Your faculty for this presentation will be Dr.  Lee Schwartzberg  from the West Clinic in Memphis, Tennessee.

Dr. Schwartzberg: This is an exciting area. In addition to getting testing for the size of their tumor and whether or not there are positive lymph nodes, patients in the United States get a Bloom-Richardson grade; this gives us additional information, although it’s variable. Perhaps most importantly, all patients get tested for the presence of estrogen and progesterone receptors and the presence of over-expression and/or amplification of HER2. Using all of these factors together, we can now start to individualize treatment options for early-stage breast cancer.

More recently, we have developed the ability to further subtype patients using gene expression testing. The most fully advanced of these tests is the Oncotype DX® 21-gene recurrence score assay. It’s an approved test that has been in the clinic now for several years. In my opinion it’s very useful in helping to determine what the best therapy is for certain groups of early-stage breast cancer.

The Oncotype DX test is approved for ER-positive, lymph node-negative patients and gives a continuous readout from the 21-gene reverse transcriptase PCR testing of expression of those genes. It has been validated to show what the 10-year risk of distant recurrence will be based on the recurrence score that is generated from those 21 genes. In practical terms it sorts patients into three groups: low-risk, intermediate-risk, and high-risk.

The benefit of the Oncotype DX test is twofold: first of all it’s prognostic, so independent of therapy, it tells what the risk of recurrence will be for that individual tumor; perhaps more importantly, it is predictive. For patients who have low recurrence scores, the test is predictive for high benefit from tamoxifen and no benefit from additional chemotherapy over tamoxifen. For patients who have high recurrence scores, it is predictive for limited benefit from tamoxifen and high benefit from chemotherapy. The intermediate group tends to have some benefit from tamoxifen and some benefit from chemotherapy, although the confidence intervals are wide there. The benefit of chemotherapy is not clearly established in the intermediate group. We can use this test to determine whether or not patients will benefit from the addition of chemotherapy to hormonal therapy in a variety of ER-positive, node-negative settings.

At San Antonio there were a number of posters that addressed how the use of the Oncotype DX test actually works in the real-world setting. A multicenter prospective study assessed the impact of the assay on patient satisfaction, anxiety, and decisional conflict for adjuvant breast cancer treatment selection.¹ In general, patients were very open to having the test done on their tumor. In addition, 27% of women had a change in their adjuvant therapy that was recommended as a direct result of the Oncotype DX score. Patients felt better and were more satisfied with the decision making when they were included in the discussion and decision to use the Oncotype DX test. The bottom line was that this was an additional, valuable tool in the dialogue between the physician and the patient in terms of reaching the best treatment decision.

A retrospective analysis assessed the impact of using the Oncotype DX test on treatment decisions in a single academic center: The University of Pittsburgh Cancer Institute.² In this case they compared the risk score as generated by the NCCN guidelines, and they found a big difference between the Oncotype test and how the NCCN sorts women into low- and high-risk groups based on clinical factors. It appeared that the Oncotype test was better able to stratify women into different classes. In addition, they did a cost-saving analysis, which suggested that because there were several women who were considered high-risk clinically but were found to be low-risk by Oncotype DX score and therefore did not receive chemotherapy, that there was a net savings by using the Oncotype test in appropriate patients; it offset the cost of the chemotherapy that was not given.

OncoEd: Does gene-expression profiling have any role in predicting response for node-positive, ER-positive early-stage breast cancer patients?

Dr. Schwartzberg: The benefit of gene-expression profiling is now well established in ER-positive node-negative patients. The obvious question that clinicians have been asking is does it also help in patients who are node-positive. One of the most important presentations from San Antonio this year addressed that issue.

