The landscape of breast cancer treatment has undergone a transformative shift with the integration of immunotherapy. Historically considered a “cold” tumor with limited immunogenicity, breast cancer management was dominated by surgery, chemotherapy, radiotherapy, and targeted therapies1. However, the advent of immune checkpoint inhibitors (ICIs) has challenged this paradigm, opening a new frontier. The initial breakthrough in triple-negative breast cancer (TNBC) demonstrated that a subset of patients could derive profound and durable clinical benefit from pembrolizumab and atezolizumab2,3. Today, precision immunotherapy aims to identify the patients most likely to respond, to convert immunologically silent tumors into responsive tumors, and to strategically combine immunotherapies with other modalities to overcome resistance. This evolution from empirical application to biomarker-driven strategies marks the critical juncture at which we stand, transitioning promising clinical trial data into refined, effective, and accessible clinical practice4. Recent key clinical studies on breast cancer immunotherapy are summarized in Table 1.
Summary of key clinical trials in breast cancer immunotherapy
Current status of immunotherapy in breast cancer
TNBC: breakthroughs in ICIs and combination strategies
TNBC accounts for 15%–20% of all breast cancers. Patients with TNBC derive limited benefit from endocrine and anti-HER2 targeted therapy due to a lack of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) expression. However, TNBC exhibits a higher tumor mutational burden (TMB) and greater number of tumor-infiltrating lymphocytes (TILs) compared to other molecular subtypes, which indicates stronger immunogenicity. TNBC also has a higher expression of programmed cell death ligand-1 (PD-L1), suggesting the presence of an active but suppressed antitumor immune response5. These characteristics have positioned TNBC as the most intensively studied breast cancer subtype undergoing immunotherapy.
The KEYNOTE-522 study represents a landmark trial in early-stage TNBC immunotherapy. This phase III randomized controlled trial enrolled 1,174 patients with high-risk TNBC to evaluate the efficacy of pembrolizumab combined with neoadjuvant chemotherapy followed by adjuvant pembrolizumab monotherapy6. The results showed that the pathologic complete response (pCR) rate was significantly higher in the pembrolizumab group than the placebo group (64.8% vs. 51.2%; P < 0.001). The final survival data showed a sustained event-free survival (EFS) benefit in the pembrolizumab group [hazard ratio (HR) = 0.65, 95% confidence interval (CI): 0.51–0.83]. More importantly, the overall survival (OS) achieved statistically significant improvement for the first time (HR = 0.66, 95% CI: 0.50–0.87; P = 0.003)7.
The KEYNOTE-355 study established the role of pembrolizumab as first-line treatment of advanced TNBC. This phase III trial enrolled 847 treatment-naïve patients with advanced TNBC who were randomized to receive pembrolizumab or placebo combined with chemotherapy8. The pembrolizumab group in the population with a combined positive score (CPS) ≥ 10 had a significantly prolonged median progression-free survival [PFS] (9.7 months vs. 5.6 months, HR = 0.65) and improved median OS (23.0 months vs. 16.1 months, HR = 0.73)9. The phase 3 TORCHLIGHT trial demonstrated that adding toripalimab to nab-paclitaxel significantly improved survival outcomes with a manageable safety profile in PD-L1-positive patients with advanced TNBC10. However, not all ICIs have achieved positive results. The IMpassion130 study initially demonstrated that atezolizumab combined with chemotherapy improved PFS and OS in PD-L1-positive advanced TNBC but the final OS analysis failed to reach the pre-specified statistical threshold11. The IMpassion131 study also yielded negative results, failing to demonstrate a PFS or OS benefit. This finding in the IMpassion131 study might represent replacement of the chemotherapy backbone drug with solvent-based paclitaxel and pre-treatment corticosteroids, which might have weakened initiation of anti-tumor immunity and the effect of PD-L1 blockade12. The IMpassion131 results were in contrast to the IMpassion130 results, which was based on albumin-bound paclitaxel. IMpassion132 involved a biologically unfavorable, early recurrent TNBC population. Atezolizumab combined with chemotherapy failed to improve OS, even in the PD-L1 positive population (median OS: 15.4 months vs. 15.9 months, HR = 1.10). This finding might be because the population is rich in treatment option heterogeneity and immune exclusion/exhaustion phenotypes13. In such cases, PD-L1 status alone is insufficient to capture the microenvironment responsive to ICIs.
Metronomic chemotherapy combined with ICIs represents a promising treatment approach. Metronomic chemotherapy (low-dose, high-frequency, continuous administration) can induce immunogenic cell death (ICD) and promote tumor vascular normalization, thereby improving immune cell infiltration. A phase II randomized study evaluated metronomic chemotherapy combined with anti-PD-1 therapy in metastatic breast cancer with preliminary results demonstrating encouraging antitumor activity14. Antibody-drug conjugates (ADCs) combined with ICIs represent another strategy of considerable interest. ADCs induce ICD through intracellular release of cytotoxic payloads, thereby promoting tumor antigen release and dendritic cell (DC) maturation, theoretically enhancing ICI efficacy. In the phase III ASCENT-04 study, sacituzumab govitecan combined with pembrolizumab as first-line treatment for patients with advanced PD-L1-positive TNBC significantly reduced the risk of disease progression or death by 35% compared to chemotherapy combined with pembrolizumab with a manageable safety profile15. Poly ADP-ribose polymerase (PARP) inhibitors combined with ICIs also represent an important area of exploration. PARP inhibitors can enhance immunotherapy sensitivity by increasing genomic instability, upregulating tumor cell PD-L1 expression, and activating interferon signaling pathways. The MEDIOLA study evaluated olaparib combined with durvalumab in patients with germline BRCA1/2 mutation (gBRCAm) metastatic breast cancer, achieving an objective response rate (ORR) of 63.3% and a median PFS of 8.2 months16.
HR+/HER2− breast cancer: unlocking the potential of immunotherapy
Hormone receptor (HR)+/HER2− is the most common breast cancer subtype, accounting for approximately 70% of all breast cancers. Due to a lower TMB, TIL infiltration, and PD-L1 expression, combined with significantly higher collagen deposition compared to TNBC, this subtype was traditionally considered insensitive to immunotherapy. However, recent studies have demonstrated that a subset of HR+/HER2− breast cancers may benefit from immunotherapy, particularly the prominent collagen-rich tumor microenvironment, which underscores the need for innovative strategies to overcome extracellular matrix (ECM) barriers in this subtype.
The CheckMate 7FL study enrolled 510 patients with high-risk HR+/HER2− breast cancer who were randomized to receive nivolumab or placebo combined with neoadjuvant chemotherapy17. The primary endpoint analysis showed that the pCR rate was significantly higher in the nivolumab group than the placebo group [24.5% vs. 13.8%, odds ratio (OR) = 2.05; P = 0.0021]. The KEYNOTE-756 study adopted a similar design, enrolling 1,278 patients with high-risk HR+/HER2− breast cancer to evaluate the efficacy of pembrolizumab combined with neoadjuvant chemotherapy18. The pCR rate was also significantly higher in the pembrolizumab group than the placebo group (24.3% vs. 15.6%; P = 0.00005). These studies demonstrated for the first time that adding ICIs to neoadjuvant chemotherapy for high-risk HR+/HER2− breast cancer can significantly improve pCR rates. However, whether improvements in pCR rates translate into long-term EFS and OS benefits requires longer follow-up. Biomarkers based on tumor stromal characteristics, such as collagen deposition, warrant further exploration to optimize patient stratification strategies.
The exploration of immunotherapy in HR+/HER2- type breast cancer lags behind that in TNBC. The combination of cyclin-dependent kinase 4/6 (CDK4/6) inhibitors with ICIs represents a strategy of considerable interest. CDK4/6 inhibitors can theoretically enhance ICI efficacy through mechanisms, including enhanced tumor antigen presentation, promotion of T-cell infiltration, and modulation of the tumor microenvironment. Multiple clinical studies involving CDK4/6 inhibitors combined with ICIs are ongoing, including the NEWFLAME (palbociclib + nivolumab) and PACE studies (palbociclib + avelumab)19. A meta-analysis demonstrated acceptable safety of CDK4/6 inhibitors combined with ICIs in HR+ breast cancer, although the efficacy data require further validation20. Notably, the combination of anti-estrogen therapy with ICIs is under exploration. Estrogen signaling can indirectly suppress antitumor immune responses by modulating immune cell function and influencing the tumor ECM. Exploratory analyses from the KEYNOTE-028 and KEYNOTE-086 studies suggested that ER-positive breast cancer patients have a limited response to ICI monotherapy but combination with endocrine therapy may improve efficacy21. Currently, multiple combination regimens, including fulvestrant plus pembrolizumab and abemaciclib plus atezolizumab, are under clinical investigation.
