TILs include innate lymphoid cells and T and B lymphoid cells (25). TILs may be used as prognostic markers in the neoadjuvant treatment of TNBC. Several studies on TILs in the neoadjuvant chemotherapy environment have demonstrated this marker's predictive and prognostic value (26-30). Compared with luminal breast cancer, high immune infiltration is more common in TNBC (31). In addition, the immunosuppressive markers programmed cell death-1 (PD-1) and PD-L1 were positively correlated with tumor-infiltrating lymphocytes (32). Higher levels of TILs are associated with an improved prognosis of TNBC, with every 10% increase in TILs being associated with a 15-25% reduction in the risk of recurrence and death (33-36). Iroquois homeobox protein 2 (IRX-2) is a natural cytokine consisting of multiple components, which may protect human T cells from tumor-induced apoptosis (37). A phase Ib study evaluated the safety and feasibility of pre-operative local IRX-2 treatment in early breast cancer and assessed intratumoral and peripheral immunomodulatory activity (38). The results suggested that IRX-2 was safe and effective in treating early breast cancer. The use of IRX-2 was associated with favorable changes in the TME, including PD-L1 upregulation and expansion of TILs. It was also indicated that modulation of TILs may be a potential strategy to improve TME immune regulation (38). There is an ongoing trial evaluating the clinical and immunological activity of pembrolizumab plus chemotherapy in combination with an IRX-2 induction regimen as a neoadjuvant therapy for TNBC (NCT04373031). Another clinical trial using IRX-2 with cyclophosphamide as a neoadjuvant treatment for TNBC is also underway (NCT02950259).

Monocytes differentiate into two subtypes of TAM, namely M1 and M2. The M1 subtype mainly has an anti-tumor effect, while the M2 subtype is related to tumor progression and immunosuppression (39) (Fig. 1). Under normal physiological conditions, the polarization of M1 and M2 phenotypes is relatively stable. However, in a tumor state, the TME promotes the differentiation of monocytes to the M2 phenotype through different factors (40). Targeting the TME to regulate the number and function of TAMs is a promising immunotherapy strategy.

Macrophage-colony stimulating factor 1 (CSF-1) regulates the migration, proliferation, function and survival of macrophages and is an important regulator of macrophage homeostasis in vivo (41,42). When CSF-1 is artificially reduced by using CSF-1 blocking antibodies or CSF-1 receptor (CSF-1R) inhibitors, the number of macrophages is rapidly reduced (43,44). LY3022855, an antibody against CSF-1R, exhibited immunomodulatory effects in advanced TNBC (45). In a mouse tumor model of breast cancer, macrophage depletion with the CSF-1R inhibitor PLX3397 enhanced tumor infiltration by CD8 T cells and enhanced the efficacy of anti-PD-1 immunotherapy (46). Cabiralizumab is also an antibody that inhibits the CSF-1R and blocks the activation and survival of macrophages. A clinical trial examining whether the use of neoadjuvant therapy with the addition of Cabralizumab and Nivolumab in combination with chemotherapy reduces TAMs, increases TILs and improves clinical outcomes in patients with early-stage TNBC is currently underway (47).

The TME has a double-edged sword role in TNBC immune regulation and tumor progression, as it is both an immunosuppressive factor and an immune response factor (48). The TME is constantly changing in cancer progression. Ways to exert the immune activation response of the TME and reduce the tumor-promoting effect require to be discovered. Given the antitumor effects of TILs, it may be feasible to increase the number of TILs in patients. For instance, the use of adoptive cell therapy to enhance TILs in patients is a potentially effective immune strategy. Highly active TILs are isolated and infused into patients for therapy after large-scale expansion and activation in vitro. In addition, controlling the direction of differentiation of TAMs so that more TAMs differentiate towards the M1 subtype is also worthy of further exploration.

A study successfully distinguished four different spatial architectures of the TME [immune desert (ID), margin-restricted (MR), stroma-restricted and fully inflamed] in order to better understand the interactions between TNBC and the TME. This study showed that tumors with low CD8+ T cell expression (including MR and ID TME subtypes) had the worst prognosis (49). Identifying the different subtypes of TME allows us to stratify TNBC, predict outcomes and determine potential treatment goals for TNBC.

