EORTC 10041/BIG 3-04 MINDACT: 70-Gene Signature (MammaPrint) in Early-Stage Breast Cancer

Primary endpoint (Survival without distant metastasis, 5-year rate): 94.7% (95% CI, 92.5 to 96.2) in the primary-test population — lower boundary of 95% CI exceeded the prespecified 92% noninferiority threshold Key secondary (Survival without distant metastasis, ITT — High Clinical Risk / Low Genomic Risk): 95.9% vs 94.4% — HR 0.78 (95% CI, 0.50 to 1.21; P=0.27) Safety signal: No safety or adverse event data were reported in this publication

📱 Clinic quick reference: Mobile version — save for point-of-care use

Clinical Bottom Line

MINDACT demonstrated that among women with early-stage breast cancer who are classified as high clinical risk but low genomic risk by the 70-gene signature (MammaPrint), forgoing adjuvant chemotherapy resulted in an excellent rate of survival without distant metastasis at five years. The trial met its primary objective: the lower boundary of the confidence interval exceeded the prespecified noninferiority threshold, confirming that the genomic test can reliably identify patients within the clinically high-risk population who derive minimal benefit from chemotherapy and can safely be spared its toxicity.

This trial fundamentally changed how clinicians approach the chemotherapy decision in early breast cancer. Before MINDACT, patients classified as high clinical risk — based on tumor size, grade, nodal status, and other traditional factors — were broadly recommended adjuvant chemotherapy. The 70-gene signature now provides a second, independent layer of risk assessment that reclassifies nearly half of clinically high-risk patients to a low genomic risk category, enabling a shared decision to omit chemotherapy. As of 2026, MammaPrint is incorporated into major clinical guidelines as a validated tool for adjuvant chemotherapy decision-making in hormone receptor–positive, HER2-negative early breast cancer.

Because this publication focused exclusively on efficacy outcomes, no safety data from the trial were reported. The practical clinical implication is straightforward: for the patient sitting in your office with a high-clinical-risk, low-genomic-risk tumor, the 70-gene signature provides Level I evidence that chemotherapy can be safely omitted — with the understanding that a small absolute difference in distant recurrence may exist and warrants transparent discussion.

Trial Overview

Study Design

• Trial name and registration: EORTC 10041/BIG 3-04 MINDACT (NCT00433589; EudraCT 2005-002625-31)
• Design: Phase 3, prospective randomized clinical utility trial, open-label, final analysis
• Randomization: 1:1 randomization within discordant-risk groups only (high clinical risk/low genomic risk, and low clinical risk/high genomic risk). Concordant-risk patients (low/low and high/high) were assigned to treatment without randomization. Stratification factors: institution, risk group, hormone-receptor status (ER-positive or PR-positive vs. ER-negative and PR-negative), nodal involvement (yes or no), age (<50 years vs. ≥50 years), HER2 status (HER2-positive vs. HER2-negative vs. unknown), axillary treatment (sentinel node only vs. dissection), and type of surgery (mastectomy vs. breast conservation).
• Setting: Early-stage breast cancer, stage T1 or T2 or operable T3, node-negative or 1–3 positive axillary nodes
• Enrollment period and sites: Between 2007 and 2011; multicenter, international

Important design context: MINDACT is not a conventional two-arm drug trial. It is a clinical utility trial of a genomic test — the 70-gene signature (MammaPrint). Patients were classified into four risk groups based on clinical risk (modified Adjuvant! Online) and genomic risk (70-gene signature). Only patients with discordant risk assessments were randomized: those at high clinical risk/low genomic risk (N=1497 at enrollment) and those at low clinical risk/high genomic risk (N=690 at enrollment) were randomized 1:1 to use either clinical risk or genomic risk to guide the chemotherapy decision. Concordant-risk patients received treatment per both assessments: low/low patients did not receive chemotherapy; high/high patients received chemotherapy.

Mechanism of Action

MammaPrint is a 70-gene expression profile assay that classifies early breast cancer tumors into low or high genomic risk categories based on gene expression patterns associated with distant metastatic potential. It is an FDA-cleared in vitro diagnostic test [2]. The assay does not have a therapeutic mechanism of action — it is a prognostic and predictive tool used to guide the adjuvant chemotherapy decision.

