Secondary Logo

Share this article on:

An Immunoscore Using PD-L1, CD68, and Tumor-infiltrating Lymphocytes (TILs) to Predict Response to Neoadjuvant Chemotherapy in Invasive Breast Cancer

McLemore, Lauren E., DO, MPH*; Janakiram, Murali, MD; Albanese, Joseph, MD, PhD*; Shapiro, Nella, MD; Lo, Yungtai, PhD*,§; Zang, Xingxing, PhD†,∥; Fineberg, Susan, MD*

Applied Immunohistochemistry & Molecular Morphology: October 2018 - Volume 26 - Issue 9 - p 611–619
doi: 10.1097/PAI.0000000000000485
Research Articles

Response to neoadjuvant chemotherapy (NAC) in invasive breast cancer (IBC) is partly regulated by the immune microenvironment. We evaluated immune checkpoint PD-L1 expression, presence of CD68+ cells of macrophage/monocytic lineage and stromal tumor-infiltrating lymphocytes (TILs) in prechemotherapy biopsies and correlated with NAC response. We studied 76 cases of IBC. Prechemotherapy biopsies with >30% TILs were considered lymphocyte-rich IBC. We performed immunohistochemistry for PD-L1 and CD68. Prechemotherapy cores showing >1% PD-L1+ immune or tumor cells were considered positive. CD68 was positive if >40% of tumor stroma contained CD68+ cells or atleast 50% of tumor cells showed infiltration by CD68+ cells. Residual Cancer burden (RCB) Score of 0/I represented excellent response to NAC and RCB II or III unfavorable response. Thirty-five patients had RCB 0/I and 41 pts RCB II/ III. TILs>30% were present in prechemotherapy biopsies in 19 pts of whom 14 showed RCB 0/I (P=0.0075). Twenty-seven cases were PD-L1+ and 20 had an RCB 0/I (P=0.0003). Twenty-two cases were CD68+ of whom 18 showed RCB 0/I (P=<0.0001) There was a significant association between TILs>30%, PD-L1+ and CD68+ expression. Using atleast one of these immunologic parameters identified 26 of 35 patients with RCB 0/I and showed a higher sensitivity for response prediction than TILs alone (40% vs. 74.3%). In conclusion we demonstrate that high numbers of CD68+ monocytic/macrophage cells and PD-L1 expression in IBC shows significant association with NAC response. An immune biomarker profile including TILs, PD-LI and CD68 is more sensitive for NAC response prediction than TILs alone.

Departments of *Pathology

Oncology

§Epidemiology and Population Health, Montefiore Medical Center and the Albert Einstein College of Medicine

Department of Surgery, Eastchester Center for Cancer Care and Montefiore Medical Center

Department of Microbiology and Immunology, Albert Einstein College of Medicine, Bronx, NY

The authors declare no conflict of interest.

Reprints: Susan Fineberg, MD, Department of Pathology, Montefiore Medical Center, 4th floor silver Zone, 111 East 210th Street, Bronx, NY 10467 (e-mail: sfinebergshear@hotmail.com).

Received July 31, 2016

Accepted November 6, 2016

The immune response plays an important role in tumor elimination as well as prevention of metastatic tumor spread, and the presence of an activated immune response, as marked by increased tumor-infiltrating lymphocytes (TILs) in invasive breast cancer, has been linked to improved survival.1 The mechanism by which breast cancer can elicit an immune response is an area of active investigation. It is known that mutated genes may result in the production of novel antigenic epitopes which can be immunogenic. In addition, proteins not normally expressed, such as the testis antigen NY-ESO-1, can be aberrantly expressed on tumor cells and elicit a humoral immune response.2 The improved prognosis associated with a high percentage of TILs in breast cancer is now well established, and efforts are underway to incorporate assessment of TILs in the pathologic assessment of all breast cancers.3,4

Therapies which specifically activate the immune system, such as blockade of the immune checkpoint pathway with anti PD-1/PD-L1 antibodies, have shown promise in clinical trials in advanced breast cancer.5,6 However, it appears that the effects of even conventional cytotoxic chemotherapies, which do not specifically target the immune system, are mediated at least in part through activation of immune pathways. Studies have shown that anthracyclines induce an immunogenic cell death which in turn activates antigen presenting cells resulting in activation of tumor specific cytotoxic T cells.7 In the neoadjuvant setting, evidence of immune activation, assessed either by the presence of TILs or gene expression arrays, are associated with a high rate of pathologic complete response (pCR).8

An immune response during neoadjuvant chemotherapy (NAC) is a complex interplay of immune cells (lymphocytes, macrophages, and monocytic cells), stroma (secretion of cytokines), and tumor cells (expression of immune check points) in the microenvironment. A high percentage of TILs is associated with high rates of response to NAC.9 Studies reporting the prognostic and predictive significance of PD-L1 expression have yielded conflicting results.10,11 In addition, although tumor-associated macrophages (TAMS) have been associated with a poor prognosis,12 very little research exist on the role of TAMS in response to NAC. What appears evident is that chemotherapy and HER2-targeted therapy are most effective in tumors with evidence of immune activation1 and methodologies which include assessment of several immune activation pathways might provide ability to predict response to NAC.