PD-1 is a co-inhibitory receptor expressed on immune cells, including T cells, B cells, dendritic cells, NK cells and TILs (54). PD-1 has two known ligands, PD-L1 and PD-L2 (55). Given that PD-L1 is more widely expressed than PD-L2 in normal and tumor cells, extensive research is being dedicated to exploring the role of PD-1/PD-L1 in the physiological and pathological immune responses and how to modify their interactions to treat cancer (56). The interactions between PD-1 and PD-L1 occur in the TME and PD-1 is highly expressed on activated T cells, while the PD-L1 is expressed on certain types of tumor cells and antigen-presenting cells (APCs). The interaction between PD-1 and PD-L1 inhibits the biological functions of T cells, including lymphocyte proliferation, cytokine secretion and CTL cytotoxicity, leading to tumor-specific T-cell failure and apoptosis, allowing tumor cells to evade T-cell immune surveillance (57).

It has been indicated that high PD-1+ immune cell infiltration was associated with significantly reduced patient survival (58). Of note, the PD-L1 expression rate in patients with TNBC was higher than that in other types of breast cancer (59). PD-L1 expression was observed in 20% of TNBC cases, meaning that targeting PD-1 or PD-L1 may benefit patients with TNBC (59). PD-1/PD-L1 inhibitors are able to specifically attenuate the inhibitory effects of PD-1/PD-L1 on activated antitumor T cells. The development of PD-1/PD-L1 inhibitor drugs has become a hot area of research on TNBC therapies.

Pembrolizumab, a PD-1 inhibitor, has been mainly used to treat melanoma (60). The feasibility of using pembrolizumab for TNBC was first proposed in 2014 at the San Antonio Breast Cancer Symposium (61). The KEYNOTE-012 study confirmed the safety and anti-tumor activity of single-agent pembrolizumab use in advanced TNBC expressing PD-L1 (62). The subjects included in the KEYNOTE-086 phase II trial were divided into cohorts A and B. Cohort A included patients with advanced metastatic TNBC (mTNBC) who had previously received treatment, while cohort B included patients with mTNBC who had not received any treatment. All patients in cohort B had PD-L1 positive tumors, while 61.8% of patients in cohort A had PD-L1 positive tumors. The results demonstrated that pembrolizumab had significant efficacy as a first-line treatment for PD-1+ mTNBC. The safety was acceptable, as patients in both cohorts were able to tolerate pembrolizumab.

In the KEYNOTE-355 clinical trial, pembrolizumab was used in combination with chemotherapy as a first-line treatment for patients with locally recurrent inoperable or metastatic TNBC. The results indicated that the combination of pembrolizumab and chemotherapy significantly improved progression-free survival (PFS) in patients with tumors expressing higher PD-1 compared with the combination of placebo and chemotherapy. The median PFS (mPFS) of pembrolizumab combined with chemotherapy was determined to be 9.7 months, and it was 4.1 months longer than that in the group treated with placebo combined with chemotherapy, which was 5.6 months (19).

Of note, as tumor PD-L1 expression increased, the therapeutic effect of pembrolizumab also increased. Furthermore, the duration of response to pembrolizumab was also prolonged with increased tumor PD-L1 expression, suggesting that the clinical benefit of pembrolizumab may be related to PD-L1 expression. In addition, compared with chemotherapy, pembrolizumab had less high-grade toxicity (63). The conclusion that may be drawn from these results is that there is a greater benefit of using an immunosuppressant as a first-line treatment option for advanced TNBC as compared with chemotherapy alone.

However, the KEYNOTE-119 trial indicated that compared with chemotherapy, pembrolizumab monotherapy as a second-line or third-line treatment for metastatic TNBC did not significantly improve overall survival (OS). This difference may have resulted from the development of drug and immune resistance in patients who received a previous systemic therapy (64).

A systematic review indicated that the PD-1 inhibitor pembrolizumab was effective in patients with early as well as advanced TNBC independent of the PD-L1 status of the cancer. The addition of pembrolizumab to chemotherapy in early TNBC was more effective than chemotherapy alone (65). The KEYNOTE-173 study reported that using pembrolizumab combined with chemotherapy in the neoadjuvant treatment of early high-risk TNBC had controllable toxicity and good anti-tumor activity. However, limitations of this study included short follow-up times, a small sample size per group and the lack of a control group (66).

