Immunotherapeutic Strategies in Breast Cancer: A Clinical Update

Abstract

Immunotherapy has been incorporated into the standard of care for a wide range of malignancies. The study of tumor-infiltrating lymphocytes has emphasized the importance of the host anti-tumor immune response in the natural history of breast cancer. Recent clinical trials have used immunotherapeutic approaches to augment this response and improve outcomes for patients with breast cancer. Here, we review several current clinical trial data that indicate checkpoint blockade may mediate clinically significant responses.

1. Introduction

Breast cancer is the most common malignancy in women in both the developing and developed world, with its global incidence on the rise[1]. Although improvements in systemic therapy have significantly improved survival[2,3], there are an estimated 270,000 new cases of breast cancer and over 40,000 deaths in the United States in 2019[4].

Recent advances in tumor immunotherapy have been demonstrated in a variety of malignancies[5,6]. However, the response to immunotherapy is not uniform across different tumor types. Notably, malignancies with a high mutational burden, such as melanoma, lung cancer, and bladder cancer, are often very responsive to immunotherapy[7]. However, breast cancer has a relatively low mutational burden[8], which may give pause to the use of immunotherapy in the treatment of breast malignancies. Yet, breast cancer is a heterogenous disease stratified clinically by multiple subtypes with varying immunologic activation[9]. Here, we review the rationale and clinical data supporting the use of checkpoint blockade immunotherapy in breast cancer and their relevant adverse events.

2. Tumor infiltrating lymphocytes in breast cancer

The importance of tumor-infiltrating lymphocytes (TILs) have been demonstrated in many solid tumor types[10], and their presence in the tumor microenvironment is associated with favorable clinical outcomes in breast cancer[9]. Breast tumors and their related stroma contain more T lymphocytes compared to normal breast tissue, where myeloid cells are more prevalent[11]. Different subtypes of breast cancer have varying TIL infiltration, with HER2-positive and triple-negative breast cancer (TNBC) containing higher relative percentages of TILs compared to estrogen receptor (ER)-positive and HER2-negative tumors[12]. Approximately 10% of HER2-positive and TNBCs show lymphocyte predominance[12], displaying 50% or more infiltration of either stromal or intratumoral lymphocytes, compared to approximately 3% of ER-positive breast cancers.

Histologically, TILs can be categorized as intratumoral or stromal, and the quantitation of TILs in one compartment usually correlates with the other[13]. Intratumor TILs are defined as lymphocytes found in tumor nests that have cell-to-cell contact with the tumor without intervening stroma. Stromal TILs are scattered within the stroma in between tumor cells and do not directly make contact with tumor cells. Although both are considered to be true TILs, the majority of TILs are usually found within the stromal compartment[13]. The quantification of TILs can be difficult given the variability in TIL distribution, which was addressed by the TILs working group recommendations published in 2014[13]. The working group recommends reporting TILs for the stromal compartment as a percentage, which should be calculated as the area of TILs over the total intratumoral stromal area. TILs should also be assessed within the borders of the invasive carcinoma and excluded if they are outside of the invasive border, around DCIS or normal breast lobules, or found in areas of crush artifact, necrosis, or previous core biopsy sites. Reporting of TILs should be performed on full sections instead of core biopsies whenever possible, with one section per patient deemed to be sufficient. Average TIL infiltration should be used for quantitation rather than focal areas of TILs. Although TILs should be scored as a continuous variable, in practice, pathologists can round up to the nearest 5%–10%.

In TNBC, improved prognosis correlates with an increasing frequency of TILs. In a pooled analysis of untreated TNBC patients, an increasing percentage of TILs was associated with improvements in invasive disease-free survival (DFS), distant disease-free survival, and overall survival (OS)[14]. Similarly, evaluation of TNBC in the adjuvant setting from the Eastern Cooperative Oncology Group (ECOG) E2197 and ECOG E1199 trials demonstrated a 14% reduction in risk of recurrence or death for every 10% increase in stromal TILs[15]. A pooled analysis of over 2,000 patients from nine adjuvant trials with early-stage TNBC also found a strong correlation between improvement in DFS and each 10% increase in TILs[16]. In an analysis of over 12,000 patients, in both ER-negative and ER-positive HER2-positive tumors, CD8⁺ T cell infiltration within the tumor was associated with a reduction in the hazard of breast cancer-specific mortality of 27%–28%[17].

