I’m very pleased with the RxPONDER results. We were proud of the accomplishments made by SWOG, the cooperative groups, our colleagues, and patients who participated in this study. This is really a practice-changing study. It settles an important question regarding the value of chemotherapy in younger and older woman with stage II breast cancer, 1 to 3 positive nodes. I think the study is changing practice in an important way and will enable thousands of women to avoid chemotherapy safely. Also, it helps patients who are younger feel more comfortable with chemotherapy, because they know that this could potentially save their life.

I should mention that these results are the first preliminary results from RxPONDER. We have about 60% of the planned events, but even with 60% of planned events, we have a statistically valid and solid finding. Of course, the final major results of RxPONDER will be published as well, once all the events have been accrued. Currently, about 60% of all events that were expected from the trial have occurred. We also plan to do some important correlative science studies, particularly to get more precise answers to the question of, what’s the contribution of the chemotherapy-induced ovarian suppression versus the chemotherapy-related cytotoxicity to the benefit that is observed?

: What additional questions do the results of SWOG-1007 pose to us for our patients in the clinic?

We touched upon a couple of them. The first being in postmenopausal women, can we now take a step further and look at strategies to avoid chemotherapy in patients who have more than 3 positive nodes? I think that will be answered in some of the ongoing studies. That’s one of the questions. The second question that pertains to premenopausal women is what is the value of ovarian function suppression in this subgroup with a recurrence score less than 25 and 1 to 3 positive nodes? Will ovarian function suppression be an add-on to chemotherapy, or will it provide benefit equivalent to chemotherapy? I think that is a question that will have to be answered in a prospective trial.

Videos

Lajos Pusztai, MD, PhD, and Priyanka Sharma, MD, share closing thoughts on additional data they hope to see from the RxPONDER trial, and using these data in clinical practice for the treatment of HR+ early breast cancer.

HR+ Breast Cancer: Role of RxPONDER Data in Clinical Practice

The important question that the RxPONDER results pose is whether premenopausal woman with 1 to 3 positive nodes, who seem to have benefited from adjuvant chemotherapy regardless of the recurrence score, even below 25, should actually get a recurrence score. Since we know that it doesn’t matter what their recurrence score is, they still benefit from chemotherapy. I think it is a fair conclusion from the study that most premenopausal woman, with lymph node-positive disease that’s only 1 to 3 nodes, should safely decide to go along with chemotherapy. There is evidence that it could improve their survival: both the invasive disease-free and distant, recurrence-free survival. However, in some instances, it might be helpful to get a recurrence score, even in this population. What are these instances? For example, patients who want to see the absolute predicted benefit from chemotherapy, because of the potential risks of chemotherapy, or patient reluctance to accept it, or comorbid illnesses. Seeing the actual risk of recurrence as a predictive number, to help show the improvement chemotherapy would add, would empower a person to make a personalized decision based on their own value system, about the risks and the benefits of the chemotherapy. The value of the recurrence score, in this instance, is really providing an objective metric on which the patient could base a decision. However, I also think that for many premenopausal women, the test is not that helpful.

: I think the question comes down to how confident we are, in every situation, that the number of positive nodes is no more than 3, because the eligibility criteria for RxPONDER was 1 to 3 positive nodes. And the treatment paradigm for managing axillae has changed based on well done surgical trials. These surgical trials demonstrate that patients who have low-burden axillary disease can forego further axillary dissection because the chances of finding additional lymph node positivity is low and doesn’t negatively impact their survival. In SWOG-1007, about one-third of patients had sentinel lymph node biopsy only. In my practice, I feel comfortable in making my decisions based on sentinel lymph node biopsy only if the patient had 1 or 2 positive nodes and they fulfill the criteria for avoidance of axillary dissection. I don’t feel I need to recommend an axillary dissection for all postmenopausal women with node-positive disease.

