Abstract
The biennial Madrid Breast Cancer Conference has become one of the leading summer meetings dedicated to reviewing and discussing the most relevant advances in breast cancer research. Seminal, practice-changing results have been presented and reviewed in different sessions and symposia, embracing various aspects from molecular genetics to advances in imaging, staging and treatment in the different settings of the disease.
Oncologists need improved tools to select treatments for individual patients. Most cancer treatments benefit only a minority of the patients to whom they are administered. Prognostic and predictive factors play important roles in the treatment of breast cancer, but very few such factors are used in clinical practice and prognostic relevance remains unclear. The ability to predict which patients are most likely to benefit from a given treatment would not only save patients from unnecessary toxicity and inconvenience, but might also facilitate their receiving drugs that are more likely to help. The current standard tools used to test new therapies in early breast cancer are Phase III randomized adjuvant trials. Such trials are large, expensive and take years to achieve their outcome. Neoadjuvant or primary systemic therapy (PST) trials might offer the eventual possibility of a more rapid and less resource-demanding alternative, at least for some therapies, if short-term surrogate clinical, pathologic or biologic end points could be identified that might predict for long-term outcome in adjuvant trials. Genome-wide monitoring of gene expression using DNA microarrays makes it possible to study thousands of genes in a tumor sample with a single experiment. By looking for an association between the gene expression pattern and tumor behavior, it should be possible to identify new prognostic and predictive factors. Dr M Van de Vijver (Rotterdam, The Netherlands) and his team have previously identified a 70-gene expression profile (GEP) associated with an increased risk for developing distant metastases within 5 years. More recently, they identified consistent features in the transcriptional response of normal fibroblasts to serum and used this wound-response signature to reveal links between wound healing and cancer progression in the same tumors. By combining the 70-GEP (good-prognosis and poor-prognosis tumors) and the wound signature (activated and quiescent tumors), subgroups of patients with markedly different prognoses can be identified [1]. One serious limitation of microarray expression profiling, however, is that it is an RNA assay that requires fresh or well-preserved frozen tissue for extraction of viable RNA. Consequently, some investigators use microarrays to screen the genes for those factors associated with outcome and then develop classifiers for clinical applications that are based on subsequent studies using reverse transcription-polymerase chain reaction (RT-PCR) of the selected genes. This was the approach used and reviewed in detail by Dr S Paik (PA, USA) in developing the Oncotype Dx® risk score for patients with lymph node-negative, estrogen receptor (ER)-positive breast cancer receiving tamoxifen monotherapy. A 21-gene expression assay was established using quantitative RT-PCR from RNA extracted from archived, formalin-fixed, paraffin-embedded tissue. Using this assay, it was possible to demonstrate that approximately half of the population of women with node-negative, hormone-receptor-positive breast cancer who were treated with adequate local therapy and tamoxifen have an excellent prognosis and are unlikely to benefit from chemotherapy [2]. Ongoing work with RT-PCR analysis based on a larger number of genes also suggests that response to preoperative chemotherapy can be predicted with this approach, as described by Dr L Gianni (Milan, Italy). A total of 90 locally advanced breast cancer patients were treated with doxorubicin and paclitaxel followed by weekly paclitaxel before surgery, and adjuvant intravenous cyclophosphamide, methotrexate and 5-fluorouracil (CMF) four times thereafter. A high-throughput RT-PCR assay quantified fixed tumor tissue expression of 384 genes selected from the literature and previous gene expression studies. In total, 30 of the 87 genes that correlated with pathologic complete response (pCR) were clustered by both expression and function into three groups – an ER group, a proliferation group and an immune group. pCR was more likely with lower expression for the ER group and higher expression for the proliferation and immune groups. Multivariate analysis indicated that combinations of genes are more powerful predictors of response than single genes and addition of other genes increases the predictive power over the Oncotype Dx® 21-gene panel.
