In normal breast tissue, immune cells are generally sparse. However, in breast cancer, TNBC tends to have higher TIL density compared to other subtypes [36]. Abundant TILs generally indicate that Teff, Tm, and Trm-like cells are actively present in the TME, and the immune system is actively recognizing and attempting to fight the tumor. Breast cancer has moderate tumor mutation burden (TMB) compared to other cancers, but TNBC often exhibits higher TMB than other subtypes [37]. High TMB contributes to neoantigen generation, recognizable by the immune system, thereby increasing the likelihood of an immune response in TNBC. As activated TILs release cytokines, particularly IFN-γ, PD-L1 expression is higher in TNBC, leading to higher ICI response rates.

In contrast, ER-positive breast cancer typically has lower TIL levels, less PD-L1 positivity, and lower TMB. Recent studies reported associations between ER signaling and immune suppression. For example, ERα signaling in tumor-associated macrophages contributed to an immune-suppressive TME state by promoting CD8+ T cell dysfunction and exhaustion [38]. Another study reported that hormone signaling can suppress the immune system by decreasing immune cell infiltration [39]. As TNBC lacks hormone-driven immune suppression, the immune system can play a more active role in tumor recognition and attack. For these reasons, breast cancer immunotherapy has primarily focused on and developed for TNBC.

Advances in tumor immunology and molecular targeted therapy have dramatically evolved metastatic TNBC treatment strategy. Metastatic TNBC treatment requires a multifaceted approach tailored to patient's tumor biology and history. Conventional chemotherapies aim for rapid tumor shrinkage, but resistance and side effects necessitate balancing long-term sustainability with quality of life. For the second line and beyond, poly(ADP-ribose) polymerase (PARP) inhibitors (e.g., olaparib, talazoparib) exploit DNA repair deficiencies to effectively control tumor growth, extending progression-free survival (PFS) and representing personalized treatment in BRCA1/2-mutated TNBC. Antibody drug conjugates (ADCs) link cytotoxic drugs to tumor-targeting antibodies such as trastuzumab deruxtecan (HER2-positive or HER2-low) [40]and sacituzumab govitecan (TNBC) [41] improved objective response rates (ORR) and progression-free survival (PFS) compared to chemotherapy.

As the first-line treatment, ICIs (e.g., pembrolizumab, atezolizumab) combined with chemotherapy achieve excellent clinical outcomes, particularly in PD-L1-positive TNBC. Both KEYNOTE‑355 (using pembrolizumab with one of three chemotherapy regimens: nab‑paclitaxel, paclitaxel, or gemcitabine plus carboplatin) and IMpassion130 (using atezolizumab with nab‑paclitaxel) reported significant improvements in progression‑free survival (PFS). [42, 43]. Pembrolizumab also provided significant OS benefit [44]. Based on the results of these clinical trials, pembrolizumab (or atezolizumab if available) with chemotherapy is now the standard of care for PD-L1-positive metastatic TNBC.

In contrast, IMpassion131 (using atezolizumab with paclitaxel) showed no significant benefits [45]. Reasons for differing results remain unclear but likely stem from interconnected factors. Subtle differences in patient clinical characteristics (age, prior treatment, PD-L1 status) might exist. Prior treatments/conditions may affect the immune system/TME, influencing atezolizumab efficacy. Chemotherapy agent choice may be critical. Nab-paclitaxel (IMpassion130) is albumin-bound, not requiring corticosteroid premedication, whereas paclitaxel (IMpassion131) does. Corticosteroids, necessary for paclitaxel hypersensitivity prevention, can suppress immune function, potentially diminishing PD-L1 blockade's essential immune activation. Moreover, nab-paclitaxel reportedly exerts a more favorable immunomodulatory profile (enhancing antigen presentation, preserving T-cell function, boosting mast cells), leading to more robust anti-tumor responses [46]. Differences in drug delivery/penetration between paclitaxel and nab-paclitaxel may also cause distinct TME effects.