It has been previously demonstrated that patients who have a complete pathologic complete response (pCR) have improved and OS compared to those whose tumors that do not demonstrate a pCR[18,19]. A study investigating TILs in over 1,000 core biopsies, including all histologic subtypes from two neoadjuvant studies (GeparDuo and GeparTrio), found that the presence of TILs was a predictor of pCR. Lymphocyte-predominant cancers, defined here as tumors with 60% or more TILs, had pCR rates of 40%–42% compared to 3%–7% in tumors without TILs[20]. In HER2-positive tumors from the NeoALTTO trial, where patients were randomly assigned neoadjuvant Lapatinib and/or Trastuzumab, TILs greater than 5% was associated with higher pCR regardless of treatment arm[21]. Interestingly, in those who did not achieve pCR but had 40% or more TILs within their tumors, three-year event-free survival (EFS) was excellent at 97%[21]. A more recent study including 3,771 patients treated with neoadjuvant chemotherapy from six randomized trials from the German Breast Cancer Group demonstrated that higher TILs were predictive of response to neoadjuvant chemotherapy in all molecular subtypes[22]. Although the pCR rate was 28% in TIL high luminal HER2-negative tumors compared to 6% in TIL low luminal HER2-negative tumors, a 10% increase in TILs was associated with shorter OS in this breast cancer subtype[22]. Further elucidation of the immune environment and its interaction with cancer cells and endocrine therapy in luminal HER2-negative breast cancer is needed.

A study of 193 TNBC samples found that those with high lymphocyte counts and good prognosis had a significantly lower mutation and neoantigen counts compared to TNBCs with low lymphocyte counts and poor prognosis[23]. The study authors hypothesized that TNBCs with an abundant presence of immune cells may be under a more rigorous immunosurveillance that effectively eliminates more immunogenic clones. Similarly, a study of intratumor genetic heterogeneity across multiple tumor types, including breast cancer, found that a higher level of intratumor genetic heterogeneity was associated with fewer tumor-infiltrating immune cells and worse survival[24]. There is also evidence that in TNBC, the pCR rate is higher in tumors that have a small number of clones with a high mutational burden per clone[25], suggesting there may be greater immunoediting of subclones in these tumors. This improvement in chemosensitivity could potentially be attributed to effective immunoediting of chemo-resistant clones. PD-1/PD-L1 blockade has been shown to decrease tumor clonal heterogeneity while improving tumor response[26], which is suggestive of opportunities to modulate the immune system to increase tumor immunoediting for improved disease control. Checkpoint blockade may alter the immune environment to allow better response to chemotherapy based on data from the GeparNuevo randomized trial, where an improvement in pCR was seen only in the subset of patients receiving durvalumab (a PD-L1 inhibitor) before receiving neoadjuvant chemotherapy[27]. Whether or not durvalumab is increasing immunoediting of the tumor leading to better responses to chemotherapy remains to be determined. Together, these findings emphasize the potential role for immunotherapy, and particularly modulation of the lymphocyte response to malignancy, in HER2-positive and TNBC.