To further determine the benefits of pembrolizumab and chemotherapy in neoadjuvant treatment for TNBC, the I-SPY2 study was conducted as a validated randomized phase 3 neoadjuvant clinical trial. Among the 249 patients enrolled, 69 received pembrolizumab combined with paclitaxel and 180 received paclitaxel combined with doxorubicin (Adriamycin) and cyclophosphamide, the standard neoadjuvant chemotherapy regimen. The results suggested that adding pembrolizumab increased the pCR rate from 20 to 60% in patients with TNBC, accounting for an improvement of 40% in comparison with the standard treatment (67). The KEYNOTE-522 trial expanded the number of study subjects. A total of 1,174 patients with TNBC received neoadjuvant therapy, of which 784 received pembrolizumab combined with chemotherapy and the other 390 received placebo and chemotherapy. The results indicated that in the neoadjuvant treatment of early TNBC, the pCR rate of patients who received pembrolizumab combined with chemotherapy was 64.8% and was significantly higher than the rate of 51.2% in those patients who received placebo combined with chemotherapy. Of note, this result was independent of PD-L1 expression (68,69). Based on these results, in July 2021, the FDA approved pembrolizumab combined with chemotherapy as a neoadjuvant treatment for early high-risk TNBC, with continued pembrolizumab monotherapy as adjuvant therapy after surgery.

The fourth interim analysis of the KEYNOTE-522 trial indicated that the median follow-up time was 39.1 months. The estimated 36-month event-free survival (EFS) was 84.5% (95% CI: 81.7-86.9) in the pembrolizumab + chemotherapy group and 76.8% (95% CI: 72.2-80.7) in the placebo + chemotherapy group. EFS was significantly improved in the pembrolizumab + chemotherapy group compared with the placebo + chemotherapy group. Treatment-related events or death occurred in 123 patients (15.7%) in the pembrolizumab + chemotherapy group and in 93 patients (23.8%) in the placebo + chemotherapy group (hazard ratio, 0.63; 95% CI: 0.48-0.82) (70).

Atezolizumab is a PD-L1 inhibitor that was proven to have the ability to delay disease progression in advanced metastatic TNBC (71). At present, clinical trials of neoadjuvant atezolizumab are ongoing. In the neoadjuvant therapy in TNBC with anti-PD-L1 (NeoTRIPaPDL1) study (NCT02620280), 280 patients with TNBC requiring neoadjuvant treatment were randomly divided into two groups; one received intravenous carboplatin and paclitaxel combined treatment on day one and day eight, and the other group received the same treatment with the addition of atezolizumab. The results indicated that the pCR rate was 43.5% when using atezolizumab, while that in the group without atezolizumab was comparable at 40.8% (odds ratio, 1.11) (72).

IMpassion130 is a phase 3 study with the endpoints of PFS and OS. The study was made up of patients with unresectable, locally advanced or metastatic TNBC. The results suggested that in the TNBC population with positive PD-L1 expression, atezolizumab combined with albumin-bound paclitaxel significantly improved PFS and OS compared with placebo plus albumin-bound paclitaxel. However, no such improvement was observed in a population of patients with TNBC with negative PD-L1 expression (20,73). In March 2019, atezolizumab received accelerated approval from the US FDA for use in combination with albumin-bound paclitaxel in patients with unresectable locally advanced or metastatic TNBC whose tumors express PD-L1. However, the IMpassion131 study indicated that the combination of atezolizumab and paclitaxel did not improve PFS or OS compared with paclitaxel alone (74). Based on this result, Roche voluntarily withdrew atezolizumab for the treatment of PD-L1-positive metastatic TNBC. The differences observed in Impassion130 and Impassion131 may be caused by the following factors: i) Paclitaxel, not albumin-bound paclitaxel, require to be pre-administered steroids, which may have reduced the efficacy of atezolizumab; ii) paclitaxel and albumin-bound paclitaxel had different effects on tumor-infiltrating macrophages and TILs, and albumin-bound paclitaxel may have had a stronger stimulating activity on T lymphocytes; iii) the study lacked reliable biomarkers to predict the benefits of atezolizumab (75).

The IMpassion031 study was a randomized, double-blinded, multi-center, phase 3 trial. A total of 333 patients with early-stage TNBC were randomly assigned to the atezolizumab plus chemotherapy group (n=165) or the placebo plus chemotherapy group (n=168). The results indicated that the median follow-up time of the atezolizumab plus chemotherapy group was 20.6 months [interquartile range (IQR), 8.7-24.9 months]. In comparison, the median follow-up time of the placebo plus chemotherapy group was 19.8 months (IQR, 8.1-24.5 months). In patients in the atezolizumab plus chemotherapy group, pCR was observed in 95 patients (58%; 95% CI: 50-65%). By contrast, in the placebo plus chemotherapy group, pCR was observed in 69 patients (41%; 95% CI: 34-49%) and the rate difference was 17% [95% CI: 6-27%; one-sided P=0.0044]. In the PD-L1-positive population, 53 of 77 (69%, 95% CI: 57-79%) patients in the atezolizumab plus chemotherapy group achieved pCR compared with 37 of 75 (49%, CI: 38-61%) patients in the placebo plus chemotherapy group [rate difference, 20%; 95% CI: 4-35%; one-sided P=0.021]. The addition of atezolizumab did not compromise the ability to receive chemotherapy. Commonly reported adverse events were similar between the groups and were mostly driven by chemotherapy. Regardless of the PD-L1 status, patients had achieved an improved pCR. The IMpassion031 study provided evidence for the benefits of ICIs in early TNBC (76).

