suggest that patients with ER-positive disease are likely to derive lesser benefit from taxanes. A retrospective analysis in a subgroup of patients who participated in the Intergroup 9344 trial found that patients with HER2-positive tumors benefited from paclitaxel regardless of their hormone receptor status. However, in HER2-negative patients, paclitaxel only benefited the hormone receptor–negative patients. Furthermore, data from the CALGB analyzed retrospectively showed that chemotherapy was of little or no benefit in lymph node–positive and hormone receptor–positive patients.

A meta-analysis of 13 trials including 22,903 patients identified was performed to assess survival benefits of the addition of a taxane to an anthracycline-containing regimen. The pooled analysis demonstrated improved DFS (HR 0.83, 95% CI 0.79–0.87, p < .00001) and OS (HR 0.85, 95% CI 0.79–0.91, p < .00001) with the addition of taxanes, with an absolute improvement of 5% in DFS and 3% in OS. The investigators also found that the benefit from taxanes was significant for both ER-positive and ER-negative patients and was independent of whether paclitaxel or docetaxel was used. There was no benefit seen whether taxanes were given concurrently with anthracyclines or sequentially.

In summary, addition of taxanes after anthracycline-based therapy can improve survival in patients with early-stage breast cancer, particularly those with high-risk features such as lymph node involvement. Most data suggest that benefit independent of ER status and that weekly paclitaxel is the most effective taxane regimen.

Dose Density

Chemotherapy works by first-order kinetics (half -life), meaning that a chemotherapy drug will kill a constant proportion of tumor cells, rather than constant numbers. Human cancer cells grow by nonexponential Gompertzian kinetics, which means tumors have an initial rapid growth curve which levels off as they outgrow nutrients and blood supply. Because cells are more sensitive to chemotherapy during the rapid growth phase, more frequent (dose-dense) administration of cytotoxic agents rather than increased doses was hypothesized to kill more cancer cells.

To test this hypothesis, the CALGB performed a study of 2005 women with axillary node–positive breast cancer who were randomly assigned to receive one of the following regimens: (1) sequential doxorubicin, paclitaxel, and cyclophosphamide administered every 3 weeks; (2) the same sequence administered every 2 weeks with filgrastim; (3) AC followed by paclitaxel administered every 3 weeks; or (4) AC followed by paclitaxel administered every 2 weeks with filgrastim. At a median follow-up of 36 months, there was a statistically significant improvement on DFS for the dose-dense regimens (every 2 weeks) compared with the every 3-week regimen (4-year DFS 82% vs. 75%, respectively, RR = 0.74, p = .01). OS was also improved in the dose-dense arms but did not reach statistical significance (3-year OS 92% vs. 90%, respectively, RR = 0.69, p = .013). There was no difference between sequential and concurrent arms. Notably, both ER-positive and ER-negative patients benefited from dose density, but this was particularly noted in the ER-negative subgroup (19% vs. 32% in ER-positive vs. ER-negative group, respectively, in terms of relative reduction in hazard). Toxicity was comparably between arms, although severe neutropenia was less frequent in patients who received the dose-dense regimens.

To confirm findings, a meta-analysis assessing dose density was performed. Patients treated with dose-dense regimens had better DFS (HR = 0.83, 95% CI 0.73–0.94, p = .005) and OS (HR = 0.84, 95% CI 0.72–0.98, p = .03). Benefit was only seen in ER-negative (HR = 0.71, 95% CI 0.56–0.89) and not ER-positive disease (HR = 0.92, 95% CI 0.75–1.12). Again, dose density was not associated with an increase in treatment related events.

Although dose density has been accepted as a standard of care in patients with TNBC, controversy still exists in ER-positive disease. This likely stems from the biological fact that most ER-positive tumors tend to be more indolent in nature (i.e., luminal A–like); however, special consideration should be made when treating patients with aggressive phenotype ER-positive disease because these tend to be more chemoresponsive (i.e., luminal B–like).