3. Checkpoint blockade in breast cancer

The success of checkpoint blockade has become a cornerstone of cancer immunotherapy in the treatment of a wide range of human malignancies[5,6]. Targeting antigens like cytotoxic T lymphocyte antigen 4 (CTLA-4), programmed cell death protein 1 (PD-1), and programmed death ligand 1 (PD-L1) that act as checkpoints on the immune system have become the focus of many recent clinical trials[5,6]. The basis of checkpoint blockade is founded on restoring the natural anti-tumor immunity of the host immune system. A series of stepwise events are required for effective host immunity against cancer cells. In brief, tumor neoantigens are processed by antigen presenting cells, such as dendritic cells, which are presented on major histocompatibility complex (MHC) class I and MHC class II molecules to T cells. With appropriate co-stimulation, the T cells become effectors that can kill tumor cells expressing the cancer-specific antigens. Ideally, these activated T cells recognize cancer cells within the tumor via their T cell receptors (TCRs) and initiate tumor-specific killing[28]. However, various immunosuppressive factors in the immune microenvironment of tumors may prevent effector cells from functioning against the tumor. Many of these tolerogenic mechanisms recapitulate normal host mechanisms that protect against autoimmunity such as the expression of checkpoint molecules and accumulation of immunosuppressive regulatory T cells[29]. PD-1 is expressed by activated T cells upon TCRs binding, and upon engagement by its ligand PD-L1, inhibits T cell effector functions[6,30,31]. PD-L1 is found to be expressed on a wide range of cell types, including activated hematopoietic cells, epithelial cells, and tumor cells[32].

PD-L1 expression varies amongst subtypes of breast cancer. In the more immunogenic subtypes of HER2-positive and TNBC, PD-L1 expression is more commonly found. In primary breast cancers, PD-L1-positivity was found in 11% of luminal tumors compared to 33% of basal-like tumors and 56% of HER2-positive tumors[33]. A study using the Cancer Genome Atlas (TCGA) RNA sequencing data demonstrated a higher expression of PD-L1 in TNBC compared to the other subtypes[34]. Using tumor microarrays, PD-L1-positivity was found in 19% of TNBCs[34]. This value likely underestimates the degree of clinically relevant PD-L1-positivity, given the study’s definition of positivity as cell surface membrane staining >5%, rather than the more frequently used cut-off of >1% in clinical trials. In a study of 167 patients with HER2-positive breast cancer, approximately 50% stained positive for PD-L1 on tumor cells, with a significant association between PD-L1 expression in tumor cells and TIL infiltration[35]. Not unexpectedly, disruption of the PD-1/PD-L1 inhibitory pathway by blocking monoclonal antibodies has shown efficacy against breast cancer in multiple clinical trial settings, which we will detail further below[36].

3.1. Use of checkpoint blockade in advanced and/or metastatic breast cancer

The KEYNOTE-012 phase Ib clinical trial in 2014 was the first to use checkpoint blockade in breast cancer[37]. The study comprised 32 patients with metastatic TNBC with PD-L1 positive tumors and used monotherapy with pembrolizumab, a PD-1 inhibitor. Outcomes included an objective response rate (ORR) of 19% among the 27 patients who were able to be included in the efficacy analysis based on RECIST v1.1, with a median progression-free survival (PFS) of 1.9 months. In this small study, increasing the likelihood of response was found with increasing PD-L1 expression. This and other key clinical trials utilizing checkpoint blockade in the metastatic setting are summarized in Table I. The PANACEA/KEYNOTE-014 phase II study was conducted on 52 patients with HER2-positive breast cancer which progressed on trastuzumab. For the 40 patients with PD-L1 positive tumors, the ORR of pembrolizumab was 15% compared to no responders among the PD-L1 negative patients[38].

Table I:

Key clinical trials of checkpoint inhibitors in metastatic and/or advanced breast cancer

Trial	Drug	Phase	Setting	Size	ORR	PFS	OS
KEYNOTE-012[37]	Pembrolizumab	Ib	Metastatic PD-L1+ TNBC	N = 27	19%	Median 1.9 mo.	Median 11.2 mo.
PANACEA/KEYNOTE-014[38]	Pembrolizumab	Ib/II	Progression on trastuzumab, HER2+	N = 52	15% (PD-L1+)	Median 2.7 mo. (PD-L1+)	NA
JAVELIN[39]	Avelumab	I	Metastatic	N = 168	3%; 17% (PD-L1+)	Median 5.9 mo.	Median 8.1 mo.
IMpassion[40]	Atezolizumab + nab-paclitaxel	III	Metastatic TNBC	N = 902	56%; 59% (PD-L1+)	Median 7.2 mo.	Median 25.0 mo.