I think the answer to that question is becoming clearly “yes” with some provisos. One of the most important presentations from San Antonio was an oral presentation of the Intergroup study 0100 that has previously been the subject of multiple analyses.³ This study looked at patients who received tamoxifen alone or tamoxifen plus CAF chemotherapy. For the overall group, it showed a benefit for receiving chemotherapy over tamoxifen alone for hormone receptor-positive patients and node-positive disease. In this particular analysis, they went back and subjected patients who were on that trial to the 21-gene recurrence score assay. Overall, 40% of the patients who were on the parent trial had tumor blocks that were available to analyze by the Oncotype DX test. The results in this subset of patients were similar to the results in the larger study population: there was an overall benefit to CAF and tamoxifen over tamoxifen alone. When they then did the Oncotype test and looked at the risk groups, 40% of the patients were in the low-risk group, 28% were in the intermediate-risk group, and 32% were in the high-risk group. This is a different distribution compared with the prior studies of node-negative disease in the sense that more patients in this trial were classified as high-risk; this is not too much of a surprise because all of the patients here had node-positive disease. The main result of this analysis was that the 21-gene recurrence score was first of all prognostic for disease-free survival and overall survival in the tamoxifen arm, paralleling the data that has been previously seen in the node-negative patients. Also, and very importantly, for patients who had a low recurrence score (RS), there was no additional benefit to CAF compared with tamoxifen alone. These results are virtually identical to what we are seeing in the node-negative patients. Conversely, patients who had a high RS showed a strong benefit from the addition of CAF to tamoxifen. These results were statistically significant. In the intermediate-risk group, there was no statistically significant improvement with chemotherapy, but numerically, patients who received CAF-T did somewhat better than those who received tamoxifen only.

In the low RS group, 10-year DFS was 60% with tamoxifen alone and 64% with CAF-T. In the high RS group, 10-year DFS was 43% with tamoxifen alone compared with 55% with CAF-T; that was a strong benefit. Therefore the 21-gene recurrence score assay was not only prognostic, but also predictive for the benefit of chemotherapy, and this was statistically significant in a Cox multivariate analysis. The RS was also predictive for OS in this trial; there was no benefit to CAF in the low RS group in either the first five years or over the entire time period, but there was a strong impact of CAF in the high RS group, both in the first five years and carrying over the entire 10 years of follow-up. In the high RS group, 10-year OS was 51% among patients receiving tamoxifen alone compared with 68% among patients receiving CAF-T. The conclusions of the authors were that this 21-gene recurrence score was prognostic for tamoxifen-treated patients with positive nodes, that chemotherapy benefit is predicted when the RS is high, and that a low RS may define a group of women with positive nodes who do not appear to benefit from anthracycline-based chemotherapy.

There is one other trial with node-positive patients that had previously been presented at ASCO and was updated at San Antonio, and that is an analysis of ECOG 2197.⁴ This trial design was different from the Intergroup trial in that it did not compare tamoxifen to tamoxifen and chemotherapy; the arms of this trial compared two different types of chemotherapy and all patients received tamoxifen if they were hormone receptor-positive. They selected a group of patients for the Oncotype DX test and they did centralized testing for ER, PR, HER2, and grade. What they found was that for hormone receptor-positive patients with 0-3 positive nodes treated with a standard adjuvant chemo-hormonal therapy, they had comparable outcomes for 1-node positive versus 0 nodes. The RS correlated positively with recurrence, and a low RS was predictive of an excellent outcome at five years, with only a 3% recurrence rate if one positive node was detected. It should be kept in mind that everyone received chemotherapy in this case. A low RS was found in almost half of these patients who were hormone receptor-positive. The authors concluded that the RS provides additional information that is complementary to the clinical pathologic features that are normally reported on breast cancer patients and that are integrated into the analysis provided by Adjuvant! Online, another model that predicts DFS after adjuvant therapy.

OncoEd: Can you comment on the differences between Oncotype DX and the 70-gene assay, MammaPrint®?

