Written by Dr. Roupen Odabashian MD, FRCPC
Oncologist | Founder, MeDucation AI | Updated October 2026
The verdict: DESTINY-Breast05 asked whether fam-trastuzumab deruxtecan-nxki (T-DXd) is better than trastuzumab emtansine (T-DM1) as adjuvant therapy for patients with HER2-positive early breast cancer who have residual invasive disease after neoadjuvant therapy. The answer, as presented, is yes on efficacy: T-DXd showed a statistically significant improvement in invasive disease-free survival (iDFS) over T-DM1[1]. The cost is interstitial lung disease (ILD), which occurred more often with T-DXd than with T-DM1, including fatal cases[2]. The FDA approved this adjuvant indication on May 15, 2026[1]. Whether T-DXd should now replace T-DM1 in your patient depends on how high-risk that patient actually is, and that nuance is exactly what boards will test.
Conflict of interest, up front: I'm a practicing hematologist-oncologist and I founded MeDucation AI, which sells a physician-written heme/onc question bank and other board-prep and clinical tools. Nothing below changes based on that. Where a trial result is still conference-presented rather than peer-reviewed in full, I say so.
Question | Short answer |
|---|---|
What did DESTINY-Breast05 test? | T-DXd vs T-DM1 as adjuvant therapy in HER2-positive early breast cancer with residual invasive disease after neoadjuvant therapy[1] |
Primary endpoint result | 3-year iDFS 92.4% (T-DXd) vs 83.7% (T-DM1), HR 0.47, p<0.0001[1] |
Headline toxicity | Adjudicated drug-related ILD occurred more often with T-DXd than with T-DM1, including fatal ILD events[2] |
Regulatory status | FDA approved for this adjuvant indication May 15, 2026, with breakthrough designation[1] |
Does it replace T-DM1 for everyone? | No, per invited discussant Sara Tolaney, MD: T-DM1 should remain standard for operable, node-negative residual disease[3] |
What question did DESTINY-Breast05 actually ask?
DESTINY-Breast05 asked whether T-DXd is superior to T-DM1 as adjuvant treatment in patients who still have residual invasive disease after neoadjuvant HER2-directed therapy. This is a global, open-label, randomized, two-arm, phase III trial (NCT04622319) that enrolled 1,635 adults with HER2-positive breast cancer with residual invasive disease following neoadjuvant therapy[1]. Patients were randomized 1:1 to T-DXd or T-DM1 for a maximum of 14 cycles, or until recurrence or unacceptable toxicity[1].
This is the same clinical population that the KATHERINE trial addressed in 2019, patients with residual disease after neoadjuvant chemotherapy and HER2-targeted therapy, for whom adjuvant T-DM1 became standard of care after showing a reduction in recurrence risk and an overall survival benefit compared with trastuzumab alone[3]. DESTINY-Breast05 is the trial that tests whether a newer antibody-drug conjugate can do better than that established standard in the same setting[2].
Eligibility required centrally confirmed HER2-positive disease (IHC 3+ or ISH-positive) with residual invasive disease in the breast and/or axillary lymph nodes after at least 16 weeks of neoadjuvant chemotherapy that included at least 9 weeks of taxane-based therapy and 9 weeks of trastuzumab with or without pertuzumab, plus an ECOG performance status of 0 or 1[2][4]. The trial specifically enrolled a high-risk population: inoperable disease before neoadjuvant therapy (cT4, N0-3, M0 or cT1-3, N2-3, M0), or operable disease with axillary node-positive residual disease after neoadjuvant treatment[2].
What were the DESTINY-Breast05 results?
Outcome | T-DXd | T-DM1 | Statistic |
|---|---|---|---|
Patients randomized | 818 | 817 | 1,635 total; interim analysis[6] |
3-year iDFS | 92.4% (95% CI 89.7 to 94.4) | 83.7% (95% CI 80.2 to 86.7) | HR 0.47 (95% CI 0.34 to 0.66), p<0.0001[6][7] |
3-year DFS | 92.3% | 83.5% | HR 0.47 (95% CI 0.34 to 0.66)[6] |
3-year distant recurrence-free interval | 93.9% | 86.1% | HR 0.49 (95% CI 0.34 to 0.71)[6] |
3-year brain metastasis-free interval | 97.6% | 95.1% | HR 0.64 (95% CI 0.35 to 1.17)[6] |
3-year overall survival | 97.4% | 95.7% | HR 0.61 (95% CI 0.34 to 1.10), immature[6][7] |
Deaths at the iDFS analysis | 18 | 29 | 47 patients (2.9%) overall; interim analysis[1][7] |
Those numbers come from the interim analysis reported with a median study duration of about 30 months and 153 iDFS events, representing 74% of the planned information fraction[2]. The overall survival data are immature (only 2.9% of patients had died at the time of the IDFS analysis)[1][2], so the OS hazard ratio of 0.61 should be read as directional, not definitive[2]. The iDFS benefit held consistently across prespecified subgroups, including age, race, geographic region, hormone receptor status, type of HER2-targeted neoadjuvant therapy, and timing of radiotherapy[2].
