Current and future immunotherapy for breast cancer

Abstract

Substantial therapeutic advancement has been made in the field of immunotherapy in breast cancer. The immune checkpoint inhibitor pembrolizumab in combination with chemotherapy received FDA approval for both PD-L1 positive metastatic and early-stage triple-negative breast cancer, while ongoing clinical trials seek to expand the current treatment landscape for immune checkpoint inhibitors in hormone receptor positive and HER2 positive breast cancer. Antibody drug conjugates are FDA approved for triple negative and HER2+ disease, and are being studied in combination with immune checkpoint inhibitors. Vaccines and bispecific antibodies are areas of active research. Studies of cellular therapies such as tumor infiltrating lymphocytes, chimeric antigen receptor-T cells and T cell receptor engineered cells are promising and ongoing. This review provides an update of recent major clinical trials of immunotherapy in breast cancer and discusses future directions in the treatment of breast cancer.

Background

One function of the immune system is to eliminate nascent malignant cells. A hallmark of malignancy is the ability of cancer cells to either actively suppress or escape detection by the immune system. Cancer cells alter their tumor immune microenvironment (TIME), creating an immunosuppressive environment through various mechanisms such as signaling changes, restriction of antigen recognition, and induction of T cell exhaustion [1]. Immune checkpoint proteins, such as programmed-death receptor (PD-1) on T cells, B cells and antigen-presenting cells, PD ligand (PD-L1) on tumor cells, or cytotoxic T-lymphocyte associated protein 4 (CTLA4) are essential components in the host immune response to tumor cells [2]. Mutations in these proteins essentially “turn off” the immune system response to malignant cells. The function of immunotherapy delivered through immune checkpoint inhibitors (ICIs), then, is to restore the normal immune response, allowing the host immune system to destroy malignant cells. ICIs include monoclonal antibodies targeting PD-L1 (atezolizumab, avelumab, and durvalumab), PD-1 (pembrolizumab and nivolumab) or CTLA-4 (Ipilumumab) [3]. Lymphocyte Activation Gene 3 (LAG-3), a recently-discovered immune checkpoint which is found on T cells that have been activated but exhausted, has also been targeted by immunotherapy [4]. In addition, cellular therapies offer a distinct mechanism of immunotherapy with the goal of inciting a durable immunologic response to cancer antigens.

The role of immunotherapy in breast cancer has expanded in the past two decades. Breast cancer was classically considered to be immunologically “cold,” unresponsive to immune manipulations. Ongoing research hopes to identify patients who might have enhanced benefit to immunotherapy based on their individual TIME.

Antibody–drug conjugates (ADCs) are another recent advancement in breast cancer. ADCs utilize monoclonal antibodies conjugated to cytotoxic agents (referred to as the “payload”) at higher concentrations into antigen-expressing tumor cells [5]. This has the benefit of targeting traditional chemotherapies to malignant cells, reducing the risk of off-target systemic toxicities. Some ADCs exhibit the bystander effect, in which cells adjacent to the antigen-expressing tumor cell are also killed by the payload [6]. Common ADC targets in breast cancer include TROP2, HER2, and HER3. While ADCs are not defined as immunotherapy, this new class of agents has been safely combined with ICIs in clinical trials [7]. Research into combinations of ADCs with ICIs is ongoing.

Cancer vaccines are being investigated for both prevention and treatment of breast cancer. A principal target of cancer vaccines are tumor-associated antigens (TAAs), which are proteins that are expressed in both normal tissue and malignant cells [8]. Examples of TAAs in breast cancer include HER2, Muc-1, and CEA [8, 9]. Cancer vaccines stimulate CD4+ T helper cells and CD8+ cytotoxic T lymphocytes to engage with TAAs to induce a specific adaptive immunologic response and to initiate immunologic memory to protect against further exposure to the antigens. A benefit of cancer vaccines is the low adverse effects associated with treatment, though studies have found minimal clinical activity. Possible mechanisms of resistance include downregulation of activated tumor antigens and antigen-presenting cells by the TIME, as well as the destruction of activated T-cells [10].

The role of immunotherapy in breast cancer has expanded in the past two decades. Breast cancer was classically considered to be immunologically “cold,” unresponsive to immune manipulations. Ongoing research hopes to identify patients who might have enhanced benefit to immunotherapy based on their individual TIME (Fig. 1).

Fig. 1

Mechanisms of Current and Upcoming Immunotherapy in Breast Cancer. Therapies in bold are FDA approved. Created in BioRender. https://BioRender.com/t90y911

Across multiple cancer types, responsiveness to ICIs has been shown to correlate with higher PD-L1 expression on tumor cells, as determined by immunohistochemistry (IHC) [11]. Two of the most common methods of reporting levels of PD-L1 expression are the combined positive score (CPS) and immune-cell score (IC). CPS measures the proportion of PD-L1 positive cells in both tumor and immune cells out of all cells in a tumor sample, as assessed by Dako assay 22C3 and reported as a maximum score out of 100 [12]. IC is defined by the proportion of tumor area occupied by PD-L1-staining immune cells regardless of staining intensity and is assessed by Ventana assay SP142 [12, 13]. Studies of pembrolizumab typically utilize CPS, while studies of atezolizumab most commonly use IC. Concordance between assays is suboptimal, with IC typically reporting a more conservative estimate of PD-L1 positivity [14]. Standard cutoffs for PD-L1 positivity (PD-L1+) include CPS ≥ 1 and IC ≥ 1% [12,13,14]. PD-L1 positivity is used clinically to determine benefit of pembrolizumab in mTNBC [15].

Another biomarker that may predict response to ICIs in breast cancer is the presence of tumor infiltrating lymphocytes (TILs), such as CD4+ T cells, CD8+ T cells, NK cells, and Th1, which cause an anti-tumor effect through IFN-γ signaling [16, 17]. Recent clinical trials suggest an enhanced benefit for immunotherapy in patients with high levels of TILS, but this is not yet a standard biomarker in clinical practice. Biomarkers for immunotherapy response primarily utilized outside of breast cancer include high tumor mutational burden (TMB), defined as the number of non-inherited mutations per million base pairs, as well as deficiencies in mismatch repair genes (dMMR) which causes increase mutations in microsatellites, termed microsatellite instability-high (MSI-H), though these biomarkers are typically employed in gastrointestinal cancers [18,19,20]. Negative biomarkers such as TAMs, MDSc, CD4+ TRegs, B2M, HLA-A deletions and JAK1/2 mutations may decrease a cancer’s response to immunotherapy but are not used in clinical practice [21,22,23].

Immune-related adverse events (irAEs) are thought to be caused by disruptions in self-tolerance, or a non-specific autoinflammatory reaction [24]. The most common irAEs are cutaneous (maculopapular rash, pruritus, or lichenoid dermatitis), followed by gastrointestinal (diarrhea and colitis), then endocrine (hypothyroidism, hypophysitis or adrenal insufficiency) [25]. Rare but fatal irAEs include pneumonitis, myocarditis, hepatitis, and enteritis [25]. Management of irAE typically includes high-dose corticosteroids, with immunosuppressive therapies in severe cases [26]. Permanent discontinuation of immunotherapy is required for all grade 4 irAEs with the exception of endocrinopathies that can be controlled with hormone replacement [26].

Resistance to immunotherapy is a complex process that is possible due to both intrinsic and extrinsic tumor effects. This can occur due to multifactorial changes in the TIME. Causes of resistance include activation of immune checkpoint pathways that suppress any response from cancer cells as well adaptation within the immune recognition cascade to prevent a response by the immune system [27]. These are often referred to as “cold tumors” as they are difficult to be penetrated by immune cells. There is also the possibility of acquired resistance which occurs after an initial response to immunotherapy, and development of genetic mutations leading to resistance. There are also extrinsic factors such as patient characteristics and lifestyle that can lead to resistance as well. Strategies to overcome resistance to immunotherapy is an ongoing investigation. While there is significant improvement in responses with combination dual immunotherapy, addition of targeted therapy, or PD-L1 down regulators/ inhibitors that are being used in other cancers, there needs to be further research in breast cancer to identify ways to overcome immunotherapy resistance [28].

