Gene Expression Assays in Early-Stage Breast Cancer

Oncology (Williston Park). 32(10):510-2.

Joseph A. Sparano, MD

Results from the first clinical trial to incorporate a gene expression assay-which may potentially facilitate clinical decisions in women with early-stage breast cancer-were recently published in the New England Journal of Medicine in July.[1] The trial, known as the Trial Assigning Individualized Options for Treatment, or TAILORx, examined adjuvant therapy for early-stage, estrogen receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)-negative breast cancer.

Q: First, can you describe the design of the study, the treatment arms, and how this 21-gene expression assay was incorporated into the trial?

DR. SPARANO: TAILORx is the largest cancer treatment trial ever conducted. It was sponsored by the National Cancer Institute, and the primary purpose of the trial was to integrate the 21-gene assay called Oncotype DX into the clinical decision-making process or paradigm for patients with early-stage breast cancer, based on what we knew at the time about the prognostic and predictive information that the assay told us. We designed the trial to address the questions that remained unanswered at that time, including the potential benefit of adjuvant chemotherapy and how to manage patients with a midrange recurrence score on the 21-gene panel assay.

The target population included patients with ER-positive, HER2-negative, axillary node–negative breast cancer. Together, this represents about 50% of all breast cancers and 8% of all cancers in the United States. Adjuvant chemotherapy is typically recommended for women who have ER-positive breast cancer with a tumor of at least 1 mm, according to a prior position statement issued by the National Institutes of Health in 2000. This was based on prior evidence indicating that even patients who had a relatively low risk of recurrence could potentially benefit from adjuvant chemotherapy. What we embarked upon in the trial was to take patients who met established clinical criteria for a recommendation, or who at least should be considered for adjuvant chemotherapy in accordance with National Comprehensive Cancer Network guidelines, and to either assign or randomize their treatment based on their Oncotype DX recurrence score. Patients who had a recurrence score of 10 or less (17%), which is considered low or very low, were assigned to endocrine therapy alone and followed. Those with a recurrence score of 26 or higher (15% to 20%), which is considered high, were assigned to chemotherapy plus endocrine therapy. The remaining two-thirds of patients who had a recurrence score between 11 and 25 were assigned to either chemotherapy plus endocrine therapy (standard arm) or endocrine therapy alone (experimental arm). The trial was designed to determine whether endocrine therapy alone was inferior or noninferior to chemotherapy plus endocrine therapy.

Q: What are the key results of this trial, and was anything particularly surprising?

DR. SPARANO: The key results were as follows: in 2015, we reported in the New England Journal of Medicine that patients with a low recurrence score (defined in the study as between 0 and 10) had a 1% risk of recurrence at 5 years with endocrine therapy alone. Obviously, this indicates that these patients would be unlikely to benefit from chemotherapy, and this information was integrated into the 8th edition of the American Joint Committee on Cancer staging and prognostic staging beginning in January of this year. Updated results from this most recent analysis indicated that at 9 years, there was about a 3% distant recurrence rate. Again, this confirms that chemotherapy is unlikely to benefit this population.

The second major finding, which involved the primary endpoint of invasive disease–free survival, was that endocrine therapy was not inferior to chemotherapy in patients who had an Oncotype DX recurrent score of 11 to 25, with very similar clinical outcomes at 5 years and 9 years. At 9 years, the rate of invasive disease–free survival was 83.3% in the endocrine therapy arm and 84.3% in the chemotherapy/endocrine therapy arm, which was not significantly different. There were also similar rates of local, regional, and distant recurrence at 9 years, so the study did meet its primary endpoint.

One of the findings that was of some surprise was that there was an interaction between age and potential chemotherapy benefit. For example, a woman age 50 or younger who had a recurrence score between 21 and 25 did derive some benefit from chemotherapy. Specifically, there was about a 7% difference in distant recurrence rates at 9 years for patients who received chemotherapy, so some benefit was seen in this age group in the 21-to-25 score group. In the overall group, however, there was no benefit, and, certainly, there was no benefit in women over the age of 50 who had a recurrence score of between 11 and 25.

Q: Are these trial results particularly practice-changing?

DR. SPARANO: I think practice has already changed since the introduction of this assay to clinical practice in 2004. There has definitely been a decline in the use of chemotherapy, largely in patients who had a recurrence score in the lower range of this study. For those who had a recurrence score in the middle range, which was the group that was randomized in this study, there was still some variability in chemotherapy use, but chemotherapy use had declined. What this study did, however, was provide much more definitive information about potential benefit or lack of benefit in patients who had a midrange recurrence score.

Q: Will the results alter the way clinicians and their patients make decisions about which adjuvant therapy to use?

DR. SPARANO: I think at this point, clinicians can make decisions and treatment recommendations with a much higher level of evidence-really, an unprecedented amount of evidence-and we have a greater degree of precision than we’ve ever had.

