June 4, 2019 -- Adding clinical risk factors such as tumor size and histologic grade to Oncotype DX scores for predicting cancer recurrence could help guide the treatment of women with early breast cancer, particularly younger women, according to results presented at the American Society of Clinical Oncology (ASCO) meeting and published in the New England Journal of Medicine on June 3.
The update to the prospective, long-term Trial Assigning Individualized Options for Treatment (TAILORx) of almost 10,000 newly diagnosed breast cancer patients stresses the importance of assessing clinical risk, based on tumor size and histologic grade, in addition to results from the 21-gene Oncotype DX assay (Genomic Health) in women with early-stage, hormone receptor-positive/human epidermal growth factor receptor 2-negative, node-negative breast cancer.
The new analysis provides additional insight into how to manage patients age 50 or younger using recurrence risk predictive scores to show which breast cancer patients may be treated with endocrine therapy alone versus those who also need chemotherapy.
In this age group, the rate of recurrence at nine years was less than 5% regardless of clinical risk for those who had a low recurrence score (defined as 0-10) and were treated with endocrine therapy alone. For those with low clinical risk and an intermediate score (defined as 11-25), the recurrence rate was 4.7%. But for those with an intermediate Oncotype DX score and high clinical risk, the recurrence rate was 10%.
"The integration of genomic and clinical information may provide a more accurate estimation of prognosis for individual patients than could be provided by either the genomic or clinical information alone," reported lead investigator Dr. Joseph Sparano, associate chair of clinical research at Montefiore Medical Center, and colleagues.
Building a more complete, but complex, picture of TAILORx
The new analysis was based on a prespecified secondary end point in TAILORx, a prospective, noninferiority study that evaluated distant recurrence rates in patients treated with endocrine therapy alone (mostly tamoxifen) or endocrine therapy with add-on adjuvant chemotherapy. The study was sponsored by the U.S. National Cancer Institute and designed and led by the Eastern Cooperative Oncology Group and the American College of Radiology Imaging Network (ECOG-ACRIN). Results were published about a year ago in the New England Journal of Medicine (July 12, 2018, Vol. 379:2, pp. 111-121).
Oncotype DX scores are based on the expression of 21 genes -- 16 are cancer-related and five are reference genes -- and the risk of recurrence is rated on a scale of 1 to 100. Testing has been incorporated into professional treatment guidelines.
In TAILORx, patients with intermediate recurrence scores were randomized to treatment with endocrine therapy or endocrine therapy plus chemotherapy. The study's central conclusion in July 2018 was that patients with a score of 11 to 25 had a similar rate of distant recurrences at nine years of 5%, regardless of the treatment group.
The finding promised to spare newly diagnosed patients from the effects of chemotherapy. However, an exploratory analysis at the time found that some women who were 50 years or younger and who had recurrence scores between 16 and 25 did benefit from the addition of chemotherapy, which complicated interpretation of the data.
Sparano and colleagues' latest analysis assessed how clinical pathological features might be used to further evaluate the risk of recurrence on top of the Oncotype DX assay. Clinical risk was evaluated using an algorithm deployed in the Microarray in Node-Negative Disease May Avoid Chemotherapy (MINDACT) trial. Low clinical risk was defined based on one of the following three criteria and high risk on the lack of these characteristics:
"Classic pathologic information and outcome results were also used to refine models based on classic information and genomic tests," Sparano and colleagues noted.
For the 6,496 women with intermediate risk scores based on the recurrence assay, the clinical risk score provided significant prognostic information: Women with high risk scores had a hazard ratio of 2.42 for recurrence, compared with those with low risk scores (p < 0.001), the researchers reported. But clinical risk alone was not predictive of risk for recurrence.
The researchers also looked at age and menopausal status in women on the upper end of the recurrence risk score intermediate range, i.e., from 16 to 25.
