The New England Journal of Medicine July 12, 2018

Adjuvant Chemotherapy Guided by a 21-Gene Expression Assay in Breast Cancer (TAILORx)

Sparano JA, Gray RJ, Makower DF, et al.

Bottom Line

In women with HR-positive, HER2-negative, axillary node-negative early breast cancer and an intermediate 21-gene recurrence score (11 to 25), endocrine therapy alone was noninferior to chemoendocrine therapy for invasive disease-free survival.

Key Findings

1. Endocrine therapy alone was noninferior to chemoendocrine therapy for invasive disease-free survival in patients with a midrange recurrence score of 11 to 25 (Hazard Ratio 1.08; 95% CI, 0.94 to 1.24; P=0.26) [10.1.8].

2. At 9 years, the invasive disease-free survival rate was 83.3% in the endocrine-therapy group compared to 84.3% in the chemoendocrine-therapy group.

3. Freedom from distant recurrence at 9 years was similarly high in both groups: 94.5% with endocrine therapy alone versus 95.0% with chemoendocrine therapy.

4. Overall survival at 9 years was almost identical between the two arms (93.9% for endocrine therapy alone vs. 93.8% for chemoendocrine therapy).

5. An exploratory subgroup analysis found that women 50 years of age or younger with a recurrence score of 16 to 25 derived a benefit from chemotherapy, with a lower rate of distant recurrence.

Study Design

Design

RCT

Open-Label

Sample

10,273

Patients

Duration

7.5 yr

Median

Setting

Multicenter, International

Population Women aged 18-75 with HR-positive, HER2-negative, axillary node-negative early breast cancer, with tumor size 1.1-5.0 cm (or 0.6-1.0 cm and intermediate/high grade), and an intermediate 21-gene recurrence score of 11 to 25.

Intervention Endocrine therapy alone

Comparator Chemoendocrine therapy (adjuvant chemotherapy plus endocrine therapy)

Outcome Invasive disease-free survival (IDFS)

Study Limitations

• The trial allowed for physician choice of the specific endocrine and chemotherapy regimens, which introduces potential heterogeneity and unmeasured confounding.

• In the premenopausal subgroup analysis (age 50 or younger), it is difficult to determine whether the chemotherapy benefit observed for scores 16-25 was due to a direct cytotoxic effect or simply secondary to chemotherapy-induced early menopause [7.1.8].

• Late recurrences are common in HR-positive breast cancer; thus, follow-up beyond 9 years is necessary to definitively assess late recurrence outcomes.

Clinical Significance

TAILORx was a landmark, practice-changing study demonstrating that the vast majority of women (up to 70%) with HR-positive, HER2-negative, node-negative early breast cancer can safely avoid the toxicity of adjuvant chemotherapy. It provided level 1 evidence cementing the 21-gene recurrence score assay as a critical standard-of-care tool to guide precision de-escalation of breast cancer treatment.

Historical Context

Prior to TAILORx, retrospective analyses of trials like NSABP B-14 and B-20 showed that patients with an Oncotype DX score <11 did well with endocrine therapy alone, and those with a score >25 benefited significantly from chemotherapy. However, management for patients in the intermediate range (scores 11-25) remained a clinical dilemma. TAILORx was designed prospectively to resolve this uncertainty and represents the largest randomized breast cancer treatment trial ever conducted.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

What is the biological rationale for using a 21-gene expression assay (Oncotype DX) to guide chemotherapy decisions in HR-positive, HER2-negative breast cancer, and how do endocrine therapy and chemotherapy target these tumors differently?

Key Response

This question tests foundational knowledge of tumor biology and pharmacology. The 21-gene assay quantifies the expression of genes related to estrogen signaling, cellular proliferation (e.g., Ki-67), and invasion. Endocrine therapy specifically targets the hormone-receptor dependency of the tumor (e.g., via estrogen receptor blockade or estrogen deprivation), whereas chemotherapy broadly targets rapidly dividing cells. The assay helps determine if a tumor is driven primarily by estrogen (responsive to endocrine therapy) or high proliferation (responsive to chemotherapy).

Resident

Based on the TAILORx findings, how would your management algorithm differ for a 60-year-old postmenopausal woman versus a 45-year-old premenopausal woman, both presenting with HR-positive, HER2-negative, node-negative breast cancer and an Oncotype DX recurrence score of 20?

Key Response

This assesses clinical application of a crucial subgroup analysis in TAILORx. For women older than 50 with a recurrence score of 11 to 25, chemotherapy can be safely omitted as endocrine therapy alone is noninferior. However, in women 50 years of age or younger, there was an observed chemotherapy benefit for invasive disease-free survival in those with scores of 16 to 25. Therefore, the 60-year-old would receive endocrine therapy alone, while chemotherapy should be strongly considered for the 45-year-old.

Fellow

In the TAILORx subgroup of women aged 50 or younger with recurrence scores of 16-25, there was a benefit to adding chemotherapy. Is this benefit likely due to the direct cytotoxic effects of the chemotherapy, or is it mediated by chemotherapy-induced ovarian suppression (CIOS)? How does this debate influence your decision to recommend chemotherapy versus ovarian suppression plus an aromatase inhibitor?