Dr L Pusztai (TX, USA) reviewed key methodologic issues related to gene expression-based predictors in the discovery of breast cancer [3,4]. Unsupervised analysis of global gene expression has identified molecularly distinct classes of breast cancer associated with different clinical outcomes. Even if the molecular class of breast cancer (e.g., basal, luminal and normal-like) does not predict prognosis or response more accurately than conventional clinical variables, it gives a biologic framework to understand why these associations between phenotype and clinical outcome occur. The GEP could also assist in identifying molecular differences between responders and nonresponders and in discovering predictive signatures in preclinical models (e.g., cell lines and transgenic mice) that may apply to human cancers. Several early Phase I pharmacogenomic studies suggest that prediction of pCR to neoadjuvant chemotherapy is possible. Whether these predictors are regimen specific or just markers of general chemotherapy sensitivity is yet to be determined. Large-scale validation of these results is needed and is currently underway. Dr G Hortobagyi (TX, USA) reviewed some of the most challenging areas for gene expression profiling of human breast cancer. GEP with currently available platforms includes a number of genes or gene segments of uncertain function (expressed sequence tags). These provide an excellent opportunity to assess the functional value of these genes and improve our knowledge of the biology behind them. Large-scale breast cancer studies have reported the identification of signature genes with the potential to predict clinical outcome. One of the most confounding findings, when comparing the signature lists, is the virtualy complete lack of agreement in the included genes. Differences in performances related to the different platforms adopted can only partially explain these results. The major challenges are still in developing and utilizing the most appropriate statistical methods for data analysis. Large, prospective, multicenter clinical trials are needed to prospectively validate the clinical utility of gene profiling for individual patients.
Moving to advances in diagnosis and staging, Dr M Cristofanilli (TX, USA) presented some of the recent applications of magnetic resonance imaging (MRI), particularly in the preoperative setting. MRI has been shown to be more accurate than standard imaging in determining the initial staging and evaluation of the extent of invasive disease in locally advanced and inflammatory breast cancer, usually treated with PST. As a result, MRI is an attractive method for assessing tumor response to PST, appropriate staging (baseline evaluation), early identification of poor responders (intermediate evaluation during treatment) and identification of residual disease for surgical planning (preoperative imaging study). Correlative studies integrating serial MRI tumor measurements with serial collection of tissue for evaluation of biomarkers could lead to identification of the most appropriate tool for prediction of pathologic response to PST. Dr R Delgado (Madrid, Spain) described the 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) findings in staging and response monitoring in breast cancer patients. FDG-PET was consistently demonstrated to be superior to computed tomography (CT) in the detection of internal mammary and/or mediastinal lymph nodal metastases. Dual modality CT–PET performed better than traditional tools in the evaluation of tumor recurrence, particularly in the evaluation of lytic bone metastases and in identifying widespread disease in patients who otherwise appear to have an isolated, potentially curable, local–regional relapse. CT–PET may improve the accuracy of PET alone in the evaluation of treatment response by directly defining metabolic and morphologic changes. Larger prospective studies are needed to delineate effective strategies and guidelines for incorporating PET imaging into standard clinical practice.
Occult dissemination is the main cause of recurrent metastatic cancer in patients who have undergone resection of their primary tumor. Recently, the Cell Search System® (Veridex, LLC) has been used for the isolation and enumeration of circulating tumor cells. The results, described by Dr M Cristofanilli (TX, USA), indicate that in metastatic breast cancer, the level of circulating tumor cells before a new therapy is initiated and, even more importantly, the level measured at the first follow-up visit, are useful predictors of progression-free and overall survival [5]. This technology is becoming a standard tool for the real-time assessment of prognosis and response to treatment, offering the possibility of collecting the cells after sorting for evaluation of biomarkers (e.g., RT-PCR, GEP). A recent acquisition is represented by the use of glycan arrays. Malignant transformation and tumor progression are associated with the specific changes in the complex surface carbohydrates known as tumor-associated carbohydrate antigens. Production of autoantibodies against these abnormal carbohydrates during cancer progression is expected. A robust printed glycan array was recently fabricated that employs a library of over 200 well-defined structures comprising carbohydrate sequences of N-glycans, O-glycans, glycolipids and glycoproteins. This printed glycan array was used to simultaneously detect multiple specific antiglycan autoantibodies in sera from breast cancer patients.
Reviewing advances in local treatment, Dr H Bear (Richmond, VA, USA) discussed the great deal of information now available regarding the use of lymphatic mapping and biopsy of the sentinel lymph nodes (SLNs) in light of the National Surgical Adjuvant Breast and bowel Project (NSABP) B-32 and B-27 trial results. From the clinical point of view, three areas are in need of further investigation:
The significance of occult micrometastases found in SLNs by immunohistochemistry
The routine use of SLN biopsy in patients with ductal carcinoma in situ (DCIS)
The role and timing of SLN biopsy in breast cancer patients receiving neoadjuvant chemotherapy
Validation of useful indicators to accurately predict which patients with a positive SLN have no other nodes involved, and could therefore avoid complete axillary lymph node dissection, is eagerly awaited.