Triple negative breast cancer (TNBC)

TNBC is an aggressive subtype of breast cancer, accounting for 12–15% of all breast cancer in the United States [29]. In comparison to hormone receptor positive (HR+) breast cancer, TNBC is known to have higher levels of PD-L1 expression, with a median PD-L1 CPS of 7.5 and 50% expressing ≥ CPS 10 [30]. In addition, TNBC has higher levels of TILs and TMB, which are associated with an enhanced response to immunotherapy [18, 22, 23]. Over the past decade, clinical trials in TNBC have incorporated ICIs in combination with traditional cytotoxic chemotherapy with favorable results. In the metastatic setting, response to ICIs is dependent on the degree of PD-L1 expression, whereas in the neoadjuvant/adjuvant setting benefit to ICIs is seen regardless of PD-L1 positivity. This is thought to be explained, at least in part, by decreasing PD-L1 expression and TILs over the course of disease progression and with higher burden of disease [31].

Advanced and metastatic TNBC

The first studies of immunotherapy in breast cancer were conducted as monotherapy in treatment-refractory, PD-L1+ advanced/metastatic TNBC (a/mTNBC). The 2016 KEYNOTE-012 phase 1b trial showed an encouraging overall response rate (ORR) of 18.5% for patients with PD-L1 ≥ 1% mTNBC treated with single-line pembrolizumab [32]. This was followed by the phase II KEYNOTE-086 of single-agent pembrolizumab in mTNBC, which found an ORR of 21% when used in the first line for PD-L1 positive disease, and a more modest ORR of 5.7% after the first line in any level of PD-L1 expression [33, 34]. These results suggested a benefit from incorporating ICIs in PD-L1+ disease, with an enhanced benefit earlier in the treatment course. The phase III KEYNOTE-119 then compared single-agent pembrolizumab to chemotherapy of choice (capecitabine, gemcitabine, eribuilin, or vinorelbine) in 622 patients in the second- or third-line setting. While no difference in progression free survival (PFS) or overall survival (OS) was seen, a graded ORR to pembrolizumab was seen with higher PD-L1 CPS, with the most benefit seen in patients with CPS ≥ 10 [35, 36]. In a phase II trial across all solid malignancies with MSI-H or MMR disease including thirteen patients with breast cancer, single-agent pembrolizumab had a median ORR of 30.8% and PFS of 3.5 months [20].

Pembrolizumab in combination with chemotherapy is now the standard first-line therapy for PD-L1+ a/mTNBC based on the KEYNOTE-355 trial. This phase III study of 847 patients compared first-line standard chemotherapy with pembrolizumab or placebo. Patients with PD-L1 CPS ≥ 10 receiving pembrolizumab exhibited a statistically significant improvement in PFS (9.7 months vs. 5.6 months, HR (hazard ratio) 0.65) and OS (23.0 months vs. 16.1 months, HR 0.73) [15, 37], with a trend towards significance for patients with CPS ≥ 1. The PFS benefit was more pronounced in patients receiving paclitaxel or nab-paclitaxel rather than gemcitabine/carboplatin. Patients with a disease-free interval (DFI) of more than 6 months after curative treatment for early-stage TNBC were included. In an exploratory analysis, patients with early recurrence (6–12 months after curative-intent treatment) had less benefit from immunotherapy in combination with chemotherapy (HR 1.44) [15]. The FDA approved pembrolizumab for first-line advanced/metastatic TNBC in 2020 and this regimen is now the standard of care in the first line for patients with PD-L1+ a/mTNBC (Table 1).

Table 1 Current and upcoming clinical trials of immune checkpoint inhibitors in advanced/metastatic triple negative breast cancer

Atezolizumab as monotherapy in the first-line setting for a/mTNBC was found to have an ORR of 24% [38] and an ORR of 39.4% when combined with nab-paclitaxel after 0–2 lines of prior therapy [39]. The phase III IMpassion130 study randomized 902 patients with advanced PD-L1+ TNBC at least one year from curative-intent therapy or de novo metastatic disease to nab-paclitaxel with atezolizumab or placebo. Patients who received nab-paclitaxel with atezolizumab had a significant PFS benefit compared to patients who received nab-paclitaxel with placebo (7.2 months vs. 5.5 months, HR 0.80) in the intention-to-treat (ITT) analysis and in the PD-L1 > 1% subgroup (7.5 months vs. 5.0 months, HR 0.62) [11]. While these results did not translate into a significant OS benefit in the ITT analysis, a median 9.5-month OS benefit was seen in the PD-L1 > 1% subgroup (HR 0.62, significance not evaluated due to hierarchical testing plan) [11].

Atezolizumab in combination with other chemotherapy backbones has found less success. The IMpassion131 phase III trial investigated paclitaxel + atezolizumab or placebo in 651 patients, with no difference in PFS or OS in either the ITT to PD-L1 subgroup [40], though it is noted that the control group experienced an unprecedented OS of 22.8 months, which may have impacted the results. Overall, the reasons for discrepant results between IMpassion130 and IMpassion131 are unclear. Possible explanations include different formulations between paclitaxel and nab-paclitaxel, pretreatment with steroids in the IMpassion131 study, or differences in the tumor microenvironment in each study population [40, 41]. Though the FDA initially approved atezolizumab based on the results of IMpassion 130, the approval was removed after the IMpassion 131 results were reported.

The possible benefit of atezolizumab in early-relapsing TNBC, a high-risk population defined as relapse less than twelve months after last chemotherapy or surgery for early-stage disease, was evaluated in the phase III IMpassion132 study. In this study, 354 patients without prior immunotherapy were randomized to chemotherapy of physician’s choice plus either atezolizumab or placebo. Initially, patients with any PD-L1 status were included, which was later restricted to patients with PD-L1 > 1%. Two thirds of patients had a DFI of < 6 months. No difference in median disease-free interval or OS was seen [42]. These results, in combination with the exploratory analysis of relapse < 12 months in KEYNOTE-355, suggest that some patients with quickly relapsing TNBC may have an intrinsic resistance to immunotherapy. However, as discussed in greater depth below, the standard of care for first-line therapy in early-stage TNBC now includes immunotherapy, and as such few patients will reach the early-relapsed setting without prior immunotherapy.

The Atract1B phase II trial challenged the view that ICIs only have benefit in PD-L1 positive disease. This trial investigated paclitaxel, atezolizumab and bevacizumab (a VEGF-inhibitor) in the first line for advanced TNBC, with 97% of patients having PD-L1 negative disease. Median PFS was 11.0 months, with an ORR of 63%, including thirteen complete responses and 50 partial responses [43]. Bevacizumab in combination with nivolumab and paclitaxel was investigated in the first line of patients with metastatic HR + /HER- or TNBC in the NEWBEAT phase II trial, with an ORR of 70% (59% in TNBC, 74% in HR + /HER-) [44].

Clinical trials of dual ICI therapy in mTNBC have shown some clinical benefit but also raise concerns of higher rates of toxicity. Durvalumab in combination with tremelimumab, an anti-CTLA-4 antibody, was found to have an ORR of 43% in TNBC in a pilot study of 7 patients [45]. The DART/SWOG S1609 phase II trial of ipilumumab with nivolumab found an ORR of 18%, though all patients who had an initial response continued to respond nearly 3 years later. All responders developed adrenal insufficiency [46].

The use of Poly (ADP-ribose) polymerase inhibitors (PARPi) in combination with ICIs for patients with BRCA-mutated disease has shown promise in phase II trials. Olaparib with durvalumab had a 63.3% ORR in patients with heavily pretreated BRCA-mutant HER2-negative metastatic breast cancer (mBC) with an 80% 3-month disease control rate [47]. However, a trial of olaparib with or without atezolizumab in BRCA-mutant a/m TNBC found no PFS or OS benefit for combination therapy but did have more adverse effects [48]. The TOPACIO/KEYNOTE-162 trial evaluated patients with a/mTNBC with any PD-L1 status. Patients were treated with niraparib and pembrolizumab, with higher ORR seen in patients with BRCA-mutated disease (47%) compared to BRCA-wild type (11%), with updated PFS and OS not yet reported [49].