 The result seen in the TAILORx and RxPONDER studies is highly consistent. Remember in TAILORx, the group that benefited was the younger woman, less than 50 years old. In RxPONDER, it was the premenopausal women, the majority of which were also less than 50. If the chemotherapy is beneficial, it could be for 2 reasons. One is the cytotoxic effect of chemotherapy on micrometastatic disease. Basically, the chemotherapy kills the micrometastatic cells through its cytotoxicity. The other well-recognized effect of chemotherapy is on the ovarian function of younger women. Chemotherapy causes ovarian function suppression in perimenopausal women, or women in their late 40s. The ovarian suppression is oftentimes permanent.They undergo a premature chemotherapy-induced menopause. In younger women, especially in their early 30s, chemotherapy-induced menopause is oftentimes reversible and only lasts for a year or two. We also know that in younger women, in the combined moderated hormonal therapy, there is ovarian suppression plus tamoxifen, which is the drug that was given in most of the patients in RxPONDER because, remember, this study was conducted between 2010 and 2017. Tamoxifen with ovarian suppression is better than tamoxifen alone. Ovarian suppression can be induced by Lupron, or leuprolide, or GnRH [gonadotropin-releasing hormone] analogues, but it also could be induced by chemotherapy. The question is, does the chemotherapy-induced ovarian function, or the cytotoxicity of chemotherapy, drive the benefit in these 2 trials? This is unknown. Of course, this is a very important question because chemotherapy is more toxic than ovarian suppression with another drug, such as a GnRH analogue. My opinion is that both effects of chemotherapy probably contribute to the benefit. It’s not necessarily a mutually exclusive question of either this or that. I think both effects contribute, and the relative contribution may vary by age. I think, in younger women in their 30s, it’s the cytotoxic effect that is probably more important. In older women, or perimenopausal women, who are just 1 or 2 years away from their natural menopause, the chemotherapy-induced menopause may be an important driver. However, we don’t really know this, and there are additional studies that are planned behind the RxPONDER study, where we want to get serum markers to try to get more precise answers to this question. But ultimately, new clinical trials will be required to really settle this question.

: How should we take these results in women with 4 or more positive nodes, especially the postmenopausal women? Can we take the results from SWOG-1007 and apply it to those women?

We do not have data from trials on this question since SWOG-1007 only included patients with N1 disease. I think this is a question that is being answered in another large trial, OPTIMA, which is enrolling patients with a higher clinical risk disease. Currently, I do not take these results to patients with N2 or higher disease. If they have higher clinical risk disease, even if they’re postmenopausal, I am recommending chemotherapy. I’m not translating this to higher nodal disease in my practice yet. I think we need to await more data on that.

The question is whether these findings have any relevance in the neoadjuvant setting. Remember, both TAILORx and RxPONDER are adjuvant trials. Patients had their tumor removed, and then their pathological assessment shortly after, for positive or negative disease, with the precise understanding of how many nodes are involved and how big the tumor was. However, these results do have some implications for the use of neoadjuvant chemotherapy in early stage breast cancer. Many studies have shown that pathological complete eradication of tumors almost never happens in ER-positive disease that has a recurrence score of less than 25. This goes along with the observation from both of these studies that by and large, a recurrence score below 25 is not associated with significant chemotherapy sensitivity. I think neoadjuvant chemotherapy has some limited value in ER-positive disease. It certainly can downstage the tumor, or shrink the cancer, and make the operation easier. Or an inoperable patient could be converted to operable by neoadjuvant chemotherapy. However, complete pathological eradication of the cancer is very rare.

I think neoadjuvant chemotherapy works best in the very high recurrence score spectrum, recurrence scores above 30. An alternative to neoadjuvant chemotherapy in ER-positive patients is neoadjuvant endocrine therapy. Endocrine therapy can also accomplish many of the same benefits that chemotherapy would in this setting, such as downstaging the tumor, reducing the size of the cancer. I think for patients who have a lower recurrence score on the core biopsy, neoadjuvant endocrine therapy is probably a better choice than neoadjuvant chemotherapy. On the other hand, women with a high recurrence score who would require chemotherapy anyway, neoadjuvant chemotherapy could be meaningful in the sense that it makes their surgery smaller because the cancer shrinks. Ultimately, they would require the same chemotherapy anyway postoperatively, because of the high recurrence score.

Questions to consider when approaching the treatment of HR+, HER2-breast cancer based on the RxPONDER trial results.