"We found that a chemotherapy benefit was most evident at 45 years of age with a recurrence score of 16 to 25 in premenopausal women and waned at younger and older ages, and with menopause, consistent with an effect due to chemotherapy-induced premature menopause," Sparano et al wrote.
The data could also be used to precisely target subgroups of younger women who could benefit from additional therapy beyond tamoxifen, the researchers wrote. It's possible that younger women with scores between 16 and 25 could benefit from the addition of ovarian function suppression and an aromatase inhibitor, they added.
"Although the potential pitfalls of a subgroup analysis to identify more effective therapies in trials with a superiority design have been well described and the exploratory analyses presented here were not adjusted for multiple comparisons, caution is warranted when withdrawing potentially lifesaving therapy on the basis of a noninferiority trial such as TAILORx, especially when the findings are biologically plausible and supported by population-level data, as described here," Sparano and colleagues wrote.
The latest findings indicate that the chemotherapy benefit that had been observed for women younger than 50 with risk scores in the 16 to 25 range "may be due to a castration effect associated with cytotoxic therapy," the ECOG-ACRIN research group explained in a June 3 statement. The July 2018 results had suggested that 30% of women -- including this subgroup -- could benefit from add-on chemotherapy, but the update suggests there would be other treatment options.
Dana-Farber specialist Dr. Harold Burstein, PhD, described the TAILORx presentation as "arguably [the] most important clinical data in breast cancer" and suggested the results were practice-changing in a tweet from the ASCO meeting.
TAILORX update from @jsparano Arguably most important clinical data in breast cancer. “Benefit” of chemo in women <age 50 with oncotype scores of 16-25 is almost certainly all due to ovarian suppression effects of chemotherapy.— Hal Burstein, MD (@DrHBurstein) June 3, 2019
"TAILORx continues to elevate Oncotype DX to a new global standard with increasing utilization and reimbursement," Genomic Health noted in its statement about the data.
When precision medicine gets messy
However, an accompanying editorial in NEJM suggested that the update shows the challenges of using test results from large randomized studies of diseases with heterogeneous populations and highlights the unfortunate lack of precision in precision medicine.
The hazard ratios for those with high clinical risk versus low risk were two to three times higher regardless of the treatment arm they were assigned, noted David Hunter, a professor of epidemiology at the University of Oxford, and Dr. Dan Longo, deputy editor of NEJM and a hematologist at Brigham and Women's Hospital.
The authors also highlighted the finding that for women age 50 or younger with an intermediate risk score from 11 to 25, the rate of disease recurrence at nine years was 12.3% for endocrine therapy alone versus 6.2% for adjuvant chemotherapy. Adding ovarian suppression and an aromatase inhibitor might provide a risk reduction similar to adding chemotherapy in these women, but this would mean following practice-changing suggestions "made on the basis of a subgroup analysis of a secondary objective in the trial, published as a follow-up to the report of the primary objective," wrote Hunter and Longo, who also questioned why the information wasn't part of the initial report.
"The promise of 'precision' medicine has collided with the rather messier world of using all available evidence to try and make educated guesses to improve patient outcomes," they wrote.
As a study of early-stage disease, the trial has a lot of value, but it's unclear if follow-up studies will be done given the large size, high cost, and long timelines required, they noted. And in the decade or more it would take to complete such studies, advice still needs to be given to women facing treatment decisions.
Would newer tests be better?
The authors also noted the limitations of the test itself, as it dates from early days of analysis of tumors more than 15 years ago; new panels have been developed, but the one used in the study has become established in clinical practice.
"It seems likely that a contemporary approach to breast cancer prognosis incorporating the much larger-scale gene expression, genomic copy-number alteration, and epigenetic data now available would lead to an at least incrementally superior panel," Hunter and Longo wrote.
Distinguishing results that need to be acted on will not be precise, they concluded.
"Medicine in the molecular era will be no more 'precise' than in prior eras -- evidence synthesis, clinical judgment, and patient preference all factor in," Hunter and Longo wrote.