Twenty Years of Systemic Therapy for Breast Cancer

To get a perspective on how breast cancer was treated 2 decades ago, a good starting place would be the National Cancer Institute's Consensus Development Conference, held in September 1985.[1] The senior author of this paper, then a very junior assistant professor, attended this first attempt at codifying the use of adjuvant therapy for breast cancer. The themes of the conference informed a great deal of what would follow in the next 2 decades. Adjuvant therapy, the central theme of the conference, was still quite a new concept. The suggestion that an assault on microscopic metastatic breast cancer might eradicate disease and provide curative potential remained controversial. Indeed, papers would still be published questioning the efficacy of this approach several years following the conference.[2] Adjuvant therapy had developed from two great strands of research. The first was the development of systemic chemotherapy. In the 2 decades preceding the conference, numerous chemotherapeutic agents had been developed for clinical use and had been shown to have activity both individually and in combination. Preclinical work performed by Skipper and colleagues at the Southern Research Institute had laid down general principles for the use of systemicchemotherapy and had suggested its potential for eradicating small volumes of disease.[3] Promising early adjuvant chemotherapy trials performed in the United States by the National Surgical Adjuvant Breast and Bowel Project (NSABP, under the direction of Bernard Fisher) and at the Istituto Tumori Nationale (under the direction of Bonnadonna and Veronesi) had led to a profusion of adjuvant chemotherapy trials.[4,5] The other great strand of research involved hormonal therapy. Although it had been recognized since the 1890sthat many breast cancers were estrogen- dependent, it was not until the 1960s that a true mechanistic basis for this sensitivity was advanced with the discovery of the estrogen receptor.[ 6,7] By the mid-1970s, it became obvious that the presence of the estrogen receptor correlated with hormonal sensitivity in the advanced setting.[8] The same decade saw the development of target-specific compounds such as tamoxifen(Drug information on tamoxifen), which quickly demonstrated activity in advanced breast cancer. Tamoxifen entered adjuvant trials, most prominently in Europe with theNolvadex Adjuvant Trial Organization (NATO) trial.[9,10] It is important to recognize that adjuvant therapy has, from its infancy, been driven by theory (eg, the Skipper-Schabel studies) and by a progressive unfolding of our understanding of biology (eg, the estrogenreceptor complex). What was astonishing, in retrospect, was the optimism of early randomized adjuvant trials, which demanded statistically implausible benefits. This optimism resulted in the development of numerous small, underpowered trials, many of which were individually negative, or were positive for progression-free but not overall survival. These studies were salvaged by what became the third great strand of adjuvant therapy, first manifested at the 1985 Consensus Development Conference: the application of higher statistical methodology (what would later be transformed into bioinformatics) to clinical trials. Indeed, that conference saw the first use of statistical meta-analysis for clinical trials in breast cancer.[11] This meta-analysis, performed by Richard Peto's group at Oxford, demonstrated both the modest nature of the benefits achieved and their undeniable reality. Hormonal Therapy: The First Targeted Therapy Many physicians, particularly those in the United States, remained skeptical regarding the ultimate benefits of adjuvant hormonal therapy. It was considered by many to be insufficiently aggressive. Hormonal therapy was "cytostatic" and could not be expected to kill the enemy. Looking back 2 decades later, there is no question but that the agent with the single greatest public health impact in all of oncology (at least to date) has been tamoxifen, which has saved countless lives around the world. But this was not apparent, or at least not obvious, at the time. Early opinions on adjuvant hormonal therapy involved several telling errors. One such miscalculation was the belief that tamoxifen had little or no benefit in premenopausal women. It was not until the 1995 meta-analysis that accumulating data laid this dangerous error to rest.[12] Another error was the belief by some that adjuvant hormonal therapy might have real benefits in estrogen-receptor- negative patients, which was finally laid to rest by the 2000 Oxford Overview. In retrospect, both errors resulted from an inadequate understanding of estrogen-receptor biology, as well as from inadequate quality control for estrogen-receptor testing married to underpowered individual studies. The very use of the term "antiestrogen" represented another area of misunderstanding. It became clear that tamoxifen was a selective estrogen-receptor modulator (SERM) acting as an estrogen in some organs (eg, bone and uterus), with both positive effects (as antiosteoporotic agent) and harmful outcomes (uterine carcinoma).[13-15] Preclinical data even suggested that resistance to tamoxifen might involve mutational events allowing breast cancer cells to recognize tamoxifen as a form of estrogen.[16] Aromatase Inhibitors