ORR = objective response rate; PFS = progression-free survival; OS = overall survival; TNBC = triple-negative breast cancer; mo. = months.

PD-L1 inhibition has also proven to be beneficial in several clinical trials. The JAVELIN trial was a phase Ib study using avelumab (a PD-L1 inhibitor) in 168 patients with metastatic breast cancer with progression on standard therapy[39]. Although the overall ORR was 3%, a trend towards a higher ORR of 17% was seen in patients with PD-L1-positive tumors. The key trial that accelerated the approval of atezolizumab (another PD-L1 inhibitor) in the treatment of metastatic TNBC was the IMpassion130 trial. This phase III randomized trial included 902 patients with untreated metastatic TNBC randomly assigned to atezolizumab plus nab-paclitaxel or placebo plus nab-paclitaxel[40]. ORR was 56% in the atezolizumab group compared to 46% in the placebo group, and was 59% vs. 43% in PD-L1 positive only tumors. The median PFS was 7.2 months in the atezolizumab group compared to 5.5 months in the placebo group. The median OS was 21.3 months in the atezolizumab group and 17.6 months in the placebo group, with a median OS of 25 months in PD-L1 positive tumors in the atezolizumab group. Atezolizumab, in combination with nab-paclitaxel, received approval in 2019 for the first-line treatment of metastatic TNBC with PD-L1 positivity, described as 1% or more of the tumor-infiltrating immune cells in the tumor area staining for PD-L1.

3.2. Use of checkpoint blockade in the neoadjuvant setting

Recent data have also shown efficacy of combination checkpoint blockade and chemotherapy in the neoadjuvant setting in achieving pCR. The KEYNOTE-173 phase Ib trial evaluated the use of pembrolizumab plus neoadjuvant chemotherapy in high-risk, early-stage TNBC[41]. Pembrolizumab was combined with six separate chemotherapy regimens in a total of 60 patients. The chemotherapy regimens included either nab-paclitaxel or paclitaxel, and all except one included carboplatin. All regimens were followed by doxorubicin and cyclophosphamide. For all patients, the overall pCR rate was 60%, with two cohorts involving nab-paclitaxel and carboplatin having a pCR of 80%. OS at 12 months in the carboplatin-inclusive groups was 98% compared to 80% in the group without carboplatin. Increased pre-treatment PD-L1 positivity and TILs were associated with higher pCR rates. The key neoadjuvant breast cancer clinical trials are highlighted in Table II.

Table II:

Key clinical trials of checkpoint inhibitors in the neoadjuvant setting

Trial	Drug	Phase	Setting	Size	pCR
KEYNOTE-173[41]	Pembrolizumab + nab-paclitaxel or paclitaxel +/− carboplatin +AC	Ib	Early-stage TNBC	N = 60	60%
KEYNOTE-522[42]	Pembrolizumab or placebo + paclitaxel, carboplatin, doxorubicin or epirubicin, cyclophosphamide	III	Untreated stage II-III TNBC	N = 602	65% (pembro) vs. 51% (placebo)
I-SPY2[43]	Pembrolizumab + paclitaxel, doxorubicin, cyclophosphamide	II	Stage II-III, HER2-negative	N = 250	44% (pembro) vs. 17% (no pembro)
GeparNuevo[27]	Durvalumab or placebo + nab-paclitaxel, epirubicin, cyclophosphamide	II	Primary non-metastatic TNBC	N = 174	61% (durval) vs. 41% (placebo) in window cohort only
IMpassion031[44]	Atezolizumab or placebo + nab-paclitaxel, doxorubicin, cyclophosphamide	III	Stage II-III TNBC	N = 333	58% (atezo) vs. 41% (placebo)

TNBC = triple-negative breast cancer; pCR = pathologic complete response.

