The anthracyclines doxorubicin(Drug information on doxorubicin) (A) and epirubicin(Drug information on epirubicin) (E) are among the most active agents for breast cancer. As described in the review by Dr. Henderson, many clinical trials demonstrate that the use of the anthracyclines as adjuvant chemotherapy offers a consistent survival advantage compared with older regimens such as cyclophosphamide(Drug information on cyclophosphamide), methotrexate(Drug information on methotrexate), and 5-fluorouracil (CMF). The Early Breast Cancer Trialists’ Collaborative Group overview of polychemotherapy, which examined data from more than 14,000 patients, showed that anthracycline use is associated with an absolute incremental improvement of 4.2% in all-cause mortality at 10 years.[1] However, the anthracyclines are associated with toxicities, and these must be considered when carrying out risk-benefit calculations for individual patients. In general, the short-term adverse effects of anthracyclines, such as nausea and vomiting, can be controlled with modern anti-emetics. Of potentially greater concern is the risk of long-term toxicities, specifically the small but measurable risks of secondary leukemias and congestive heart failure (CHF). A key research goal is to develop regimens that retain the benefits of anthracycline-containing regimens but not the risks.
Omission of Anthracyclines for Patients with HER2-Normal Early Breast Cancer
One possible approach might be the outright substitution of taxanes for anthracyclines. Although the taxanes are associated with other short-term toxicities, including sensory neuropathy and a risk of life-threatening hypersensitivity reactions, a potential advantage of taxanes is that they have not been associated with an increased risk of leukemia and do not appear to cause CHF.
This substitution strategy has been investigated in several clinical trials; however, results have been inconsistent. In US Oncology study 9735, which included 1,016 patients unselected for estrogen receptor (ER), progesterone(Drug information on progesterone) receptor (PR), or HER2 status, 4 cycles of the non-anthracycline regimen docetaxel(Drug information on docetaxel), 75mg/m2, and cyclophosphamide (DC) were compared to 4 cycles of conventionally scheduled doxorubicin and cyclophosphamide (AC).[2] At a median follow up of 7 years, DC was associated with both an increase in disease-free survival (DFS) from 75% to 81% (P = .033) and an increase in overall survival (OS) from 82% to 87% (P = .032). However, it should be noted that approximately half of the patients in this study had node-positive disease, and for such patients, 4 cycles of conventionally scheduled AC is possibly suboptimal treatment. Furthermore, although DC does not appear to be associated with an increased risk of cardiotoxicity, the toxicities of this regimen may still be considered substantial by some investigators. In the 9735 study, there were 2 deaths during treatment in the DC arm, compared with none in patients treated with AC.[3] Specific toxicities were also greater: the incidence of febrile neutropenia was twice as high with DC as with AC (5% vs 2.5%)—and this may be an underestimation of the risk of febrile neutropenia in community practice, since in one recent study, the incidence of febrile neutropenia with DC in this setting was almost 12% in the absence of primary prophylaxis with growth factor support.[4]
Additional data supporting the substitution of taxanes for anthracyclines are pending. For example, Cancer and Leukemia Group B (CALGB) has compared 4 or 6 cycles of paclitaxel(Drug information on paclitaxel) (dose-dense) to AC using the same schedules.[5] As Dr. Henderson has pointed out, DC may be equivalent to an optimal anthracycline regimen. However, we are concerned that every-3-weeks AC × 4 is no longer the appropriate comparator arm in treatment of high-risk disease. This particular regimen has never outperformed CMF, and the dose-dense schedule could be superior.
A subtle issue is whether the anthracyclines and taxanes are additive, rather than equivalent. If so, then the goal of replacing the anthracycline might be the wrong one. There is a significant amount of evidence that the two drug classes are additive. For example, Swain et al have recently demonstrated that in the treatment of node-positive breast cancer, AC → D is superior to DAC × 4, which is essentially DC with doxorubicin.[6] The point here is that “full”-dose AC × 4 followed by full-dose docetaxel (100 mg/m2) is superior to the lower-dose concurrent combinations tested in the 4-cycle arms (whether AD or DAC). This suggests that full-dose AC matters. A similar conclusion is supported by a Japanese study testing AC followed by a taxane vs a taxane alone.[7] Poole et al have also suggested that anthracyclines are still necessary in the treatment of early breast cancer, since E → CMF was superior to CMF alone.[8] Finally, Burnell et al have shown that CEF (cyclophosphamide, epirubicin, and 5-fluorouracil), as well as dose-dense EC → paclitaxel (P), is better than the older regimen of every-3-weeks AC → P.[9] The subtle point is that if one version of an anthracycline-containing regimen is shown to be superior, then the anthracycline must be contributing to the effectiveness of the regimen as a whole.