How to get around this problem became the next dominant theme in hormonal therapy. Early clinical investigations with the antiseizure medication aminoglutethimide (Cytadren) had demonstrated that the drug inhibited steroidogenesis, and that it could be used as a form of second-line hormonal therapy. Doctors and patients (the senior author of this paper will attest) cordially detested aminoglutethimide, a relatively toxic hormonal agent. Detailed investigations of its biochemical effects suggested that its breast cancer benefits related not-as originally suspected-to its adrenal effects, but rather to its effects on peripheral aromatization. This led to the development of agents specifically inhibiting peripheral aromatization of androgens to estrogen. During the 1990s, these aromatase inhibitors followed what by now was a standard pathway from the laboratory to clinical trials in advanced disease[17-19] to the adjuvant (microscopic metastatic) disease setting.[ 20-22]. At every clinical step, the aromatase inhibitors proved somewhat superior to tamoxifen, culminating in the 2004 American Society of Clinical Oncology (ASCO) technology assessment panel's recommendation to incorporate aromatase inhibitors into the adjuvant therapy of all postmenopausal estrogen-receptor- positive breast cancer.[23] Timing of Therapy
Hormonal therapy not only switched drug classes, it also switched disease settings. Data accumulated during the 1980s suggested that the estrogen receptor drove the progression from premalignancy to invasive cancer in many breast cancer patients, and that (in preclinical models) blocking the estrogen receptor could prevent this progression. This led to the development of severalproof-of-concept trials, conducted in the 1990s, testing tamoxifen chemoprevention in high-risk patients as well as tamoxifen in patients undergoing breast-conserving surgery for ductal carcinoma in situ.[24,25] Both approaches succeeded, though not without real and continuing practical concerns associated with tamoxifen's SERM-related toxicities. Chemotherapy: The Interplay of Theory and New Agents The initial adjuvant chemotherapy trials reported at the 1985 Consensus Development Conference used agents developed in the 1960s and 1970s. Chemotherapy then seemed to go into a kind of stasis. From doxorubicin(Drug information on doxorubicin)'s approval in 1977 until paclitaxel(Drug information on paclitaxel)'s appearance in the early 1990s, the US Food and Drug Administration (FDA) failed to approve a single new chemotherapeutic agent for breast cancer. Instead, chemotherapy research in breast cancer focused on territorial expansion and a sort of escalation previously reserved for the nuclear arms race. Territorial expansion involved the progressive movement of chemotherapy into lower-risk (ie, lymphnode- negative) disease as well as into the realm of large tumors (neoadjuvant or preoperative chemotherapy). Adjuvant trials demonstrated the benefits of chemotherapy, first in estrogen-receptor-negative and then in estrogen-receptor-positive, lymphnode- negative patients.[26,27] In the latter, however, the benefit was sufficiently modest to provoke angst in oncologists' offices worldwide. Neoadjuvant chemotherapy, initially developed as a means of rendering inoperable tumors operable, now became a means of improving the rates of breast conservation. Initial theory, based on preclinical animal models, had suggested that preoperative therapy would result in improved patient survival. Clinical trials failed to support this hypothesis: The contest of adjuvant and neoadjuvant chemotherapy resulted in a draw.[28] Theory-Based Approaches
Science has always been driven by the tension between theory and newtechnology. In the late 1980s and early 1990s, the lack of new chemotherapeutic agents led to the temporary triumph of theory-based approaches. Variations in dose and schedule, developed based on preclinical animal models and computer modeling, led to three related therapeutic approaches: high-dose chemotherapy, dose intensification, and dose densification. High-dose chemotherapy was driven both by theory (the hypothesis that standard chemotherapy failures were a consequence of inadequate peak dosing) and technology (the development of supportive care technology embodied in autologous stem cell transplantation and hematopoietic growth factors). There were early successes in small trials, both in the metastatic and adjuvant settings.