One of the key differences in TAILORx vs previous trials was how recurrent score ranges were defined. For example, in the TAILORx study, we defined low as a score of 0 to 10 rather than 0 to 17; this is the group of patients who were assigned to endocrine therapy alone, and we did that because we wanted to be absolutely sure that we were not undertreating these patients. Patients who had a recurrence score of between 11 and 25 were defined as midrange, while the intermediate recurrence score range was originally defined as between 18 and 30. The reason we made this change was also to make sure that we were not undertreating patients who would or would not be randomized to chemotherapy/endocrine therapy.

The 10-year distant relapse–free survival rate of about 5% observed in the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-20 trial-the original trial in which patients were randomized to chemotherapy/endocrine therapy or endocrine therapy alone-was irrespective of chemotherapy use for patients with a recurrence score between 11 and 25 and turned out to be almost identical to what we observed in TAILORx. This supported our decisions regarding changing the recurrence score ranges that we used in TAILORx compared with the ranges that were initially defined as low, intermediate, or high.

Q: Are there any other details from the trial that you think are particularly important that may be overlooked from just reading the trial result publication? Can any results be interpreted in different ways by different clinicians?

DR. SPARANO: I think one of the most important and misunderstood aspects of the trial relates to what I described earlier. By this, I mean the recurrence score ranges we used in the TAILORx trial compared with the ranges that were initially reported in the NSABP B-14 validation study conducted in patients with ER-positive, node-negative disease treated with 5 years of tamoxifen, and in the NSABP B-20 trial, in which patients with ER-positive, node-negative disease were randomized to receive chemotherapy plus tamoxifen or tamoxifen alone. When we designed the trial back in 2004 and 2005, we actually re-analyzed the B-20 data using these different recurrence score cutoffs, defining a score of 11 to 25 as midrange, and we did that for several reasons. First, the original B-20 cohort included patients who had HER2-positive disease, about 12% of the population. Almost all of these patients have a high recurrence score. Since we were designing this trial exclusively for a HER2-negative population, we needed to adjust the recurrence score ranges to actually reflect the population to whom we were offering the trial. This was one of the reasons that we dialed down the recurrence score at the upper range of a high score from 31 or higher to 26 or higher. That was a key point. Plus, when we analyzed the B-20 data, we found a similar chemotherapy benefit for both of these ranges. A similar benefit from chemotherapy was also seen after excluding HER2-positive patients, regardless of whether the high score cutoff was 26 or higher or 31 or higher. And, finally, if a score of 26 or higher was used as the high score cutoff point rather than 31 or higher, more patients who could potentially benefit from chemotherapy will be captured, therefore minimizing the potential for chemotherapy undertreatment. I think that this is one of the most critical and misunderstood issues about the trial. In addition, I think the results of the trial will change the definitions of low-, intermediate-, and high-risk scores. Eventually, I think the recurrence score will become more or less a binary test used to identify a group of patients who have a high recurrence and will benefit from chemotherapy, as well as those with a low or intermediate score who won’t benefit from chemotherapy.

Q: Lastly, in the context of this trial, do you think the way the 21-gene expression assay is utilized in the clinic has evolved?

DR: SPARANO: Yes, absolutely. I think the next potential application is expanding the use of this assay to patients who have node-positive disease. This has already been tested in the RxPONDER trial, which enrolled patients with 1 to 3 positive axillary nodes and ER-positive, HER2-negative disease. Individuals with a recurrence score of 25 or lower were randomized to chemotherapy plus endocrine therapy, the standard arm, or endocrine therapy alone, the experimental arm. The trial has been completed and we are awaiting the results. However, some early evidence from the prospective Plan B trial is that women with a recurrence score of 11 or less with up to 3 involved axillary lymph nodes do well with endocrine therapy alone, suggesting that some women with small-
volume–disease involving one axillary node may be spared chemotherapy.

Other ongoing trials are integrating other gene expression assays into clinical decision making, and their results will also be important. Lastly, ongoing efforts to develop tools that integrate information from both the recurrence score-which captures both prognostic and predictive information-and clinical and pathologic features such as age, tumor size, and grade, may be able to provide more accurate and refined estimates for prognosis (risk of recurrence with endocrine therapy alone) and prediction (chemotherapy benefit when chemotherapy is added to endocrine therapy). I think that we will see these tools becoming available in the near future. One such tool, called the RSPC (Recurrence Score-Pathology-Clinical), is available right now. This is an online tool for providers available on the Genomic Health website, and it allows clinicians to plug in the recurrence score, as well as age, tumor size, and tumor grade, to help provide patients with more accurate estimates about their individual risk of recurrence. A revision of this tool is being developed now, and it may be able to provide predictive information regarding chemotherapy benefit.

Financial Disclosure: The author has no significant financial interest in or other relationship with the manufacturer of any product or provider of any service mentioned in this article.

References:

1. Sparano JA, Gray RJ, Makower DF, et al. Adjuvant chemotherapy guided by a 21-gene expression assay in breast cancer. N Engl J Med. 2018;379:111-21.