Patient Population

• Key eligibility: Women aged 18–70 years with histologically confirmed primary invasive breast cancer (stage T1 or T2 or operable T3). Initially, all patients required lymph-node–negative disease; as of August 2009, the protocol was amended to allow enrollment of women with up to three positive axillary nodes.
• Biomarker selection: Patients were classified by 70-gene signature (MammaPrint) for genomic risk and modified Adjuvant! Online for clinical risk. There was no biomarker-based selection for enrollment — all comers meeting eligibility criteria were enrolled and classified.
• Sample size flow: 11,288 patients screened → 6,693 enrolled → classified into four corrected risk groups → 1,497 high clinical/low genomic at enrollment (749 randomized to chemotherapy, 748 randomized to no chemotherapy) → 690 low clinical/high genomic at enrollment (344 randomized to chemotherapy, 346 randomized to no chemotherapy). The primary-test population comprised 644 patients. A total of 75 patients (1.1%) were found to be ineligible.

Baseline Characteristics

Baseline characteristics are presented by the four corrected risk groups, as reported in the trial.

Values are n (%) unless otherwise specified. The overall discordance rate between clinical and genomic risk assessments was 32%.

Treatment Protocol

Important Design Context

This trial did not compare a single experimental drug to a single control drug. Treatment assignment was determined by risk classification strategy (clinical risk vs. genomic risk), and chemotherapy was administered or withheld accordingly.

High Clinical Risk / Low Genomic Risk — Chemotherapy (Clinical Risk Arm) (n=749 randomized)

Adjuvant chemotherapy (anthracycline-containing regimen or docetaxel-plus-capecitabine, per optional randomization) plus endocrine therapy if hormone-receptor positive.

• Dose and schedule: Not reported in this publication
• Treatment duration: Not reported in this publication
• Median treatment duration: Not reported in this publication
• Adherence: 85%

High Clinical Risk / Low Genomic Risk — No Chemotherapy (Genomic Risk Arm) (n=748 randomized)

No adjuvant chemotherapy; endocrine therapy if hormone-receptor positive.

• Dose and schedule: Not applicable (no chemotherapy)
• Treatment duration: Not reported in this publication
• Adherence: 89%

Low Clinical Risk / High Genomic Risk — Chemotherapy (Genomic Risk Arm) (n=344 randomized)

Adjuvant chemotherapy plus endocrine therapy if hormone-receptor positive.

• Dose and schedule: Not reported in this publication
• Treatment duration: Not reported in this publication
• Adherence: 80%

Low Clinical Risk / High Genomic Risk — No Chemotherapy (Clinical Risk Arm) (n=346 randomized)

No adjuvant chemotherapy; endocrine therapy if hormone-receptor positive.

• Dose and schedule: Not applicable (no chemotherapy)
• Treatment duration: Not reported in this publication
• Adherence: 88%

Overall adherence to the chemotherapy assignment across discordant groups was 86%.

Efficacy Outcomes

Primary Endpoint: Survival Without Distant Metastasis (5-Year Rate)

Definition: Survival without distant metastasis was defined as the time until the first distant metastatic recurrence or death from any cause.

Analysis population: Primary-test population — high clinical risk / low genomic risk patients randomized to use genomic risk who adhered to no chemotherapy, excluding those with risk changes (N=644).

Statistical method: Noninferiority test — whether the lower boundary of the 95% CI for the 5-year rate of survival without distant metastasis would be 92% or higher, at a one-sided significance level of 0.025.

Median follow-up: 5.0 years. Data cutoff: March 1, 2016.

Analysis maturity: The criteria for the primary analysis were met: the percentage of patients with 5-year follow-up was 60% (>33%), and the standard error for the rate of survival without distant metastasis at 5 years was 0.0094 (<0.01).

Result (primary-test population, no chemotherapy): 94.7% (95% CI, 92.5 to 96.2)

The lower boundary of the 95% CI (92.5%) exceeded the prespecified noninferiority threshold of 92%. The primary objective of the study was achieved.

Key Secondary Endpoints

Survival Without Distant Metastasis — High Clinical Risk / Low Genomic Risk (ITT, Chemotherapy vs. No Chemotherapy)

Analysis population: Intention-to-treat population (high clinical risk and low genomic risk at enrollment; N=749 chemotherapy, N=748 no chemotherapy)

Formally tested: Yes

Chemotherapy (Clinical Risk): 95.9% at 5 years (95% CI, 94.0 to 97.2) No Chemotherapy (Genomic Risk): 94.4% at 5 years (95% CI, 92.3 to 95.9) Comparison: HR 0.78 (95% CI, 0.50 to 1.21; P=0.27)

The absolute difference was small and not statistically significant.