HR+ Breast Cancer: Impact of RxPONDER on Clinical Practice

: These results have, in my opinion, established for us that postmenopausal women with 1 to 3 positive nodes with hormone receptor-positive, HER2-negative disease who have a recurrence score of less than 25, would not benefit from chemotherapy. It spares a lot of these women toxic treatment, and we can safely forgo chemotherapy in postmenopausal women. Whereas, in premenopausal women, we do see a benefit of chemotherapy, with an absolute difference ranging from 3% to 6.5%, depending on the different subgroups in the premenopausal subset. I think that led to the adoption of this in the NCCN [National Comprehensive Cancer Network] guidelines, which now recommend use of this assay in women with 1 to 3 positive nodes. It was already recommended for chemotherapy decision-making for women with node-negative disease.

For postmenopausal women with a recurrence score of less than 25, chemotherapy can be omitted. I think it’s the premenopausal subgroup where people have a lot more questions. We see the benefit of chemotherapy when it’s added to endocrine therapy from the SWOG-1007 trial. It is not known whether that benefit is totally or partially driven by the ovarian function suppression effect of chemotherapy, and whether the same benefit would have been observed if all patients received ovarian function suppression in this trial. I think this is a question that will have to be addressed in another prospective randomized study. We do know from the SOFT and TEXT trials that the absolute benefit of the addition of ovarian function suppression to endocrine therapy was much larger in women who received chemotherapy, from 5% all the way up to 15%, based on the risk. It was much lower in women who did not receive chemotherapy, so who had clinically low-risk disease. That benefit from the addition of ovarian function suppression in SOFT was about 1% to 2% in women who did not receive chemotherapy.

In our clinical practice, we are using both treatments, not necessarily using one instead of another, in our highest-risk patients. I think there are positive data for premenopausal women with lymph node-positive disease, looking at efficacy of ovarian function suppression in isolation, without chemotherapy. SOFT included only about 10% of patients who had lymph node-positive disease who did not receive chemotherapy. The majority of patients in SOFT who had lymph node-positive disease ended up receiving chemotherapy as part of their standard management, and similar treatment was noted in TEXT. I think this is a question that all of us would like to have an answer to, and it will require another prospective trial. That goes back to the NCCN guidelines, where for the premenopausal subgroup, for those with node-positive disease, 1 to 3 positive-nodes, and a recurrence score of less than 25, the guidance is a little more, I would say, “wait.” So, you can do chemotherapy or ovarian function suppression; we don’t yet know how much of the benefit of chemotherapy in SWOG-1007 was derived from ovarian function suppression. I don’t think that trial will provide us that answer.

We know the Oncotype DX recurrence score is prognostic. We see that from TAILORx. We also see it from SWOG-1007, that the higher the score, the higher the risk. We also see from 1007 that in premenopausal women, those randomized to chemotherapy plus endocrine therapy did better than those who weren’t randomized to chemotherapy and just received endocrine therapy. What 1007 showed was the benefit of chemotherapy was not tied to the actual recurrence score value. Higher value did not translate into more benefit, and vice versa. The clinical utility of the assay is still totally there, and I do not think the SWOG-1007 results question that aspect.

The wonderful thing about medicine is that you see changes that are continuously happening. What is the current consensus about the use of adjuvant chemotherapy in lymph node-positive disease? This consensus is changing as we speak. The most recent NCCN guidelines remarkably, effectively, and efficiently already incorporate the RxPONDER results. They recommend that for many women who have ER-positive, 1 to 3 lymph node-positive, invasive breast cancer, the recurrence score could be used or should be used to help guide the value of adjuvant chemotherapy.

Experts in breast cancer comment on NCCN guideline recommendations as they reflect the outcomes of the RxPONDER trial for HR+ early breast cancer.