Dr. Schwartzberg: There are other tests that are being developed to evaluate the expression of genes that are important in prognosis and prediction of response to specific breast cancer therapies. The Oncotype DX test is done on formalin-fixed paraffin-embedded tissue, which has the advantage that it can be done on archival samples and can be done by RT-PCR, which remains accurate in that setting. Another approach is to look at gene microarray technology, and that is what is done in another test, MammaPrint, which is done on 70 genes; interestingly, almost a completely different gene set from those found in the 21-gene recurrence score assay. This test is done on fresh-frozen tissue, and because most American pathology groups and surgeons are not set up to store fresh-frozen tumor tissue at this moment, MammaPrint has limited practical applicability. However, from a scientific perspective it remains very interesting indeed.

A study among patients with 1-3 positive lymph nodes assessed how well the 70-gene signature determined outcome.⁵ Of 106 patients, 43 were classified as good-prognosis based on the gene signature and 63 were classified as poor-prognosis. Node-positive patients with a good-prognosis signature had a 98% survival at 10 years and 95% distant DFS at 10 years compared with 54% OS for poor-prognosis signature patients and 70% distant DFS at 10 years. It appears that the 70-gene prognosis signature can identify low risk groups in the 1-3 positive lymph node group as well.

Another study of the 70-gene signature involved 241 patients, of whom 41% fell into the good prognosis profile and 59% into the poor-prognosis profile.⁶ Eight-year follow-up showed significant difference in disease outcome between the good and poor profile. However, these sets included patients who were treated with hormonal therapy or chemotherapy or both. They don’t have the degree of analytic rigor that the Oncotype studies did where they were looked at different therapeutics within the context of the clinical trial.

OncoEd: Among early-stage breast cancer patients who are receiving an aromatase inhibitor, what can be done to minimize bone complications?

Dr. Schwartzberg: The biggest problem with aromatase inhibition is the fact that because it depletes all the estrogen in the body, it has a significant effect on normal tissues that are dependent on estrogen and perhaps most importantly, on the bone, as the development of structural bone is dependent on having estrogen. There has been both theoretical and objective evidence that aromatase inhibitors (AIs) reduce bone mineral density and can potentially increase the risk of fractures. There has been a lot of study looking at ways to predict which patients who receive AIs are at risk of developing bone-related problems and at ways to appropriately intervene with patients on AIs to prevent any bone-related problems. Most of those interventions have looked at the use of bisphosphonates. One of the most important sets of presentations at San Antonio this year related to the use of a variety of agents, particularly oral and IV bisphosphonates, to prevent or reduce the impact of AIs on bone health.

One presentation provided updated results from the Z-fast study.⁷ The Z-fast study assessed zoledronic acid, a potent IV bisphosphonate, given at a dose of 4mg every six months to postmenopausal patients receiving AIs. They were either started on it right away or when their bone density fell, as evaluated by DEXA scan. They showed very positive results, indicating that use of zoledronic acid every six months preserved and actually improved bone mineral density of patients on AIs as compared with patients who did not receive it and who had a progressive loss in bone mineral density.

Another study was presented by the Austrian Breast Cancer Trial Group.⁸ This study looked at using ovarian suppression in premenopausal women along with either tamoxifen or the AI anastrozole. As part of a sub-study, some patients also received zoledronic acid. Among patients who did not receive zoledronic acid, there was a significant drop in bone mineral density in the lumbar spine for patients who received either tamoxifen plus ovarian suppression or anastrozole plus ovarian suppression. Among patients who did receive zoledronic acid, bone mineral density was improved at five years in both the tamoxifen and anastrozole arms. The overall difference between zoledronic acid and no-zoledronic acid was 10% net improvement in bone density at five years.

There was also a poster entitled “Practical guidance for the prevention of aromatase inhibitor-associated bone loss in women with breast cancer.”⁹ This was based on metaanalyses of risk factors for bone loss that included not only AI but other known factors that were both patient-specific and historical. Their recommendations were that all patients on an AI should receive calcium and vitamin D supplements, and that patients with a baseline T-score by DEXA scan of less than 2 should receive zoledronic acid 4 mg every 6 months IV. They also recommended zoledronic acid for AI-treated patients with any two of the following risk factors:

• T-score less than 1.5, consistent with osteopenia
• Age over 65
• Low BMI
• Family history of fracture
• Personal history of fragility fracture after age 50
• Oral corticosteroid use after six months
• Current smoking

They recommended that treatment should continue for at least two years, pending the long-term results of the Z-fast study and other studies that are looking at long-term usage of zoledronic acid in patients receiving AIs. I think these are useful recommendations and ones that we certainly need to incorporate into our routine practice so that we can preserve the benefit of AIs without the common potential side effects.