Optimizing maintenance regimens for patients with initial response to chemotherapy is an active area of research. The DORA phase II study evaluated the role of maintenance olaparib with or without durvalumab in patients with aTNBC who responded to platinum-based chemotherapy. Patients experienced a median PFS of 4.0 months versus 6.1 months, with benefit seen regardless of BRCA or PD-L1 status [50]. The KEYLYNK-009 phase II trial investigated the efficacy of maintenance pembrolizumab and olaparib compared to pembrolizumab and chemotherapy in patients with recurrent inoperable or mTNBC who responded to induction pembrolizumab and chemotherapy. No difference in PFS after completion of induction therapy (5.5 months vs. 5.6 months) or OS (25.1 months vs. 23.4 months) was seen, though there was a trend toward improved PFS for patients with BRCA-mutated disease [51]. Interestingly, no improvement in patient-reported outcomes was seen for patients who were maintained on a chemotherapy-free regimen in comparison with standard pembrolizumab and chemotherapy [52]. Another study of maintenance immunotherapy with durvalumab in comparison to chemotherapy found a 7.1 month OS benefit with durvalumab in an exploratory analysis of patients with TNBC [53]. Overall, these studies suggest an emerging role for chemotherapy-free maintenance for patients who have an initial response to chemotherapy.

Sacituzumab govitecan (SG), an ADC consisting of an anti-TROP2 antibody linked to a topoisomerase I inhibitor, was compared to physician’s choice of treatment in the second or third line of a/m TNBC in the phase III ASCENT trial, with a significant improvement in PFS (4.8 months vs 1.7 months, HR 0.41) and a 4.9 month absolute OS benefit (11.8 vs 6.9 mo, HR 0.51), leading to early termination for efficacy [54, 55]. SG gained FDA approval for mTNBC after two prior therapies in April 2020 [54]. SG in combination with atezolizumab is being compared to the IMPassion030 regimen of nab-paclitaxel + atezolizumab in the front line for PD-L1+ a/m TNBC in the MORPHEUS-pan BC trial, with preliminary data showing an encouraging ORR of 76.7% versus 66.7% and immature PFS data of 12.2 months versus 5.9 months (HR 0.27) [7].

Datopotamab deruxtecan (Dato-DXd) is another anti-TROP2 antibody linked to a topoisomerase inhibitor that has shown activity in TNBC. This ADC was investigated in the phase I TROPION-Pan Tumour 01 study, which found an ORR of 31.8% in patients with heavily pretreated a/m TNBC [56]. Dato-DXd is being studied in combination with an ICI in the phase Ib/II BEGONIA trial, with arm 7 of this trial investigating Dato-DXD with durvalumab in the first line for a/m TNBC. Early results found an ORR of 79% ORR, with 47% of patients having an ongoing response at 11.7 months and response seen regardless of level of PD-L1 expression [57]. Other ADCs being investigated in mTNBC include enfortumab vedotin, which found an ORR of 195 and PFD of 3.5 months in heavily-pretreated mTNBC in the phase II EV-202 trial [58]. Another anti-TROP2 ADC of note is Sacituzumab tirumotecan, which was found to significantly improved OS versus chemotherapy of physician’s choice in the second line for a/m TNBC [59].

Early Stage TNBC

Neoadjuvant treatment of early stage TNBC aims to reduce the extent of surgical excision for operable tumors or convert inoperable tumors to operable tumors. The treatment goal is to attain a pathological complete response (pCR), defined as the eradication of invasive cancer from the breast and lymph nodes (ypT0/is, ypN0) at the time of surgery [60]. Patients who achieve pCR experience significantly improved disease-free survival (DFS) and OS outcomes, thus it is a common clinical end point in neoadjuvant trials [61]. Patients failing to achieve pCR are classified according to the degree of residual cancer burden (RCB). The degree of RCB is also prognostic, with higher RCB scores prognostic for worse event-free survival (EFS) [62].

Combining chemotherapy with ICI increases the rate of pCR for patients with TNBC in comparison to standard neoadjuvant chemotherapy regimens. Unlike the metastatic setting in which PD-L1 is predictive of response to ICIs, the development of predictive biomarkers in the neoadjuvant setting is elusive. As such, there is currently no indication for PD-L1 testing outside of clinical trials as the current evidence shows benefit for ICIs for all early TNBC in the neoadjuvant setting, regardless of PD-L1 status.

Pembrolizumab is FDA approved for neoadjuvant therapy of early-stage TNBC. The phase 1b KEYNOTE-173 trial demonstrated safety and preliminary efficacy of pembrolizumab in the first line with neoadjuvant chemotherapy for early-stage, high-risk TNBC, with higher PD-L1 CPS and TILs significantly associated with higher rates of pCR [16, 63]. Data from the phase II I-SPY2 trial established the benefit of pembrolizumab added to standard neoadjuvant chemotherapy. Twenty-nine patients with TNBC > 2.5 cm and any nodal status were included. Patients in the experimental arm were treated with pembrolizumab with weekly paclitaxel followed by dose-dense (dd) doxorubicin and cyclophosphamide (AC), with estimated pCR rates of 60% vs. 22% for patients treated with paclitaxel followed by AC [64]. Another I-SPY2 regimen evaluated paclitaxel with pembrolizumab in an anthracycline-free regimen, but did not reach target pCR rates [65].

The landmark phase III, double-blind KEYNOTE-522 trial evaluated 1174 patients with cT1, N1-2 or cT2-4, N0-2 TNBC with the goal of investigating the efficacy of neoadjuvant and adjuvant pembrolizumab. Patients were randomized 2:1 to receive neoadjuvant pembrolizumab or placebo with paclitaxel and carboplatin (PC) followed by AC or epirubicin and cyclophosphamide (EC) (every-3-week dosing). After surgery, patients in the study group continued pembrolizumab to complete one total year of treatment. The pembrolizumab regimen was associated with a pCR rate of 64.8% vs. 51.2% in the placebo group, representing a treatment difference of 13.6 percent [66]. Moreover, the risk of recurrence was significantly lower in the pembrolizumab group (HR 0.63). More benefit was seen for patients with node-positive disease (treatment difference 20.6% [8.9–31.9] vs. 6.3% [− 5.3–18.2]), with no difference in response based on PD-L1 status. An updated 5-year EFS showed continued benefit for pembrolizumab with EFS rates of 81.2% versus 72.2% [67]. Improved EFS was seen even for patients with RCB-I and RCB-II after neoadjuvant chemotherapy, though patients with RCB-III after neoadjuvant therapy with pembrolizumab had worse 3-year EFS than patients who received placebo (26.2% vs. 34.6%, HR 1.24 [0.69–2.23]) [68]. This decrease in survival was driven by a higher rate of local recurrence. Five-year OS was 86.6% versus 81.7% [69]. Immune-related adverse effects occurred in 33.5% of patients receiving pembrolizumab, most commonly hypothyroidism (15.1%), skin reactions (5.7%) and adrenal insufficiency (2.6%) [70]. Based on this study, the FDA granted approval for neoadjuvant and adjuvant pembrolizumab in 2021. The KEYNOTE-522 regimen is the current standard of care for early stage TNBC.

Given the high rates of pCR and EFS seen with the Keynote-522 regimen of pembrolizumab with PC + AC, the NeoPACT trial evaluated the role of de-escalating anthracyclines in the neoadjuvant setting in an effort to decrease anthracycline toxicities. In this phase II trial, patients receiving carboplatin with docetaxel and pembrolizumab had a pCR rate of 58%. Patients achieving pCR experienced an impressive 3-year EFS of 98%, with 3-year EFS 86% overall [71]. The currently-enrolling phase III SWOG2212 / SCARLET (NCT05929768) trial will compare the KEYNOTE-522 regimen with the NeoPACT regimen in patients with T2-4, N0, M0 or T1-3, N1-2, M0 TNBC with a primary endpoint of EFS, with the goal to establish an optimal chemotherapy backbone.