A subgroup breakdown presented at the 2025 San Antonio Breast Cancer Symposium also showed the benefit held regardless of HER2 expression level: in IHC 3+ disease, the 3-year iDFS rate was 91.8% with T-DXd versus 83.2% with T-DM1 (HR 0.49)[4], and in ISH-positive disease it was 96.2% versus 86.5% (HR 0.35)[4]. Treatment completion rates were similar between arms, 72.3% with T-DXd and 76.3% with T-DM1 completed all 14 planned cycles[2].
How does the ILD risk compare between T-DXd and T-DM1?
Adjudicated drug-related ILD occurred more often with T-DXd than with T-DM1[2]. Most cases in both arms were grade 1 or 2, but the T-DXd arm had seven grade 3 events (0.9%) and two grade 5, fatal, events (0.2%)[2]. Treatment-emergent adverse events led to death in three T-DXd patients (0.4%), two of them from ILD or pneumonitis, and in five T-DM1 patients (0.6%) from various causes[2].
Safety measure | T-DXd | T-DM1 |
|---|---|---|
Any-grade adjudicated drug-related ILD | 9.6%[6][7] | 1.6%[6][7] |
Grade 3 ILD | 0.9% (7 events)[7] | 0%[7] |
Grade 5 ILD | 0.2% (2 events)[7] | 0%[7] |
Grade 3 or higher treatment-emergent adverse events | 50.6%[6][7] | 51.9%[6][7] |
Discontinuation due to adverse events | 17.9%[6] | 12.7%[6] |
Discontinuation due to drug-related ILD or pneumonitis | 10.8%[6] | 2.5%[6] |
Decreased platelet count, any grade | 21.2%[6] | 49.8%[6] |
Decreased neutrophil count, any grade | 31.6%[6] | 14.4%[6] |
Nausea, any grade | 71.3%[6] | 29.3%[6] |
The timing of radiotherapy, sequential versus concurrent with study drug, did not change the incidence of ILD in either arm[4]. Investigator-reported radiation pneumonitis was common in both arms because of the intensive serial chest CT monitoring built into the protocol, but no grade 3 or higher pneumonitis occurred in either arm[4][3]. As invited discussant Sara Tolaney, MD, framed the tradeoff: "In balance, however, benefits far outweigh risks in high-risk patients"[3], while cautioning that quality-of-life impact, generalizability given limited North American enrollment, and mature overall survival data remain unknown[3].
How does DESTINY-Breast05 compare with the KATHERINE trial?
KATHERINE established T-DM1 as the adjuvant standard for residual disease in 2019, and DESTINY-Breast05 is the trial that challenges it with a head-to-head comparison in the same clinical niche.
Feature | KATHERINE (2019) | DESTINY-Breast05 |
|---|---|---|
Comparison | T-DM1 vs trastuzumab[3] | T-DXd vs T-DM1[3] |
Population | Residual invasive disease after neoadjuvant therapy[3] | High-risk residual disease; the discussant described it as a much higher-risk population than KATHERINE[3] |
Headline efficacy | About a 50% reduction in the risk of invasive recurrence or death vs trastuzumab[3] | 3-year iDFS 92.4% vs 83.7%, HR 0.47[6] |
CNS recurrence effect | T-DM1 did not reduce CNS recurrence rate vs trastuzumab[2] | Numerically fewer CNS recurrences with T-DXd; brain metastasis-free interval data still immature[2] |
Who is actually eligible for adjuvant T-DXd right now?
The FDA-approved adjuvant indication is for adult patients with HER2-positive (IHC 3+ or ISH-positive) breast cancer who have residual invasive disease following neoadjuvant treatment with trastuzumab (with or without pertuzumab) and taxane-based treatment, using an FDA-authorized test[1]. Two companion diagnostics, the PATHWAY anti-HER-2/neu (4B5) antibody and the VENTANA HER2 Dual ISH DNA Probe Cocktail, were approved alongside it for identifying eligible patients[1]. The recommended adjuvant dose is 5.4 mg/kg every three weeks for a maximum of 14 cycles, unless there is disease recurrence or unacceptable toxicity[1].
Monitoring matters as much as the prescribing decision. The trial protocol required baseline chest CT in all patients, additional CT after completion of radiation and the first dose of study drug for patients on sequential radiotherapy, and further scans at cycles 3, 7, and 11, then at about 40 days of follow-up[4]. For grade 1 ILD, treatment was interrupted with corticosteroids considered, and T-DXd could resume once ILD resolved to grade 0[5]. For grade 2 or higher ILD, T-DXd was permanently discontinued[4]. The approved label carries a boxed warning for ILD and pneumonitis, plus warnings for neutropenia and left ventricular dysfunction[1].