Immunotherapy in the neoadjuvant setting is now the standard of care for early-stage TNBC. The KEYNOTE-522 trial examined neoadjuvant pembrolizumab plus chemotherapy (including platinum) followed by adjuvant pembrolizumab in high-risk early TNBC. Pembrolizumab significantly improved pathological complete response (pCR) rate, prolonged event-free survival (EFS) and OS. In contrast, in IMpassion031, adding atezolizumab to chemotherapy significantly improved pCR and showed a trend toward better prognosis, although the prognosis improvement was not statistically significant [9]. In KEYNOTE-522, pembrolizumab was given both before and after surgery, which helped sustain immune activation whereas IMpassion031 used atezolizumab only as neoadjuvant therapy, possibly limiting the long-term immune benefit. Pembrolizumab, a PD-1 inhibitor, blocks immune suppression more broadly and can activate a more robust, sustained antitumor response. Atezolizumab, a PD-L1 inhibitor, might provide only a transient boost in immunity. Differences in the chemotherapy regimens (KEYNOTE522: carboplatin/paclitaxel then anthracycline/cyclophosphamide [AC]; IMpassion031: nab-paclitaxel then AC) may influence immune activation and its synergy with immunotherapy. These differences likely explain why KEYNOTE-522 showed improvements in both short- and long-term outcomes, while IMpassion031 only showed a significant pCR improvement.

In the ALEXANDRA/IMpassion030 trial, early-stage TNBC patients received adjuvant chemotherapy (paclitaxel then dose-dense AC) with or without atezolizumab, then maintenance atezolizumab. However, atezolizumab did not enhance the effect of adjuvant chemotherapy [47]. IMpassion030 used adjuvant (postoperative) immunotherapy only, whereas KEYNOTE‐522 and IMpassion031 applied the checkpoint inhibitors in a neoadjuvant setting (or both neoadjuvant and adjuvant in KEYNOTE-522). From an immunological perspective, the timing of immunotherapy appears critical. In neoadjuvant settings, the presence of the primary tumor provides a continuous source of tumor antigens, enabling effective priming of naïve T cells through DC presentation. This not only reactivates pre-existing Tex—many of which express PD-1—but also promotes the generation of new T cell clones, thereby broadening the T cell repertoire and enhancing anti-tumor responses [48, 49]. These immunological events occur both within the tumor and TDLNs, maximizing the immune activation window. In contrast, adjuvant immunotherapy is administered after tumor resection, when the source of tumor antigens is largely eliminated. This limits new T cell priming and may reduce the overall efficacy of ICIs. Accordingly, neoadjuvant immunotherapy may offer superior anti-tumor effects in early-stage TNBC by harnessing a more diverse and robust T cell response [50]

The KEYNOTE-522 was designed to administer adjuvant pembrolizumab regardless of the response to neoadjuvant immunotherapy. As patients achieving pCR demonstrated excellent clinical outcomes, it remains to be determined whether this adjuvant treatment provides extra benefits for all patients. In GeparNuevo trial [51], while EFS improved with preoperative durvalumab and chemotherapy, no postoperative therapy was given. This suggests long-term benefits might be achievable with preoperative ICI alone. Notably, patients achieving pCR showed 100% three-year distant disease-free survival and OS. We do not clearly know if adjuvant ICI is required for patients achieving pCR.