HER2+ breast cancer: exploring immune synergistic effects
HER2+ breast cancer accounts for 15%–20% of all breast cancers. The success of anti-HER2 targeted therapy has transformed this subtype from having the poorest prognosis to becoming the subtype with the most abundant treatment options. Interestingly, HER2+ breast cancer exhibits increased TIL infiltration and a relatively high TMB, suggesting potential sensitivity to immunotherapy. More importantly, anti-HER2 monoclonal antibodies can mediate immune cell-mediated tumor killing through antibody-dependent cell-mediated cytotoxicity (ADCC), providing a theoretical rationale for combination with ICIs.
The PANACEA study was the first phase II trial to evaluate pembrolizumab combined with trastuzumab in trastuzumab-resistant HER2+ metastatic breast cancer22. The ORR was 15% and the disease control rate (DCR) was 25% in PD-L1-positive tumors, while nearly no responses were observed in PD-L1-negative tumors, suggesting that PD-L1 expression is an important predictor of efficacy. Unfortunately, this study was terminated early at the time of the interim analysis because the addition of atezolizumab failed to improve PFS. Similarly, the IMpassion050 study evaluated atezolizumab combined with neoadjuvant chemotherapy plus dual anti-HER2 therapy in HER2+ early breast cancer and also failed to meet the primary endpoint23. These negative results suggested that simply adding ICIs to existing anti-HER2 regimens may be insufficient to provide additional benefit. Indeed, HER2+ breast cancer is highly sensitive to anti-HER2 therapy with limited room for further improvement and other suppressive factors are present in the tumor microenvironment.
ADCs combined with ICIs represent an emerging direction in HER2+ breast cancer immunotherapy.
Trastuzumab deruxtecan (T-DXd) induces ICD through the release of the cytotoxic payload deruxtecan, which promotes tumor antigen release and DC maturation, and theoretically enhancing tumor immunogenicity. In addition, the "bystander effect" of T-DXd can kill heterogeneous tumor cells with low HER2 expression, further releasing neoantigens. The DESTINY-Breast08 study explored T-DXd combined with various treatment regimens (including ICIs) in HER2-low breast cancer24. T-DXd combined with endocrine therapy demonstrated favorable efficacy and safety in the HR+/HER2− subgroup. The T-DXd plus ICI cohort is ongoing.
Core strategies for precision immunotherapy
The success of precision immunotherapy hinges on robust biomarkers that accurately predict therapeutic response. Classic biomarkers, such as PD-L1, TMB, and TILs, have provided a foundational framework for patient selection, particularly in TNBC25. Microsatellite instability-high (MSI-H) remains a potent predictor across cancer types but is rare in breast cancer26. Nevertheless, the predictive power is imperfect, revealing the complexity of the tumor immune microenvironment. Emerging biomarkers are refining our understanding of therapeutic responsiveness in patients with TNBC. Multi-omic classifications, such as the Fudan classification, which integrates genomic and immunologic features, offer a more holistic view of tumor biology27. Furthermore, the intratumoral and gut microbiome is emerging as a novel predictive and prognostic factor, influencing both systemic immunity and local therapeutic responses28. Based on the interplay between the extracellular matrix and TILs, we propose a novel immuno-collagenic classification for predicting responses to immunotherapy across various solid tumors, including TNBC29. The future lies in composite biomarker models that synthesize these diverse data streams.
A central challenge in breast cancer, particularly the HR+/HER2− subtype, is the presence of a profoundly immunosuppressive tumor microenvironment. Effective immunotherapy in this setting therefore requires strategies capable of converting immunologically “cold” tumors into inflamed tumors through rational combinations that enhance antigen presentation, T-cell priming and trafficking, and relieve stromal immune barriers. In this regard, growing evidence underscores the importance of the ECM, especially collagen deposition, as a critical determinant of immune exclusion30. Transforming growth factor-beta (TGF-β) signaling and cancer-associated fibroblasts (CAFs) emerge as key orchestrators of this process, shaping a stromal architecture that limits cytotoxic T-cell infiltration. Consistent with this concept, therapeutic interference with TGF-β signaling has been shown to alleviate stromal immune exclusion and enhance the efficacy of immune checkpoint blockade31.