Survival Without Distant Metastasis — Low Clinical Risk / High Genomic Risk (ITT, Chemotherapy vs. No Chemotherapy)

Analysis population: Intention-to-treat population (low clinical risk and high genomic risk at enrollment; N=344 chemotherapy, N=346 no chemotherapy)

Formally tested: Yes

Chemotherapy (Genomic Risk): 95.8% at 5 years (95% CI, 92.9 to 97.6) No Chemotherapy (Clinical Risk): 95.0% at 5 years (95% CI, 91.8 to 97.0%) Comparison: HR 1.17 (95% CI, 0.59 to 2.28; P=0.66)

No significant difference in distant metastasis-free survival was observed in this group.

Per-Protocol Analyses — High Clinical Risk / Low Genomic Risk

Analysis population: Per-protocol population (N=636 chemotherapy, N=592 no chemotherapy)

Of note, in the per-protocol analysis, disease-free survival favored chemotherapy (HR 0.64; P=0.03), whereas survival without distant metastasis and overall survival did not reach statistical significance.

Per-Protocol Analyses — Low Clinical Risk / High Genomic Risk

Analysis population: Per-protocol population (N=224 chemotherapy, N=254 no chemotherapy)

No statistically significant differences were observed across any endpoint in the low clinical risk / high genomic risk per-protocol population.

Exploratory Endpoints

Survival Without Distant Metastasis by Four Risk Groups (All Enrolled Patients)

Among all 6,693 enrolled patients:

• Low clinical / low genomic risk: 97.6% at 5 years (highest)
• High clinical / high genomic risk: 90.6% at 5 years (lowest)
• Discordant groups: approximately 95% at 5 years (intermediate)

These data confirmed a clear gradient of risk stratification by combining clinical and genomic assessment.

70-Gene Signature — Multivariate Analysis

In a multivariate analysis adjusting for chemotherapy use, clinical risk, and patient and tumor characteristics, the 70-gene signature was independently associated with survival without distant metastasis: HR 2.41 (95% CI, 1.79 to 3.26; P<0.001) for high genomic risk versus low genomic risk.

Chemotherapy Reduction Potential

Overall, 3,356 patients were categorized as being at high clinical risk, and 2,398 were categorized as being at high genomic risk. The difference between the two strategies would be 958 patients (14.3%) — patients who would receive chemotherapy by clinical risk assessment but not by genomic risk assessment. Among high-clinical-risk patients, using the 70-gene signature would lead to a reduction in the use of adjuvant chemotherapy in 1,550 of 3,356 patients (46.2%).

The rate of survival without distant metastasis would have been 95.0% with the clinical-risk strategy alone and 94.7% with the genomic-risk strategy alone.

Safety

No safety or adverse event data were reported in this publication. The paper focused exclusively on efficacy outcomes (survival without distant metastasis, disease-free survival, overall survival) and the clinical utility of the 70-gene signature.

Standard metrics including any TEAE, grade ≥3 TEAE, serious AE, treatment-related AE, discontinuation due to AE, dose reduction, dose interruption, and treatment-related death were all not reported in this publication.

Deaths

Deaths were reported only as overall survival events in the per-protocol populations:

• High Clinical Risk / Low Genomic Risk — Chemotherapy (per-protocol): 10 deaths
• High Clinical Risk / Low Genomic Risk — No Chemotherapy (per-protocol): 18 deaths
• Low Clinical Risk / High Genomic Risk — Chemotherapy (per-protocol): 5 deaths
• Low Clinical Risk / High Genomic Risk — No Chemotherapy (per-protocol): 8 deaths

Total deaths across all enrolled patients were not reported in this publication. Treatment-related deaths were not reported in this publication.

Subgroup Analyses

Subgroup analyses were performed within the high clinical risk / low genomic risk population. These were prespecified exploratory analyses and were not powered for formal statistical comparisons. Results should be interpreted as hypothesis-generating.

Node-Negative Subgroup (High Clinical Risk / Low Genomic Risk)

5-year survival without distant metastasis:

• Chemotherapy: 95.7% (95% CI, 93.0 to 97.4)
• No chemotherapy: 93.2% (95% CI, 90.1 to 95.4)

Node-Positive Subgroup (High Clinical Risk / Low Genomic Risk)

5-year survival without distant metastasis:

• Chemotherapy: 96.3% (95% CI, 93.1 to 98.1)
• No chemotherapy: 95.6% (95% CI, 92.7 to 97.4)

ER-Positive / HER2-Negative / Node-Negative Subgroup (High Clinical Risk / Low Genomic Risk)

5-year survival without distant metastasis:

• Chemotherapy: 95.5% (95% CI, 92.5 to 97.3)
• No chemotherapy: 93.9% (95% CI, 90.6 to 96.1)

Hazard ratios, interaction p-values, and subgroup sample sizes were not reported for these subgroup analyses. Across all subgroups examined, 5-year survival without distant metastasis exceeded 93% regardless of chemotherapy use, with small absolute differences between the chemotherapy and no-chemotherapy groups.

