The Frontier Edge: New Appreciation of Triple Negative Disease – A Review (Part 1)
Three Myths of Triple Negative Disease
It has been standard wisdom to date to say, and believe, (1) that patients with triple negative disease (ER- / PR- /HER- tumors) intrinsically have a poor prognosis relative to those with endocrine-positive disease (ER+ and/or PR+) or HER2+ disease; (2) that triple negative tumors are responsive to, and hence treatable by, only chemotherapy (this is known as the "triple-negative paradox", coined by Dr. Lisa Carey at UNC); and (3) that there is some essential association between triple negative disease and BRCA1-deficient breast carcinoma. All of these judgments are unwarranted, the first by virtue of resting on a misunderstanding of the pattern and velocity of recurrence in triple negative disease, and hence, a half-truth, as I discuss below, and the second by virtue of being "just plain wrong" which I'll expand upon momentarily in this introduction; the third is unwarranted by virtue of what's called lack of corollary, which I expand upon below, but in essence means that from the fact that most BRCA1-deficient carcinomas are triple negative (this is true), it does not follow (in reverse) that most triple negative disease exhibits BRCA-deficiency (this is false).
Myth 1: Chemotherapy Only
As to the first myth, the point to remember is that there are three classes of oncotherapy (cancer therapy), not two. There is endocrine therapy (aka, hormonal therapy) for endocrine-responsive disease (ER+ and/or PR+) which includes the SERM tamoxifen, the aromatase inhibitors (AIs), the pure antiestrogen (technically, SERD) fulvestrant (Faslodex), and ovarian suppression (oophorectomy if surgical, and via LHRH/GnRH analogs like goserelin (Zoladex), leuprolide (Lupron) if medical); all endocrine therapy is cytostatic therapy, that is tumor cell growth-inhibitive and hence anti-proliferative, without direct tumor cell kill activity. Then there is chemotherapy which is definitionally cytotoxic (tumor cell kill) that includes of course a very broad array of agents and regimens, from traditional to new generation (an example of the latter being the just approved epothilone agent ixabepilone (Ixempra) ).
But there is also the third class of oncotherapy, biological therapy, which is neither cytostatic like endocrine therapy nor cytotoxic like chemotherapy, using biological agents or "biologics", that target intrinsic molecular (signaling) pathways underlying fundamental onco-processes like carcinogenesis, tumorigenesis, angiogenesis, metastasis, cell adhesion and motility, etc. So there is monoclonal antibody (MoAb) therapy such the anti-HER2 MoAb trastuzumab (Herceptin) and there is also TKI (tyrosine kinase inhibitor) biological therapy such as dual-TKI anti-HER1/HER2 therapy such as lapatinib (Tykerb) , and in addition to MoAb (monoclonal antibody) and TKI ((tyrosine kinase inhibitor) biological therapies, there are an extraordinary range of other biologics that can and have been leveraged in breast cancer, and many of these have demonstrable value in the treatment of triple negative disease. These include the anti-VEGF MoAb bevacizumab (Avastin) that is both antiangiogenic, and chemotherapy-synergistic; the EGFR-inhibitor MoAb cetuximab (Erbitux) and the SM-TKI (small molecule-TKI) dasatinib (Sprycel) ; as well as PARP inhibitors, mTOR inhibitors, and HSP (heat shock proteins) , all found active against triple negative disease (all discussed below).
So the full truth is that oncotherapy for triple negative disease can deploy chemotherapy and appropriate triple-negative-targeting biological therapy, and these can be combined into a regimen "backbone" of chemobiotherapy (biological + chemotherapy); and example would be the ABX-BEV combination chemobiotherapy regimen, that is, chemotherapy via nab-paclitaxel (Abraxane) + bevacizumab (Avastin) anti-VEGF biological therapy. This is one of many deployable and effective therapies against triple negative tumors. So in sum, as Lisa Carey at UNC, an expert in triple negative disease, said in her exceptional presentation on the Biology and Therapy of Basal Breast Cancer [click on link to download as pdf] at the Controversies in Breast Cancer conference I recently attended, triple negative disease is challenging to treat, but it's also highly treatable; at this point in time, triple negative disease therefore no longer needs to be defined by limited options, by an increasingly broad spectrum of highly motivated and targeted therapeutic interventions, much the way a new breed of forward-thinking breast oncologists no longer accepts the inevitability of metastatic disease mortality (the great Neil Rosen at Memorial Sloan-Kettering Cancer Center recently at the same conference observed radically – but no longer a lonely voice - that he believes in curing metastatic disease, as I do myself].
Myth 2: Prognosis
Now let me first address the second myth and the issue of prognosis in triple negative disease. This has been critically elucidated recently by Rebecca Dent's team at Sunnybrook who demonstrated that triple negative disease exhibits a unique recurrence pattern and that not only is there a very sharp decline in recurrence risk of triple negative disease after the fourth year post-diagnosis, but that the risk of distant recurrence falls to absolute zero! - unheard of in any other type of breast cancer - from eight years and after (and is in any event extremely small, almost negligible, even from five years forward), and in addition, although local recurrence is a risk factor for later distant recurrence among women with all other types of breast carcinomas, this does NOT hold true for triple-negative tumors – it was found that any local recurrence in triple negative disease is not associated with increased metastatic risk.
All in all, we are finally beginning to arrive at a new understanding that triple negative disease is more a matter of a qualitatively different pattern of recurrence and risk, rather than as traditionally thought, a radically different low-prognostic disease entity. Indeed, Marina Cazzaniga and her colleagues at Treviglio Hospital, Italy found in the NORA study contrary to other observations, that triple negative patients did not have worse prognosis, in terms of disease-free (DFS) or overall survival (OS), than others in the total cohort of 3515 patients treated in 77 cancer centers in Italy from to 2000 to 2003. The NORA study used a median follow up of 27 months, while we know from the Dent findings that the distant recurrence risk peaks at approx. 36 months. Therefore survival past the 3 year peak would appear a seminal hurdle for ultimate survival and mortality in triple negative disease.
Myth 3: Triple Negative Disease and BRCA Deficiency
The final myth I wish to address is the garbled association between triple negative disease, BRCA1-deficient breast carcinoma and the basal molecular subtype. What's critical to note here is that:
(1) Although ER-positive tumors fall predominantly into the molecular subtypes called luminal A or luminal B, a small percentage of basal-like and HER2+/ER- tumors also appear to be classified as ER positive, confirmed in the research of Charles Perou at UNC which found that 78% of basal-like tumors are indeed triple negative, and interestingly some 6% of these basal-like tumors actually are ER+, something that of course triple negative cannot be, definitionally. Thus, although “triple negative” is often used as a surrogate identifier for the basal-like tumor subtype, this is not the whole truth, and would lead to a misclassification of a non-trivial proportion of ER+ and/or PR+ tumors as triple negative.
(2) Although approx. 80% - 90% of women with a BRCA1 gene mutation, and about 14% of women with a BRCA2 gene mutation, are triple negative, AND that most BRCA1 gene mutations exhibit the basal-like pattern, it is NOT the case, although widely misunderstood, that the preponderance of triple negative disease is associated with BRCA-deficiency (either BRCA1 or BRCA2 gene mutation): the incidence of triple negative disease (relative to all breast carcinomas) is roughly 12.5%, while only about 3.3% of breast cancer patients in the US carry a BRCA1-mutated gene, so that that only a very small percentage of women with triple negative breast cancer are BRCA1-deficient.
What is true, and what is clinically important for reasons suggested below, is that the vast preponderance of women with triple negative disease, and also the vast preponderance of women with BRCA-1-deficiency, exhibit the basal phenotype. This is the clinically significant insight as it indicates (1) that triple negative disease and BRCA1-deficient breast cancer share features and behavior associated with the basal-like molecular subtype, and (2) that therapies effectively targeting the basal carcinoma molecular subtype should be highly effective in both the treatment of triple negative disease, and the treatment of BRCA1-deficient breast carcinomas.
