This article is more than 5 years old. Information may no longer be current.
TAILORx ‘shows path’ to de-escalate therapy for some women, raises treatment questions for others
Click Here to Manage Email Alerts
Editor’s Note: In this guest commentary, Douglas Yee, MD, discusses results from and implications of the TAILORx trial, presented at this year’s ASCO Annual Meeting.
Adjuvant therapy of breast cancer has unquestionably resulted in improved survival for women with operable breast cancer.
NSABP B-20 — published in 2004 in The Lancet — showed treatment of women with ER-expressing tumors benefited from both chemotherapy and tamoxifen. RFS improved when both therapies were administered compared with tamoxifen alone.
Still, most oncologists knew only a small number of women benefited from both therapies; for many women, tamoxifen alone was sufficient.
However, a method to distinguish endocrine therapy-sensitive tumors from chemotherapy-sensitive tumors was not available. Thus, we erred on the side of overtreatment and we gave both therapies.
When molecular-biological techniques were developed to measure gene expression in formalin-fixed paraffin-embedded tumors, the 21-gene recurrence score assay (Oncotype DX, Genomic Health) was born.
Researchers used this test — initially developed to determine prognosis by measuring 18 genes and three reference genes — on archived specimens from randomized clinical trials testing tamoxifen in the adjuvant setting, including the NSABP B-20 trial. When applied to these specimens, researchers defined a range of low-, intermediate- and high-risk scores to show benefit from adjuvant tamoxifen.
Specimens with low-risk scores clearly derived little benefit from chemotherapy and those with high-risk scores derived little benefit from tamoxifen; however, benefit was uncertain for the intermediate-risk scores, defined as 18 to 30.
Study results
TAILORx was created to provide prospective validation of the Oncotype DX score.
Researchers determined scores for women with ER-positive, node-negative tumors from 1 cm to 5 cm.
To be cautious, the ranges of low, high and intermediate risk were shifted to a more conservative range, and patients with low or high scores were not randomly assigned to treatment. Those with low-risk scores received endocrine therapy, and those with high-risk scores received chemotherapy followed by endocrine therapy.
The results of the low-risk arm were published in 2015 in The New England Journal of Medicine, showing excellent outcomes with hormone therapy alone.
Sparano and colleagues presented the results of the remaining arms of the trial at this year’s ASCO Annual Meeting, with simultaneous publication in The New England Journal of Medicine.
Researchers randomly assigned patients with intermediate-risk tumors — defined as scores of 11 to 25 — to hormone therapy alone or chemotherapy followed by hormone therapy.
After median follow-up of 7.5 years, chemotherapy offered no benefit to patients with intermediate-risk scores.
The trial — designed as a noninferiority trial — showed endocrine therapy was noninferior to chemotherapy plus endocrine therapy for any of the outcomes measured: invasive EFS, distant RFS, relapse-free interval or OS.
This was good news for a majority of women with node-negative, ER-positive breast cancer; however, there are some important caveats.
In a subset analysis, women aged younger than 50 years did not appear to enjoy the same benefits from endocrine therapy alone. The value of chemotherapy in this younger group of women was most striking for scores greater than 16. In this subgroup, women with higher scores appeared to benefit from chemotherapy.
Further, women with scores above 25 had poor outcomes compared with all other women in the study, despite receiving chemotherapy. These findings suggest new strategies for these high-risk patients are needed.
Specific populations
Overall, these results show the low-risk group can be defined as a score up to 25, especially among women aged older than 50 years.
These results are like those reported in 2016 in The New England Journal of Medicine on the MINDACT study. In this study, Cardoso and colleagues used the 70-gene assay (MammaPrint, Agendia) to determine low or high risk. There is no intermediate range using the MammaPrint assay.
MINDACT also used clinical risk as a prognostic indicator. Results showed women with low-risk tumors — determined by either genomic or clinical risk — had no benefit from chemotherapy.
Researchers also measured clinical risk in TAILORx. A little more than 26% were considered clinical high risk. In this subgroup, women with low recurrence scores — less than 25 — also derived no benefit from chemotherapy.
Thus, data from TAILORx show us how to safely de-escalate therapy among women with node-negative, ER-positive breast cancers.
Although data from the RxPONDER trial — using Oncotype DX to randomly assign node-positive patients to endocrine therapy or chemotherapy plus endocrine therapy — are not yet reported, it is notable MINDACT included patients with lymph node involvement and showed there was no benefit from chemotherapy for these clinical high-risk patients with low genomic risk scores.
Future paths forward
TAILORx has important implications for the future.
First, it is possible the low-risk range might be different for specific subgroups of women — most notably younger and presumably premenopausal women, almost one-third of the study population.
Supplementary material provided in the publication suggests some of these premenopausal women received inappropriate therapy — aromatase inhibitor alone, tamoxifen plus an aromatase inhibitor, or ovarian functional suppression alone. Only about 10% of women received ovarian functional suppression plus an aromatase inhibitor in the endocrine therapy-alone arm.
Because chemotherapy frequently suppresses ovarian function in this age group, perhaps the benefits of the chemotherapy are due to chemotherapy-induced ovarian suppression in the group of women who received chemotherapy. The rate of ovarian functional suppression in the combined arm is not reported, but recent studies — such as data from Francis and colleagues, published this year in The New England Journal of Medicine — have shown ovarian functional suppression with either tamoxifen or an aromatase inhibitor for premenopausal women is superior to tamoxifen alone.
Second, it is clear we need improved therapies for high-risk women. This group of women who were assigned chemotherapy plus endocrine therapy had an inferior OS rate at both 5 years and 9 years.
It must be remembered that the Oncotype DX score was initially designed to predict prognosis, and it seems to do that for high-risk patients; these women had poor outcomes despite receiving state-of-the-art treatment. New strategies, such as cyclin-dependent kinase inhibitors or improved cytotoxic therapies, are needed for this group of patients.
The genomic risk predictors — both Oncotype DX and MammaPrint — have shown us a path to safely de-escalate therapy for women with node-negative, ER-positive breast cancer.
Despite some caveats regarding younger women, the results of TAILORx will allow us to assure women that those with Oncotype DX scores up to 25 can safely avoid chemotherapy.
References:
Cardoso F, et al. N Engl J Med. 2016;doi:10.1056/NEJMoa1602253.
Cobleigh MA, et al. Clin Cancer Res. 2005;doi:10.1158/1078-0432.CCR-05-0735
Fisher B, et al. Lancet. 2004;doi:10.1016/S0140-6736(04)16981-X.
Francis PA, et al. N Engl J Med. 2018;doi:10.1056/NEJMoa1803164.
Paik S, et al. J Clin Oncol. 2006;doi:10.1200/JCO.2005.04.7985.
Sparano J, et al. Abstract LBA1. Presented at: ASCO Annual Meeting; June 1-5, 2018; Chicago.
Sparano JA, et al. N Engl J Med. 2015;doi:10.1056/NEJMoa1510764.
Sparano JA, et al. N Engl J Med. 2018;doi:10.1056/NEJMoa1804710.
For more information:
Douglas Yee, MD, is professor of medicine and pharmacology in the division of hematology, oncology and transplantation at University of Minnesota, as well as director of Masonic Cancer Center. He also is a HemOnc Today Editorial Board Member. He can be reached at yeexx006@umn.edu.
Disclosure: Yee reports no relevant financial disclosures.