Adding atezolizumab immunotherapy to capecitabine for the adjuvant treatment of triple-negative breast cancer (TNBC) in patients with residual disease after neoadjuvant chemotherapy does not offer a significant survival benefit, suggested a Korean trial.

There was nevertheless a numerical improvement seen in both invasive disease-free and overall survival with atezolizumab plus capecitabine vs capecitabine alone, reaching a trend toward a benefit in patients with tumors positive for PD-L1 expression.

In Hae Park, MD , a hemato-oncologist at Guro Hospital, Korea University College of Medicine in Seoul, Republic of Korea, underlined that the trial was designed before the results of KEYNOTE-522 were published. The author presented the study results at the ESMO Breast Cancer 2026.

That trial showed a benefit from adding pembrolizumab to neoadjuvant chemotherapy in this setting, leading to the FDA granting approval in 2021 to pembrolizumab in combination with chemotherapy for neoadjuvant treatment, continued as a single agent for adjuvant treatment, following surgery for patients with high-risk, early-stage TNBC.

“Our findings should be interpreted in the context of a population that did not receive neoadjuvant immunotherapy,” Park clarified, adding: “Whether adjuvant immune checkpoint addition plus capecitabine adds a benefit on top of neoadjuvant pembrolizumab requires further study.”

Park said that patients with TNBC with residual disease after neoadjuvant chemotherapy have a high risk for recurrence, at a 5-year disease-free survival (DFS) rate of 56%-70%, and that the “optimal adjuvant strategy for this population is an unmet clinical need.”

She noted that the CREATE-X trial demonstrated that adjuvant capecitabine improved DFS in this patient population, “establishing capecitabine as a standard post-neoadjuvant option.”

The A-BRAVE trial also suggested that adjuvant immunotherapy with avelumab after neoadjuvant chemotherapy was associated with DFS and overall survival benefits, although Park pointed out that that trial lacked an active comparator.

Park and colleagues conducted the MIRINAE trial to examine the combination of atezolizumab and capecitabine in the adjuvant setting in patients with TNBC with residual disease after neoadjuvant chemotherapy.

The study involved patients with TNBC with measurable residual disease of at least 1 cm and/or at least one positive lymph node after anthracycline- and taxane-based neoadjuvant chemotherapy.

In all, they randomly assigned 280 patients to 18 cycles of atezolizumab plus eight cycles of capecitabine (n = 139) or eight cycles of capecitabine alone (n = 141).

The baseline characteristics were very similar across the two treatment arms. The median patient age was 48-49 years, and approximately 45% were postmenopausal. The majority (72%-74%) had stage II-III disease, and 62% were negative for PD-L1 expression.

Park reported that, after a median follow-up of 47.4 months, there was no significant difference in invasive DFS (IDFS) between the groups, at a 5-year rate with atezolizumab plus capecitabine of 70.3% vs 66.5% with capecitabine alone, or a hazard ratio of 0.84 (P = .42).

There was also no significant difference in overall survival, at a rate of 79.3% at 5 years with combination adjuvant therapy vs 77.1% with capecitabine alone, or a hazard ratio of 0.81 (P = .45).

The results were similar across prespecified subgroups, including when stratifying the patients by age, menopausal status, and disease stage.

Park showed, however, that there was a signal for an IDFS benefit with atezolizumab plus capecitabine in patients positive for PD-L1 expression, at a hazard ratio vs capecitabine alone of 0.67 (P = .32) vs 0.92 (P = .75) in PD-L1-negative patients.

Similar results were seen for overall survival, with atezolizumab plus capecitabine showing a potential benefit over capecitabine alone in PD-L1-positive patients, at a hazard ratio of 0.57 (P = .32) vs 0.93 (P = .81) in PD-L1-negative patients.

As expected, the combination of atezolizumab plus capecitabine was associated with a higher rate of immune-related adverse events, seen in 38.1% of patients vs 2.8% with capecitabine. There was nevertheless only one grade ≥ 3 immune-related adverse event: a case of type 1 diabetes in the combination arm.

Invited discussant Giampaolo Bianchini, MD, head of breast oncology at the Department of Medical Oncology of the IRCCS San Raffaele Hospital in Milan, Italy, reminded the audience that, in light of the KEYNOTE-522 results, perioperative pembrolizumab combined with neoadjuvant chemotherapy is the standard of care for intermediate/high-risk TNBC.

However, he emphasized that there are lessons to be learned from the current findings, which add to those of the A-BRAVE and IMpassion030 trials. “The key message,” said, Bianchini, who was not involved in the study, “is that, more or less, they are negative,” so immunotherapy in this setting does “not provide a clinical benefit.”

Bianchini also noted that “we have a signal of activity” in these trials, but that perhaps they have been focused on the wrong type of immunotherapy.

He explained that anti-PD-L1 therapies “are not exactly the same” as those targeting PD-1. Consequently, results of the phase 3 SWOG S1418/BR006 of pembrolizumab in this setting are “eagerly awaited.”

Bianchini said, the current findings won’t impact clinical practice, “because they are hypothesis-generating,” but that they “highlight the urgent need for precision immunology in breast cancer.”

Just because the results were negative overall, it “doesn’t mean that there aren’t patients that can benefit from that specific strategy,” he said.

The study was supported by the National R&D Program for Cancer Control through the National Cancer Center, funded by the Ministry of Health & Welfare, Republic of Korea.

Park declared having relationships with Norvatis and Roche Korea.

Bianchini declared having relationships with AstraZeneca, MSD, Roche, Menarini, Novartis, Eli Lily, Daïichi Sankyo, Seagen, Tethis, Gilead, Pfizer, Exact Sciences, Takeda, Standard BioTools, and Helsinn.