Background
Inflammatory breast cancer (IBC) is a rare and aggressive malignancy. In the United States (US), IBC accounts for 2–6% of all patients with breast cancer [
1,
2]. Although IBC is a relatively rare clinical subtype of locally advanced breast cancer, it is responsible for approximately 10% of breast cancer-associated deaths annually in the US, which translates into about 4000 deaths per year [
3,
4]. Approximately 20–30% of IBC patients present with distant metastasis (stage IV disease) at diagnosis, compared to 6–10% of patients with breast cancer that is not inflammatory (non-IBC) [
5]. Although recent trends indicate an improvement in survival in IBC patients, the prognosis remains worse than non-IBC cases. The median overall survival (OS) for patients with stage III IBC is 4.75 years, compared to 13.40 years in those with non-IBC, and for stage IV disease OS is 2.27 years in IBC patients versus 3.40 years in non-IBC patients [
6,
7]. IBC patients tend to be younger compared to other breast cancer patients, with a median age at diagnosis of 52 years compared to 57 for non-IBC patients [
8].
The principal clinical symptoms of IBC are breast erythema, edema, peau d’orange, and dermal lymphatic invasion [
9]. Despite its name, IBC is not associated with a profuse inflammatory response. Rather, the characteristic redness and swelling of the breast are due to obstruction of lymphatic channels in the dermis by tumor cells [
4,
10].
The current consensus regarding clinical management of IBC includes neoadjuvant systemic therapy (chemotherapy or chemotherapy plus targeted therapy), modified radical mastectomy and level I and II ipsilateral axillary node dissection, post-mastectomy radiotherapy of the chest wall and nodal basin, and adjuvant targeted therapy and hormonal therapy [
11]. In the case of stage IV de novo IBC, primary systemic therapy is also recommended, but the decision to use surgery and radiation therapy should be evaluated using a multidisciplinary approach, particularly in those patients who have significant clinical response to systemic therapy [
11]. Taxane, doxorubicin, and cyclophosphamide in neoadjuvant systemic therapies have been recommended for stage III primary breast cancer, but there is a clear lack of clinical trials specifically for IBC [
11]. For HER2
+ disease, dual anti-HER2 therapy (pertuzumab and trastuzumab) combined with chemotherapy is recommended [
11].
Data on risk factors for IBC are limited, and the contributions of hereditary versus environmental or life-style factors remain poorly understood [
12]. Although IBC, like non-IBC breast cancers, is a heterogeneous disease and can occur as any of the five molecular subtypes, the disease is most commonly either HER2 overexpressing or triple-negative (TN) [
13]. TN breast cancer, which is defined by absence of estrogen and progesterone receptors, and a lack of HER2 overexpression, has a poorer prognosis than other subtypes [
14]. Importantly, the TN phenotype breast cancer (non-IBC) has been associated with higher expression of PD-L1 on tumor cells [
15]. The interaction of the PD-1 receptor on T cells with its ligand, PD-L1 on tumor- and immune-infiltrating cells, suppresses T cell-mediated immune responses and may play a role in immune escape by human tumors [
16]. The extensive accumulation of tumor emboli in the lymphatic vessels of IBC patients supports the notion that the host immune surveillance system is suboptimal or that the tumor cells have adopted immune escape mechanisms to avoid detection by the host. Several recent studies have provided evidence for immune responses toward IBC, suggesting that patients may benefit from immunotherapies, such as PD-1/PD-L1 blocking antibodies [
17‐
22]. Nonetheless, there is insufficient information on peripheral blood leukocyte immune-phenotypes in IBC patients. In the present work, we utilized flow cytometry and immunohistochemistry analysis to investigate immune parameters of metastatic IBC patients and compared them to healthy female volunteer donors. In addition, we studied the expression of PD-L1 and PD-1 in tumor biopsies of these patients with metastatic IBC. Our study aims to determine whether cellular components of the immune system are altered in IBC patients, thereby contributing to the pathogenesis of the disease.
Discussion
In the present work, we studied the immune profile of 14 IBC patients with stage IV disease and compared to 11 age-similar healthy controls. The analysis of fresh peripheral blood showed that the most notable parameter that differed in the immune cells of the patients with metastatic IBC was lymphopenia. Our data showed significant deficits in numbers of T, NK, and B lymphocytes in peripheral blood of stage IV IBC patients. Although we cannot rule out that some of the lymphopenia in our patient population may have resulted from prior chemotherapy, we show that lymphocyte subset counts from our two treatment naïve IBC patients were generally consistent with treated non-TN IBC patients and nearly always below the medians of our healthy controls, indicating a disease-related impact. A previous study by Reuben and Lee also demonstrated that patients with metastatic IBC had lymphopenia associated with significantly lower CD4
+ T and B cells, but higher counts of monocytes in peripheral blood as compared to healthy controls, whereas these cell counts were similar to healthy controls in IBC patients with non-metastatic disease [
28]. Mego et al. also reported more severe decreases in absolute lymphocyte count of IBC patients with metastatic disease [
29].