The subsequent phase III trial KEYNOTE-522 randomized patients with untreated stage II or stage III TNBC to neoadjuvant pembrolizumab plus paclitaxel and carboplatin or placebo plus paclitaxel and carboplatin[42]. Both groups also received doxorubicin or epirubicin and cyclophosphamide. A total of 602 patients with newly diagnosed TNBC with T1cN+ or T2 or greater N0–N2 disease with any PD-L1 status underwent randomization. The overall pCR rate was 65% in the pembrolizumab and chemotherapy group compared to 51% in the placebo and chemotherapy group. PD-L1 positive tumors treated with pembrolizumab and chemotherapy had a pCR of 69% compared to 55% in those treated with placebo and chemotherapy. The difference in response with pembrolizumab was seen even in the PD-L1 negative group, with 45% in the pembrolizumab group achieving pCR compared to 30% in the placebo group. EFS at 18 months was 91% in the pembrolizumab group and 85% in the placebo group, with the most common event being distant recurrence. A major strength of this trial is the comparison to the control group treated with a chemotherapy regimen including a platinum agent, which allowed for comparison of the addition of pembrolizumab to a neoadjuvant chemotherapy regimen that has been associated with the highest rate of pCR in early TNBC.

An open-label, multicenter, adaptively randomized phase II platform trial known as the I-SPY2 study evaluated multiple investigational arms in parallel for high-risk stage II and III breast cancer. Recently published results of 250 patients with HER2-negative breast cancer were randomized to pembrolizumab plus neoadjuvant chemotherapy (paclitaxel followed by doxorubicin and cyclophosphamide) or neoadjuvant chemotherapy alone[43]. By subtype, pCR rates for pembrolizumab and neoadjuvant chemotherapy compared to neoadjuvant chemotherapy alone were 44% vs. 17% overall, 30% vs. 13% for HR-positive and HER2-negative tumors, and 60% vs. 22% for TNBC. Although no significant differences in three-year EFS were found due to short follow-up time and few patients, those who did achieve pCR had good outcomes with three-year EFS of 93%.

The GeparNuevo phase II randomized trial evaluated the addition of durvalumab or placebo to standard neoadjuvant chemotherapy in 174 patients with TNBC[27]. The standard neoadjuvant chemotherapy regimen was nab-paclitaxel followed by dose-dense epirubicin and cyclophosphamide. Patients with primary non-metastatic TNBC with cT2-cT4a-d disease were included. One-hundred and seventeen patients participated in the window-phase where durvalumab was given two weeks before the start of chemotherapy. However, as the time to start of chemotherapy (mean 48 days) was felt to be too long by the Independent Data Monitoring Committee, the rest of the patients received durvalumab or placebo plus chemotherapy at the same time on day 1. PD-L1 positivity was present in 87% of patients. The effect of durvalumab on pCR was seen only in the window cohort, where pCR was 61% in the durvalumab group compared to 41% in the placebo group. In both treatment and placebo arms, the pCR rate increased with higher TILs.

The recently published IMpassion031 trial is a phase III double-blind randomized trial evaluating the efficacy and safety of atezolizumab or placebo plus nab-paclitaxel followed by doxorubicin and cyclophosphamide in patients with untreated stage II-III TNBC[44]. One hundred and sixty-five patients were assigned to the atezolizumab and chemotherapy arm and 168 to the placebo and chemotherapy arm. A significant improvement in pCR rate was demonstrated in the atezolizumab group compared to placebo (58% vs. 41%). Forty-six percent of patients were positive for PD-L1. In PD-L1 positive patients, 69% in the atezolizumab group had a pCR compared to 49% in the placebo group. Even in PD-L1 negative patients, 48% vs. 34% of patients had pCR in the atezolizumab vs. placebo groups. Improvement in response in the PD-L1 negative group was similarly seen in the KEYNOTE-522 trial[42]. In contrast, the IMpassion130 trial demonstrated an improvement in OS only in the PD-L1 positive group in the setting of metastatic TNBC[45]. The biologic mechanism for these findings remains to be defined. Differences in observed outcomes may be due to variations in chemotherapy regimens, PD-L1 assays, or intrinsic immunologic differences between early-stage versus metastatic TNBC. There is evidence demonstrating a relative decrease in immune activation in metastatic breast cancer compared to non-metastatic[46,47], which may in part lead to improved responses in early stage breast cancer even in the presence of <1% PD-L1 positivity. Although the KEYNOTE-522 trial accounted for tumor cell and immune cell PD-L1 positivity, the IMpassion031 trial only included tumor-infiltrating immune cells in their assessment of PD-L1 expression, suggesting the possibility that PD-L1 positivity on tumor cells may play a yet undefined role in tumor response. Interestingly, in the IMpassion031 trial, pCR rate was higher in patients with positive regional lymph nodes[44]. Similarly in the KEYNOTE-522 trial, patients with higher nodal stage had better response to checkpoint blockade compared to node negative disease, while there was no difference in pCR rates when stratified by T stage[42]. There is some evidence that PD-L1 expression in lymph node metastasis is higher than in primary breast tumors[48], which may explain in part the improved response to PD-1/PD-L1 inhibition seen in higher nodal stages. Further research is warranted to understand how regional nodal metastases impacts anti-tumor immunity and the potential use of this clinical variable as a stratification factor for the consideration of neoadjuvant immunotherapy.

The currently recruiting KEYNOTE-756 study is evaluating pembrolizumab compared to placebo in the setting of neoadjuvant chemotherapy and adjuvant endocrine therapy for the treatment of early-stage ER-positive, HER2-negative breast cancer[49]. The primary study endpoints are pCR rate and EFS. As the clinical trial data to date has been largely focused on TNBC given its higher tumor immunogenicity, we eagerly await the results of the KEYNOTE-756 trial.

4. Toxicities

An understanding of potential adverse events associated with checkpoint inhibitors will be essential in improving patient care, especially given the promise of several inhibitors like atezolizumab and pembrolizumab in the neoadjuvant setting. Adverse events in the phase III IMpassion130 trial evaluating atezolizumab plus nab-paclitaxel vs. placebo plus nab-paclitaxel relatively tolerable[45]. Hypothyroidism of any grade occurred in 18% of patients in the atezolizumab group compared to 5% in the placebo group. Similarly, hyperthyroidism occurred in 5% of patients in the atezolizumab group compared to 1% in the placebo group. The median time to onset was 3.8 months for both hypothyroidism and hyperthyroidism. Pneumonitis was more common in the atezolizumab group (4% vs. <1%), with a median time to onset of 4.7 months. Rash occurred in 34% in the atezolizumab group compared to 26% in the placebo group. Hepatitis and colitis were more frequent in the atezolizumab group at 2% and 1% compared to <1% for in the placebo group, with a mean time of onset 4.6 months for hepatitis and 6.7 months for colitis. Adrenal insufficiency is a rare but reported event in the atezolizumab group occurring in less than 1% of patients, compared to no patients in the placebo group. Median time to onset is 4.9 months. Given that multiple clinically significant adverse events occur at a median time to onset of four months or later, including hepatitis, colitis, adrenal insufficiency, and pneumonitis, clinical vigilance and understanding of the temporal relationship with administration of checkpoint blockade will be critical.

The KEYNOTE-522 trial comparing pembrolizumab to placebo in the setting of neoadjuvant chemotherapy demonstrated trends in the adverse effects profile similar to that of atezolizumab[42]. The pembrolizumab group experienced higher rates of hypothyroidism (14% vs. 3%) and hyperthyroidism (5% vs. 1%), as well as adrenal insufficiency (2% vs. 0%) compared to the placebo group. Rash occurred in 22% of patients in the pembrolizumab group and 15% of patients in the placebo group. Severe skin reactions were also higher (4% vs. 1%), including more incidents of grade 3 or higher adverse events (4% vs. 0%). A higher incidence of infusion reaction (17% vs. 11%) was also reported. Overall, both inhibitors have a relatively tolerable side effect profile with few numbers of serious adverse events. It is important to note that in the KEYNOTE-522 trial, immunotherapy did not change the likelihood of a patient having surgery, with 96.7% in the pembrolizumab arm and 97.4% in the control.