Triple Negative Disease: The Molecular Era
The recent breakthroughs in the molecular classification and profiling using DNA microarray analysis of breast cancers has demonstrated that breast tumors can be classified according to their genetic profile into well-defined subtypes and this has served to enrich our understanding of triple negative disease, showing, as it has, associations with loss of expression of the androgen receptor and E-cadherin and P-cadherin, positive expression of basal cytokeratins CK5 and CK17 (basal phenotype), p53, vimentin, a high MIB1 labeling index, vascular-endothelial growth factor (VEGF), and in addition appears to be strongly EGFR-driven. And Torsten Nielsen in Vancouver showed a relationship between c-KIT expression and the basal-like breast cancer subtype, with the majority of c-KIT-positive breast tumors belonging to the basal-like breast cancer subtype. Moreover, these tumors share clinical features and gene expression profiles with tumors in patients who inherit germline mutations in the breast cancer predisposition gene BRCA1 and tumors arising in patients with BRCA1 mutations tend to exhibit a very similar histological phenotype to basal-like tumors, including similar gene expression profiles with BRCA1 tumors (which heavily fall into the basal-like category). The fact that basal-like tumors tend to have very high expression levels of VEGF suggests molecularly targeting VEGF should return special benefit in triple negative disease from the anti-VEGF agent bevacizumab (Avastin). Furthermore, recent research has observed an increased frequency of the triple negative phenotype in African-American patients.
Genotoxic (DNA-Damaging Agents) for Triple Negative Disease
But the real question is what practical in-the-clinic lessons can we draw from all these considerations of molecular classification and underlying molecular pathways? It turns out that an especially important insight culled from molecular profiling, with the potential to dramatically change our notions of the optimal treatment of triple negative disease, is that basal-like and triple negative tumors, many of which as I've already indicated, are associated with BRCA1 mutation, are particularly sensitive to genotoxic modalities, that is to those that are damaging to DNA, in part because the BRCA1 pathway activity appears to be significantly impaired in many triple negative tumors. Some examples of genotoxic modalities includes DNA-damaging chemotherapy - which critically prevent the tumors from reproducing – and these include platinum compounds like carboplatin and cisplatin, as well as the classical alkylating agents like cyclophosphamide (Cytoxan), and the antineoplastic antibiotic anthracycline agents doxorubicin (Adriamycin) and epirubicin (Ellence), and Mitomycin C (MTC / Mitomycin / Mutamycin), which is also an antineoplastic antibiotic widely used in Japan but less well-known in breast oncology in the US. But it's important to note that it is not only chemotherapeutic agents that are DNA-damaging; radiation therapy is also genotoxic, suggesting that additional locoregional radiotherapy beyond the standard deployment may be of particular benefit to triple negative patients. And another non-chemotherapeutic intervention which is genotoxic is the class of biological agents known as PARP inhibitors (to be discussed further below). The practical upshot is that triple negative tumors are now known to be especially sensitive to genotoxic agents, listed in summary form and discussed further below.
List of Triple-Negative Sensitive Genotoxic Agents
Mitomycin C (MTC / Mitomycin / Mutamycin)
What About Taxanes?
Note that all anthracyclines are genotoxic and hence DNA-damaging, but the antimicrotubular taxanes, classed as mitotic spindle poisons, such as docetaxel (Taxotere) and paclitaxel (Taxol) are non-genotoxic. However, this does not mean they are not active in triple negative disease. Quite the contrary: Roman Rouzier found that these basal-like tumor are more sensitive (with a 45% pathologic complete response (pCR)) to taxane/anthracycline regimens in the form of paclitaxel- and doxorubicin-containing preoperative chemotherapy than the luminal and normal-like cancers which only sustained a 6% responsive. I should note here that another neoadjuvant study - the infamous "The Triple Negative Paradox" study of Lisa Carey at UNC – is often cited as suggesting that the clinical response (pCR) to doxorubicin and cyclophosphamide was considerably higher in patients with triple negative tumors than in those without. However, this study strikes me as somewhat methodologically compromised, as over a third of the triple negative group failed to receive any chemotherapy, and of those patients who did less than half received adjuvant anthracycline and taxane chemotherapy, casting doubt on the methodological robustness of the conclusions. Nonetheless, the weight of the evidence strongly supports both taxane and anthracycline regimens as beneficial in the treatment of triple negative disease.
New Insights about Platinum Sensitivity
An important set of data is typified in the results of the Harvard team of Chee-Onn Leong and Leif Ellisen who found that triple negative cancers independently share the cisplatin sensitivity of BRCA1-associated tumors (even in those without BRCA mutations), a sensitivity that is mediated by activation of a proapoptotic (inducing programmed cell suicide) molecular pathway p53 family member, and from this and other studies it appears that p53 is what fundamentally mediates the apoptosis induced by DNA-damaging agents.