In addition to a structural role, CAFs display marked phenotypic heterogeneity with distinct immunoregulatory functions. Specific CAF subsets are preferentially enriched in HR+/HER2− tumors and contribute to immune suppression by fostering regulatory T-cell-dominated niches32. Single-cell and spatial profiling studies further support the notion that TGF-β-associated fibroblast programs spatially coordinate T-cell exclusion and dysfunction, thereby reinforcing resistance to immunotherapy33,34. Importantly, recent data suggest a novel therapeutic axis in HR+/HER2− breast cancer, whereby anti-progestin therapy modulates CAF activity and remodels the ECM, providing a potential molecular rationale for combining endocrine therapy with immunotherapy35. Collectively, these findings highlight stromal reprogramming, particularly disruption of the collagen barrier, as a promising strategy to overcome immune exclusion in this disease.
Translational and clinical challenges
The path to precision immunotherapy remains challenging because primary and acquired resistance limit durable clinical benefit and reflect a complex interplay between tumor-intrinsic programs and host-related systemic factors36. In parallel, immune-related adverse events (irAEs) constitute a major barrier to broader clinical application, contributing substantially to treatment morbidity, and in rare cases, mortality. irAEs span a wide clinical spectrum, affecting multiple organ systems and often necessitating treatment interruption or immunosuppressive intervention, thereby complicating long-term disease control37. Although corticosteroids and selective immunosuppressive agents remain the cornerstone of irAE management, emerging evidence suggests that targeted cytokine modulation, particularly IL-6 receptor blockade, may mitigate toxicity without necessarily compromising, and potentially even enhancing, antitumor immunity38.
Accordingly, the identification of reliable biomarkers to predict irAE risk has become a key translational priority. Genetic susceptibility, including specific human leukocyte antigen (HLA) genotypes, has been linked to organ-specific irAEs, while pre-existing autoantibodies and dynamic cytokine profiles further reflect an underlying predisposition to immune toxicity39. In addition to host-intrinsic factors, the gut microbiome has emerged as an important modulator of ICI-associated inflammation, with distinct microbial signatures correlating with immune-related colitis40. Integrating these genetic, serologic, immunologic, and microbial parameters into a unified framework may enable personalized risk stratification and proactive toxicity management, ultimately improving the therapeutic index of immunotherapy in breast cancer.
Despite these challenges, the horizon of immunotherapy in breast cancer is expanding. Novel agents, such as bispecific antibodies and cellular therapies, are entering clinical investigation, offering new mechanisms of action41. Artificial intelligence (AI) and multi-omics technologies are poised to revolutionize precision medicine. A multi-omics machine learning model has been developed that integrates clinical data, digital pathology, and genomic and transcriptomic features to predict pCR to neoadjuvant chemotherapy for breast cancer. Notably, the model identified pre-existing immune infiltration features as key predictors of treatment response42. Dynamic monitoring is becoming a prevailing trend. Liquid biopsy based on circulating tumor DNA (ctDNA) enables tracking of clonal evolution and early signs of resistance, while serial imaging combined with radiomics may non-invasively assess changes in the tumor microenvironment43,44. These tools support a goal-directed therapeutic strategy, allowing therapeutic adjustments before clinical progression becomes evident.
Conclusions
Precision immunotherapy for breast cancer has evolved into an indispensable component of clinical management. This journey began with landmark success in TNBC and is now progressively expanding to challenge other subtypes. The critical developmental pathway has been delineated. Specifically, the pathway fundamentally relies on the guiding role of biomarkers, the driving effect of combination therapies, and the conceptual framework provided by a deep understanding of the tumor microenvironment. For example, promising strategies targeting the ECM in HR+/HER2− breast cancer exemplify the innovative thinking required to overcome biological barriers. Looking ahead to the next 3–5 years, progress will likely hinge on moving beyond conventional checkpoint blockade toward immune-redirecting and biomarker-steered strategies. Bispecific antibodies and ADC–ICI combinations may broaden benefit by actively recruiting/activating effectors and priming tumors for checkpoint blockade, while spatial multi-omics/AI biomarker and microbiome information strategies should be able to optimize patient selection and enhance treatment efficacy.
Conflict of interest statement
No potential conflicts of interest are disclosed.
Author contributions
Conceived and designed the analysis: Yongmei Yin and Xiang Huang.
Wrote the paper: Jie Mei, Kai Yang, and Xinkang Zhang.
- Received December 19, 2025.
- Accepted February 5, 2026.
- Copyright: © 2026, The Authors
This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.