ATEMPT

Clinical question: In stage I HER2-positive breast cancer, can one year of adjuvant T-DM1 provide excellent disease control with less clinically relevant toxicity than paclitaxel/trastuzumab?

Registry: NCT01853748 · One underlying study; aliases and follow-up publications consolidated.

Randomized phase II · 3:1 allocation · Evidence cut-off: 8 November 2026

Design, population and eligibility

Design

Randomized phase II study; co-primary objectives were clinically relevant toxicity with T-DM1 and 3-year iDFS in the T-DM1 arm. Not powered for efficacy superiority between arms.

Sample size

497 initiated protocol therapy: 383 T-DM1 and 114 paclitaxel/trastuzumab.

Eligibility

Centrally confirmed stage I HER2-positive invasive breast cancer after surgery; selected small, predominantly node-negative disease.

Follow-up / maturity

Five-year follow-up median 5.8 years in the published analysis; longer follow-up publications should be linked as same study.

Treatment and comparator

Intervention: T-DM1 3.6 mg/kg IV every 3 weeks for 17 cycles (approximately one year).

Comparator: Paclitaxel 80 mg/m² weekly for 12 weeks with trastuzumab 4 mg/kg loading then 2 mg/kg weekly, followed by trastuzumab 6 mg/kg every 3 weeks to complete one year.

Primary endpoint: Clinically relevant toxicities and iDFS in the T-DM1 arm.

Key efficacy outcomes

5-year iDFS: 97.0% (95% CI 95.2–98.7) with T-DM1; 91.1% in TH arm Between-arm efficacy comparison was not powered.

5-year recurrence-free interval: 98.3% (95% CI 97.0–99.7) with T-DM1

5-year OS / BCSS: 97.8% / 99.4% with T-DM1

Clinically relevant toxicity: 46% T-DM1 vs 47% TH; p=0.83

The T-DM1 arm had 97.0% 5-year iDFS. Clinically relevant toxicity was not lower (46% vs 47%). The study was not designed or powered to establish T-DM1 superiority over TH.

Absolute outcome visualization

Five-year iDFS and clinically relevant toxicity as reported in the randomized study.

iDFS · T-DM1

97.0%

iDFS · TH

91.1%

CRT · T-DM1

46%

CRT · TH

47%

Bars show source-reported endpoint values; no outcomes are reconstructed from median survival. Single-arm outcomes are not comparative effects.

NNT / NNH

No comparative NNT: the trial was not powered for an efficacy comparison between T-DM1 and TH; the 5-year iDFS estimates should not be interpreted as a superiority result.

No NNH for overall clinically relevant toxicity: 46% vs 47%, p=0.83. Toxicity profiles differed by type rather than total incidence.

Toxicity, grade 3–5 harms and treatment burden

Clinically relevant toxicities occurred in 46% with T-DM1 and 47% with TH. T-DM1 was associated with less neuropathy and alopecia; thrombocytopenia, transaminase elevation and treatment discontinuation are relevant T-DM1 burdens. Grade 3–5 values must be taken from the full toxicity tables and are not extrapolated here.

Both strategies planned one year of therapy. T-DM1 requires 17 three-weekly infusions; TH requires 12 weekly paclitaxel/trastuzumab visits followed by trastuzumab to one year. Discontinuation and dose modification differ by toxicity profile.

Quality of life

Patient-reported outcomes indicated less neuropathy and alopecia and better work productivity with T-DM1; no overall reduction in clinically relevant toxicity.

Subgroups and applicability

Stage I HER2-positive disease; outcomes are most applicable to small node-negative tumours. Do not extrapolate to larger, node-positive or high-risk disease where escalation evidence differs.

Limitations

Randomized phase II with 3:1 allocation; co-primary design did not establish efficacy superiority. Five-year outcomes in the TH arm are descriptive. Not a definitive head-to-head efficacy trial.

Clinical interpretation

ATEMPT establishes excellent 5-year outcomes with adjuvant T-DM1 in selected stage I HER2-positive disease, but it did not demonstrate fewer clinically relevant toxicities than TH and does not prove superior efficacy.

Contemporary-practice note (8 November 2026): historical trial efficacy is distinguished from present-day treatment selection. Comprehensive guideline-level freshness review remains pending for this batch; apply current disease-specific guidance and multidisciplinary context.

Primary sources and related publications

Missing values are not imputed; meeting abstracts are identified as such.