Extending these findings John Chia's team conducted a retrospective analysis to determine the response rates of such patients treated with paclitaxel and carboplatin (TC) chemotherapy, finding that TC induces a high response rate in patients with metastatic / recurrent triple negative disease, even for patients with prior exposure to taxanes and moreover, and impressively, even for those with large volume disease.
Collectively, therefore data from preclinical and clinical studies indicate that both BRCA1 and triple negative tumors have unique sensitivities to platinum agents such as cisplatin and carboplatin, as well as to the genotoxic biological agents, the poly(ADP-ribose)polymerase (PARP) inhibitors, and these observations are helping to guide a new series of clinical trials, and at least as importantly, helping to hone and optimize the treatment of triple negative disease, and suggest for instance that adding platinum agents to taxane chemotherapy may induce high levels of efficacy for triple negative disease.
What About HDCT (High-Dose Chemotherapy)?
Triple negative disease has also been found significantly responsive to high-dose chemotherapy (HDCT) : the superiority has been shown of the (1) FEC + HDCT and (2) EC + HDCT regimens in triple negative patient subsets, as well as EP (epirubicin and paclitaxel) plus filgrastim, followed by three HDCT courses epirubicin, preceded by the cardioprotective agent dexrazoxane and paclitaxel) in high-risk triple negative breast cancer patients.
Finally, and most recently the efficacy of high-dose chemotherapy (HDCT), followed by autologous stem cell transplantation, versus dose-dense chemotherapy (DDCT) was compared, with a significantly better outcome (overall survival (OS) and event-free survival (EFS)) for patients in the basal-like, as well as the HER-2, subgroups who received HDCT in contrast to patients in largely endocrine (hormonal)-responsive) clusters who did not benefit from HDCT. Thus several studies converged to suggest the efficacy and sensitivity of high-dose chemotherapy against triple negative tumors.
HSP (Heat Shock Proteins)
Another novel insight culled from our recently gained understanding of the molecular nature and underlying pathways of triple negative disease is represented by the recent results of a study by Jose Moyana at the Robert H. Lurie Comprehensive Cancer Center and colleagues, who found that a small heat-shock protein / HSP (called alpha-basic crystalline) is commonly expressed in triple negative tumors and that this HSP overexpression increased cell migration and invasion, among other molecular activity, via the MEK/ERK pathway, suggesting that inhibition of the underlying MEK/ERK pathway may be an effective therapy for these types of basal-like breast tumors. In this connection there is a Pfizer-sponsored clinical trial exploring the novel MEK inhibitor PD-325901 in certain solid tumors including breast cancer.
I'll also note here that aspirin is known to itself be a potent MEK/ERK inhibitor suggesting a potential role in triple negative disease if further confirmed (as demonstrated early in the research of Zhongyan Wang and Peter Brecher at Boston University, Nina Vartiainen in Finland, among many others following). In addition, along with aspirin, NSAIDs like ibuprofen, and COX inhibitors (like celecoxib (Celebrex)) are independently of benefit in breast cancer risk reduction[16 – 19], a benefit that may be shared by natural COX inhibitory curcuminoid components of curcumin, which is activity in the regulation of COX-2, EGFR, VEGF, PI3K/Akt, MEK/ERK, p53, c-Myc, NF-kappaB, Bcl-2, e-cadherin, and apoptotic pathways all known to be critically involved in breast carcinomas in general and in triple negative disease in particular, as well as HER2 (ErbB2) [20 – 36], and some of which are also regulated by the activity of the EGCG (epigallocatechin-3 gallate) component of green tea[31 – 36].
Anti-VEGF / Antiangiogenic Chemobiotherapy
One of the best-evidenced highly effective regimens for basal-like / triple negative carcinoma would be the ABX-BEV combination chemobiotherapy regimen, that is, nab-paclitaxel (Abraxane) + bevacizumab (Avastin) . The now near-legendary results from Kathy Miller's ECOG-E2100, used what I would consider a somewhat weaker but similar regimen, the difference being that they used standard paclitaxel (Taxol) rather than nab-paclitaxel (Abraxane), yet even with this, it yielded a doubling of median progression-free survival (PFS), a larger absolute improvement than that seen with seminal trastuzumab trials. Given the efficacy data of Abraxane over standard paclitaxel, the ABX + BEV should therefore add an order of magnitude of further improvement without adding significant toxicity, indeed resulting in a more tolerable regimen. Furthermore, a subset analysis of ECOG-E2100 patients with triple negative disease suggested that this population benefited more from the paclitaxel + bevacizumab regimen than did hormone-responsive patients, so it appears to be a rare instance of a 3NEG-targeted regimen. The soon to open BEATRICE international trial is examining the benefit of adding bevacizumab (Avastin) to standard chemotherapy in triple negative disease.