OncoEd: What are your comments on the data presented on the monoclonal antibody denosumab and bone density among women treated with an AI?

Dr. Schwartzberg: One of the most interesting presentations at San Antonio was a late-breaking abstract of the use of denosumab a new monoclonal antibody in clinical trial as an alternative treatment for the preservation of bone density in selected groups of patients.¹⁰ Denosumab is a novel agent that inhibits bone resorption through its effect on the RANK ligand, an important and central pathway in the development of bone. This study looked at denosumab in women with non-metastatic breast cancer receiving aromatase inhibition therapy. The women had some degree of osteopenia, defined as a T-score of less than -1 but greater than -2.5. They were excluded if they had osteoporosis or prior oral bisphosphonates. It was a placebo-controlled study. The drug was delivered subcutaneously ever six months. Compared with placebo there was a 5.5% difference in lumbar spine bone mineral density at 12 months after initiation of denosumab, which increased to 7.6% difference at 24 months; this was highly statistically significant. Similar results were seen in the analysis of the total hip bone density and distal third of the radial density. Analysis of adverse events showed minimal toxicity attributable to denosumab. The degree of arthralgias, pain in extremity, back pain, and fatigue were virtually identical in both arms. Twelve percent had constipation compared with 9% in the placebo arm. The conclusion was that denosumab, a new monoclonal antibody with a novel mechanism of action, provided consistent increases in bone mineral density over the duration of the 24 month study and was very-well tolerated. There are ongoing clinical trials with denosumab as well.

OncoEd: In light of the recent controversy over erythropoeisis stimulating agents, where do we stand in the management of chemotherapy-induced anemia for breast cancer patients?

Dr. Schwartzberg: There was little new information at San Antonio about this topic. However, much has come out over the last month regarding the use of ESAs in breast cancer. There are now two studies that suggest that patients who receive these agents in a manner that is not consistent with the label—that is to drive hemoglobin levels into the high normal or high range, above a hemoglobin of 12—may experience adverse effects associated with the use of these agents. These adverse effects may include a potential impact on survival, and possibly DFS, in ways that are not completely clear. There are no data to suggest that patients with breast cancer who use ESAs in a manner specified per the label—that is to start when the patient’s hemoglobin drops below 12 and typically below 11, and maintaining hemoglobin in the range of 11-12—have any adverse outcomes. Furthermore, patients who receive ESAs on-label have a reduction in their risk of requiring transfusions and an improvement in QOL as measured by a variety of QOL instruments.

OncoEd: Can you comment on the status of dose-dense therapy for breast cancer patients, and have we made progress to minimize dose delays and dose reductions that impact survival?

Dr. Schwartzberg:Dose-dense therapy remains the standard of care for the treatment of certain subsets of breast cancer in the adjuvant setting, and that is based on particularly the CALGB 9741 study, but other trials that have also substantiated the benefit of the dose-dense approach. In order to deliver dose-dense chemotherapy, granulocyte colony stimulating factor support is necessary, as the bone marrow must be prompted to recover faster than it would normally in order to achieve the kinetics of more frequent dosing of chemotherapy. No significant new data on dose-dense chemotherapy were presented at San Antonio this year, but there continue to be refinements in supportive care. We know that we can replace daily G-CSF with once per cycle peg-figrastim given along with dose-dense chemotherapy on an every 14-day basis without any significant negative impact. In general, colony stimulating factors have proven to be very effective in supportive care to minimize dose delays and dose reductions in the adjuvant setting, where there is substantial evidence that any significant dose delays or reductions could potentially impact long-term survival.