Our results expand upon previous studies by showing that the numbers of all subpopulations of CD4
+ helper T cells (naïve, central memory, effector memory, and effector) were significantly lower in stage IV IBC patients, whereas reductions within CD8
+ cytotoxic T lymphocytes were more concentrated in the memory subsets. CD4
+ T lymphocytes enhance tumor antigen-specific immune responses by producing cytokines, and CD8
+ T cells provide key adaptive anti-tumor immunity through their production of IFN-γ and cytolytic activity [
30]. While naïve T cells have not yet been activated and exist in a resting state, effector T cells are short-lived and exhibit low proliferative capacity, but elicit potent functional responses toward an antigenic target [
31]. The memory subsets provide more rapid and robust secondary responses upon re-exposure to antigens, whereupon they differentiate to an effector state. Central memory T cells have the most potent proliferative capability and longevity, but weakest functional responsiveness, while effector memory cells have intermediate properties between central memory and effector T cells [
31].
It should be noted that lymphopenia is more common in patients with a variety of advanced tumors, as compared to those with localized disease [
32]. In particular, reductions in peripheral CD4
+ T lymphocytes are commonly observed in advanced cases of pancreatic cancer, melanoma, non-Hodgkin lymphoma, sarcoma, hepatocellular carcinoma, and breast cancer [
32]. Lymphopenia, particularly the low frequency of CD4
+ T cells, in patients with advanced cancer has also been shown to correlate with performance status, unfavorable prognostic factors, and worse survival [
32,
33].
Our data support the findings of Reuben and Lee, which reported that the CD4/CD8 ratio of non-metastatic IBC patients was significantly higher than in metastatic IBC patients, because the non-metastatic IBC patients had a significantly greater reduction in CD4
+ helper T lymphocytes than healthy controls [
28]. In contrast, Mego et al. did not observe any alterations in CD4/CD8 T cell ratio between patients with metastatic or non-metastatic IBC and healthy normal donors [
29]. Of note, the percentage of CD4
+ T cells that were FoxP3
+ and CD25
high (regulatory T cells; Treg) did not differ between our IBC patient and healthy control cohorts (data not shown), in contrast to the Mego et al. report of decreased numbers of Tregs in metastatic IBC patients compared to non-metastatic IBC patients and healthy controls [
28].
While our study showed reduced numbers of peripheral NK cells in metastatic IBC, Reuben and Lee found no significant reductions in NK cells of metastatic or non-metastatic IBC patients, as compared to healthy controls [
28]. NK cells constitute approximately 5% of the lymphocytes in healthy human peripheral blood and are involved in controlling tumor progression and metastases in a variety of contexts [
34]. Two major NK cell subsets are found in human subjects that can be distinguished by their levels of CD56 expression, namely CD56
dim and CD56
bright [
35]. CD56
bright NK cells make up approximately 2–10% of total NK cells in peripheral blood, are less mature, more apt to leave the vasculature, more efficient at producing cytokines, and less cytolytic than CD56
dim cells. The predominant cytolytic targets of NK cells are rare cells that have downregulated expression of class I MHC (MHC-I), which is normally expressed on healthy nucleated cells of the body [
36]. MHC-I loss is a common mechanism by which tumors and virus-infected cells can evade recognition by cytolytic T cells, and NK cells can thereby overcome this potential immunologic evasion mechanism [
27]. A counterbalance of signals from activating and MHC-I-binding inhibitory receptors on NK cells regulate their responsiveness [
37]. Non-IBC tumors commonly express ligands for the NK cell activating receptors, DNAM-1 and NKG2D, which can increase their susceptibility to attack [
38]. CD56
dim NK cells express the activating receptor CD16 (low-affinity FcγRIIIA) and mediate cytotoxicity by the directed exocytosis of perforins and granzymes from cytolytic granules, which perforate the target cell plasma cell membrane and trigger apoptosis, respectively [
39]. A previous report described increased frequencies of immature and non-cytolytic NK cells in advanced non-IBC patients [
40], although such a shift was not evident in our study. Instead, our data showed that IBC patients had higher expression levels of CD16, granzyme B, and perforin on their CD56
bright NK cells, suggesting that these immature cells are undergoing accelerated maturation to replenish the diminished mature CD56
dim population. In contrast to our observed increase in CD16 expression in stage IV IBC, a previous report showed decreased expression of activating NK cell receptors, including CD16, during progression of non-IBC, while inhibitory receptors increased and this correlated with decreased NK cell function, at least partially due to TGF-β1 in the TME [
41].