5. Future direction

Data regarding the efficacy of checkpoint blockade in the treatment of breast cancer in multiple clinical settings have rapidly accumulated. Multiple trials have shown the efficacy of checkpoint blockade in both the metastatic and neoadjuvant settings, which in large part, are focused on TNBC. Analysis of HR-positive and HER2-negative tumors in the I-SPY2 trial showed an improved response with the addition of pembrolizumab to neoadjuvant chemotherapy[43]. These results are encouraging, but more studies in the ER-positive and HER2-positive subtypes that address their unique tumor microenvironment are needed. These studies also stress the need for a better understanding of the contribution of PD-L1 positivity to the anti-tumor effect of checkpoint blockade. The KEYNOTE-522 trial demonstrated an increase in pCR rates with pembrolizumab, even in PD-L1 negative tumors. The mechanism of action of PD-1 and PD-L1 inhibitors in this setting needs to be further elucidated.

The success of the durvalumab arm in the window cohort in the GeparNuevo trial poses some interesting questions. The administration of durvalumab before the initiation of chemotherapy resulted in improved pCR rates. The authors report that an increase of intratumoral TILs in post-window samples trended towards being predictive of pCR in the durvalumab arm, although this was not statistically significant[27]. It is conceivable that administering checkpoint blockade first may lead to more effective immunoediting of chemo-resistant subclones within the tumor, allowing for improved response to chemotherapy. More studies are needed to clarify mechanisms of action and to define the optimal sequencing of immunotherapeutic agents with chemotherapy.

Although PD-1/PD-L1 blockade has been generally well-tolerated, there remains an increased risk for significant adverse events, including endocrinopathies, hepatitis, pneumonitis, colitis, and renal toxicity[42,50]. As not all patients with TNBC have an equal prognosis, and patients with stage I TNBC have been reported to have breast cancer-specific survival of 95% at four years compared with 53% for stage III[51], thoughtful patient selection becomes an important part of maximizing the benefit of immunotherapy for subsets of patients with TNBC. Also, the KEYNOTE-522 trial demonstrated a 21% increase in pCR with pembrolizumab in node-positive patients compared to a 6% increase in pCR in node-negative patients[42], suggesting that pembrolizumab may be more beneficial in patients with more advanced disease. More data are needed on the risks and benefits of checkpoint blockade in patients who have early-stage TNBC and/or predisposition to autoimmune disorders.

6. Conclusions

As TILs provide a strong prognostic value in breast cancer, the success of checkpoint blockade in a select group of patients has not been entirely unexpected. However, single-agent checkpoint inhibitors are not as effective as combination therapy. Atezolizumab in combination with nab-paclitaxel was the first checkpoint inhibitor to be approved in breast cancer, with the described indication for the first line treatment of PD-L1 positive metastatic TNBC. Thereafter, multiple trials have demonstrated improvements in pCR rates in non-metastatic TNBC with the addition of a single-agent checkpoint inhibitor to standard neoadjuvant chemotherapy. There are several ongoing clinical trials in the neoadjuvant setting, including the evaluation of outcomes in ER-positive cancers. We eagerly await the results of these trials, as well as guidance from regulatory agencies, which will help further elucidate the role of immunotherapy in surgical patients with breast cancer.

Synopsis for Table of Contents:

Tumor-infiltrating lymphocytes have prognostic significance in triple-negative breast cancer. Checkpoint blockade has shown promise in both the advanced setting and the neoadjuvant setting in triple-negative breast cancer.