OncoEd: Was there any important news about adjuvant hormonal therapy to come out of San Antonio?

Dr. Schwartzberg: I think the most important news on adjuvant hormonal therapy was the 100-month update of the ATAC trial.¹¹ ATAC enrolled postmenopausal women with invasive breast cancer and compared anastrozole for five years with tamoxifen for five years. It was a very large trial with approximately 3,000 patients in each arm. At nine years 21.8% of patients in the tamoxifen group have had a recurrence compared with 17% in the anastrozole group; the hazard ratio of 0.76 was highly statistically significant. The percentage of patients who had distant recurrence at nine years was also greater in the tamoxifen arm compared with the anastrozole arm. Contralateral breast cancer had a hazard ratio of 0.6 for patients receiving anastrozole versus tamoxifen at nine years of follow-up. All-cause death was not significantly different between the two arms with a hazard ratio of 0.97. Serious AEs were less common on treatment for anastrozole compared with tamoxifen, except for fractures. Interestingly, although there were approximately one-third more fractures in the anastrozole arm compared with tamoxifen while the patients were on therapy during the first five years, during the next four years there were virtually identical numbers of fractures. This suggests that the negative effect of anastrozole on bone disappears after the drug is discontinued, as opposed to the significant benefit in reducing the risk of breast cancer recurrence; that is an important distinction.

OncoEd: What do you consider the most urgent priorities for future clinical trials in the adjuvant setting?

As we have discussed throughout this interview, the theme of subsetting patients is perhaps even more important in the adjuvant setting than in the metastatic setting. We already subset patients and treat patients with ER-positive disease somewhat differently than ER-negative disease, and with the use of Oncotype DX and other gene prognostic tests, we will be doing that more in the future. The results already suggest that we can extend those observations from node-negative to node-positive patients, which would include the vast majority of patients who present for adjuvant therapy.

We have good therapies for HER2-positive patients with trastuzumab, but an important new trial is launching to compare trastuzumab to the other new approved agent, lapatinib. It will be a very large four-arm trial, looking at chemotherapy with trastuzumab versus chemotherapy with lapatinib versus combination chemo plus trastuzumab alone, lapatinib alone, a combination of trastuzumab and lapatinib, or sequential trastuzumab and lapatinib. That will be a very important trial to help advance the treatment of HER2-positive patients.

We clearly need targeted therapy against the triple-negative group, which has a very poor-prognosis, even in the adjuvant setting.

An issue was raised at San Antonio this year about the benefit of anthracyclines. Updated results from a study of TC versus AC as adjuvant chemotherapy now show a survival benefit with TC.¹² This suggests that one could omit the anthracycline without deleterious effect, at least in relatively low-risk women who need chemotherapy. An interesting analysis by Dr. Slamon suggests that the benefit of anthracyclines is confined to patients who overexpress topoisomerase 2, which is the major target of anthracyclines.¹³ It appears that topoisomerase 2 overexpression is tied to overexpression of HER2, as the two genes lie relatively close to one another on chromosome 17. In Dr. Slamon’s analysis of patients from the BCIRG006 study who had tissue samples provided, patients who had topoisomerase 2 overexpression always had HER2 overexpression as well. Therefore HER2 can be looked at as a surrogate marker for topoisomerase 2 overexpression. However, topo2 overexpression only occurs in one-third of patients with HER2 overexpression. So we may be seeing an era in the future where anthracyclines are used less commonly, particularly in patients who don’t have topoisomerase 2 overexpression. That needs to be tested in prospective clinical trials.

The use of the gene expression profiling is already well underway, and these are some of the most important adjuvant trials that are now ongoing around the world. In the United States the TAILORx study is taking women who have hormone receptor-positive, node-negative disease and performing the Oncotype DX test on them. Those with low risk scores receive hormonal therapy only, and those with high risk scores receive chemotherapy and hormonal therapy. The women that have intermediate risk scores are randomized to receive hormonal therapy plus or minus chemotherapy, and that will prospectively validate, if true, the benefit of the Oncotype DX score in that group of patients. A similar trial is being done in Europe with the 70-gene MammaPrint assay; that study is called the MINDACT trial.

