Atezolizumab added nothing to HER2-positive breast cancer therapy in IMpassion050

Adding the immunotherapy atezolizumab to standard treatment for HER2-positive early breast cancer did not improve how long patients stayed free of their cancer, according to the final results of the phase 3 IMpassion050 trial [s1]. The study had already reported, at its primary analysis, that atezolizumab did not raise the rate of complete tumour clearance before surgery; the final data now confirm that the drug also failed to translate into a longer-term survival advantage [s1].

The trial tested a strategy that has reshaped other cancers — adding a checkpoint inhibitor, which releases a brake on the immune system, to existing therapy — in a setting where its value was unproven. HER2-positive breast cancer is already treated with potent antibodies against the HER2 protein, and the question was whether layering immunotherapy on top would add anything. IMpassion050 says it did not.

What the trial did

IMpassion050 (NCT03726879) randomly assigned patients with high-risk, HER2-positive early breast cancer 1:1 to atezolizumab or placebo, each given alongside neoadjuvant pertuzumab, trastuzumab, and chemotherapy before surgery [s1][s2]. After surgery, patients continued atezolizumab or placebo with pertuzumab and trastuzumab for up to a year; those with residual disease could switch to the antibody-drug conjugate trastuzumab emtansine [s1]. The trial was funded by F. Hoffmann-La Roche, the drug's manufacturer — a commercial sponsorship worth noting given that the result is negative, which is a point in the data's favour rather than against it [s1].

The primary endpoint had been pathologic complete response — no residual invasive cancer at surgery. At that first analysis, the complete-response rates were similar with and without atezolizumab in both the intention-to-treat and the PD-L1-positive populations, and safety matched other atezolizumab combinations [s1]. The secondary endpoints reported in this final analysis included event-free survival (EFS) and disease-free survival (DFS), the measures that reveal whether an early-response signal — had there been one — holds up over years rather than months [s1].

What it found

At the data cut-off of 24 August 2023, 411 of 454 patients remained on study (206 of 226 in the atezolizumab arm and 205 of 228 on placebo), with median follow-up of 44.2 and 43.4 months [s1]. Three-year event-free survival was 91.4% with atezolizumab versus 89.0% with placebo — a gap that did not reach significance, with a stratified hazard ratio of 0.90 (95% confidence interval 0.50 to 1.59) [s1]. Among the patients who had surgery and post-neoadjuvant therapy, three-year disease-free survival was 92.9% with atezolizumab versus 88.5% with placebo (hazard ratio 0.71, 95% CI 0.38 to 1.32) [s1].

Side effects of special interest — the immune-related toxicities typical of checkpoint inhibitors — were more common with atezolizumab, affecting 59.4% of patients versus 47.0% on placebo, though none were fatal [s1]. The most frequent adverse events included radiation skin injury (24.0% versus 19.4%), arthralgia (20.7% versus 17.1%), and diarrhoea (20.7% versus 13.4%) [s1].

How to read it

Both survival comparisons had confidence intervals comfortably crossing 1, so neither supports a real benefit: the EFS interval runs from a 50% reduction in events to a 59% increase, and the DFS interval is similarly inconclusive [s1]. The trial's authors did note numerically higher survival with atezolizumab in some subgroups — including PD-L1-negative, hormone-receptor-positive, and stage T2 disease — but these are exploratory slices of a trial whose main comparisons were flat, and chasing them would invert the discipline that makes a negative trial informative [s1]. The authors also observed that the numerical disease-free-survival improvements with atezolizumab appeared regardless of whether a patient had achieved a pathologic complete response — a pattern that, in a trial with flat primary results, is better read as noise than as a hidden benefit [s1].

The cleaner reading is that atezolizumab did not improve outcomes and did add toxicity. That is a useful result: it spares patients a drug that would have brought immune-related side effects without a demonstrated payoff, and it adds HER2-positive breast cancer to the list of settings where checkpoint inhibition, so transformative elsewhere, has not delivered [s1]. Related coverage has examined a neoadjuvant atezolizumab trial in triple-negative breast cancer, an antibody-drug conjugate paired with immunotherapy in the same disease, and a prominent null immunotherapy result in lung cancer.

What to watch

The broader puzzle is why checkpoint inhibitors help so clearly in triple-negative breast cancer yet have struggled to prove themselves in the HER2-positive subtype, where antibody therapy already engages the immune system [s1]. For patients and clinicians weighing options today, IMpassion050's message is simple: standard HER2-targeted treatment remains the backbone, and adding atezolizumab to it is not supported by this trial.

This article describes research and is not medical advice. Breast cancer treatment decisions belong with treating oncologists.

Sources

• Atezolizumab in HER2-positive early breast cancer: final results of the phase III IMpassion050 trial — ESMO Open, 24 September 2026
• IMpassion050 trial registration (NCT03726879) — ClinicalTrials.gov