T-DXd cuts death risk 31–42% in HER2-low breast cancer

Why DESTINY-Breast04’s Long-Term Data Matters for HER2-Low Metastatic Breast Cancer

The 32-month long-term follow-up dataset from DESTINY-Breast04 is the most complete picture available of trastuzumab deruxtecan (T-DXd) efficacy in HER2-low metastatic breast cancer — a patient population historically underserved by HER2-directed therapies. Before T-DXd, HER2-low expression (defined by immunohistochemistry scores of 1+ or 2+ without gene amplification) was not considered an actionable therapeutic target. DESTINY-Breast04 changed that, and the extended follow-up data now confirm that the initial survival gains are sustained and consistent across biologically distinct subgroups.

The trial’s primary analysis had already demonstrated statistically significant improvements in progression-free survival and overall survival. The long-term update, with its 32-month median follow-up, addresses the critical clinical question of whether those benefits hold: whether early response translates into sustained survival advantage, and whether the drug’s efficacy is genuinely broad across hormone receptor status and ER expression level. According to ASCO reporting standards for oncology trials, landmark survival rates at 24 and 36 months are among the most clinically meaningful endpoints for characterising durability of benefit in metastatic settings.

What makes the DESTINY-Breast04 long-term dataset particularly valuable from an innovation intelligence perspective is its granularity: the trial reports efficacy not just for the overall cohort and the pre-specified HR+ population, but also for the exploratory HR− cohort and two ER expression subgroups — ER-low-positive (IHC 1%–10%) and ER >10%. This level of subgroup reporting enables a rigorous assessment of whether T-DXd’s mechanism of action, centred on its antibody-drug conjugate payload delivery, is truly independent of the tumour’s hormonal biology.

Overall Survival: Hazard Ratios and What They Reveal Across Patient Populations

T-DXd extended median overall survival by 6.1 months in the overall cohort (22.9 months versus 16.8 months for TPC), with a hazard ratio of 0.69 (95% CI 0.55–0.86) — a 31% reduction in the risk of death that is both statistically significant and clinically meaningful. The HR+ cohort, which represented the primary analysis population, showed a near-identical survival benefit: median OS of 23.9 months versus 17.6 months (HR 0.69; 95% CI 0.55–0.87).

The exploratory HR− cohort produced the most numerically striking OS result: a median of 17.1 months with T-DXd versus just 8.3 months with TPC (HR 0.58; 95% CI 0.31–1.08), corresponding to a 42% reduction in death risk. While the confidence interval crosses 1.0 in this smaller exploratory subgroup — reflecting reduced statistical power — the directional consistency with the HR+ primary analysis is clinically noteworthy. The 24-month OS rate in the HR− cohort was 32.6% with T-DXd versus 11.8% with TPC, a more than two-fold difference at that landmark.

“In the HR− cohort, the 24-month overall survival rate was more than two times greater with T-DXd (32.6%) compared with TPC (11.8%) — a landmark difference that underscores the depth of benefit even in this exploratory population.”

Progression-Free Survival Across Every Subgroup: The PFS Hazard Ratio Story

T-DXd’s progression-free survival advantage over TPC is even more pronounced than its OS benefit, with PFS hazard ratios below 0.40 across all five populations studied. In the overall cohort, median PFS by investigator assessment was 8.8 months with T-DXd versus 4.2 months with TPC — a hazard ratio of 0.36 (95% CI 0.29–0.45), meaning T-DXd reduced the risk of disease progression or death by 64% relative to TPC.

The HR+ cohort PFS mirrored the overall result: 9.6 months versus 4.2 months (HR 0.37; 95% CI 0.30–0.46). In the HR− exploratory cohort, median PFS was 6.3 months versus 2.9 months (HR 0.29; 95% CI 0.15–0.57) — a 71% reduction in progression risk. Across the ER expression subgroups, the ER >10% population showed a PFS of 9.6 months versus 4.3 months (HR 0.37; 95% CI 0.29–0.47), while the ER-low-positive (IHC 1%–10%) subgroup recorded the most favourable PFS hazard ratio of any group: 8.3 months versus 2.3 months (HR 0.27; 95% CI 0.14–0.52).

The consistency of PFS hazard ratios — all between 0.27 and 0.37 across five distinct patient populations — is a statistically unusual finding in oncology trials. It suggests that T-DXd’s mechanism of action, which relies on HER2-mediated internalisation of its cytotoxic payload rather than on downstream signalling pathway dependence, delivers a reproducible anti-tumour effect regardless of whether the tumour also expresses hormone receptors. This is consistent with the established pharmacology of antibody-drug conjugates as described by EMA in its ADC regulatory guidance framework.

PFS2 and the Durability Question: Does T-DXd Compromise Subsequent Therapy?

PFS2 — defined as the time from randomisation to disease progression on the next line of therapy or death — is a critical endpoint for assessing whether a first-line or second-line treatment impairs the effectiveness of what follows. In the overall cohort, T-DXd achieved a PFS2 of 15.4 months versus 9.7 months with TPC. In the HR+ cohort, PFS2 was 15.5 months versus 10.5 months, and in the HR− cohort (exploratory), 12.9 months versus 6.7 months.