Anthracyclines in HER2-Positive Early Breast Cancer
For patients with early breast cancer that overexpresses HER2, the addition of trastuzumab(Drug information on trastuzumab) (Herceptin) to chemotherapy has improved DFS and OS, as demonstrated in 4 large (and 1 smaller) randomized controlled trials. [10-13] In these studies, almost all patients (about 90%) were treated with anthracyclines. Furthermore, long-term follow-up of these studies appears to show that cardiotoxicity is relatively infrequent and is generally associated with a favorable outcome.[14-16] In fact, only one of the adjuvant trastuzumab trials, Breast Cancer International Research Group (BCIRG) study 006, has incorporated a non-anthracycline regimen with trastuzumab, using docetaxel and carboplatin(Drug information on carboplatin) as the chemotherapy backbone (DCarboH).[13] At 5 years, patients treated with DCarboH had DFS and OS of 81% and 91%, respectively, compared with 84% and 92%, respectively, for AC-docetaxel and trastuzumab. We agree with Dr. Henderson that this difference is not statistically significant, but since the study is not powered for a non-inferiority comparison, we cannot assume that DCarboH is equivalent to AC-docetaxel and trastuzumab. Hence, there is strong evidence from 5 clinical trials that the addition of trastuzumab to anthracycline-based regimens improves survival, but there is much less evidence suggesting that the alternative non-anthracycline approach is as effective.
Dose-Dense Anthracycline-Based Regimens
For patients with high-risk early breast cancer, irrespective of HER2 status, a further therapeutic advance is the development of dose-dense regimens. The improved efficacy of this approach has been demonstrated in CALGB study 9741.[17] These benefits have subsequently been confirmed in some, but not all, clinical trials.[18, 19] Furthermore, this improved efficacy does not appear to be associated with an increased risk of cardiotoxicity. Perhaps unexpectedly, in the CALGB 9741 study, patients treated with dose-dense anthracycline-based chemotherapy have experienced half the cardiac events of patients treated with conventionally scheduled chemotherapy.[20] Furthermore, dose-dense anthracycline-based chemotherapy can safely be combined with anti-HER2 therapy with no apparent increased risk of cardiotoxicity.[21-23] Again, the fact that an anthracycline-containing regimen can be superior to a conventional anthracycline-containing control suggests that the drug is contributing to outcomes.
Future Directions: Biomarkers
There is extensive evidence that for unselected patients with high-risk breast cancer, the anthracyclines offer a survival benefit and are an important component of the most active chemotherapy regimens.[1] In the future, we hope that their use will evolve so that these agents target those patients most likely to benefit. Preliminary attempts at identifying biomarkers for anthracycline sensitivity have focused on the topoisomerase IIα (TOP2A) gene, their putative target near the HER2 gene on chromosome 17; results have been mixed.[13, 24] A possible explanation is the lack of consistent assessment of either TOP2A protein expression by immunohistochemistry or gene amplification, as suggested by Dr Henderson. High-resolution techniques, such as comparative genomic hybridization (CGH) or Representational Oligonucleotide Microarray Analysis (ROMA), both of which allow a more precise definition of these genes, may provide the granularity needed to delineate TOP2A status and help individualize therapeutic approaches.[25] However, as Dr. Henderson notes, the evidence for molecular profiling is insufficient to guide our selection of anthracycline-based treatments. We therefore believe that clinicians should treat otherwise healthy patients who have high-risk disease with evidence-based anthracycline-based or anthracycline-taxane combinations.
Financial Disclosure: Dr. Morris has received honoraria from Roche-Genentech. Dr. Hudis has received research funding from Onyx and Merck. Dr. Dang has received research funding from Roche-Genentech.