Neoadjuvant atezolizumab has also shown promise in TNBC. Neoadjuvant atezolizumab demonstrated a pCR benefit when added to an anthracycline-free regimen of carboplatin and paclitaxel in a phase II trial of 67 patients [72]. The phase III IMpassion031 study of 333 patients with cT2-4, N0-3 TNBC found a significant pCR benefit for neoadjuvant atezolizumab with nab-paclitaxel followed by ddAC (pCR 58% vs. 41%). In the PD-L1 cohort, rates of pCR were significantly increased with atezolizumab (68.8% vs. 49.3%) [73]. EFS data is pending. The NeoTRIPaPDL1 / Michelangelo phase III study evaluated carboplatin and nab-paclitaxel with and without atezolizumab followed by surgery and adjuvant AC in 280 patients, without a pCR benefit seen in the atezolizumab group (48.6% vs. 44.4%) [74]. Unlike the KEYNOTE-522 study, the NeoTRIP study included patients with N3 disease, with 88% of all patients having node-positive disease, which may explain lower overall rates of pCR. A multivariate analysis found PD-L1 expression to significantly increase rates of pCR (OR 2.08, 95% CI 1.64–2.65) [74]. EFS data is pending.

Based on the OS benefit seen with maintenance durvalumab in mTNBC [53], the phase II GeparNuevo trial investigated neoadjuvant durvalumab with chemotherapy. Patients with cT2-4d, N0-3 TNBC were randomized to a window period of either durvalumab or placebo, followed by the same treatment combined with nab-paclitaxel, followed by dd EC with durvalumab or placebo. While there was no significant pCR benefit (53% vs. 44%, OR 1.45 [0.80—2.63] [22], durvalumab did show an increased 3-year DFS (85.6% vs. 77.2%, HR 0.48, [0.24–0.97]), supporting the hypothesis of long-term benefits of early ICI without adjuvant ICI [75]. In a subgroup analysis, TMB > 10% and the presence of TILs predicted treatment response, with pCR rates of 82% seen in patients with high TMB and TILs in comparison to pCR rates of 28% in patients with low TMB and TILs [18]. Data from I-SPY2 supports further investigation of neoadjuvant paclitaxel, durvalumab and olaparib followed by AC in early-stage TNBC, as this trial found pCR rates of 47% compared to 27% in the standard therapy arm [76].

Biomarkers to identify patients who are likely to respond to neoadjuvant immunotherapy are needed. Novel biomarkers such as DetermaIO utilize RNA sequencing to produce a score which predicts pCR of early stage TNBC when treated with immunotherapy [77]. A pooled analysis of 343 patients treated in one of 5 immunotherapy arms of the I-SPY2 trial identified an immune classifier, called ImPrintTN, in hopes of identifying which patients with early-stage TNBC may not benefit from immunotherapy. In the 28% of patients who were ImPrintTN+ , 74% of patients achieved a pCR. In patients who were ImPrintTN-, only 16% of patients achieved a pCR [78]. Further validations of this biomarker is needed, but it suggests that a proportion of patients with early-stage TNBC may be able to avoid immunotherapy when it is unlikely to have a benefit.

Neoadjuvant ADCs in early TNBC is an area of active research. The phase II NeoSTAR trial found a pCR rate of 30% with SG monotherapy [79]. The SOLTI TOT-HER3 trial studied neoadjuvant patritumab deruxtecan as a single dose in a window-of-opportunity phase I trial found an ORR of 35% [80].

The Neo-N phase II trial investigated the effect of lead-in vs. concurrent neoadjuvant immunotherapy. Patients with early-stage TNBC were randomized to A) lead-in nivolumab followed by nivolumab with carboplatin and paclitaxel, or B) up-front nivolumab, carboplatin and paclitaxel followed by nivolumab monotherapy. No difference in pCR was seen between the two arms (50.9% vs. 54.5%) [81]. Notably, 66.7% of patients with high TILs and 70.6% of patients with PD-L1 positive disease achieved pCR, delineating a potentially efficacious anthracycline-sparing neoadjuvant regimen [81]. EFS data is pending.

Dual neoadjuvant ICI therapy with combined PD-1 and CTLA-4 agents in early-stage TNBC has been investigated in two phase II trials. The BELLINI trial treated 31 patients with TILs ≥ 5% with neoadjuvant nivolumab ± ipilumumab followed by chemotherapy or surgery. Evidence of immune activation (defined as doubling of CD8 + T cell or IFN-γ) was seen in 58% of patients [82]. Of the three patients who underwent surgery without neoadjuvant chemotherapy, one pCR and one near-pCR was seen. All patients with radiographic response had TILs > 40% [82]. The CHARIOT trial is a phase II, single arm trial of patients with stage III TNBC with RCB  ≥ 15 mm or ≥ 10 mm with node-positive disease after neoadjuvant AC. Patients were treated with neoadjuvant paclitaxel, ipilumumab and nivolumab followed by adjuvant nivolumab. In this high-risk population, overall pCR rates were 24.4%, with pCR rates of 44.4% in the PD-L1 positive subset [83]. Recently presented EFS and OS data showed a remarkable 100% 3-year EFS and OS in the PD-L1 and/or TIL high subset, even though the minority of patients achieved a pCR [83, 84].

Investigations into adjuvant ICIs in TNBC have had limited success. The phase III IMpassion030 / ALEXANDRIA study evaluated the effect of adjuvant atezolizumab with paclitaxel followed by atezolizumab with AC or EC compared with chemotherapy alone in 2199 patients with stage II-III TNBC who underwent upfront surgery. After a median follow up of 25.3 months, the study was halted after a futility analysis found the study was unlikely to meet its primary endpoint of improved iDFS vs. chemotherapy with a HR of 1.12 (0.87–1.45) [85]. When this study was designed, it was unclear if neoadjuvant vs. adjuvant chemotherapy with immunotherapy would provide better outcomes for patients with TNBC. Mounting evidence now points towards focusing on upfront systemic treatment including immunotherapy for early-stage TNBC, with a tailored approach to adjuvant immunotherapy (Table 2).

Table 2 Current and upcoming clinical trials of immune checkpoint inhibitors in early triple negative breast cancer

Vaccines in TNBC and across subtypes

The majority of breast cancer vaccine clinical trials have focused on the metastatic setting. The Theratrope trial was a phase III clinical trial of 1028 patients with any subtype of mBC who were treated with sialyl-TN, a Muc1 epitope, conjugated to keyhole limpet hemocyanin (KLH) protein vs. KLH alone, with a primary endpoint of time to progression [86]. Patients were given a priming dose of cyclophosphamide 3 days prior to vaccine administration. Though the treatment group did produce anti-mucin antibodies, no difference in time to progression or overall survival was seen [86]. Adagloxad simolenin, a Globo-H epitope conjugated to KLH with cyclophosphamide, was evaluated in a phase II trial of patients with mBC [87]. No difference in overall PFS was seen in comparison to the placebo group, though patients who achieved higher anti-GloboH titers did have an improved PFS [87]. The upcoming phase III GLORIA (NCT03562637) trial will investigate this vaccine in patients with TNBC expressing Globo-H in the adjuvant setting [88].

Virus-based breast cancer vaccines have also been investigated in phase I and II studies, with an objective of infecting antigen-presenting cells to enhance the immunologic response against malignant cells. PANVAC is a poxvirus vaccine that encodes transgenes for Muc-1, CEA, and three co-stimulatory molecules [9]. Phase I trials of PANVAC demonstrated safety and immunoreactivity, with one patient with mBC achieving a complete response [9]. The phase II study combined PANVAC with docetaxel, with a nearly-significant improvement in PFS (7.9 months vs. 3.9 months, HR 0.65, p = 0.09) compared to docetaxel alone [89]. While most viral vaccines have focused on heavily pretreated patients, two vaccines have moved into the neoadjuvant space. Pelareorep, a type III reovirus, in combination with paclitaxel, was found to nearly-significantly increase overall survival in the metastatic setting (17.4 months vs. 10.4 months, HR 0.65, p = 0.1) [90]. Pelareorep is now being studied in the neoadjuvant setting in early-stage TNBC, with preliminary data demonstrating effective priming of an adaptive immune response [91]. A recent phase II study of neoadjuvant intra-tumoral talimogene iaherparepvec (T-VEC) + paclitaxel followed by AC in stage II-III TNBC found a pCR rate of 45.9%, with a 2-year DFS rate of 89%, with no recurrences in patients with RCB 0 or 1 [92].