Leveraging the "Right" Taxane
Furthermore, Spanish researchers Socorro María Rodríguez Pinilla and colleagues recently showed that CAV1 (caveolin-1) expression, a gene overexpressed with tumor progression, is associated with a triple negative phenotype in both sporadic and hereditary breast cancer. I consider the clinical impact of this insight to be substantial, given another finding from, among others, Neil Desai at American BioScience that finding being that the albumin-bound particles of nab-paclitaxel (Abraxane) preferentially deliver paclitaxel to tumors by exploiting a molecular pathway (which is called transcytosis) involving caveolin-1 (CAV-1).
In fact, nanoparticle drug carriers like nab-paclitaxel (Abraxane) preferentially accumulate in tumor beds and tissues, yielding increased antitumor activity and intratumor concentrations. Thus it appears that nab-paclitaxel (Abraxane) binds to albumin receptors inside the tumor blood vessel, resulting in the "freed" paclitaxel penetrating and killing tumor cells via microtubule binding.
So what is the upshot of all these molecular activities and interactions? Well, increased intratumoral accumulation, because nab-paclitaxel (Abraxane) appears to exploit caveolin-1 (CAV-1) to deliver more active drug (paclitaxel) selectively to tumors. This suggests that breast cancer patients with higher CAV-1 expression such as those with triple negative disease are likely to gain higher efficacy with nab-paclitaxel (Abraxane) due to CAV-1 activation, and this is a molecular advantage over the other standard formulation taxanes (paclitaxel (Taxol) and docetaxel (Taxotere) which exhibit no comparable CAV-1 specific activity. It strikes me that therefore this greatly hones our targeting of the underlying molecular pathways of triple negative disease and provides a rough but suggestive evaluation metric that nab-paclitaxel (Abraxane) might be more optimal in this context than standard taxanes for the treatment of triple negative tumors. I observe further that the TKI dasatinib (Sprycel) – discussed further below - is also active against CAV1, making it to some extent triple negative-specific ("triple negative-targeting").
Enhancing the ABX-BEV Regimen Further
And as indicated above in our discussion of the platinum agents, adding such a platinum agent like carboplatin (yielding ABX + BEV + CARBO) is a increasingly deployed practice (growing out of some preclinical work, and of findings from trastuzumab (Herceptin) trial data, as well as from triple negative populations with inherited BRCA1/2 mutations).
In the introduction above, I noted that EGFR over-expression in triple negative and basal-like breast carcinoma is now well-established, as is therefore the therapeutic value of EGFR-inhibition, given that triple negative tumors are EGFR-signaling dependent, highly expressed in at least 50% of all such tumors. Preclinical evidence from Zyhiyuan Hu and colleagues with the Lineberger Comprehensive Cancer Center at UNC and Stefano Calza at the Swedish Karolinska Institutet and his coresearchers, suggests that the molecular profile of triple-negative breast cancer is characterized by a unique signature that includes EGFR gene overexpression, suggesting an important role for monoclonal antibodies (MoAbs) binding the extracellular ligand-binding domain such as cetuximab (Erbitux). Rebecca Clark-Snow at the University of Kansas is exploring in clinical trial the value of the EGFR inhibitor erlotinib (Tarceva) added to chemotherapy for triple negative disease.
EGFR-Targeted Therapies: Cetuximab (Erbitux)
It's been determined that combinations of cetuximab + carboplatin are highly synergistic at low doses of each drug, according to the preclinical research of Katherine Hoadley, along with Lisa Carey[45,46] at UNC, who showed (1) that of all breast cancer subtypes, basal-like tumors are both the most sensitive to EGFR inhibitors and carboplatin individually, (2) that the combination was synergistic as well, not just additive, and that (3) the EGFR-RAS-MEK pathway may be a requisite event for basal-like tumor formation, guiding targeted therapy. In addition, the Bali-1 trial is examining the benefits of cetuximab + cisplatin in triple negative disease.