Non–Anthracycline-Containing Regimens

Anthracyclines can have potent anticancer effects, but they can have significant long-term toxicity. A dreaded consequence of use of adjuvant anthracyclines for breast cancer is the risk of bone marrow neoplasms. The overall rate of marrow neoplasms has been found to be as high as 0.46 per 1000 person-years in patients treated with adjuvant chemotherapy. Anthracyclines are also associated with cardiotoxicity, which can be acute or chronic. Acute cardiotoxicity is rare, reported in approximately 3.2% of patients treated with anthracyclines, and may include arrhythmias, acute congestive heart failure, myocarditis, and myocardial infarction. Chronic cardiotoxicity can occur subclinically (OR = 6.25, 95% CI 2.58–15.13) or be clinically significant (OR = 5.43, 95% CI 2.34–12.62) most frequently manifesting as an irreversible heart failure. To test whether nonanthracycline regimens could be as effective as anthracycline-based regimens US Oncology performed an adjuvant trial in women with stage I to III breast cancer patients who were randomized to either doxorubicin at 60 mg/m ² and cyclophosphamide at 600 mg/m ² given every 3 weeks for four cycles (AC) and docetaxel at 75 mg/m ² and cyclophosphamide at 600 mg/m ² given every 3 weeks for four cycles (TC). With a median follow-up of 7 years, the DFS rate was significantly superior for TC compared with AC (81% vs. 75%, respectively; HR = 0 .74; 95% CI 0.56–0.98, p = .033), and OS were the same (87% vs. 82%, respectively; HR = 0.69, 95% CI 0.50–0.97, p = .032). Both regimens were well tolerated, with TC having a higher incidence of febrile neutropenia and AC having three long-term fatal toxicities (one patient with heart failure and two with marrow events). From the results of this trial, TC emerges as a valuable regimen in the adjuvant treatment of breast cancer and may replace AC as a “standard of care” for women who are not considered eligible or appropriate for a combination of an anthracycline and taxane. For high-risk disease in which patients are eligible for anthracycline-based therapy, combination anthracycline and taxane-based therapy is preferred. Currently, a phase III study is comparing TC with combination anthracycline and taxane-based therapy ( clinicaltrials.gov identifier NCT00493870).

Bone Marrow Transplant

The rationale behind the use of high-dose chemotherapy (HDC) necessitating stem cell support is based on the fact that because some chemotherapy could kill some cancer cells, more chemotherapy should kill more cancer cells and thus increase the cure rate. The dose-limiting event of chemotherapy was thought to be myelotoxicity; therefore harvesting the bone marrow before treatment with HDC and reinfusing it after treatment is complete was thought to allow for safe administration of HDC. Investigation of this strategy, however, was impaired by the political and social climate of the times as well as one of the largest scandals in medical research history, resulting in the needless overtreatment of thousands of patients with breast cancer and delays in obtaining definitive data.

In the early 1980s the identification of human immunodeficiency virus (HIV) and acquired immunodeficiency syndrome (AIDS) spurred the development of anti-HIV medications and put into question the ethics of drug development when there are patients with serious or life-threatening diseases. This led the US Food and Drug Administration (FDA) to develop measures of providing such patients new treatments as early as possible. Breast cancer advocates had likewise formed powerful groups and were at the forefront of these issues. Despite a review of HDC with autologous stem cell transplant (ASCT) in breast cancer published in 1992, which stated that although responses were impressive, there were insufficient data to conclude whether this was superior to or worse than conventional dose chemotherapy, this procedure became increasingly popular outside of clinical trials. Furthermore, at the 1992 American Society of Clinical Oncology (ASCO) Annual Meeting, a randomized control trial from South Africa was presented demonstrating significant response rates and improvements in survival. These circumstances delayed accrual of patients to definitive phase III studies. It was not until the 1999 ASCO Annual Meeting that studies demonstrating that HDC with ASCT offered no benefit compared with standard adjuvant regimens were finally presented. The Philadelphia Bone Marrow Transplant Group found that bone marrow transplant did not improve survival in patients with metastatic breast cancer. Furthermore, results from the CALBG 9082, Southwestern Oncology Group (SWOG) 9114, and NCIC MA-13 studies did not demonstrate a benefit in high-risk early-stage breast cancer. In addition, the highly publicized research misconduct found after an audit in the trials presented by the South African group at the 1992 ASCO meeting demonstrated why there had been discordant results. Thus after almost 2 decades and thousands of women being needlessly transplanted, this practice finally came to an end.

Despite the harm and controversy surrounding HDC with bone marrow transplantation, there are valuable lessons to be learned. The most important is the crucial importance of level of evidence when prescribing a therapy; evidence-based medicine is a standard of care in all of medicine, and especially in oncology. Exposure to toxic medications in the absence of known benefit should be considered with extreme caution and preference given to enrollment into clinical trials. In addition, although mechanisms to grant early approval of promising medications are important, there must be a strategy to complete definitive studies. A case in point is the antivascular endothelial growth factor receptor (VEGF) inhibitor, bevacizumab, in metastatic breast cancer. Although initial early-phase studies showed promise, which led to accelerated approval, the definitive studies were able to be conducted, and when results showed no benefit, approval was repealed.