Dr AP Calvo (Madrid, Spain) reviewed state of the art radiotherapy treatment options for patients with early stage invasive breast cancer treated with breast-conserving surgery. Whole-breast irradiation still remains the radiation standard of care in early stage invasive breast cancer after breast-conserving surgery. However, intraoperative radiation therapy is an appropriate technical alternative to delivering partial breast irradiation (PBI), together with high-dose rate brachytherapy and/or external irradiation precision techniques (e.g., three-dimensional conformal radiotherapy, intensity-modulated radiation therapy). Intraoperative radiation therapy implies delivery of a high, single dose of radiation to a limited intrasurgical anatomic area. Accelerated PBI appears to be a promising alternative to whole-breast irradiation in selected patients. Different accelerated PBI techniques can be used, such as intraoperative electrons, catheter-based interstitial brachytherapy, mammosite balloon brachytherapy or external-beam partial irradiation. However, some concerns remain, particularly regarding potential late adverse effects and differences among techniques.
Moving to the chemoprevention arena, Dr J Cuzik (London, UK) reviewed the available evidence and the data from the ongoing chemoprevention trials in different risk populations. Four trials have now reported on the use of tamoxifen for the prevention of breast cancer, and one has reported on the use of raloxifene. A consensus is now emerging that 30–40% of breast cancers can be prevented by tamoxifen. Tamoxifen, although accepted by US investigators as the standard of care for risk reduction in high-risk women, is not viewed as such in Canada and Europe, mainly due to its toxicities. The newer selective estrogen-receptor modulator (SERM), raloxifene, was shown to decrease breast cancer incidence by 65% in postmenopausal women with osteoporosis as a secondary end point in the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, and appeared to be more effective in prevention than tamoxifen. Motivated by these results, NSABP initiated an equivalency trial comparing the efficacy and toxicity of these SERMs. Raloxifene does not have the gynecologic problems of tamoxifen, due to the selectivity of its SERM action, but it still leads to an increase in thromboembolic events. Data from six adjuvant trials in postmenopausal women with early breast cancer provide a compelling rationale for exploring the use of aromatase inhibitors (AIs) in the prevention setting. AIs have proven to be superior to tamoxifen in decreasing new contralateral tumor incidence as secondary efficacy end points in all published studies, suggesting that 70–80% of ER-positive breast cancers can be prevented with these drugs. The AIs are also devoid of gynecologic and thrombotic complications, but do lead to bone mineral loss and increased fracture rates in the absence of additional bone-sparing therapy. An important question will be the effectiveness of bisphosphonates in arresting and/or reversing bone loss associated with the almost complete depletion of estrogen that occurs with AIs. No impact for receptor-negative cancers is expected from all these studies, however, and this remains a challenging area for future trials [6].
PST is under active clinical investigation for several potential treatment benefits: facilitating breast conservation, downstaging primary tumors and lymph nodes, early assessment of response to chemotherapy and obtaining prognostic molecular information on the basis of the pathologic response to therapy. pCR in the breast (NSABP trial) and lymph nodes (according to the definition of the MD Anderson Group) is the main goal of PST. Recently, prospective trials have demonstrated that patients achieving a pCR have significantly improved disease-free survival (DFS) and overall survival (OS) compared with patients who do not. Therefore, the pCR rate has become one of the most important intermediate trial end points in assessing the efficacy of new adjuvant chemotherapy regimens. Although a pCR portends a very favorable outcome for the majority of patients, 13–25% will still develop recurrent disease. Dr G Hortobagyi (TX, USA) reviewed the 20 years' MD Anderson experience in this setting. The recent introduction of taxanes has improved the pathologic response rate by 25–34% [7], and an almost threefold increase in pCR (up to 65%) with the addition of trastuzumab to chemotherapy was achieved without significant cardiac toxicity [8]. Current studies are aimed at identifying useful predictors of pCR and developing individualized medicine for patients with primary breast cancer.