The PFS2 data are particularly important for interpreting the OS results. If T-DXd’s OS benefit were driven primarily by the drug’s own activity — and subsequent therapies were equally effective regardless of prior treatment — PFS2 would be expected to converge toward TPC values. The fact that PFS2 remains substantially longer in the T-DXd arm suggests one of two things: either patients on T-DXd remain in better overall condition at the time of second progression, enabling more effective use of subsequent therapies; or T-DXd’s mechanism does not cross-sensitise tumours to resistance against the drug classes used in subsequent lines. The long-term data support the latter interpretation, with investigators noting that OS results were consistent with the primary analysis and that investigator-assessed PFS in the updated analysis aligned with BICR-assessed PFS from the primary analysis.

ER Expression Level and T-DXd Efficacy Independence: What the Subgroup Data Show

One of the most clinically significant findings from the long-term DESTINY-Breast04 dataset is the near-identical T-DXd efficacy observed across the ER >10% and ER-low-positive (IHC 1%–10%) subgroups. In the ER >10% population, median OS with T-DXd was 24.0 months and median PFS was 9.6 months. In the ER-low-positive (IHC 1%–10%) population, median OS with T-DXd was 22.3 months and median PFS was 8.3 months — values that are clinically comparable despite the substantial difference in ER expression level between the two groups.

By contrast, TPC outcomes differed substantially between the two ER subgroups: median OS was 18.9 months in the ER >10% group but only 10.2 months in the ER-low-positive group — a 8.7-month gap that likely reflects the greater chemotherapy sensitivity or endocrine therapy responsiveness of higher-ER-expressing tumours under standard-of-care treatment. This divergence in TPC outcomes, combined with the convergence of T-DXd outcomes, produces markedly different OS hazard ratios: 0.72 (95% CI 0.56–0.93) in the ER >10% group versus 0.40 (95% CI 0.20–0.79) in the ER-low-positive group. The ER-low-positive subgroup OS hazard ratio of 0.40 — a 60% reduction in death risk — is among the most striking single-subgroup findings in the trial.

“The similar median OS and PFS values for T-DXd across the ER >10% and ER-low-positive subgroups suggest that T-DXd’s efficacy is independent of the ER level in the tumour — a finding with direct implications for patient selection and treatment sequencing.”

This pattern aligns with the biological rationale for HER2-directed ADC therapy: the drug’s efficacy is determined by HER2 surface expression enabling payload delivery, not by the tumour’s downstream signalling dependencies. According to published pharmacological frameworks from NIH/NCI, ADC activity in low-antigen-expressing tumours depends critically on the bystander killing effect of the free payload — a mechanism that is inherently independent of receptor pathway biology downstream of the target antigen.

Depth, Consistency, and What the Confidence Intervals Collectively Confirm

Reading the DESTINY-Breast04 long-term dataset as a whole, three properties of T-DXd’s clinical benefit emerge with particular clarity: depth, consistency, and durability. Each is supported by specific features of the hazard ratios and confidence intervals across the five patient populations.

Depth of Benefit

Depth is most directly expressed in the magnitude of the hazard ratios. OS hazard ratios of 0.69 in the overall and HR+ cohorts, 0.58 in the HR− cohort, 0.72 in the ER >10% subgroup, and 0.40 in the ER-low-positive subgroup all represent clinically meaningful reductions in death risk. PFS hazard ratios are more striking still, ranging from 0.27 to 0.37 — reductions in progression risk of 63–73%. These are not marginal improvements; they represent a fundamental shift in the natural history of HER2-low metastatic breast cancer when T-DXd is used.

Consistency Across Subgroups

Consistency is demonstrated by the narrow spread of hazard ratios across biologically distinct populations. The PFS HRs cluster between 0.27 and 0.37; the OS HRs between 0.40 and 0.72. The trial investigators confirmed that T-DXd showed efficacy benefits across diverse demographic subgroups in the overall and HR+ cohorts, and that efficacy was consistent by IHC status within the overall cohort. The exploratory subgroup analyses — HR−, ER >10%, and ER-low-positive — all directionally favour T-DXd, with the ER-low-positive subgroup producing the most favourable hazard ratios of any group in the trial. This consistency across hormone receptor status and ER expression level, as noted in trial reporting consistent with WHO clinical trial reporting standards, is a hallmark of a mechanism-driven rather than tumour-biology-dependent effect.

Durability of Benefit

Durability is confirmed by the 32-month median follow-up, the alignment between updated investigator-assessed PFS and primary BICR-assessed PFS, and the sustained PFS2 advantage. The OS results at 32-month follow-up are consistent with those from the primary analysis, indicating that the initial survival separation between T-DXd and TPC Kaplan-Meier curves has been maintained rather than converging over time. The extended PFS2 data further support the conclusion that T-DXd’s benefit is not limited to its own treatment period but extends through subsequent lines of therapy — a property that distinguishes truly effective treatments from those that merely delay progression at the cost of future treatment options. This type of durable benefit is increasingly recognised by regulatory bodies including the FDA as a key criterion for meaningful clinical benefit in metastatic oncology.

Taken together, the DESTINY-Breast04 long-term dataset provides a comprehensive evidence base establishing T-DXd as a highly effective treatment for HER2-low metastatic breast cancer. The breadth of the benefit — across five patient populations, at two landmark timepoints, across three efficacy endpoints — makes this one of the most robustly characterised efficacy profiles in recent metastatic breast cancer clinical trial history. For researchers, oncologists, and innovation intelligence professionals tracking the antibody-drug conjugate landscape, these data represent a foundational reference point for understanding where the therapeutic bar now stands in this indication. Detailed patent and pipeline analysis tools available through PatSnap‘s innovation intelligence platform enable deeper exploration of the competitive landscape surrounding T-DXd and next-generation HER2-directed ADCs.