What do the boards actually test on this trial?
Know the comparator and the population precisely. DESTINY-Breast05 is T-DXd vs T-DM1, specifically in patients with residual invasive disease after neoadjuvant therapy, not a frontline or metastatic comparison[1][2]. Confusing this with DESTINY-Breast11 (the neoadjuvant pCR trial) or with metastatic HER2-positive trials is a classic distractor trap.
Know the headline number and its direction. 3-year iDFS 92.4% vs 83.7%, HR 0.47[1]. Boards will give you the absolute numbers and ask you to identify the correct hazard ratio range or the correct endpoint (iDFS, not OS).
Know that ILD is the defining toxicity tradeoff. ILD occurred more often with T-DXd than with T-DM1, with rare but real grade 5 events[2]. Expect a vignette describing new cough or dyspnea on T-DXd, where the correct answer is to hold the drug and start systemic corticosteroids per the grade of ILD, not to continue with closer monitoring alone[5].
Know how KATHERINE fits historically. T-DM1 became standard after showing a 50% reduction in recurrence risk versus trastuzumab alone in residual disease, published in NEJM in 2019[3], and that it did not meaningfully reduce CNS recurrence[2][3]. DESTINY-Breast05 builds directly on that standard rather than replacing it outright.
Know that overall survival data are still immature. Only 2.9% of enrolled patients had died at the time of the presented interim analysis[1][2], so test-writers may ask you to correctly state that OS benefit is not yet established even though iDFS clearly favors T-DXd.
If you are building a study plan around trials like this rather than just memorizing isolated numbers, this is the kind of distinction that benefits from active recall rather than passive rereading. Our spaced-repetition flashcard approach and the explanations in our physician-written question bank are built specifically to force you to separate trial design, endpoint, and toxicity profile the way an exam question will.
FAQ
Is T-DXd now the standard of care after neoadjuvant therapy for all HER2-positive patients with residual disease?
No. The trial's own invited discussant recommended adjuvant T-DXd specifically for node-positive-at-surgery or inoperable-at-baseline disease (T3/T4 or N2/N3), while recommending that T-DM1 remain standard for operable, node-negative residual disease[3]. Eligibility for the FDA-approved indication is broader than that risk-stratified recommendation, so the choice between the two drugs is a clinical judgment, not an automatic substitution.
What is the single biggest safety difference between T-DXd and T-DM1 in this setting?
Interstitial lung disease. Adjudicated drug-related ILD occurred in 9.6% of patients on T-DXd vs 1.6% on T-DM1, including two grade 5 events (0.2%) on T-DXd[6][7]. Discontinuation for drug-related ILD or pneumonitis was 10.8% with T-DXd vs 2.5% with T-DM1[6]. T-DM1 has its own pattern: decreased platelet counts of any grade occurred in 49.8% of T-DM1 patients vs 21.2% of T-DXd patients[6].
Has the FDA approved T-DXd for this adjuvant indication?
Yes, as of May 15, 2026, the FDA approved T-DXd for adjuvant treatment of adults with HER2-positive breast cancer with residual invasive disease following neoadjuvant trastuzumab-based and taxane-based treatment[1]. Before that date, T-DXd was not approved in the neoadjuvant or adjuvant early-stage setting[2].
Is the overall survival benefit with T-DXd established?
Not yet. At the time of the presented interim iDFS analysis, only 2.9% of enrolled patients had died across both arms, and the OS hazard ratio of 0.61 (95% CI 0.34-1.10) did not reach statistical significance at that maturity[1][2]. This is a conference-presented interim result, and longer follow-up is needed before an OS claim can be made confidently.
How did DESTINY-Breast05 enroll a higher-risk population than KATHERINE?
It required high-risk disease: either inoperable at presentation (cT4 with any N and M0, or cT1 to cT3 with N2 or N3 and M0), or operable disease with residual node-positive disease after neoadjuvant therapy (ypN1 to ypN3)[4]. The invited discussant described it as a much higher-risk population than KATHERINE, which is why she recommended that T-DM1 remain the standard for operable, node-negative residual disease[3].
For fellows building board prep around trials exactly like this one, our question bank comparison coverage and our ABIM blueprint breakdown cover how heavily breast cancer trial data is weighted on the medical oncology exam. If you want a tool that pushes new practice-changing data like DESTINY-Breast05 into your inbox automatically, OncoPulse lets you schedule alerts by topic, including FDA approvals and guideline updates, so you are not relying on catching a conference abstract months after it drops.
MeDucation AI's heme/onc question bank is physician-written and reviewed by board-certified hematologist-oncologists, never AI-generated, with explanations that cover why each correct answer is right and why every distractor is wrong, plus direct PubMed links so you can go straight to the primary data on trials like DESTINY-Breast05 and KATHERINE. If trials like this one are where your board prep is weakest, that is exactly what the bank is built to fix.