Of note, among high-clinical-risk patients with 1 positive lymph node, the 70-gene signature indicated low risk in 505 (63.0%) of 801 patients, and among those with 2–3 positive nodes, 226 (55.8%) of 405 were classified as low genomic risk.

Key Comparator Trials

Cross-trial comparisons are limited by differences in patient populations, trial designs, and endpoints. These data are presented for context, not for direct statistical comparison.

Contextual Notes

MINDACT and TAILORx address the same fundamental clinical question — can a genomic assay identify patients who can safely omit adjuvant chemotherapy? — but they use different genomic tests (70-gene signature vs. 21-gene recurrence score), different populations, and different endpoint definitions. MINDACT enrolled all molecular subtypes including HER2-positive and triple-negative patients, whereas TAILORx was restricted to HR+/HER2−/node-negative disease. RxPONDER extended the Oncotype DX question into the node-positive population, a group partially overlapping with the MINDACT high-clinical-risk cohort.

The key differentiating feature of MINDACT is its clinical utility design — it directly tested whether using the 70-gene signature to override the clinical risk assessment would result in acceptable outcomes, rather than simply demonstrating prognostic accuracy. This pragmatic design answers the clinician's actual question: can I safely withhold chemotherapy from this patient?

Grey Zones and Unanswered Questions

• Patients with >3 positive nodes were excluded. Only 8 patients (0.1%) in the entire trial had ≥4 positive nodes. The 70-gene signature's ability to identify low-risk patients in heavily node-positive disease remains unresolved, and clinicians should not extrapolate these results to that population.

• HER2-positive patients were included but not analyzed separately in sufficient detail. HER2-positive tumors comprised 9.5% of the cohort (638 patients), and the interaction between genomic risk classification and anti-HER2 therapy effect was not addressed. Given that HER2-directed therapy fundamentally changes the treatment paradigm, the relevance of a low genomic risk result in a HER2-positive patient is unclear.

• Five-year follow-up may be insufficient for HR-positive breast cancer. Late recurrences in ER-positive disease are well documented beyond 5 years. The 5-year DMFS data reported here may not capture the full impact of forgoing chemotherapy, particularly in the hormone receptor–positive subgroup that dominates the high clinical / low genomic risk population (98.1% ER/PR-positive).

• The absolute benefit of chemotherapy was small but not zero. In the per-protocol analysis of the high clinical risk / low genomic risk group, the 5-year DFS difference favored chemotherapy (93.3% vs. 90.3%; HR 0.64; P=0.03). Whether this approximately 3% absolute DFS difference — which includes locoregional events and second primaries — is clinically meaningful to individual patients is a shared-decision question the trial cannot answer.

• No safety data were reported. The toxicity burden of adjuvant chemotherapy is well established but trial-specific adverse event data were not presented, preventing a formal risk-benefit assessment within the MINDACT population.

Clinical Implications

Rationale and Clinical Context

Before MINDACT, the adjuvant chemotherapy decision in early breast cancer relied on clinical-pathological factors — tumor size, grade, nodal status, receptor status, and clinical risk calculators such as Adjuvant! Online. A substantial proportion of clinically high-risk patients were recommended chemotherapy despite uncertainty about individual benefit. The 70-gene signature offered an orthogonal, biology-driven risk assessment. MINDACT was designed to determine whether using this genomic test to override clinical risk assessment in discordant cases would lead to acceptable outcomes.

The trial's primary finding — that high-clinical-risk / low-genomic-risk patients who forgo chemotherapy have a 5-year DMFS rate of 94.7% with a lower CI boundary above 92% — established Level I evidence that the 70-gene signature can safely guide the chemotherapy de-escalation decision.

Monitoring and Long-Term Follow-Up

With a median follow-up of 5.0 years, these results should be considered in the context of late recurrence risk in ER-positive disease. Long-term follow-up data from MINDACT (subsequently published) are essential for confirming that the chemotherapy-sparing strategy remains safe beyond the initial reporting window. Patients managed without chemotherapy based on low genomic risk should continue standard breast cancer surveillance per institutional guidelines.