Our study also provides further evidence that the immune system in some metastatic IBC patients has responded to the tumor at some stage of cancer development. We showed that IBC tumors from a subset of stage IV patients expressed moderate to high levels of PD-1 (18.2% of patients) and PD-L1 (36.4% of patients) on infiltrating immune cells by IHC analysis. In addition, we found a positive correlation between PD-L1 expression and PD-1 expression in our IBC tumor biopsies. Our results are consistent with Bertucci et al., who reported overexpression of PD-L1 mRNA in 38% of IBC patient tumors that were associated with increased B and CD8
+ T cell gene expression signatures [
18]. Similarly, two groups recently found that expression of PD-L1 in IBC tumor samples correlated with higher stromal tumor-infiltrating lymphocytes (sTIL) that were highly enriched in CD20
+ B cells, and their combined presence was associated with better response to neoadjuvant therapy [
17,
22]. We have expanded upon these results by further showing that expression of PD-1 and PD-L1 in our stage IV IBC samples correlated significantly with infiltration of CD20
+ B cells and a trending correlation was noted for infiltration of CD3
+ T cells. The invasion of the tumor stroma by sTIL is often associated with a better prognosis in ER-negative non-IBC [
42]. Although breast cancer is considered moderately immunogenic, the presence of neoantigens seems to elicit an immune response, and infiltrating immune cells play an essential role in the host-defense mechanism against ER-negative non-IBC in both adjuvant and neoadjuvant studies [
43,
44]. Van Berckelaer et al. recently showed that PD-L1 expression on sTIL was more frequently observed in IBC than non-IBC except in the Her2
+ subtype [
22]. PD-L1 expression on immune cells was seen in 38.6–42.9% of the IBC patients, and it was significantly higher than in non-IBC patients [
22]. Hamm et al. also showed that some IBC tumors had high infiltration of CD8
+ cells expressing PD-L1, and these had genetic profiles predictive of greater incidence of potential neoantigens [
21].
As an extension of the previous studies, we also found correlations between PD-L1 expression in the stage IV IBC tumors and immune parameters in peripheral blood. Our results showed that the expression levels of PD-L1 in tumor tissues correlated positively with expression levels of cytolytic granule components (perforin and granzyme B) in peripheral blood CD4
+ T and CD56
dim NK cells. The higher levels of these granule components are indicative of a previously activated state and consistent with the higher levels in immature CD56
bright NK cells, as compared to healthy controls. Furthermore, we found that higher expression of PD-L1 in the tumors also correlated with a shift from reduced percentages of naïve to increased frequency of effector memory CD8
+ T cells in peripheral blood. Taken together, these results suggest that the CD4
+ T and NK cells have been activated and effector memory CD8
+ T cells have at some point expanded in response to expression of the immunosuppressive ligand in the tumor in a subset of the metastatic IBC patients. Despite these activation events, these immune cells have declined in overall numbers and presumably progressed to the classical exhausted state after chronic exposure to tumor. T cell exhaustion is a phenotype defined by poor effector function, such as reduced secretion of IL-2, IFN-γ, and TNF-α [
45]. Reuben and Lee have further found that CD8
+ T cells from the blood of patients with non-metastatic IBC had enhanced IFN-γ production responses upon T cell receptor (TCR)-stimulation, while IFN-γ production declined again toward levels of healthy donors in patients with metastatic IBC [
28]. These results suggest that CD8
+ T cell responsiveness may be enhanced in non-metastatic IBC patients. All of these observations imply that immunotherapy should be considered as a potential treatment for those patients exhibiting increased expression of PD-1 and/or PD-L1 in tumor and/or increased numbers of cytolytic effector memory T cells in peripheral blood. In such patients, blockade of the PD-1/PD-L1 inhibitory axis has the potential to reactivate antigen-experienced, exhausted T cells toward the tumor and thereby might improve clinical outcome. In fact, four clinical trials are currently testing the efficacy of PD-1 or PD-L1 blockade in IBC patients (NCT03515798, NCT02411656, NCT03742986, and NCT03202316).
The data reported in our study support the concept of progressive immune dysfunction as IBC advances to highly metastatic clinical behavior. In fact, IBC is associated with early metastatic dissemination as suggested by higher numbers of circulating tumor cells (CTCs) compared to other forms of breast cancer. It has been shown that IBC patients with higher numbers of CTCs also have a more compromised immune status, which is characterized by reduced percentages of CD4
+ helper T cells, higher percentage of Treg cells, and reduced cytokine-producing CD8
+ T cells [
29]. Peripheral blood immune cells can contribute to an unfavorable environment for CTC survival, since innate and adaptive immune mechanisms are purportedly responsible for controlling tumor dissemination and, perturbations in the immune surveillance could favor an environment conducive for the survival and dissemination of CTCs, ultimately leading to cancer progression [
46]. We can hypothesize that the lower level of immune surveillance in lymphopenic metastatic IBC patients could facilitate additional metastasis, but the ultimate causes of immune dysfunction in IBC need to be better defined. To improve mechanistic understanding, Reddy et al. recently showed greater macrophage infiltration in IBC tumors than in other breast cancers [
47], and they further showed that IBC tumors with increased infiltration of mast cells were associated with poorer clinical responses to neoadjuvant chemotherapy [
48]. Also, Valeta-Magara et al. recently showed that IBC tumors produce chemokines and cytokines that recruit monocytes and polarize macrophages to the M2 phenotype, which are immunosuppressive and tumor-promoting [
49]. Thus, the accumulating evidence suggests that M2 macrophages and the PD-1/PD-L1 axis contribute to the immunosuppressive TME in IBC, although more studies are clearly needed.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.