NCCN Guideline Updates for HR+ Operable Breast Cancer

More recently, we presented the results of the RxPONDER trial. The study was designed as a complementary trial to the TAILORx. Remember, TAILORx asked the question about the value of adjuvant chemotherapy in lymph node-negative, ER-positive patients with a recurrence score of 11 to 25. TAILORx asked a similar question for women with 1 to 3 positive lymph nodes, classified as stage II breast cancer. The study also asked an additional question related to whether the recurrence score itself has a chemotherapy sensitivity prediction value. That is, as the recurrence score increases, the relative benefit from chemotherapy would increase. It’s important to understand the difference between absolute and relative benefit from a particular treatment. The absolute benefit is the percentage improvement in patients who avoid a particular event. The percentage improvement increases as the risk increases. So, if you have a 5% risk of recurrence, you cannot possibly have a 10% absolute improvement in recurrence because the absolute risk is already smaller than that: it’s just 5%. The relative risk shows the relative improvement if a marker interacts with the treatment in a statistical sense. That would imply that as the marker increases, its relative benefit increases. So, instead of having a relative 5% improvement in the outcome, you get 10%, 20%, 30% improvement as the marker increases. That was an important goal of TAILORx, to test this relationship between chemotherapy sensitivity and the recurrence score. Of course, it also tested the question of whether chemotherapy improves the survival—the invasive disease-free survival or distant recurrence-free survival—of women with 1 to 3 positive nodes, with a recurrence score of less than 25. We included patients with recurrence scores of less than 11 to test the relationship between recurrence score value and chemotherapy sensitivity.

: One of the things in the trial in terms of the standard endocrine therapy the patients received that perhaps may not completely parallel or match what we do in practice today, is the use of ovarian function suppression. A small fraction of women received ovarian function suppression in this study, as reported at 6-month follow-up. Those numbers might change as the trial reports on use of ovarian function suppression at 12 or 18 months, especially in women who have received third-generation chemotherapy. Only about 15% or so of women received ovarian function suppression in the endocrine therapy arm and this number was less than 5% in the chemotherapy arm. Based on the SOFT and TEXT trials, our use of ovarian function suppression in women with high-risk disease, especially the ones who receive chemotherapy in clinical practice, has increased compared to what it was when the trial enrolled patients in SWOG-1007.

The data was released at the San Antonio Breast Cancer Symposium last year, in 2020. The result was released prematurely since it’s not the final analysis. It was the third interim analysis. The data safety monitoring committee recommended that we publish the results. The role of the data safety monitoring committee is to determine, at predetermined points, whether a study continues to be safe. It also has resulted in a practice-changing result that the public should be aware of. The data safety monitoring committee at the third interim analysis, which was preplanned by the trial statistics, realized that there is a very significant interaction between chemotherapy benefit and menopausal status. Premenopausal women had a very significant improvement in their invasive disease-free and distant recurrence-free survival from chemotherapy, across the entire range of recurrence score values. The two-thirds of the patients in the trial who were postmenopausal didn’t benefit at all. In fact, there was a slight hint that the chemotherapy may be detrimental. The data safety monitoring committee felt that the trial actually answered its question, and the information could potentially save some younger women’s lives. It could also avoid, for many older women, the unnecessary burden and toxicity of chemotherapy. That was the reason the trial results were announced.

The practice-changing results and implications from the RxPONDER trial for early-stage HR+ breast cancer is discussed.

HR+ Breast Cancer: Practical Implications of RxPONDER

: TAILORx was focused on patients with lymph node-negative disease. We’ve had data from a previous study, SWOG-8814, again, retrospective data, that suggested predictive value of a recurrence score for chemotherapy benefit in postmenopausal women with lymph node-positive disease. That was the premise for the design for SWOG-1007, also known as the RxPONDER trial. That trial aimed to evaluate the value of the Oncotype DX recurrence score in patients with lymph node-positive, hormone receptor-positive, HER2-negative breast cancer and the benefit of chemotherapy in these patients. The primary objective of SWOG-1007 was to determine the effect of chemotherapy on invasive disease-free survival [IDFS] in patients with 1 to 3 positive lymph nodes, so those with N1 disease and a recurrence score of less than 25. Another objective was to assess whether the effect of chemotherapy was dependent on the actual recurrence score value. The SWOG trial randomized patients with a recurrence score of 0 to 25 and included both pre and postmenopausal women. This randomization was stratified based on the actual recurrence score, whether it was 0 to 13 or 14 to 25, menopausal status, and the type of axillary surgery. Patients received either endocrine therapy of physicians’ choice, or chemotherapy followed by endocrine therapy.