De-Escalation Considerations

MINDACT provides the evidentiary framework for chemotherapy de-escalation in early breast cancer. Among clinically high-risk patients, the 70-gene signature reclassified 46.2% as low genomic risk, potentially sparing nearly half of this population from chemotherapy and its associated toxicities (hematologic, cardiac, neuropathic, fertility-related). The chemotherapy reduction potential — 958 patients (14.3%) of the entire cohort — represents a substantial shift in treatment patterns.

The de-escalation decision should be individualized. The per-protocol DFS data suggest a small chemotherapy benefit even in the low-genomic-risk group (HR 0.64; P=0.03), which should be discussed transparently. Patient preferences regarding small absolute risk reductions versus chemotherapy toxicity should drive the final decision.

Unanswered Questions

The two most practice-relevant open questions are: (1) Do these results hold with longer follow-up, particularly for ER-positive disease where late recurrences are common? (2) How should the 70-gene signature result be integrated with other clinical factors (such as high nodal burden, very young age, or triple-negative histology) that were underrepresented or excluded from this trial?

Regulatory and Guideline Status

Regulatory

• FDA: MammaPrint (70-gene signature) received FDA clearance as an in vitro diagnostic device (510(k)) in 2007 for use in breast cancer patients. Following MINDACT, the FDA cleared an updated indication in 2017 to include patients with 1–3 positive lymph nodes. Regulatory status should be verified with current FDA labeling.
• EMA: MammaPrint is CE-marked and available for clinical use in the European Union. Regulatory status should be verified with current EMA guidance.

⚠️ Regulatory status verified as of March 2026. Confirm current approval status and labeled indications before clinical use at FDA.gov.

Guidelines

• NCCN: The NCCN Clinical Practice Guidelines in Oncology: Breast Cancer include MammaPrint as one of several validated gene expression assays for adjuvant therapy decision-making in HR-positive, HER2-negative early breast cancer. The guidelines reference MINDACT as supporting evidence for omitting chemotherapy in patients with low genomic risk [3].
• ASCO/ESMO: Both ASCO and ESMO guidelines incorporate the 70-gene signature as an acceptable tool for guiding adjuvant chemotherapy decisions in appropriate early breast cancer populations.

Companion Diagnostics

The 70-gene signature (MammaPrint) is itself the diagnostic test evaluated in this trial. It is performed on tumor tissue and classifies patients as low or high genomic risk. No companion diagnostic in the traditional pharmacogenomic sense is required, as this trial tested the clinical utility of the assay rather than a therapeutic agent.

About the Author

Andrew Stevenson is the founder and systems architect of kill-cancer.com, a clinical intelligence platform delivering structured, source-traced oncology trial analysis to practicing clinicians. He holds 17 Google technical certifications in data systems, automation, and applied AI — the engineering foundation behind the extraction and verification pipeline that produces every article on this platform. Every number traces directly to its source publication. Zero calculation. Zero editorializing. Zero hallucination.

Five siblings lost to cancer built the urgency. The engineering builds the trust.

📧 andrew@kill-cancer.com 🌐 kill-cancer.com 💬 kill-cancer.com/forum

Disclaimer

This article is intended for healthcare professionals only. It is not medical advice and should not be used as a substitute for professional clinical judgment. Treatment decisions should be made in consultation with qualified healthcare providers based on individual patient circumstances.

All trial data presented in this article are sourced directly from the published clinical trial report and its supplementary materials. Numbers are reproduced exactly as reported; no calculations, derivations, or estimates have been performed.

Trial results are presented as reported in the source publication. Updated data, label changes, or guideline revisions published after the source article may alter clinical applicability.

Comparator trial data presented in Section 8 are sourced from their respective published reports and are provided for contextual purposes only. Cross-trial comparisons have inherent limitations and should not be interpreted as direct statistical comparisons.

References

1. Cardoso F, van 't Veer LJ, Bogaerts J, et al. 70-Gene Signature as an Aid to Treatment Decisions in Early-Stage Breast Cancer. N Engl J Med. 2016;375:717-729. doi:10.1056/NEJMoa1602253
2. MammaPrint 510(k) Summary. U.S. Food and Drug Administration. Accessed March 2026.
3. NCCN Clinical Practice Guidelines in Oncology: Breast Cancer. Version 2.2026. Accessed March 2026.
4. 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;379:111-121. doi:10.1056/NEJMoa1804710
5. Kalinsky K, Barlow WE, Gralow JR, et al. 21-Gene Assay to Inform Chemotherapy Benefit in Node-Positive Breast Cancer. N Engl J Med. 2021;385:2336-2347. doi:10.1056/NEJMoa2108873