As always, San Antonio is a wonderful opportunity for any participant to gain a deep appreciation of the entire gamut of breast cancer research and clinical development. Certainly, it gives clues into where the field may be heading in the future. There is emerging evidence that breast cancer stem cells may be the ultimate target for curing breast cancer, and there was intriguing preliminary data along those lines. I think the future looks very bright based on what we’ve seen at this year’s San Antonio.

References:

¹ Mumby PB, Los SS, Norton J et al. Prospective multi-center study of the impact of the 21-gene recurrence score assay on patient satisfaction, anxiety and decisional confluct for adjuvant breast cancer treatment selection. Presented at the 30th Annual San Antonio Breast Cancer Symposium, San Antonio, TX, December 13-16, 2007. Abstract 1092.

² Liang H, Brufsky AM, Lembersky BB, Rastogi P, Vogel VG. A retrospective analysis of the impact of oncotype DX low recurrence score results on treatment decisions in a single academic breast cancer center. Presented at the 30th Annual San Antonio Breast Cancer Symposium, San Antonio, TX, December 13-16, 2007. Abstract 2061.

³ Albain K, Barlow W, Shak S et al. Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal, node-positive, ER-positive breast cancer (S8814,INT0100). Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 10.

⁴Goldstein L, Ravdin P, Gray R et al. Prognostic utility of the 21-gene assay compared wtih Adjuvant! in hormone receptor (HR) positive operable breast cancer with 0-3 positive axillary nodes treated with adjuvant chemohormonal therapy (CHT): an analysis of intergroup trial E2197. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 63.

⁵ Mook S, Rutgers EJT, Peterse JL et al. The Amsterdam 70-gene signature predicts outcome in breast cancer patients with 1-3 positive axillary lymph nodes. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 1064.

⁶ Mook S, Schmidt MK, Viale G et al. Breast cancer patients with 1-3 positive lymph nodes and a low risk 70-gene profile have an excellent survival. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 50.

⁷ Brufsky A, Bosserman L, Caradonna R et al. The effect of zoledronic acid on aromatase inhibitor-associated bone loss in postmenopausal women with early breast cancer receiving adjuvant letrozole: the Z-FAST study 36 month follow-up. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 27.

⁸ Gnant M, Mlineritsch B, Luschin-Ebengreuth G et al. Bone mineral density (BMD) at 5 years after diagnosis in premenopausal patients with endocrine-responsive breast cancer, after 3 years of adjuvant endocrine treatment with goserelin and tamoxifen or anastrozole or both treatments in combination with zoledronic acid new results from ABCSG-12. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 26.

⁹ Hadji P, Appro M, Brufsky A et al. Practical guidance for the prevention of aromatase inhibitor-associated bone loss in women with breast cancer. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 504.

¹⁰ Ellis G, Bone HG, Chlebowski R et al. A phase 3 study of the effect of denosumab therapy on bone mineral density in women receiving aromatase inhibitors for non metastatic breast cancer. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 47.

¹¹ Forbes JF, Cuzick J, Buzdar A et al. ATAC: 100-months median follow-up (FU) shows continued superior efficacy with no excess fracture risk for anastrozole (A) compared with tamoxifen (T) after treatment completion. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 41.

¹² Jones S, Holmes F, O’Shaughnessy J. Extended follow-up and analysis by age of the US Oncology Adjuvant trial 9735: docetaxel/cyclophosphamide is associated with an overall survival benefit compared to doxorubicin/cyclophosphamide and is well-tolerated in women 65 or older. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 12.

¹³ Slamon DJ, Machey J, Robert N et al. Role of anthracycline-based therapy in the adjuvant treatment of breast cancer: efficacy analyses determined by molecular subtypes of the disease. Presented at the 30th Annual San Antonio Breast Cancer Symposium. San Antonio, TX, December 13-16, 2007. Abstract 13.