Key Response

This explores a highly nuanced, ongoing debate in breast oncology. Retrospective analyses suggest the chemotherapy benefit in young women with intermediate scores may actually be an endocrine effect driven by CIOS causing amenorrhea, rather than a cytotoxic effect. A fellow must integrate this concept with data from the SOFT/TEXT trials, weighing whether aggressive endocrine therapy (ovarian suppression + AI) could spare these patients from systemic chemotherapy toxicity while achieving similar outcomes.

Attending

When counseling a highly anxious patient with a recurrence score of 24 (the high end of the intermediate range), how do you navigate the shared decision-making process to omit chemotherapy, balancing the population-level non-inferiority data of TAILORx with the patient's individual fear of recurrence?

Key Response

This question focuses on the art of clinical practice and risk communication. While TAILORx provides robust population-level evidence for non-inferiority in the 11-25 range, a score of 24 approaches the high-risk threshold where chemotherapy is clearly beneficial. The attending must demonstrate how to translate complex statistical boundaries into patient-centered care, validating patient anxiety while confidently explaining continuous risk variables and the diminishing marginal returns of cytotoxic therapy.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The TAILORx trial utilized a non-inferiority design for the intermediate-risk cohort (score 11-25). Critically evaluate the chosen non-inferiority margin and the event-driven analysis approach. How might unmeasured biological heterogeneity, such as varying proportions of Luminal A versus Luminal B subtypes within this intermediate score band, affect the statistical power to detect a true difference in the high-intermediate end?

Key Response

This addresses advanced research methodology. Evaluating non-inferiority margins requires understanding whether the acceptable margin of difference is clinically meaningful versus purely statistical. Furthermore, because the 21-gene score is a continuous variable categorized into bands, biological heterogeneity exists within the 11-25 group. A PhD-level critique examines whether the trial was adequately powered to detect non-proportional hazards or specific subgroup interactions at the upper limits of the intermediate range where Luminal B tumors may cluster.

Journal Editor

As a peer reviewer evaluating the TAILORx manuscript, how would you assess the threat to validity posed by varying adherence and early discontinuation rates of assigned endocrine therapies over the 9-year follow-up? Does differential adherence bias a non-inferiority outcome towards or against the null?

Key Response

This tests critical appraisal skills regarding trial validity. In non-inferiority trials, poor adherence to the study intervention (in this case, either endocrine therapy alone or chemoendocrine therapy) biases the results toward the null (i.e., making the treatments look more similar than they are), which artificially inflates the likelihood of declaring non-inferiority. A rigorous reviewer would flag real-world adherence to endocrine therapy as a major methodological factor requiring robust sensitivity analyses (e.g., per-protocol versus intention-to-treat).

Guideline Committee

Based on the TAILORx trial, what Level of Evidence and Grade of Recommendation should ASCO and NCCN assign to omitting chemotherapy in postmenopausal women with HR+/HER2- node-negative disease and Oncotype scores of 11-25? Furthermore, how should guidelines synthesize this finding with the subsequent RxPONDER trial data to create a unified testing algorithm for early-stage breast cancer?

Key Response

This question targets guideline development and evidence synthesis. TAILORx provides Level I evidence (large, randomized, phase III trial) supporting a strong recommendation to omit chemotherapy in postmenopausal, node-negative patients with scores 11-25. The committee must map this into practice guidelines alongside RxPONDER (which showed postmenopausal women with 1-3 positive nodes and scores 0-25 can also omit chemotherapy). Synthesizing these creates a cohesive clinical pathway based on menopausal status, nodal status, and recurrence score.

Clinical Landscape

Noteworthy Related Trials

2006

NSABP B-20 Genomic Analysis

n = 651 · JCO

Tested

Chemotherapy plus tamoxifen

Population

Women with node-negative, ER-positive breast cancer

Comparator

Tamoxifen alone

Endpoint

10-year distant recurrence-free survival

Key result: Patients with a high Recurrence Score derived substantial benefit from chemotherapy, while those with a low score derived minimal to no benefit.

2016

MINDACT Trial

n = 6,693 · NEJM

Tested

Chemotherapy decision guided by 70-gene signature (MammaPrint)

Population

Women with early-stage breast cancer at high clinical risk but low genomic risk

Comparator

Chemotherapy decision guided by clinical-pathological criteria

Endpoint

5-year distant metastasis-free survival

Key result: Among women with high clinical risk and low genomic risk, the 5-year survival without distant metastasis was 94.7% without chemotherapy, indicating it could be safely omitted.

2021

RxPONDER Trial

n = 5,015 · NEJM

Tested

Endocrine therapy alone

Population

Women with HR-positive, HER2-negative breast cancer with 1-3 positive nodes and Recurrence Score of 25 or lower

Comparator

Chemoendocrine therapy

Endpoint

Invasive disease-free survival

Key result: Postmenopausal women with 1-3 positive nodes and a recurrence score of 25 or lower did not benefit from adding chemotherapy to endocrine therapy, whereas premenopausal women experienced a significant benefit.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis