Immune checkpoint inhibition has transformed the treatment of triple-negative breast cancer, but its role in hormone receptor-positive, HER2-negative (HR+/HER2−) metastatic breast cancer remains uncertain. As tumors progress through endocrine therapy and CDK4/6 inhibition, however, their biology may evolve, raising the possibility that selected endocrine-resistant tumors could become more susceptible to immune-based treatment.
At ESMO 2026, the phase III KEYNOTE-B49 trial will evaluate this question by comparing pembrolizumab plus chemotherapy with placebo plus chemotherapy in patients with PD-L1-positive, endocrine-resistant HR+/HER2− locally recurrent inoperable or metastatic breast cancer who are candidates for chemotherapy.
Ahead of the presentation, OncoDaily IO spoke with Dr. Matthew Kurian, a triple board-certified hematologist/oncologist at St. Elizabeth Healthcare and Assistant Professor of Medicine at the University of Kentucky, whose clinical focus includes breast malignancies. Dr. Kurian leads the breast and genitourinary oncology clinical trial program at St. Elizabeth and has research interests spanning precision oncology, ctDNA-based disease monitoring, artificial intelligence, and emerging cancer therapeutics.
ESMO Congress 2026 Breast Cancer: 10 Abstracts to Watch
Why Revisit Immunotherapy in HR+/HER2− Breast Cancer?
HR+/HER2− breast cancers have historically been considered relatively immune-cold compared with triple-negative disease, contributing to the limited activity of checkpoint inhibition observed in previous metastatic studies. However, HR+/HER2− disease is biologically heterogeneous, and the immune landscape may not remain static as tumors acquire resistance to endocrine therapy.
Dr. Kurian explained that endocrine-resistant disease may contain immunologically distinct subsets:
“Historically, most HR+/HER2− breast cancers have traditionally been considered relatively immune-cold, but emerging data suggests that endocrine-resistant tumors may have a more complex immune microenvironment. Interestingly, some resistant tumors demonstrate increased CD8+ T-cell infiltration and interferon signaling, suggesting an inflamed, but potentially dysfunctional immune state. Whether these changes develop as a consequence of endocrine resistance or reflect pre-existing tumor biology remains unclear. It’s possible that chemotherapy could help overcome some of this immune dysfunction, providing a rationale for combining it with checkpoint blockade. However, I suspect this opportunity is limited to a subset of patients rather than endocrine-resistant disease broadly.”
This distinction is important for interpreting KEYNOTE-B49. The trial is not simply testing whether immunotherapy can be extended from triple-negative to HR+/HER2− breast cancer. It is testing checkpoint inhibition in a population whose tumors have already progressed through endocrine-based treatment and reached a point at which chemotherapy is clinically appropriate.
KEYNOTE-B49: Testing Pembrolizumab at the Transition to Chemotherapy
KEYNOTE-B49 (NCT04895358) is a randomized, double-blind, placebo-controlled phase III trial evaluating pembrolizumab plus chemotherapy versus placebo plus chemotherapy in approximately 800 patients with HR+/HER2− locally recurrent inoperable or metastatic breast cancer.
Eligible patients have centrally assessed PD-L1 CPS ≥1, documented progression on previous endocrine therapy, and no previous chemotherapy for metastatic disease. Prior treatment generally includes at least two endocrine-based regimens in the metastatic setting, including at least one containing a CDK4/6 inhibitor, although the protocol includes additional eligibility pathways for patients with early endocrine resistance.
Patients are randomized 1:1 to pembrolizumab 200 mg intravenously every three weeks or placebo, combined with investigator’s choice of paclitaxel, nab-paclitaxel, liposomal doxorubicin, or capecitabine. Pembrolizumab or placebo can continue for up to 35 cycles, approximately two years.
Randomization is stratified by PD-L1 CPS 1–9 versus ≥10, presence of visceral metastases, and chemotherapy class. The primary endpoints evaluate PFS by blinded independent central review and OS in the PD-L1 CPS ≥10 and CPS ≥1 populations.
Is PD-L1 Enough?
The biomarker strategy may become one of the most important aspects of KEYNOTE-B49. Although all enrolled patients have PD-L1 CPS ≥1, HR+/HER2− breast cancer encompasses multiple molecular states, and PD-L1 expression may capture only one component of immunotherapy sensitivity.
Dr. Kurian believes a broader biomarker strategy will ultimately be necessary:
“I think PD-L1 CPS is the tip of the iceberg. Prior metastatic studies have not consistently demonstrated an association between PD-L1 expression and benefit from immunotherapy. More recent data suggest that immune gene signatures and intrinsic molecular subtype may be equally, if not more, important. The SOLTI-1716 TATEN study is particularly interesting, demonstrating a 61% response rate with pembrolizumab and paclitaxel in patients with HER2-enriched or basal-like tumors following CDK4/6 inhibitor progression. Although these findings require validation, they suggest that identifying non-luminal, immune-inflamed tumors may ultimately be more useful than relying on PD-L1 expression alone.”
TATEN provides an important complementary signal. Rather than selecting patients by PD-L1 alone, the study focused on HR+/HER2− metastatic tumors classified as HER2-enriched or basal-like after progression on CDK4/6 inhibition. Although the study was small and requires validation, it supports the hypothesis that conventional receptor classification may conceal biologically distinct tumors with greater susceptibility to chemoimmunotherapy.
For KEYNOTE-B49, the relationship between efficacy and PD-L1 level will therefore be particularly informative. A benefit concentrated in CPS ≥10 disease would support a more restricted biomarker-selected strategy, while limited discrimination by CPS could strengthen the rationale for integrating intrinsic subtype, immune signatures, or other biomarkers into future trials.
How Does Prior CDK4/6 Inhibition Affect the Immune Landscape?
Most patients reaching chemotherapy in contemporary HR+/HER2− metastatic breast cancer have already received CDK4/6 inhibition. This distinguishes KEYNOTE-B49 from earlier immunotherapy studies and introduces another biological question: whether previous CDK4/6 exposure modifies subsequent sensitivity to checkpoint blockade.
Dr. Kurian emphasized that the relationship is unlikely to be straightforward:
“This is a question I don’t think we know for sure yet at this time, as CDK4/6 inhibitors appear to have both favorable and unfavorable effects on the immune microenvironment. Early exposure can enhance antigen presentation, increase interferon signaling, and promote T-cell activity. However, prolonged treatment and acquired resistance may lead to greater immunosuppression, including changes in macrophage populations and T-cell dysfunction. I think the duration of prior CDK4/6 inhibitor therapy and the biology of resistance may be more relevant than simply whether a patient received one. Whether these changes meaningfully influence subsequent pembrolizumab benefit remains to be determined.”
The duration of previous endocrine sensitivity and the mechanisms driving resistance may therefore prove more informative than prior CDK4/6 exposure alone. Translational analyses from KEYNOTE-B49 could help determine whether these characteristics identify populations with different sensitivity to pembrolizumab.
Which Patients Could Actually Need Chemoimmunotherapy?
The treatment landscape after endocrine resistance has expanded substantially. Depending on tumor biology and previous therapy, patients may still have effective endocrine-targeted options, while antibody-drug conjugates have increasingly changed treatment sequencing. Even if KEYNOTE-B49 is positive, pembrolizumab plus chemotherapy will therefore need to find its place within this broader therapeutic landscape.
Dr. Kurian sees the strategy as particularly relevant when endocrine-based therapy is becoming less appropriate:
“I would be most interested in patients with early endocrine resistance, rapid disease progression, or symptomatic visceral involvement where chemotherapy is already being considered. Patients with more aggressive tumor biology, particularly those with HER2-enriched or basal-like molecular subtypes and evidence of immune activation, may be especially relevant. That said, we now have several effective endocrine-targeted therapies, and I would be hesitant to move directly to chemotherapy–immunotherapy in patients who still have reasonable endocrine-based options given the difference in toxicity profiles. The challenge will be identifying those who are unlikely to derive meaningful benefit from additional endocrine treatment and may instead benefit from an immunotherapy-based approach.”
This is an important boundary around the trial’s potential implications. KEYNOTE-B49 is not testing pembrolizumab as an alternative to endocrine therapy for all patients progressing after CDK4/6 inhibition. It specifically addresses patients who have reached the transition to chemotherapy.
Does the Chemotherapy Backbone Matter?
KEYNOTE-B49 permits paclitaxel, nab-paclitaxel, liposomal doxorubicin, or capecitabine, allowing the study to reflect several chemotherapy approaches used in metastatic disease. Different cytotoxic agents can affect tumor-cell death, antigen release, immune-cell populations, and other components of the tumor microenvironment, providing biological reasons to consider whether the chemotherapy partner could influence checkpoint inhibition.
These mechanisms, however, should not be assumed to translate into clinically different pembrolizumab effects. Randomization is stratified according to chemotherapy class, and any efficacy differences between chemotherapy backbones will need to be evaluated through prespecified subgroup analyses, confidence intervals, and formal interaction testing rather than comparisons of response rates alone.
What Would Make KEYNOTE-B49 Practice-Changing?
Because HR+/HER2− metastatic breast cancer now has multiple therapeutic options, statistical significance alone may not be sufficient to establish a new treatment strategy. The magnitude of PFS benefit, OS, durability of response, toxicity, and ability to identify the patients deriving benefit will all influence the clinical relevance of KEYNOTE-B49.
Dr. Kurian outlined what he will be looking for in the ESMO results:
“Given the negative experience with immunotherapy in metastatic HR+/HER2− breast cancer, I think the bar for KEYNOTE-B49 is fairly low given no previous movement in the HR+ metastatic space. A meaningful PFS improvement would be encouraging, but an overall survival benefit would make a much stronger argument for changing practice. I would also be particularly interested in whether benefit is concentrated among patients with CPS ≥10, as we have seen in triple-negative disease, or extends to those with lower PD-L1 expression. We need to move beyond just measuring PDL1, so I hope there is more emphasis on more novel biomarkers. The toxicity profile and consistency of benefit across chemotherapy backbones will also matter. With the growing number of endocrine-targeted therapies and ADCs available, a modest PFS improvement alone may not be enough to justify adding pembrolizumab. Ultimately, the study needs to demonstrate a clinically meaningful benefit in a population we can reliably identify.”
The distinction between statistical and clinical significance will be particularly important. Pembrolizumab introduces immune-mediated toxicity and potentially prolonged treatment exposure, while patients with HR+/HER2− metastatic disease increasingly have alternative biomarker-directed therapies and ADCs available. The therapeutic index will therefore need to be considered alongside efficacy.
What to Watch at ESMO 2026
The first question from KEYNOTE-B49 will be whether pembrolizumab improves PFS and OS in the prespecified PD-L1 populations. Beyond the primary endpoints, the magnitude of absolute benefit, objective response rate, duration of response, and maturity of the OS data will help establish whether any improvement is clinically meaningful.
The comparison between CPS ≥10 and the broader CPS ≥1 population may be particularly informative. Results according to visceral disease, chemotherapy backbone, and previous treatment characteristics could further refine patient selection, although subgroup findings will require caution if numbers are small or interaction tests are not significant.
Safety will be equally important, including grade ≥3 adverse events, immune-mediated toxicities, treatment discontinuation, and treatment-related deaths. Translational analyses could ultimately be especially valuable if they determine whether PD-L1 CPS alone captures the population benefiting from pembrolizumab or whether intrinsic subtype, immune gene-expression profiles, or other features provide greater discrimination.