Moving to the neoadjuvant endocrine approach in postmenopausal women, Dr M Dowsett (London, UK) reported on the mixed results of the IMmediate Preoperative Anastrozole, tamoxifen or Combined with Tamoxifen (IMPACT) trial. This study was designed to test the hypothesis that the 12-week end point of clinical response and a 2-week biologic end point of change in proliferation as assessed by binding affinity (Ki)67 staining of neoadjuvant tamoxifen compared with anastrozole and in combination before surgery might predict for outcome in the Arimidex, Tamoxifen, Alone or Combination (ATAC) adjuvant therapy trial. There were no significant differences in response rate in the intent-to-treat population between patients receiving tamoxifen, anastrozole or the combination. No significant relationship between the decrease in Ki67 and clinical response was observed, but mean changes in the Ki67 growth index in this trial were parallel to the differences in DFS in the ATAC adjuvant study. More data are clearly needed to validate Ki67 as a reliable determinant of response in primary hormonal therapy trials [9].
In the postsurgical setting, adjuvant treatment recommendations for patients with stage I or II breast cancer are based on the clinician's estimates of the likely risk of breast cancer relapse and death and the likely benefit of adjuvant therapy. To address this problem, the decision tool Adjuvant®, available online, was created by Dr P Ravdin (TX, USA) and reviewed in some detail at the meeting. The major strength of this tool is that it allows the doctor and patient to review, in a quantitative sense, the benefits and risks of different options. The main limitation is that there is limited knowledge of long-term efficacy for many therapies. The impact of new drugs or therapeutic modalities should be implemented. In addition, it is not clear how nodal status, as assessed by sentinel node mapping and biopsy, should be incorporated. Other pathologic prognostic factor data (e.g., HER-2 overexpression) should also be included. However, Adjuvant makes reliable predictions for the average patient. It is indeed a useful and validated tool (in independent series of patients) to aid physicians and patients in making adjuvant systemic treatment decisions for early tage breast cancer [10].
Dr P Valagussa (Milan, Italy) reviewed the 30-year experience at the Milan National Cancer Institute, presenting the long-term follow-up data of adjuvant chemotherapy research programs. The results of the first CMF studies demonstrate that the significant advantage in both relapse-free (29% reduction in relative risk [RR] for relapse) and overall survival (21% reduction in RR for death) persists over the years. Subsequently, different studies to test the effectiveness of sequential noncrossresistant regimens containing anthracyclines were designed. The updated 20-year results with sequential adriamycin followed by CMF in a high-risk subset could probably be explained by an increased density of anthracycline, delivered at full dose within the first 9 weeks of treatment. Recently, the association of paclitaxel and doxorubicin (AT) followed by CMF was tested as AT regimen (27%) was further increased by CMF (52%). This overview confirmed that the departure from conventional, locoregional treatment alone in operable breast cancer at risk of relapse, conceived and initiated 30 years ago by Gianni Bonadonna, has substantially contributed to the treatment and understanding of breast cancer today [11].
In recent years, the field of cancer therapy has witnessed the emergence of novel targeted strategies that inhibit specific cancer pathways and key molecules in tumor growth and progression. Among them, one class of compounds that has shown great progress are those targeting tyrosine kinases (TKs), their ligands and signal transducers. There are more than 90 known protein kinase genes in the human genome. Blocking pathogenic TKs resulted in major therapeutic success in breast cancer with trastuzumab. Dr J Baselga (Barcelona, Spain) reported results of the HERceptin® Adjuvant (HERA) trial: an international, Phase III study investigating treatment with trastuzumab every 3 weeks for 12 or 24 months versus no treatment (observation) in patients who had previously undergone a range of surgical, chemotherapy and/or radiotherapy interventions. The results for the 1 year trastazumab arm versus observation were quite impressive, with a statistically significant improvement in 2 year DFS: 85.8 versus 77.4% with a 46% reduction in disease relapse. No impact on OS was detectable in the short-term follow-up period. Benefits are independent of patients' baseline characteristics (e.g., node number, hormone receptor status and type of chemotherapy). Trastuzumab-treated patients had a higher, but still low, incidence of symptomatic primary cardiac end point as defined per protocol (0.5 vs 0%). These results are in line with the 52% decrease in disease recurrence observed in two North American trials with almost the same study design and populations, enrolling over 3300 patients presented at the American Society of Clinical Oncology (ASCO) this year.