Upcoming clinical trials in TNBC

In the a/m setting, multiple phase III studies are evaluating ADCs with ICI in the first line. For patients with PD-L1+ disease, the upcoming ASCENT-04 (NCT05382286) trial will investigate first-line SG with pembrolizumab vs. treatment of physician’s choice with pembrolizumab for PD-L1+ disease. The TROPION-Breast-05 (NCT06103864) trial will assess Dato-DXd with durvalumab against the KEYNOTE-355 regimen (chemotherapy + pembrolizumab). These studies will determine the optimal first-line therapy for patients with PD-L1+ a/mTNBC. Atezolizumab in combination with ladoratizimab vedotin (LV) is being studied in in the first line for patients with ICI-naïve a/m TNBC in another arm of the phase I/II MORPHEUS trial (NCT03424005).

For patients with PD-L1- a/m TNBC disease, upcoming phase III trials to determine the optimal first-line therapy include the ASCENT-03 (NCT05382299) trial, which is investigating SG compared to treatment of physician’s choice and TROPION-Breast 02 (NCT05374512), which investigates Dato-DXd against treatment of physician’s choice. Phase II studies such as SACI-IO TNBC (NCT04468061) study will assess SG with or without pembrolizumab in the first line while SNGLVA-002 (NCT03310957) trial will investigate first-line LV with pembrolizumab.

Deescalating neoadjuvant chemoimmunotherapy in early TNBC based on high level of TILs is being investigated in the upcoming phase II NeoTRACT (NCT05645380) trial. Patients with TILs ≥ 5% will be treated with carboplatin, docetaxel and pembrolizumab, while patients with low TILs < 5% will receive the standard Keynote-522 regimen.

Several upcoming trials will clarify the role of adjuvant ICIs, with an emphasis on tailoring therapy based on success of neoadjuvant treatment. The Optimice-pCR (NCT05812807) trial is a phase III trial that will clarify whether patients who achieve pCR can be spared adjuvant immunotherapy. Patients with early-stage TNBC who achieved pCR with combination pembrolizumab and chemotherapy will be randomized to adjuvant pembrolizumab vs. observation, with a primary outcome of recurrence-free survival.

For patients with RCB after neoadjuvant therapy, two upcoming phase III trials will evaluate the benefit of one year of adjuvant ICI vs. observation. The SWOG1418 / NRGBR0006 (NCT02954874) will study pembrolizumab in patients with > 1 cm RCB or positive lymph nodes, with co-endpoints of iDFS overall and in a PD-L1+ subgroup. The A-BRAVE (NCT02926196) trial is studying avelumab in either A) patients who underwent upfront surgery followed by adjuvant chemotherapy, or B) patients with RCB after neoadjuvant therapy.

Combinations of ICI with VEGF inhibition is undergoing further investigation, based on the success of the AtractIB study. The BELLA (NCT04739670) phase II trial is investigating atezolizumab, bevacizumab, carboplatin and gemcitabine in patient with early-relapsing PD-L1+ TNBC, while the MARIO-3 (NCT03961698) phase II trial is investigating atezolizumab, bevacizumab, nab-paclitaxel and eganelisib (A PI3Kγ inhibitor) in the first line for a/m TNBC.

Clinical trials evaluating ADCs, with or without ICIs, in the adjuvant setting are also underway. The SASCIA (NCT04595565) phase III trial is investigating adjuvant SG monotherapy in comparison to treatment of physician’s choice of therapy for patients with residual TNBC or HR+ /HER2- disease after neoadjuvant therapy and surgery. Similarly, the ASCENT-05/Optimice-RD (NCT05633654) study is testing SG with pembrolizumab against treatment of physician’s choice in the adjuvant setting. The TROPION-Breast-03 (NCT05629585) trial is investigating adjuvant Dato-Dxd with or without durvalumab versus standard of care therapy for patients with RCB, including both patients who did and did not receive neoadjuvant immunotherapy. These studies will clarify the benefit of adjuvant ICI for patients with RCB-III disease, given the poor outcomes seen in that subgroup in the KEYNOTE-522 study.

In the neoadjuvant setting, the TROPION-Breast-04 (NCT06112379) trial will compare the BEGONIA regimen of Dato-DXd with durvalumab against the KEYNOTE-522 regimen in the neoadjuvant setting for patients with stage II-III TNBC or HER2-low disease, with the hope of de-escalating toxicities of chemotherapy in the curative setting. Cohort 2 of the NeoSTAR (NCT04230109) trial will investigate SG with pembrolizumab in another de-escalated neoadjuvant regimen.

HR positive, HER2 negative breast cancer

The majority of breast cancers express estrogen and/or progesterone receptors, collectively termed hormone receptor positive (HR+). While HR+ disease has been demonstrated to respond, at least initially, to endocrine therapies, investigations into immunotherapy for the most common subtype of cancer have found less success. This may be explained, in part, by lower expression of PD-L1 and TILs, with only 15% of HR+ breast cancer expressing PD-L1 CPS > 10.[30] There are no FDA-approved immunotherapies in HR + breast cancer to date. Nonetheless, efforts are ongoing to optimize treatment regimens with some encouraging results.

Advanced and metastatic HR+ /HER2- breast cancer

The first trials of immunotherapy in HR+ /HER2- disease evaluated single-agent pembrolizumab in heavily pretreated, PD-L1 CPS ≥ 1 mBC with an ORR of 12%[93]. Combination chemotherapy and immunotherapy was investigated in a phase II trial of eribulin with or without pembrolizumab in patients who had received 0–2 prior lines of chemotherapy and at least two lines of hormonal therapy, which did not show an ORR, PFS or OS benefit.[94].

Following the success of the anti-TROP-2 ADC SG in TNBC, the TROPiCS-2 trial investigated SG vs treatment of physician’s choice in patients with endocrine-resistant HR+ /HER2- MBC in the third or later line. Patients who received SG had a significantly improved PFS (5.5 vs 4.0 months) [95] and OS (14.1 vs 11.2 months) [96], which led to FDA approval of SG in the third or later line of HR+ /HER2- MBC in February 2023 [95].

The SACI-IO HR+ phase II trial investigated SG with or without pembrolizumab in HR+ /HER2- mBC with any PD-L1 status after ≥ 1 endocrine therapy, 0–1 lines of chemotherapy and no prior immunotherapy or ADC in the metastatic setting. No PFS benefit was seen in the combination therapy arm (8.1 months vs. 6.2 months, HR 0.81) with immature OS data suggesting no difference at 12.5 months median follow-up [97]. In the PD-L1 CPS ≥ 1 subgroup, a non-significant 4.4 month increase in PFS was seen with combination SG and pembrolizumab, though OS data is immature [97].

Studies of Dato-DXd in HR+ /HER2- MBC have also shown promising results. The HR+ /HER2- arm of the phase I Pan-Tumour 01 study found an ORR of 26.8% in heavily pretreated patients who received Dato-DXd monotherapy, with a PFS of 8.3 months [56]. The phase III TROPION-Breast 01 trial compared Dato-DXd to physician’s choice of chemotherapy in the second or third line, with an improved PFS of 6.9 months versus 4.9 months (HR 0.63) but no demonstrated overall survival benefit according to a press release [98, 99]. Other ADCs in this space include enfortumab vedotin, which found an ORR of 15.6% with a PFS of 5.4 months in heavily pretreated patients in the phase II EV-202 study [58], as well as patritumab deruxtecan, which found a 3-months RR of 28.6% in the second line in the phase II ICARUS-BREAST-01 trial [100].

The AIPAC study is investigating LAG-3 modulation in breast cancer. This phase II trial randomized patients with HR+ /HER2- mBC that has developed resistance to endocrine therapy to paclitaxel with eftilagimod alpha, a LAG-3 inhibitor, or paclitaxel alone. While there was no significant difference in PFS or OS overall, patients younger than 65 did have a significant 7-month OS benefit, and patients with increased CD8 count 6 months after treatment had significantly improved OS [4].