Durvalumab is another PD-L1 inhibitor. GeparNuevo (NCT02685059) was a multi-center, prospective, randomized, double-blinded, placebo-controlled phase II clinical trial. The study consisted of 174 patients with early-stage TNBC that received durvalumab or placebo combined with chemotherapy. The results failed to prove a significant difference in pCR between the two cohorts. Of note, in an unplanned analysis, the cohort of patients who received additional durvalumab therapy prior to surgery achieved a higher pCR (61.0 vs. 41.4%) and it was indicated that immunotherapy enhanced the anti-tumor activity of cytotoxic drugs (77).

CTLA-4 is a co-inhibitory receptor only expressed on T cells (78). As a homolog of CD28, CTLA-4 is able to bind with a higher affinity to CD80 and CD86 (79). The immune system requires CTLA-4′s checkpoint function to prevent uncontrolled immune responses and autoimmune responses (50). Two signals are required for T-cell activation; the first is the presentation of an antigen by major histocompatibility complex I and II on APCs to be recognized by T-cell receptors (TCR) (80) and the second signal is the binding of CD28 receptors on T cells to the B7 ligand on the APC. However, the presence of CTLA-4 competes with CD28 for the B7 binding site, inhibiting the second activation signal, resulting in the failed activation of T cells (Fig. 2) (81,82). Allison won the Nobel Prize for proposing to fight cancer by blocking CTLA-4 (83). The high expression of CTLA-4 in breast cancer cells is associated with poor prognosis (84,85). Among all types of breast cancer, the expression of CTLA-4 is highest in TNBC (86). Inhibition of CTLA-4 is able to prevent T-cell suppression and enhance the anti-tumor activity of T cells, which is a potential immune-related target for treating TNBC. The US FDA has approved the combination of the anti-CTLA-4 antibody ipilimumab with the anti-PD-1 antibody nivolumab for treating cancers, such as melanoma, lung cancer and colorectal cancer (87,88). There are currently no CTLA-4 inhibitors approved to treat breast cancer.

Compared with the rapid development of PD-1/PD-L1 inhibitors to treat TNBC, CTLA-4 inhibitor development has been slow due to the severe side effects caused by CTLA-4 inhibition. A randomized, double-blinded, phase 3 trial enrolled 906 patients undergoing complete resection of stage IIIB, IIIC or IV melanoma. Half of the patients received nivolumab and the other half received ipilimumab. Severe (grades 3-4) immune-related adverse events (irAEs) were reported in 14.4% of patients in the nivolumab group and 45.9% in the ipilimumab group; 9.7 and 42.6% of patients, respectively, discontinued treatment due to the adverse events. At >100 days after treatment, two deaths (0.4%) related to toxic effects were reported in the ipilimumab group (89). A clinical trial (NCT03546686) is underway to determine the effects of pre-operative treatments of cryoablation, ipilimumab and nivolumab on the 3-year EFS of females with TNBC after taxane-based neoadjuvant chemotherapy, but the safety of CTLA-4 inhibitors also requires to be studied.

OV immunotherapy is a cancer treatment that directly lyses tumor cells through selective viral replication in tumor cells, leading to the release of soluble antigens, danger signals and type I interferons (IFNs), thereby driving the collective anti-tumor immune reactions (99,100). In a tumor mouse model, a study observed that the immune stimulation of type I IFNs produced after OV treatment was necessary for the optimal induction of anti-tumor immunity and increased the number of tumor-specific regulatory T cells. It has been observed that in colon and ovarian cancer models, OV attracted effector T cells into the tumor and induced PD-L1 expression on cancer cells (101).