[29,30] These trials, in turn, led to the development of large phase III investigations, which were marred by outright scientific fraud[31,32];well-conducted trials were typically negative for overall survival in both the metastatic and adjuvant setting.[33,34] The dose-intensity hypothesis, originally proposed by Hryniuk and colleagues, approached the question of dose from the standpoint of time, calculating dose of chemotherapy (typically in mg/m²) over time (typically per week). Retrospective analyses suggested a new way forward toward the goal of cure[35,36]; prospective randomized trials investigating increased dose intensity for cyclophosphamide(Drug information on cyclophosphamide), doxorubicin, and paclitaxel all failed to establish a survival benefit once one had reached standard doses of chemotherapy.[37-40] Dose density represented a third approach to hypothesis-driven alterations of dose and schedule. The dosedensity approach was based on the Norton-Simon hypothesis, which suggested that therapy results in a rate of regression in tumor volume proportional to the rate of growth expected for an unperturbed tumor of that size.[41] A logical mathematic consequence of this hypothesis is that more frequent (hence dose-dense) chemotherapy will result in increased tumor cell kill. The dose-density approach was confirmed in one large phase III trial (Cancer and LeukemiaGroup B [CALGB] 9741), with a small but statistically significant improvement in relapse-free and overall survival.[42] A more recent update of this trial suggests that the benefits of therapy were primarily restricted to estrogen-receptor-negative patients. It is worth pointing out that none of the three theory-driven approaches to chemotherapy described above had any specific foundation in breast cancer biology, although clinical scientists in the breast cancer world were quick to embrace and test all three theories. The modest gains from these theory-driven approaches appear to mark the limits of alterations in chemotherapy dose and schedule in breast cancer. New Drugs
By the early 1990s, however, theory began to give way to novel chemotherapeutics. Starting with the taxanes (paclitaxel and docetaxel(Drug information on docetaxel) [Taxotere]), the decade saw the introduction of several new agents, including capecitabine(Drug information on capecitabine) (Xeloda) and gemcitabine(Drug information on gemcitabine) (Gemzar). Adjuvant taxane chemotherapy has added modest but real improvements in relapse-free and overall survival in lymph-node-positive breast cancer.[40,43] The new decade has seen both capecitabine and gemcitabine enter our therapeutic armamentarium through randomized trials demonstrating a survival benefit in metastatic breast cancer.[44,45] These agents have now entered randomized controlled trials in the adjuvant setting. More recently, variants on taxanes (such as the epothilones and albumin-bound nanoparticle paclitaxel [Abraxane]) have been extensively studied in the setting of advanced disease. Targeting HER2 The discovery of the estrogen receptor in the late 1960s established a plausible biologic mechanism for breast cancer growth and invasion. By the late 1970s, however, it was clear that the estrogen receptor failed to explain all breast cancer growth.Indeed, and seemingly paradoxically, estrogen-insensitive tumors appeared to be both more biologically and clinically aggressive. In the early 1980s, the work of researchers therefore shifted to another family of growth factor receptors, the human epidermal growth factor-HER, or erbB-receptors. By the late 1980s, data began to accumulate suggesting that HER2 (erbB2) played an important role in a substantial fraction of breast cancers. Overexpression of HER2 was found to occur in 15% to 25% of breast tumors and correlated with a more aggressive tumor phenotype.[46] HER2 is one of four related receptor tyrosine kinases. All members of this family share common structural elements, including an extracellular ligand-binding domain and an intracellular tyrosine kinase domain. HER2 exerts its oncogenic effect through the formation of heterodimers with both the epidermal growth factor receptor (HER1, erbB1) and neu differentiation factor (NDF) receptors (erbB3 and erbB4).[47] These HER2-containing heterodimers lie at the head of a complex signal transduction cascade that regulates cell proliferation, survival, adhesion, migration, and differentiation. Trastuzumab