Considering Chemotherapy

As previously discussed, chemotherapy works best on cells that are rapidly dividing rather than those that have a slower growth curve. This basic biological principle helps guide the benefits patients may, or may not, derive from chemotherapy. Standard pathologic examination can frequently lend clues. Patients with more aggressive appearing tumors (i.e., high grade) or that have high proliferation indices (high ki67) generally tend to respond better to chemotherapy than those that are low grade with low ki67. Numerous gene expression assays have been developed and studied to help both define clinical behavior and benefit from adjuvant chemotherapy.

Tamoxifen

The selective ER modulator tamoxifen is a cornerstone therapy for patients with ER-positive breast cancer who are either premenopausal or postmenopausal and not a candidate for an aromatase inhibitor (AI). The effects of 5 years of adjuvant tamoxifen is potent; at 15 years, it decreases the risk of recurrence from 46.1% to 33.0% (log-rank 2 p < .00001) and decreases breast cancer mortality from 32.7% to 23.6% (log-rank 2 p < .00001). Numerous studies have investigated the use of tamoxifen beyond 5 years including the Scottish Adjuvant Tamoxifen Trial (SATT), NSABP-14, and the joint ECOG E4181 and E5181 analysis, which failed to consistently demonstrate a DFS or OS benefit. These studies were generally small, and ER status was either negative or unknown in many patients. For more than 2 decades, 5 years of adjuvant tamoxifen was considered a standard of care until two large studies found 10 years was superior. The Adjuvant Tamoxifen: Longer Against Shorter (ATLAS) trial enrolled 6846 patients with ER-positive disease who were treated with 10 versus 5 years of tamoxifen. Ten versus 5 years of treatment resulted in a significant reduction in recurrence (18% vs. 20.8% respectively, RR = 0.84, 95% CI 0.76–0.94), breast cancer mortality (9.7% vs. 11.5%, respectively, p = .01), and improvements in OS (18.6 vs. 21.1%, respectively, p = .01). The Adjuvant Tamoxifen: To Offer More? (aTTom) trial reported that 2755 women with ER-positive breast cancer allocated to 5 versus 10 years of tamoxifen had a similar benefit. The aTTom study found that compared with 5 years, 10 years of tamoxifen resulted in a significant reduction in recurrence (16.7 vs. 19.3%, RR = 0.85, 95% CI 0.76–0.95) and a trend toward improved breast cancer mortality. Notably, in both the ATLAS and aTTom trials benefit of extended tamoxifen occurred almost 10 years after initiating treatment, suggesting that these therapies have a carryover effect in terms of disease recurrence. Furthermore, these were studies with generalized eligibility criteria and included patients from stage I to III tumors, subset analysis did not identify a specific group that benefited more, thus the results are considered to be applicable to a diverse patient population. Although the risk of pulmonary embolism is present while patients are on tamoxifen, this risk ends upon cessation of therapy; this was not the case for the risk of endometrial cancer, which persisted even after therapy ended. The ATLAS study did not demonstrate an increase in pulmonary embolism or endometrial cancer-associated deaths; however, the aTTom trial did demonstrate a small increased risk of endometrial cancer–associated death (RR 1.83, 95% CI 1.09–3.09). Limitations of these extended tamoxifen studies include the fact that most patients were postmenopausal, where AIs have become the standard of care; furthermore, how this regimen compares with others, such an AI after tamoxifen or the addition of ovarian suppression is not addressed.

Aromatase Inhibitors

As their name implies, AIs inhibit aromatase and prevent the peripheral conversion of androgens to estrogens. Two main subclasses of AIs have been approved: steroidal (i.e., exemestane) and nonsteroidal (i.e., anastrozole and letrozole). Numerous studies have pitted 5 years of tamoxifen versus 5 years of AIs, and a meta-analysis conducted by the EBCTCG involving postmenopausal women with early-stage ER-positive breast cancer found that AIs reduced recurrence particularly in the first year (RR = 0.64, 95% CI 0.52–0.78) and years 2 to 4 (RR = 0.8, 95% CI 0.68–0.93) and had lower 10-year breast cancer mortality (RR = 0.85, 95% CI 0.75–0.96). Thus for women who are postmenopausal and have hormone receptor–positive breast cancer, AIs have emerged as the adjuvant endocrine treatment of choice.