In pCR cases (tumor completely absent at surgery), tumor-specific T cell function is likely preserved, suggesting potentially limited need for adjuvant immunotherapy. However, since pCR is determined pathologically, minimal residual disease (MRD) undetectable by imaging/diagnostics cannot be entirely excluded. Although pCR is a favorable prognostic indicator, recent studies have highlighted its limitations as a surrogate marker for complete immune eradication of tumor cells. Detection of circulating tumor DNA (ctDNA) in patients with pCR suggests the presence of MRD, which may lead to future recurrence if not addressed [52]. Supporting this, ctDNA levels were significantly lower in patients who achieved pCR after neoadjuvant immunotherapy compared to non-pCR patients, yet still detectable in some cases [53]. Therefore, adjuvant immunotherapy might help eliminate systemic minimal residual disease by enhancing the activity and persistence of circulating memory T cells, potentially contributing to long-term suppression of distant recurrence. In parallel, at the surgical site, the re-activation of pre-existing Trm-like cells may play a central role in preventing local tumor relapse [54] (Fig. 3). Conversely, a cautious approach is warranted due to potential irreversible irAE risks [55, 56]. Since recurrence risk in pCR patients is low, adjuvant immunotherapy benefits may not outweigh irAE harms. Medical economics are also important; ICI is expensive, and post-surgery administration adds significant burden. Currently, based on KEYNOTE-522 protocol, maintaining adjuvant pembrolizumab after surgery is considered reasonable for TNBC patients. However, when patients achieve pCR, the treatment plan should comprehensively consider their condition, irAE risk, and economic toxicity. The OptimICE-PCR trial (NCT05812807) is planned to investigate whether adjuvant immunotherapy can be omitted after pCR. It randomizes patients achieving pCR (with ≥ 6 cycles neoadjuvant pembrolizumab) to observation or continued pembrolizumab. OptimICE-PCR results are expected to provide a definitive answer regarding adjuvant immunotherapy necessity in pCR cases.

Perioperative immunotherapy and tumor dynamics. Neoadjuvant immunotherapy can efficiently activate pre-existing effector T cells and expand tumor-specific T cell clones. These activated T cells infiltrate the tumor and differentiate into tissue-resident memory-like T cells (Trm-like cells), which localize and persist within the tumor bed. During the neoadjuvant phase, such localized immune activation contributes to significant tumor shrinkage. Importantly, Trm-like cells remain stationed at the former tumor site after surgery. If residual cancer cells begin to re-grow locally, these Trm-like cells can rapidly recognize them and initiate a robust, site-specific immune response, thereby preventing local recurrence. However, even in cases where pCR is achieved, ctDNA may still be detectable, indicating the presence of MRD elsewhere in the body. In the adjuvant phase, systemic administration of immune checkpoint inhibitors (ICIs) can stimulate a broader population of memory T cells—including circulating or lymphoid-residing memory T cells—that can detect and targeting MRD throughout the body. By supporting both the maintenance of Trm-like cells in the surgical bed and enhancing systemic memory T cell responses, ICIs can help maintain tumor dormancy. Nonetheless, if this immune surveillance network is disrupted, tumor cells may escape immunological control, leading to renewed proliferation and recurrence as distant macro-metastases. Trm-like cell, resident memory-like T cells; pCR, pathological complete response; MRD, minimal residual disease; ctDNA, circulating tumor DNA

According to KEYNOTE-522 data, patients with residual cancer burden (RCB)-I demonstrated outcomes as favorable as RCB-0 (equivalent to pCR), suggesting minimal need for adjuvant pembrolizumab. Patients with RCB-II exhibited the greatest EFS improvement (hazard ratio [HR] 0.52) among RCB categories [57]. RCB‐II represents an intermediate-risk group in which a complete response was not achieved, yet a partial response to neoadjuvant chemo- and immunotherapy was observed. In this group, the presence of residual tumor may lead to the release of tumor antigens and an increase in immunogenicity, thereby potentially re-educating or boosting the immune response. On the other hand, RCB-III prognosis remains poor (HR 1.24), with limited benefit from adjuvant pembrolizumab. RCB‐III represents a group that shows primary resistance to neoadjuvant therapy, suggesting a high degree of resistance to both chemo- and immunotherapy. Thus, relying solely on adjuvant pembrolizumab may be insufficient, and alternative treatment strategies are needed. Indeed, few physicians consider pembrolizumab alone sufficient for non-pCR patients; administration of capecitabine or olaparib (for BRCA-mutated TNBC) is becoming widespread. Furthermore, ADCs showing excellent effects against breast cancer have been developed. OptimICE-RD trial (sacituzumab govitecan + pembrolizumab) [58] and TROPION-Breast03 trial (datopotamab deruxtecan + durvalumab) [59] are currently underway. These trial results are expected to further optimize the adjuvant treatment strategy for TNBC [60].