Given the demonstrated benefit of CDK4/6 inhibition in HR+ disease, trials investigating CDK4/6 inhibitors (CDK4/6i) in combination with ICIs are of interest but have been met with safety concerns. A phase 1b study of abemaciclib with pembrolizumab with or without anastrozole in HR+ /HER2- mBC previously untreated with a CDK4/6i found an ORR of 23% in the first line and an ORR of 29% of patients who had previously received chemotherapy [101]. Rates of grade 3 adverse effects were seen in 69.2% in the untreated group and 60.7% in the previously treated group, with one case of grade-5 interstitial lung disease (ILD) in the first-line setting and higher than expected hepatotoxicity. The NEWFLAME phase II study evaluated nivolumab with abemaciclib and letrozole or fulvestrant in the first or second line of HR+ /HER2- mBC. Though the first 17 patients enrolled experienced an ORR of 54.5% in the letrozole arm and 40.0% in the fulvestrant arm, the trial was stopped early for safety [102]. Over 90% of patients experienced a grade 3 adverse effect, with one ILD-induced grade 5 event in the letrozole arm. Similarly, the Checkmate 7A8 trial of neoadjuvant nivolumab, palbociclib and anastrozole was stopped after 43% of patients discontinued treatment due to adverse events, including hepatotoxicity, neutropenia, rash and ILD [103]. The PACE phase II study of fulvestrant, fulvestrant with palbociclib, or fulvestrant with palbociclib and avelumab was studied in patients with HR + /HER2- mBC who had progressed on prior CDK4/6i and aromatase inhibitor. While it was designed to evaluate the efficacy of continuing CDK4/6i after progression, a non-significant 3.3-month PFS benefit was seen in the fulvestrant with palbociclib and avelumab arm in comparison to fulvestrant alone (HR 0.75) [104]. Grade 3 or 4 adverse events were rare, with no ILD as seen in the above trials with pembrolizumab and nivolumab.

Early-stage HR + /HER2- breast cancer

Though HR + breast cancer overall is associated with a good prognosis, treatment escalation with chemotherapy is indicated for patients with high risk of recurrence. The role for immunotherapy in combination with chemotherapy for certain high-risk subgroups is of active interest. Data from I-SPY-2 showed an improved rate of pCR with pembrolizumab concurrent with paclitaxel followed by doxorubicin and cyclophosphamide (T-AC) (pCR 30% vs. 13%) in HR + /HER2-, MammaPrint high-risk breast cancer with tumor size ≥ 2.5 cm [64]. In the phase III KEYNOTE-756 trial, the benefit of neoadjuvant and adjuvant pembrolizumab in grade 3, high-risk ER+ disease was clarified by comparing the I-SPY-2 regimen to standard T-AC in the neoadjuvant setting, followed by adjuvant endocrine therapy with pembrolizumab or placebo in the adjuvant setting. Seventy-six percent of patients were PD-L1 positive. An 8.5% improvement in pCR rates was seen in the pembrolizumab arm (24.3% vs. 15.6%, p = 0.00005) with a larger benefit seen in patients with node-positive disease, PD-L1 positivity as defined by CPS ≥ 1 (29.7% vs. 19.6%), and ER positivity < 10% [105]. Of note, in this trial clinicians were given the option of every-2-week dosing or every-3-week dosing. EFS data is pending to evaluate the benefit of adjuvant pembrolizumab.

Nivolumab in the neoadjuvant setting has also found success. The GIADA trial evaluated patients with stage II-III luminal B breast cancer, finding a pCR rate of 16.3% after therapy with EC followed by nivolumab, troptorelin and exemestane [106]. The phase III Checkmate 7Fl trial aimed to investigate neoadjuvant and adjuvant nivolumab in 1278 patients with high-risk ER+ /HER2- breast cancer. Patients were randomized to neoadjuvant paclitaxel with or without nivolumab followed by AC, followed by adjuvant endocrine therapy and nivolumab or placebo. A significant pCR advantage (24.5% vs. 13.8%) was seen, with an enhanced advantage in the 35% of patients who were PD-L1+ (44.3% vs. 20.2%). A pCR benefit was also seen in patients with ER < 50%, PR < 10%, and TILS ≥ 1%. This trial confirmed a pCR benefit with immunotherapy in ER+ disease [107]. However, the adjuvant portion of the study stopped enrollment early after adjuvant abemaciclib gained FDA approval based on data from MonarchE because CDK4/6i cannot be safely combined with ICIs due to safety concerns as discussed above. EFS data is pending.

Data from I-SPY2 also suggests that neoadjuvant durvalumab in combination with paclitaxel and olaparib may benefit patients with stage II-III HR+ /HER2- breast cancer. Patients who were high risk for recurrence by MammaPrint (either High-1 or High-2) were included. While no difference in pCR rate was seen in the high 1 group, 64% of patients with high 2 disease achieved a pCR with immunotherapy, compared to 22%s in the paclitaxel control group [76]. This finding suggests that MammaPrint High-2 could be a predictive biomarker for immunotherapy in HR+ /HER2- early breast cancer, though further validation is needed.

ADC in combination with ICI in early-stage HR+ /HER2- breast cancer has found early promising results. Neoadjuvant Dato-DXd with durvalumab for high-risk, early stage HR+ /HER2- breast cancer achieved an overall pCR rate of 50%, with rates of 79% in the immune signature subtype [108] (Table 3).

Table 3 Current and upcoming clinical trials of immune checkpoint inhibitors in hormone receptor positive breast cancer

Overall, there may be a benefit for immunotherapy in early-stage HR+ /HER2- disease, though additional investigation of biomarkers to predict response to immunotherapy is needed given alternative treatment options in this setting.

Upcoming clinical trials in HR + /HER2- breast cancer

The upcoming SWOG S2206 (NCT06058377) phase III trial will clarify the role for neoadjuvant immunotherapy without adjuvant immunotherapy in patients with ER+ /HER2- MammaPrint High-2 disease, who will receive either durvalumab with AC-T neoadjuvant chemotherapy or AC-T alone. LAG-3 inhibition is under further investigation in the phase III trial AIPAC-003 (NCT05747794), which will investigate paclitaxel with or without eftilagimod alpha in patients with endocrine-resistant, HR+ /HER2- mBC or TNBC not eligible for PD-L1 therapy. Adjuvant ADC therapy with SG vs chemotherapy for HR+ /HER2- residual disease is being investigated in the upcoming SASCIA trial (NCT04595565), while SG with or without pembrolizumab is being evaluated in the first or second-line metastatic setting (NCT04448886).

HER2 positive breast cancer

HER2 positivity, defined as IHC 3+ , is seen in approximately 20% of breast cancer, though the majority of breast cancer express HER2 to some degree [109]. HER2+ breast cancer has higher TILs, TMB, and PD-L1 expression than HER2- disease, which may correlate with an enhanced response to immunotherapy [110]. To date, no ICI has improved outcomes in comparison to standard HER2-targeted regimens in a randomized clinical trial, though early results may suggest an improvement in PD-L1 positive disease.

Advanced and metastatic HR-/HER2 + breast cancer

The PANACEA phase Ib/II trial evaluated combination trastuzumab and pembrolizumab in patients with a/m HER2+ breast cancer who had previous progression on trastuzumab. The primary endpoint of ORR in PD-L1+ disease was 15%, with no responses in the PD-L1 negative group [111]. Median PFS did not differ by PD-L1 status (2.7 months vs. 2.5 months), though 12-month OS was numerically higher in the PD-L1+ group (65% vs. 12%) [111]. T-DM1,an ADC which consists of trastuzumab linked to DM1, a cytotoxic microtubule inhibitor, is standard therapy for patients with HER2+ residual disease after neoadjuvant chemotherapy [112].

Pembrolizumab in combination with T-DM1 was investigated in a phase Ib trial in metastatic HER2+ disease, with an ORR of 20% [113]. Atezolizumab in combination with T-DM1 in a/m HER2+ disease was associated with an ORR of 35%, while atezolizumab with trastuzumab, pertuzumab and docetaxel had an ORR of 100% in a phase Ib study [114]. The KATE2 phase II study compared T-DM1 with and without atezolizumab in a/m disease after the first line with no significant improvement in PFS in the ITT analysis, though patients in the treatment arm experienced increased toxicity requiring early unblinding of the treatment arms. However, an exploratory analysis found a trend towards improvement in PFS in patients with PD-L1+ disease [115].