Postmenopausal women who are initially treated with tamoxifen derive benefit if they switch to an AI. The Intergroup Exemestane Study found that exemestane versus tamoxifen after 2 to 3 years of tamoxifen for a total of 5 years of endocrine therapy led to improvements in recurrence or death (HR = 0.76, 95% CI 0.66–0.88, p = .0001). The Italian Tamoxifen Arimidex trial, found similar benefits of the switch method compared with 5 years of tamoxifen. Completing an additional 5 years of an AI after 2 to 3 years of tamoxifen has also been found to be beneficial. The aforementioned EBCTCG meta-analysis found that this strategy reduced recurrence during years 2 through 4 (RR = 0.56, 95% CI 0.46–0.67), as well as breast cancer mortality (RR = 0.84, 95% CI 0.72–0.96). The importance of adding adjuvant AIs can further be seen in the NCIC CTG MA.17 trial, in which patients were randomized 5 years of letrozole versus placebo after completing 5 years of tamoxifen. After a median follow-up of 30 months, those treated with letrozole had improved DFS either locally or contralaterally (HR = 0.58, 95% CI 0.45–0.76, p < .001) or in terms of distant metastasis (HR = 0.60, 95% CI 0.43–0.84, p = .002), although no OS benefit was seen.

The question as to whether 10 years of AI therapy is better than 5 years is being evaluated in the MA.17 extension trial ( clinicaltrials.gov identifier NCT00754845) and NSABP B-42 ( clinicaltrials.gov identifier NCT00382070). Although the current standard of care is 5 years of adjuvant AI in early-stage breast cancer, patients with particularly high-risk disease can be considered for extended therapy; however, providers must weigh the lack of data and potential long-term toxicity of AI therapy, including to bone health and cardiovascular risks.

Ovarian Suppression

Chemotherapy-induced amenorrhea has been associated with improved outcomes in patients with early-stage breast cancer. Indeed, an EBCTCG meta-analysis demonstrated improved recurrence rates (2 p < .00001) and breast cancer mortality (2 p = .004) in patients receiving ovarian suppression compared with no ovarian suppression. However, a later meta-analysis by the EBCTCG found that the addition of ovarian suppression to tamoxifen did not improve recurrence rates (HR = 0.85, 95% CI 0.67–1.09) nor mortality (HR = 0.84, 95% CI 0.59–1.19) but in fact increased grade III toxicities, such as menopausal symptoms and sexual dysfunction. To demonstrate the efficacy of ovarian suppression, the Suppression of Ovarian Function Trial (SOFT) and Tamoxifen and Exemestane Trial (TEXT) were performed. A joint analysis investigating exemestane with ovarian suppression compared with tamoxifen with ovarian suppression demonstrated improved 5-year DFS (91.1% vs. 87.3%, respectively, HR = 0.72, 95% CI 0.60–0.86, p = .0002). These data unfortunately did not have a comparator arm without ovarian suppression. It was not until the SOFT study was presented and published a few months later that the benefit of adding ovarian suppression to tamoxifen was evaluated, and it did not provide benefit for the entire treated population (DFS HR = 0.83, 95% CI 0.66–1.04, p = .1). The study was also stratified by those who received versus did not receive chemotherapy, and secondary objectives investigated freedom from breast cancer (FFBC) rates in tamoxifen, versus tamoxifen with ovarian suppression, versus AI with ovarian suppression. Adding ovarian suppression to tamoxifen did not improve FFBC in the entire cohort (HR = 081, 95% CI 0.63–1.03) nor in patients treated with chemotherapy (HR = 0.78, 95% CI 0.6–1.02); however, patients receiving 5 years of AI with ovarian suppression did experience improved FFBC compared with tamoxifen alone in the entire cohort (HR = 0.64, 95% CI 0.49–0.83) and in those receiving chemotherapy (HR = 0.65, 95% CI 0.49–0.87). In an unplanned subset analysis in patients less than 35 years of age (n = 350, 11% of total cohort), the majority of whom were treated with chemotherapy, suggested that the addition of ovarian suppression to either tamoxifen or an AI improved FFBC compared with 5 years of tamoxifen. Importantly, the addition of ovarian suppression was associated with more side effects including those of estrogen deprivation, and the long-term effects on cardiovascular and bone health remain unknown.

Optimal Endocrine Therapy for Pre- and Postmenopausal Women

Navigating through data on various adjuvant endocrine therapy approaches can be challenging, and it is important to weigh the benefit and toxicity of therapy to each individual patient. For patients who are premenopausal and have ER-positive breast cancer, the standard of care remains tamoxifen, which may be prescribed for up to 10 years. Premenopausal patients with high-risk disease requiring chemotherapy may be considered for ovarian suppression therapy with an AI. The benefit of adding ovarian suppression to tamoxifen has only been demonstrated in a small subset analysis of patients less than 35 years of age. For patients who are premenopausal but become postmenopausal during the first 5 years of tamoxifen, additional therapy with up to 5 years of an AI should be considered. For patients who are postmenopausal, the standard of care currently remains 5 years of adjuvant AI therapy. Based on the experience with extended tamoxifen, providers may consider extended AI therapy in high-risk patients understanding the lack of data and known long-term toxicities of AIs on bone and cardiovascular health.