An effective T cell-mediated adaptive anti-tumor immune response requires two phases; the initial phase is marked by the production of anti-tumor T cells and the effect phase is marked by the destruction of cancer cells by T cells. Anti-PD-1/PDL-1 monoclonal antibodies have been indicated to work during the effect phase (102). Clinical studies using ICIs suggested that patients with existing anti-cancer immunity exhibit the strongest response (103). Therefore, a good strategy may be to simulate an anti-tumor immune response prior to applying ICIs. It has been indicated that intratumoral injections of immune agents, referred to as intratumor immunotherapy, may activate tumor-infiltrating T cells and produce an anti-tumor response (104). Intratumor immunotherapy avoids off-target toxicities and adverse effects that may accompany overall immune stimulation, since they are injected directly into the tumor. In addition, compared with systemic administration, direct intratumor immunotherapy achieved locally high concentrations of the drugs and improved the bioavailability of the immune-stimulating drugs. Therefore, those drugs are only required at low doses to induce local and systemic anti-tumor responses (102).

OVs and Newcastle disease viruses (NDVs) induced inflammatory responses when used in local intratumoral therapy for melanoma, resulting in lymphocytic infiltration and antitumor effects in tumors distant from the original injection site, without distant viral spread. Combination therapy of local NDV and systemic CTLA-4 inhibitors prevented tumor recurrence in hypoimmunogenic tumor models (105).

CF33-hNIS-ΔF14.5 is another oncolytic poxvirus. Researchers established a TNBC mouse model and injected the virus directly into the tumor. They observed that the virus increased the infiltration of CD8+ T cells into the tumor. In mice treated with a combination of virus and anti-PD-L1 antibodies, the immune regulation was significantly increased. Although CF33-hNIS-ΔF14.5 and anti-PD-L1 antibodies failed to exert a significant anti-tumor effect as a single drug, the combination of the two drugs resulted in significant anti-tumor effects. In addition, the mice did not grow new tumors after being injected with the same cancer cells after the combined treatment, indicating that immunity was developed to these cancer cells. This study demonstrated that CF33-hNIS-ΔF14.5 had a beneficial role in regulating immune cells in the TME and increased the response to PD-L1 ICIs (106).

The injection of the oncolytic MARABA virus (MG1) into TNBC tumors changed the local TME. In primary tumors, the combination of MG1 and ICIs significantly slowed the tumor growth in the TNBC model compared with the untreated or monotherapy groups (P<0.0001). After injecting MG1 directly into tumors in the fat pad, there was a reduction in lung tumor nodules, demonstrating that locally administered virus was able to reduce metastatic progression (103). The above results provided a strong theoretical basis for clinical research on combination therapies for TNBC. In addition, the combination of intratumor immunotherapy and chemotherapy in preclinical models achieved significant anti-tumor effects.

Intratumoral injections of LTX-315, an oncolytic virus, combined with doxorubicin, were administered to TNBC and a significant anti-tumor effect compared with the monotherapy was observed (P≤0.05). Imaging techniques and histological examination indicated that the combination induced strong local necrosis, followed by increased CD4+ and CD8+ immune cell infiltration into the tumor parenchymal tissue, and indicated that the TME had been remodeled after the combination treatment of LTX-315 and CAELYX^® (107).

A clinical study using the oncolytic virus Talimogene Laherparepvec (T-VEC) combined with neoadjuvant chemotherapy in the treatment of TNBC (NCT02779855), and a clinical study of T-VEC combined with neoadjuvant immunotherapy in the treatment of TNBC (NCT04185311) are ongoing.

Cancer vaccines are designed to enhance the ability of the immune system to recognize and kill cancer cells and this is accomplished by injecting cancer-specific antigens into patients to trigger an immune response against the tumor (108). DNA and peptide vaccines are two well-studied types of anti-cancer immunotherapy (109). Synaptonemal complex protein 1 (SYCP1) and acrosin binding protein (ACRBP) are two well-known cancer/testis antigens that potently activate cellular and humoral immune responses against 4T1 murine mammary tumors. A study using combination immunotherapy of polyepitopic DNA and peptide cancer vaccines consisting of SYCP1 and ACRBP in a 4T1 breast cancer animal model indicated that combined immunization significantly inhibited the growth of murine triple-negative breast tumors (110).

In vivo studies revealed that nanoparticle (NPs)-based mRNA vaccines targeting the mannose receptors on DCs successfully expressed tumor antigens in lymph node DCs. NP vaccines induced a strong and antigen-specific cytotoxic T-lymphocyte response against TNBC 4T1 cells. The combination of vaccine immunotherapy and anti-CTLA-4 monoclonal antibody significantly enhanced the anti-tumor immune response compared with the monoclonal antibody-only group (111). Whether the vaccine is feasible in combination with other immunotherapies, or even chemotherapy, requires further study.