On the other hand, epidermal growth factor (EGF) receptor inhibitors have shown modest overall activity in unselected populations of breast cancer patients. Consequently, one of the major areas of clinical research in this field is to identify useful preclinical predictors of response as illustrated by Dr C Arteaga (TN, USA). The availability of tumor tissue, made possible in preoperative trials, is mandatory in order to study features related to sensitivity and/or resistance and to determine the optimal biologic dose of these drugs. An appealing study design is to combine either anti-EGF receptor (EGFR) or anti-HER-2 growth factor inhibitors with other molecule-targeted therapies. Preclinical models have clearly demonstrated synergistic activity of the HER-2 antibody trastuzumab in combination with the EGFR small molecule TK inhibitor (TKI) gefitinib. However, the clinical results were disappointing. Interestingly, there are emerging clinical data with small molecule TKIs that target other members of the ErbB (HER)-receptor family. Lapatinib is a dual inhibitor of the EGFR and HER-2 which showed clinical activity in patients with trastuzumab-pretreated or -refractory breast cancer in Phase I–II studies [12,13].
Dr G Hortobagyi (TX, USA) reviewed present knowledge of the pathophysiology of bone metastases, particularly in breast cancer [14]. Bisphosphonates are presently considered the standard of care. Nonetheless, many outstanding questions on the optimal use of these drugs remain unanswered: when should bisphosphonate therapy be administered? Can responders and nonresponders to therapy be identified by monitoring bone resorption markers or correlating outcomes with GEPs or protein expression profiles? In the adjuvant setting, the use of bisphosphonate therapy for the prevention of bone metastases is still not indicated outside of a clinical trial, even though the direct antitumor effect of newer aminobisphosphonates and the opportunity to alter the bone microenvironment so as to interfere with the process of homing of cancer cells are sound rationales. A different indication for bisphosphonates is for management of osteoporosis, especially related to the adoption of AIs as a standard treatment for postmenopausal early breast cancer. New strategies to manage metastatic bone disease are currently under evaluation alone or in combination. Osteoprotegerin and antibodies directed to the ligand or the receptor activator of nuclear factor-κB have demonstrated substantial activity in inhibiting osteoclast activation, recruitment and differentiation. Parathyroid hormone-related protein (PTHrP) is a critical initiating factor in the process of bone resorption. Antibodies against PTHrP are being explored as therapeutic agents in advanced clinical trials. Src inhibitors might have an important role to play in inhibiting osteoclast activation. The management of bone metastases has benefited enormously from the expanded understanding of basic biologic processes related to osteoclast and osteoblast function. These advances will result in improved management of established bone metastases and, possibly, effective preventive interventions.
In his concluding talk, Dr Larry Norton (NY, USA) reviewed some aspects of fascinating mathematical models applied to the process of tumorigenesis. The Norton–Simon hypothesis, based on the Gompertzian phenomenon, led to the dose-dense approach to breast cancer chemotherapy, shown in multiple studies to achieve substantial reduction in cancer recurrence and death. Dose-dense chemotherapy seeks to maximize cell kill by maximizing the rate of drug delivery – not the level of dosing. The Gompertzian curve also provided the basis for Norton's observations on a new model of cancer metastasis and cancer stem cell biology. He postulates that self-metastasis is responsible for small clusters of cell growth that together form a tumor, with each cluster following the initial rapid growth of the Gompertzian model prior to distant metastasis. In order to bridge together different aspects of the biology of metastasis, he suggested reconsidering the gene pattern as made of obligatory genes (regulating mitosis and mitotic rate), genes relating to self-metastasis and growth of tumors at local sites and genes that relate to the ability to metastasize to distant areas. These new insights may lead us to focus on genes involved in metastasis, including metastasis to self, as therapeutic targets.
Highlights
High-throughput gene expression profiling provides a powerful tool to identify new prognostic and predictive markers for breast cancer.
Novel, affordable, noninvasive diagnostic techniques in breast cancer (e.g., circulating tumor cells evaluation and magnetic resonance imaging [MRI]/18F-fluorodeoxyglucose positron emission tomography [FDG-PET]) might identify which patients are most likely to benefit from treatment, sparing them unnecessary toxicity and inconvenience.
The integration of novel targeted treatments is a high priority in the management of patients with metastatic and, more importantly, primary breast cancer.