The primary analysis for the trial was to look at the interaction of chemotherapy and continuous recurrence score for invasive disease-free survival. If this interaction was thought to be significant, it would show that the magnitude of benefit of chemotherapy was dependent on the recurrence score. If this interaction was not significant, then the trial would aim to look at whether recurrence score and chemotherapy are independently prognostic or whether chemotherapy improves IDFS in patients with a recurrence score of 0 to 25. The trial enrolled about 5000 patients and included both pre and postmenopausal patients, like I mentioned. About two-thirds of the patients were postmenopausal and one-third were premenopausal. Almost 90% of patients had 1 to 2 nodes, so a small percentage, 10%, had 3 positive nodes. About 85% to 90% of patients had low- or intermediate-grade disease. In terms of axillary surgery, about two-thirds had a full ALND [axillary lymph node dissection], and the remaining one-third, a little over a third actually, had only sentinel lymph node biopsies. The surgery reflects some of the shifts in how the axilla is managed in patients with limited nodal disease.

What the trial showed is that in the entire trial population, there wasn’t a difference in invasive disease-free survival between chemotherapy plus endocrine therapy, or endocrine therapy alone. Also, there was no interaction between chemotherapy benefit and the actual recurrence score. The chemotherapy benefit was not dependent on recurrence score. In other words, it was not any larger in higher recurrence score categories or lower in lower recurrence score categories. What it did note was that there was a significant interaction between chemotherapy benefit and menopausal status. When you look at the pre and postmenopausal women separately, the 5-year IDFS in postmenopausal women was 92% in both the endocrine therapy arm and the chemotherapy plus endocrine therapy arm. Whereas in premenopausal women, there was significant improvement in IDFS in the chemotherapy plus endocrine therapy arm compared to endocrine therapy. The absolute difference was about 5%.

These results were reported at almost 60% of the total events expected in the trial, so more events are expected. In the presentation, the study team also provided a breakdown regarding the type of IDFS events. About 70%-plus of the events in premenopausal women were either distant recurrence or locoregional events, so the majority of the events were breast cancer related. An absolute difference in distant recurrence in premenopausal women was about 3% between chemotherapy and endocrine therapy versus endocrine therapy. When they looked at subgroups, the benefit of chemotherapy in premenopausal women was seen in all subgroups that were shown, regardless of the number of positive nodes, the actual recurrence score category, age, and grade. On the forest plot, one could see that perhaps the absolute benefit of chemotherapy in premenopausal women who were ages 50 and older was a little less than those who were less than 50. This may all go back to how menopausal status was defined, and perhaps some of those women who were over the age of 50 who were categorized as premenopausal were actually postmenopausal. Based on the postmenopausal group, we wouldn’t expect this subgroup to benefit.

Priyanka Sharma, MD, and Lajos Pusztai, MD, PhD, review recent results from the RxPONDER trial looking at the predictive value of the Oncotype DX RS on benefit of adjuvant chemotherapy in women with lymph node-positive, HR+, HER2- breast cancer.

RxPONDER Trial in HR+ Operable Breast Cancer

: Oncotype DX is one of the multigene assays that is utilized for this decision-making. Earlier data for the Oncotype DX recurrence score came from prospective retrospective trials. The clinical utility was really established by the TAILORx study, which demonstrated that for patients with lymph node-negative, hormone receptor-positive, HER2-negative breast cancer, those who had a recurrence score between 11 and 25 and were taking chemotherapy as well as endocrine therapy, versus endocrine therapy alone, did not show improvement in all comers. In subgroup analysis, it was shown that in patients younger than 50, there might be some benefits of chemotherapy, especially for those who had a recurrence score greater than 16 up to 25. That’s the landscape regarding the utility of Oncotype and why it’s widely used in decision-making for patients with hormone receptor-positive, HER2-negative breast cancer, especially those with lymph node-negative breast cancer.