The growing evidence that HER2 played a biologically important role in breast cancer prompted the development of a new therapy targeting HER2-positive disease. Trastuzumab(Drug information on trastuzumab) (Herceptin) is the humanized counterpartof the murine monoclonal antibody 4D5. It targets the juxtamembrane region of HER2, exerting its anticancer activity through several suggested mechanisms: downregulation of the HER2 receptor, inhibition of cell-cycle proliferation, activation of antibodydependant cellular toxicity, and finally, inhibition of vascular endothelial growth factor (VEGF).[48] It is not the purpose of this review to track the progress of trastuzumab in metastatic breast cancer, other than with regard to its role in the development of adjuvant HER2-targeting trials. Trastuzumab was initially demonstrated to be efficient as second- or third-line monotherapy in the phase II trials of Baselga et al[49,50] and Cobleigh et al.[51] Patients with extensive metastatic breast cancer resistant to chemotherapy received trastuzumab intravenously at weekly intervals, and the overall tumor response rates in the two trials were 11.6% and 15%, respectively, with a tolerable safety profile. The response rates seen with trastuzumab in chemotherapy-refractory breast cancer were not, in traditional terms, particularly spectacular. Furthermore, accepted wisdom in the mid-1990s had it that monoclonal antibodies were unlikely to ever play a role in the treatment of human cancer. Nevertheless, plans were made to perform a phase III trial in front-line metastatic breast cancer. Patients were randomized to receive either chemotherapy alone or chemotherapy plus trastuzumab. Initially, the chemotherapy agents selected for use were doxorubicin and cyclophosphamide. Quite unexpectedly, the combination of doxorubicin and trastuzumab proved cardiotoxic, with 27% of patients developing congestive heart failure.[52] Here, luck intervened. During the 1990s, doxorubicin moved into the adjuvant setting and paclitaxel replaced it as front-line metastatic breast cancer therapy. As accrual lagged, the study was amended to allow the use of the taxane. Paclitaxel did not induce congestive heart failure when combined with trastuzumab. When the results of this pivotal trial were presented at the 1998 ASCOmeeting, it was clear that a new therapeutic era had arrived. The addition of trastuzumab increased response rates in comparison to chemotherapy alone (50% vs 32%, P < .001), increased progression-free survival, and crucially improved overall survival. This last achievement was particularly impressive given that the trial had a crossover design that could easily have robbed it of success. HER2 Testing
Determining HER2 status proved essential to assessing patient eligibility for trastuzumab therapy and is highly predictive of benefit.[53,54] It is frightening to reflect that had trastuzumab been developed as a standard chemotherapeutic agent in an unselected population, response rates would undoubtedly have been in the single digits, and the drug thrown away. HER2 testing is now standard in all newly diagnosed patients with invasive breast cancer. The two FDA-approved tests to evaluate HER2 status are immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH). No universal protocol has been implemented to chose one method or the other. That said, traditional testing guidelines suggested the following algorithm[55]: All tumor samples must first be screened with IHC; if the test result is 0/1+, it is reported as negative; if it is 3+, it is reported as positive. If the IHC score is 2+, however, a FISH assay is recommended to validate the result. Many investigators recommend initial evaluation by FISH, considering it the gold standard for HER2 testing. The advent of adjuvant trastuzumab (discussed below) of course raises the stakes for HER2 testing. Recent analyses of HER2 testing derived from two large phase III trials (National Surgical Adjuvant Breast and Bowel Project [NSABP] B-31 and North Central Cancer Treatment Group [NCCTG] N9831) have increased concerns regarding HER2 testing. Both trials have shown community testing to be seriously discordant with central laboratory testing and immunohistochemistry to be discordant with FISH testing. Given that a false-negative HER2 test could doom a patient to die of breastcancer, and a false-positive test could expose a patient to the risk of congestive heart failure, centralized testing (in a large reference laboratory or by an institution with recognized expertise) with FISH should, in the authors' opinion, represent the appropriate means of testing for HER2. Major Trials
In September 1998, trastuzumab was approved by the FDA as firstline treatment in combination with paclitaxel, as well as second- or thirdline monotherapy in patients with metastatic breast cancer overexpressing HER2. These factors provided the rationale to move this drug into the adjuvant setting. Four major phase III adjuvant trastuzumab trials were launched: The NSABP B-31 and NCCTG N9831 trials initiated at the same time in 2000, followed in close order by the Breast Cancer International Research Group (BCIRG) and Breast International Group (BIG, also called the HERceptin Adjuvant, or HERA) trials. These trials included approximately 12,000 patients with breast cancer whose tumors overexpressed HER2 protein or have HER2. The major end points of these trials included overall and disease- free survival.