Trastuzumab deruxtecan (T-DXd), a HER2-targeted ADC, has changed the treatment landscape for metastatic breast cancer that expresses HER2 and redefined classification of HER2 expression [109]. The landmark DESTINY-Breast 03 trial showed a dramatic improvement in PFS (29 vs 7.2 months) and OS (39.2 vs 26.5 months) with T-DXd vs T-DM1 in the second-line, with 21.1% of patients in the T-DXd group experiencing a complete response [116,117,118]. An exploratory subgroup of patients with brain metastasis found an intracranial ORR of 65.7% compared to 34.3% [119]. T-DXd is approved for HER2+ MBC in the second line.

Based on these studies, in addition to studies of T-DXd in other solid tumors such as lung and colorectal cancer, T-DXd was recently granted universal FDA approval for any HER2+ a/m solid cancer after one prior treatment [120,121,122].

Other ADCs under investigation in HER2+ MBC include trastuzumab duocarmazine (T-Duo), which was compared to chemotherapy in the third or greater line, or pretreated with T-DM1, for patients with HER2+ MBC in the phase III TULIP trial [123]. An improvement in PFS (7.0 vs. 4.9 months, HR 0.63) and a trend towards improved OS (21.0 vs 19.5 months, HR 0.87 (95% CI 0.68–1.12) was seen.[123].

Advanced and metastatic HER2 low breast cancer

T-DXd was also found to have significant activity in patients who were not traditionally considered to be HER2+ (IHC 3 +) [109, 124, 125]. This redefined the spectrum of HER2 expression from a binary (positive or negative) to a spectrum including HER2-low (IHC 1+ or 2+ , in-situ-hybridization (ISH) negative) and HER2-ultralow (faint, incomplete membrane staining in $\ge$ 10% of tumor cells, less than IHC 1+) [124,125,126]. The DESTINY-Breast 04 phase II trial found T-DXd to have significantly improved PFS (9.9 vs 5.1 months) and OS (23.4 vs 16.8 months) in comparison to physician’s choice of chemotherapy for HER2-low MBC in the second or third line, leading to FDA approval in this setting [127]. Recent results from DESTINY-Breast 06, which evaluated T-DXd in patients without prior chemotherapy for HER2-low or HER2-ultralow MBC, found a 5-month PFS advantage (HR 0.63) in comparison to standard chemotherapy [126]. T-DXd in combination with durvalumab in patients with HER2-low advanced or MBC was investigated in arm 6 of the BEGONIA trial, which found an ORR of 57% with mPFS 12.6 months [128].

Early Stage HR-/HER2+ Breast Cancer

A phase II trial of neoadjuvant chemotherapy followed by pembrolizumab, trastuzumab, and pertuzumab found a pCR rate of 46% [129]. The phase III trial IMPassion050 investigated atezolizumab or placebo with neoadjuvant AC followed by paclitaxel, trastuzumab and pertuzumab, followed by adjuvant atezolizumab or placebo in 454 patients with high-risk HER2+ disease. No difference in pCR was seen in the ITT or PD-L1+ subgroup, with EFS data immature [130]. Adverse events were more common in the treatment group, with two deaths in the immunotherapy arm (alveolitis, septic shock) attributed to the treatment (Tables 4 and 5).

Table 4 Current and upcoming clinical trials of immune checkpoint inhibitors in HER2 positive breast cancer

Table 5 Current and upcoming clinical trials of antibody drug conjugates in breast cancer

HER2+ vaccines

Several HER2-directed short peptide vaccines have been developed, with E75 being the most extensively studied. An optimal regimen of NP-S with GM-CSF monthly for 6 months was determined in a phase I/II study [131], with a trend towards increased 5-year DFS in the vaccinated group vs. control. In the phase III setting, however, no difference in 3-year DFS was seen and the trial was stopped early for futility [132]. HER2 vaccines targeting AE37 have not shown any difference in DFS [133]. GLSI-100, a HER2 vaccine targeting GP2, has been investigated in the adjuvant setting for patients with HER2+ , node-positive or otherwise high-risk disease [133]. Overall, there was no difference in 5-year DFS, but patients who had disease that was HER2 3+ by IHC did have a trend toward significant improvement in DFS [133]. This vaccine is being investigated in the upcoming phase III Flamingo-01 (NCT0523916) study in the adjuvant setting for patients with HER2+ residual disease (Table 6).

Table 6 Current and upcoming clinical trials of vaccines in breast cancer

Upcoming clinical trials in HR-/HER2+ breast cancer

The ASTEFANIA (NCT04873362) phase III trial is investigating T-DM1 with or without atezolizumab in the adjuvant setting for patients with residual disease after neoadjuvant therapy, with the goal to enhance the current standard of care for patients with RCB in the adjuvant setting. KATE3 (NCT04740918) is a phase III trial evaluating the KATE2 regimen in patients with PD-L1+ disease in the first to third line a/m setting. Another phase III trial (NCT03199885) is investigating paclitaxel with trastuzumab, pertuzumab, and atezolizumab or placebo in the first line metastatic setting, with the goal of establishing a role for ICIs in a/m HER2+ breast cancer.

Bispecific antibodies and cellular therapies

Immunotherapies other than immune checkpoint inhibitors seek to induce a durable immunologic anti-cancer response. Areas of research in breast cancer include bispecific antibodies as well as cellular therapies including tumor infiltrating lymphocytes (TILs), chimeric antigen receptor T (CAR-T), and T cell receptor engineered (TCR) treatments.

Bispecific antibodies

Bispecific antibodies are engineered antibodies that simultaneously bind to two targets, typically a tumor-specific antigen on malignant cells and an immune cell [134]. Bringing these targets in close proximity engages the immune system, inciting apoptosis, enhanced immune activation signaling, or reducing immunosuppressive factors, all of which promote anti-tumor activity [134]. The first FDA approval for a bispecific antibody in oncology was blinatumomab, which targets CD3 and CD19, in B-cell acute lymphoblastic leukemia in 2017 [135]. Bispecifics have been an area of active research in breast cancer for over 30 years. TAAs targeted by bispecific therapies in breast cancer research include HER2, HER3, PD-L1, and CD3 [134].

HER2 is the most commonly studied bispecific TAA in breast cancer, in combination with another HER2 antibody, HER3, CD3, and CD16. Zanidatamab, a dual-HER2 bispecific antibody, is efficacious in both advanced and early-stage HER2+ BC. A phase I clinical trial of zanidatamab in advanced HER2+ solid cancers, including breast cancer, found an ORR of 37% with zanidatamab monotherapy [136]. Zanidatamab in combination with docetaxel had an 90.9% ORR in patients with advanced HER2+ mBC [137]. In HR+ /HER2+ patients with advanced disease, the combination of zanidatamab with palbociclib and fulvestrant achieved a median PFS of 11.3 months with an ORR of 34.5% [138]. In the neoadjuvant setting, patients with stage 1, node-negative HER2 + BC were treated with zanidatamab in a chemotherapy-free regimen, with preliminary data showing 36% of patients achieving pCR and 64% of patient achieving pCR or RCB-1 [139]. An upcoming phase III trial (NCT06435429) will investigate chemotherapy with zanidatamab or trastuzumab for patients with HER2+ disease who progress on T-DXd.

KN026, another dual-HER2 bispecific antibody, has found success in a phase I trial of patients with HER2+ mBC. Overall, a 28.1% ORR and median PFS of 6.8 months was found, though patients with HER2+ and CDK12 co-amplification had a significantly improved response (ORR of 50%, median PFS 8.2 months) [140]. This improved response is thought to be because both HER2 and CDK12 genes are located on chromosome 17, approximately 200 kb apart. KN026 in combination with KN046, an anti-PD-L1/CTLA-4 bispecific antibody, found an ORR of 50% in patients with HER2 + MBC [141]. In the neoadjuvant setting, KN026 with docetaxel achieved a pCR rate of 56.7% [142]. A trial of KN026 with palbociclib and fulvestrant (NCT04778982) was terminated due to low enrollment. An upcoming phase I clinical trial (NCT03842085) will investigate another dual HER2 bispecific antibody, MBS301, in HER2+ solid tumors.