The TAILORx trial was conducted between 2006 and 2010, and asked an important clinical question at that time: do patients with ER-positive breast cancer, early stage disease, lymph node-negative, who had a recurrence score between 11 and 25, benefit from the addition of chemotherapy to their adjuvant endocrine treatment? By this time, there was a fairly good consensus that a recurrence score of less than 11 predicts a very good prognosis. Those patients could be safely spared from chemotherapy. At the other end of the spectrum, patients with a recurrence score greater than 25 had a high risk of recurrence. Multiple, smaller, more randomized studies and retrospective analysis of randomized trials showed that they do benefit significantly from adjuvant chemotherapy. However, a very large proportion of ER-positive patients fall into the middle, in the recurrence score category of between 11 and 25, and their benefit from chemotherapy was not clear, or unknown at that time.

TAILORx randomized 6000 patients to chemotherapy plus endocrine therapy or endocrine therapy alone, and the results were announced in 2018. It showed resoundingly that overall, there is no benefit from adjuvant chemotherapy in this recurrence score group. However, there was a very strong and significant interaction with age. Patients who were 50 or younger had a significant benefit from adjuvant chemotherapy. This benefit was seen across all recurrence score categories above 16. Below 16, between 11 and 15, the prognosis of the patients was outstanding. The benefits from chemotherapy were minuscule, nonsignificant, and clinically irrelevant. This study changed practice; it established the value of adjuvant chemotherapy for younger women with lymph node-negative disease, but the recurrence score had to be between 16 and 25. It also showed that as the recurrence score increases, the prognosis gets worse, and therefore, the absolute benefit from the chemotherapy increases. In other words, a woman with a recurrence score of 16 gained a little bit of benefit from chemotherapy. Women with a recurrence score of 25 gained much more benefit, absolute benefit, from adjuvant chemotherapy.

Based on this result, recurrence score gained quite a broad acceptance. I think today, and for the past several years, the recurrence score, or a similar molecular assay, is broadly used in most patients who have lymph node-negative disease. It also started to be used, although with less data, in patients with 1 to 3 positive lymph nodes.

This is standard of care today, to run a recurrence score or a similar molecular assay for most women who have lymph node-negative disease, and have a tumor size of between 1 and 5 cm.

It’s a very important question, how risk is defined and how the clinical risk factors complement the molecular risk factors, such as the recurrence score. It’s important to understand that prognostic risk is determined by multiple individual variables that are oftentimes unrelated to each other. A set of them relates to the anatomical and pathological features of the tumor, such as the tumor size, the number of lymph nodes involved, the histological grade, age, and to some extent, the lymphovascular invasion. These are called clinical risk variables, and the best way to determine the clinical risk is when we integrate these. 

There are multivariate models that perform better than just physicians’ judgment because they are more objective and integrate these various variables in a mathematically correct way. One that has been broadly used is the PREDICT assay, a free web tool from the United Kingdom. There are other methods as well that combine tumor size and lymph node status to make a prediction about anatomical risk. Molecular assays complement this. The most accurate risk prediction is relying on a combination of both the recurrence score or a similar genomic essay, and the anatomical variables, such as tumor size and nodal status. Some of the assays, such as the Prosigna assay, incorporate tumor size to make a prediction from the beginning. The recurrence score does not take into account any anatomical variables, which are the molecular features of the tumor. However, the recurrence score can be integrated with tumor size to arrive at a more precise risk estimate for recurrence and estimation of the absolute benefit that a woman would get from chemotherapy. That’s called RSClin. This is a free tool that’s available at the website of Exact Sciences Corporation, the company that produces the recurrence score assay.

Experts in breast cancer provide a historic perspective on the use of the Oncotype DX recurrence score (RS) in patients with HR+, HER- breast cancer and review practical implications from the TAILORx trial.

Lajos Pusztai, MD, PhD, Yale Cancer Center

Articles

HR+ Breast Cancer: Role of RxPONDER Data in Clinical Practice

HR+ Breast Cancer: Impact of RxPONDER on Clinical Practice

NCCN Guideline Updates for HR+ Operable Breast Cancer

HR+ Breast Cancer: Practical Implications of RxPONDER

RxPONDER Trial in HR+ Operable Breast Cancer

HR+ Breast Cancer: Oncotype DX RS and TAILORx