• NSABP and NCCTG Trials- The NSABP B-31[57] trial was designed to enroll 2,700 HER2-positive breast cancer patients with positive nodes who had undergone either a total mastectomy or lumpectomy with irradiation and axillary dissection. This two-arm trial compared the safety and efficacy of four cycles of doxorubicin (A) at 60 mg/m² and cyclophosphamide (C) at 600 mg/m² followed by four cycles of paclitaxel (T) at 175 mg/m²/3 wk, to that of AC followed by T concurrently with trastuzumab (H) at a 4 mg/kg loading dose, then 2 mg/kg/wk. The NCCTG N9831,[57] on the other hand, was a three-arm trial enrolling 3,000 patients also with HER2-positive, node-positive disease (although the trial was later amended to include high-risk node negative disease). The chemotherapy regimen was the same as that of the NSABP trial except that the paclitaxel regimen was80 mg/m²/wk for 12 weeks. Arms A and C paralleled the two groups in the NSABP trial; however, this trial included a third arm in which trastuzumab was initiated following completion of chemotherapy.
Because NSABP B-31 and N9831 shared similar treatment arms and identical cardiac safety monitoring, and were both conducted under the auspices of the National Cancer Institute, the study groups elected to perform a joint analysis of the two trials. Parallel data from both trials were put together in a two-group study: Group 1 (the control group) combined arm 1 and arm A of NSABP and NCCTG, respectively, and group 2 combined arm 2 and arm C of the two trials. N9831's sequential-therapy arm was not included in the analysis. The first interim analysis was done in April 2005 and presented at the 2005 ASCO meeting.[57] The results were dramatically significant. Diseasefree survival was the primary end point of the joint analysis. Trastuzumab reduced relative risk of a first breast cancer event by 52% (hazard ratio = 0.48 with 2p = 3 * 10^-12). An absolute increase in disease-free survival of 12% was noted at 3 years and that of 18% at 4 years in the trastuzumab group.
• HERA Trial-Similar impressive results were also reproduced in the HERA trial,[58] which randomized nearly 5,090 patients to a three-arm comparison of 1 and 2 years of trastuzumab vs no trastuzumab in HER2- positive women with operable primary breast cancer. HERA included a significant percentage of node-negative patients-in contrast to the joint analysis of NSABP B-31 and N9831- and administered all trastuzumab regimens following locoregional therapy and adjuvant chemotherapy. As presented at the 2005 ASCO meeting by Dr. Martine Piccart, disease-free survival was again the primary end point. Trastuzumab reduced relative risk of a first breast cancer event by 46% (hazard ratio = 0.54, P < .0001), with an absolute increase in diseasefree survival at 2 years of 8%.[58]
Although cardiotoxicity continued to be an element of concern, and while many questions remain to be answered,adjuvant trastuzumab clearly now represents the standard of care for HER2- positive early-stage breast cancer.