Bispecific antibodies targeting HER2 and HER3 include MCLA-128/Zenocutuzumab and MM-111. A phase II trial of zenocutuzumab with endocrine therapy in patients with HR+ , HER2-low MBC previously treated with CDK4/6 therapy found an DCR of 45% [143], while zenocutuzumab with trastuzumab and vinorelbine in HER2+ MBC previously treated with T-DM1 found an DCR of 77% [144]. MM-111 has also proven to be safe in phase I clinical trials [145, 146], though without further investigation in phase II or beyond. HER2 and CD3 bispecific antibodies including ertumaxomab [147], p95HER2 [148], and GBR1302 [149] have been investigated, but are not moving forward in current clinical trials.

Bispecific antibody armed T cells targeting HER2 and CD3, referred to as HER2 BATs, were investigated in a phase II clinical trial of treatment consolidation with immunotherapy after chemotherapy in HER2- MBC [150]. Evidence of immune activation in both the adaptive and innate response were seen, with a median OS of 13.1 months [150]. The role of HER2 BATs in HER2+ disease was evaluated in a phase I clinical trial, which showed that patients with higher HER2+ expression (IHC 3+) had a longer OS in comparison to patients with HER2 0 to 2+ expression (57 months vs. 27 months) [151]. The combination of HER2 BATs with pembrolizumab will be investigated in an upcoming phase I/II study (NCT03272334) (Table 7).

Table 7 Current and upcoming clinical trials of bispecific antibodies in breast cancer

Adoptive cell therapy: TILS, CAR-T, and TCR

While ICIs have made tremendous impact on outcomes in breast cancer, a significant subset of patients do not benefit from ICI. This is likely due, in part, to immune escape through defects in antigen presentation and other abnormal signaling pathways [152]. Mechanisms to overcome this barrier include adoptive cell transfer, which consists of infusing antigen-specific T cells into the patient, inducing an amplified immune response. Therapies such as TILs, CAR-T, and TCR cell therapies involve harvesting T cells from the patient, allowing for ex vivo expansion with or without genetic modification, and infusing these T cells back into the patient. Since the first FDA approval of CD-19-specific CAR-T therapy for B-cell acute lymphoblastic leukemia in 2017 [153], cellular therapy has changed the treatment paradigm in hematologic malignancies.

TIL therapy involves extracting T cells from within a tumor, multiplying them without genetic modification, then infusing them back into the patient. TIL therapy gained FDA approval for relapsed/refractory metastatic melanoma in February 2024 based on an ORR of 31.5% including three complete responses in seventy-three patients treated with lifeleucel, an autologous TIL therapy, in a phase II study (NCT02360579) [154, 155], with four-year follow up confirming a median OS of 13.9 months and 20.8% of patients having an ongoing response at 48 months [156]. This success has prompted investigation of TIL therapy for other solid malignancies, including breast cancer. A phase I trial of TIL therapy with IL-2 and pembrolizumab in breast cancer showed promising results, with three out of six patients having a response including 1 complete response [157]. Upcoming trials of TIL therapy in breast cancer include a phase II trial (NCT04111510) is investigating LN-145 TIL therapy in mTNBC, a phase I/II (NCT05451784) trial of PD-L1+ TILs in TNBC, a phase Ib (NCT05576077) study of TBio4101 with pembrolizumab in solid tumors including breast cancer, and a phase II (NCT03449108) trial of LN-145-S1 with ICI in metastatic TNBC.

CAR-T cell therapy engineers T-cell receptors to target a tumor-specific antigen on the cell surface. CAR-T is a staple in the treatment of hematologic malignancies, and is an area of active research in solid tumors including breast cancer. Phase I studies to date have demonstrated tolerable safety [158,159,160] with some patients experiencing stable disease [161]. A phase 0 trial of intratumorally injected CAR T cells against c-MET, a cell-surface molecule, in patients with mBC showed evidence of an inflammatory response within tumors, but no clinical response [158]. These results led to a phase I study of intravenous CAR-T targeting cMET, with two of four patients with TNBC experiencing stable disease at day 25 [161]. Upcoming studies of CAR-T in breast cancer include a phase I/II (NCT04020575) study targeting metastatic breast cancer expressing Muc1 growth factor receptor, called Muc1*, and a phase I study (NCT02706392) of CAR-T in liquid and solid malignancies expressing ROR1.

It is postulated that the limited efficacy of CAR-T in solid tumors to date may relate to tumor antigen modulation, insufficient specificity of target antigens, T cell exhaustion, poor cell trafficking to the tumor site, and dysregulation of effector function by the TIME [162,163,164]. Limited expansion of CAR-T cells after transfer has limited efficacy while also limited toxicity. Thus, concern remains for higher on-target off-tumor toxicity with greater CAR-T cell expansion, as seen in a patient treated with ERBB2-CAR-T therapy who experienced fatal cytokine release syndrome from on-target, off-tumor effect [165]. In response, it is now understood that CAR-T therapy targets should be restricted to cell-surface antigens with limited off-tumor expression, of which there are few [166]. Other toxicities from CAR-T therapy such as macrophage activation syndrome and immune cell-associated neurotoxicity syndrome also affect normal, healthy tissue.

TCR is similar to CAR-T, but expresses human leukocyte antigen (HLA)-restricted T-cell receptors, which allow for tumor-specific binding to both membrane and intracellular proteins expressed by the major histocompatibility complex (MHC), and thus limits toxicity to other healthy cells. The SPEARHEAD-1 (NCT03132922) phase II trial of afamitresgene autoleucel (“afami-cel”), an HLA-A*02 restricted TCR targeting MAGE-4 in patients with synovial sarcoma or myxoid round cell liposarcoma, found an ORR of 37% and a duration of response of 28.1 months [167]. Afami-cel received FDA approval for MAGE-4 expressing sarcoma in August 2024. Ongoing phase I trials of MAGE-A1 and MAGE-A3 have found tolerable safety and potential clinical efficacy in various solid tumors, without reported results for patients with breast cancer [168, 169]. Another TCR therapy, letetresgene autoleucel (“lete-cel”) targeting HLA-A*02 restricted NY-ESO1, found an ORR of 40% in the IGNYTE-ESO (NCT03967223) phase II trial of patients with synovial sarcoma and myxoid round cell liposarcoma [170]. These results have laid the groundwork for expansion of TCR into other cancer types. In breast cancer, an HLA-A*O2-restricted TCR targeted p53^R175H therapy from peripheral blood lymphocytes was infused into a patient, who experienced a partial response lasting 6 months [171]. Upcoming trials of TCR in breast cancer include a phase Ib trial (NCT05989828) for patients with metastatic TNBC which expresses HLA-A*O2 restricted NY-ESO-1, a phase I trial (NCT05877599) of HLA-A*02 restricted NT-175 in solid tumors including breast cancer, and two phase I studies (NCT05483491, NCT05035407) of HLA-A*01-restricted KK-LC-1 in solid tumors including breast cancer (Table 8).

Table 8 Current and upcoming clinical trials of cellular therapy in breast cancer

Future outlook and conclusions

The use of immunotherapy for the treatment of breast cancer has expanded over the last two decades. Pembrolizumab is a premier example of an immune checkpoint inhibitor that has changed treatment paradigms for TNBC, while ongoing clinical trials investigate the role of pembrolizumab and other immune checkpoint inhibitors in patients with HR+ and HER2+ disease. Antibody–drug conjugates such as sacitizumab govitecan and trastuzumab deruxtecan have ushered in a new generation of anti-cancer therapies, with ongoing studies evaluating new antibody–drug conjugates such as datopotamab deruxtecan as well as antibody–drug conjugates in combination with immune checkpoint inhibitors. With fourteen bispecifics approved for solid tumors to date, further studies of bispecifics such as zanidatamab and zenocutuzmab in breast cancer are needed to clarify the role for these agents in the current treatment paradigm. Future studies should investigate mechanisms to enhance the T cell response, whether through dual checkpoint inhibition and/or costimulatory signals, in both bispecifics and CAR-T therapy. Upcoming clinical trials bring hope for broadening the treatment landscape and further understanding the potential benefit of combination immune checkpoint inhibitors, antibody–drug conjugates, bispecific antibodies, and/or cellular therapies. Further research regarding biomarkers beyond PD-L1 expression and other characteristics of the tumor immune microenvironment will enhance our ability to predict which patients will respond to immune checkpoint inhibitor-containing treatments, allowing for more tailored therapy for each individual patient.