Given this molecular foundation, there are now several trials exploring the potential of EGFR inhibitors in triple negative disease in the MBC (metastatic breast cancer) setting and evaluating a combination of EGFR-inhibitors + platinum agent. Of these considerable interest surrounds those using the MoAb (monoclonal antibody) required to above, cetuximab (Erbitux) , added to a platinum agent, carboplatin. One ongoing, actively recruiting, phase II clinical trial at MD Anderson under Francisco Esteva, principal investigator is randomizing patients to receive either cetuximab alone, with the addition of carboplatin upon progression, or cetuximab + carboplatin; this trial is due to report later this year. (A parallel trial is under Lisa Carey's group at UNC Lineberger Comprehensive Cancer Center, coordinated across UNC, the Mayo Clinic in Rochester Minnesota, and Baylor in Houston, deploys the same protocol as the MD Anderson trial).
Another still actively recruiting EGFR-inhibitor clinical trial is the ongoing US Oncology Research study under Joyce O'Shaughnessy evaluating weekly irinotecan (Iressa) + carboplatin with or without cetuximab in patients with MBC, and although not triple negative-restricted, I have ascertained from trial authorities that a substantial number of patients on this trial have triple-negative disease. This approach of this trial reflects the use of small-molecule TKIs (SM-TKIs) such as gefitinib (Iressa) and erlotinib (Tarceva) ) as ATP-competitors for binding to the intracellular tyrosine kinase domain, where ATP is a known binding site of EGFR so that such SM-TKIs compete with such binding, and hence blocking the activation of various downstream signaling pathways. And Cynthia Ma at Washington University is conducting another ongoing Cetuximab-Carbo(platin trial.
EGFR-Targeted Therapies: Sunitinib (Sutent)
Given that SM-TKIs are biological agents with multiple receptor targets (including VEGF, like Avastin, as well as several others involved in angiogenesis, and in cellular proliferation), and have not only been used successfully in treating GIST and renal cancer, but also breast cancer with some promise. Besides the SM-TKI gefitinib (Iressa) discussed above, there is interest in another such SM-TKI, sunitinib (Sutent) , and there is currently a large multi-center, multi-state, and international, actively recruiting Pfizer-sponsored trial of sunitinib (Sutent) in previously treated patients with advanced triple negative disease, that is locally recurrent or metastatic; one restriction is that no previous treatment with an angiogenesis inhibitor like bevacizumab (Avastin) is allowed for trial eligibility.
EGFR-Targeted Therapies: Dasatinib (Sprycel)
Another small molecule TKI (SM-TKI) is dasatinib (Sprycel) . Dasatinib is a novel oral multitargeted kinase inhibitor that targets several important oncogenic pathways, including SRC family kinases and BCR-ABL. Dasatinib is already established in the treatment of one prominent form of leukemia (CML), but is largely unknown in breast cancer with the exception of a single in vitro cell study by Richard Finn and colleagues at UCLA, which found basal-type / triple negative breast cancer cell lines to be preferentially inhibited by and highly sensitive to dasatinib, and this has been confirmed via gene signature exploration by Fei Huang at BMS (Bristol-Meyer Squibb). BMS is currently sponsoring a multi-center, actively recruiting trial of dasatinib in triple negative patients. The importance, and promise, of dasatinib lies in part on the fact that the SRC oncogenic pathway plays an important role downstream of vascular endothelial growth factor (VEGF) signaling, and so it is anticipated that dasatinib will also have antiangiogenic activity. In addition, because SRC plays an important role in osteoclast function, it is possible that dasatinib will benefit patients with bone metastases, in addition to its antiangiogenic activity.
A Note of this Contribution (Part 1)
Although I have not included source references directly in this posting to avoid excessive technical "weight", the reference numbers are included – in the form of a bracketed  number or numbers in the text – and anyone interested can access the sources themselves in the technical version of this posting available online as Issue 3 of my Breast Cancer Watch Digest newsletter, where the references are also hyperlinked to the original sources. That version also contains some additional tutorial-type material on DNA damage and PARP inhibitors, and briefly describes the methodology of the review, while this special invited presentation for the No Surrender forum has additional material relating to recent conference discussions plus some hints at findings to be reported at the upcoming SABCS 2007 this December.
Breast Cancer Watch