The aim of our study was to develop a tumor immune profile which could be used as an immune predictive signature for response to NAC in invasive breast cancer. Specifically, we assessed TILs, and used immunohistochemistry to detect both PD-L1 expression and the presence of CD68+ cells of the monocytic and macrophage lineage including TAMs in patients treated with NAC. Our goal was to correlate each of these immune parameters, individually and in combination, with response to NAC.

Back to Top | Article Outline

MATERIALS AND METHODS

Samples

Our study included 76 retrospectively identified patients with limited stage invasive breast cancer treated with NAC at our hospital between 2008 and 2015. Cases were identified through a tumor registry search for patients with localized invasive breast cancer treated with NAC as well as a search of our pathology database. Only patients for whom relevant clinical information was available and who had their diagnostic prechemotherapy core biopsy as well as their posttreatment resection at our hospital were included. In addition, only cases with adequate tissue available in prechemotherapy core biopsies for immunohistochemical studies, to include the 2 antibodies performed in this research, were included. In all cases the hematoxylin and eosin stained sections of prechemotherapy core biopsies were reviewed. Pathology reports from postchemotherapy resections were reviewed, and in select cases these slides were reviewed as well.

Patients who underwent NAC were treated with a taxane containing regimen along with carboplatinum or an anthracycline. If HER2+ they received HER2-targeted therapy with trastuzamab, pertuzamab, or both. Patients included in this study did not receive neoadjuvant hormonal therapy. In general post-NAC resection was done within 4 weeks of the last chemotherapy dose.

Clinical information including patient age, tumor size and lymph node status before NAC and chemotherapy used were obtained through review of the medical records. Pathology reports from postchemotherapy resections were reviewed and information from the pathology report was used to calculate the Residual Cancer burden (RCB) Score.13 Information from postchemotherapy pathology reports were entered in to the RCB online calculator http://www3.mdanderson.org/app/medcalc/index. RCB score of 0 (no residual-invasive disease) and RCB score of I (minimal residual disease) were both considered together as an excellent pathologic response. RCB score of II (moderate residual disease) and RCB score of III (extensive residual disease) were both considered to represent unfavorable pathologic response.

This study was approved by the institutional review board at Montefiore Medical Center and the Albert Einstein college of Medicine.

Back to Top | Article Outline

Immunohistochemistry and TILs

All tissues were fixed in 10% neutral-buffered formalin and paraffin embedded. To determine an immunoprofile for our study immunohistochemical staining on prechemotherapy core biopsies for immune checkpoint protein PD-L1 and for CD68, which is a marker for cells of macrophage/monocyte lineage, was performed. In addition, assessment of the percentage of stromal TILs was performed on prechemotherapy core biopsies. For PD-L1 immunohistochemical determination a rabbit monoclonal antibody Clone SP142 from Spring Bioscience, Pleasanton, CA, was used with a dilution of 1:200 and antigen retrieval at pH 6. For CD68 determination a mouse monoclonal antibody, Clone PG-M1 from Dako, Carpinteria, CA was used at a dilution of 1:100 with antigen retrieval at pH 6.

Assessment of TILs was performed according to the most recent recommendations of the International TILs Working Group.4 Briefly, the percentage of TILs were evaluated for the stroma within the borders of the invasive carcinoma. Only mononuclear cells were counted and an estimate was made of the percentage of stroma occupied by TILs over all fields containing invasive cancer. We scored TILs as a continuous variable and as a categorical variable with a cutoff of >30% TILs to define a lymphocyte-rich invasive breast cancer.

Evaluation for CD68+ macrophages/monocytes were also performed within the borders of the invasive tumor. CD68+ cells were noted to occupy 2 distinct compartments of the tumor; the first compartment was within the tumor stroma and the second compartment was within tumor cells. Hence, the tumor stroma and the tumor cells were scored separately for tumor-infiltrating CD68+ cells. (1) Tumor stroma—when present within the tumor stroma and if CD68+ cells occupied 40% or greater of the tumor stroma the sample was considered 3+ positive for CD68 cells. When <40% of tumor stroma was occupied by CD68+ cells a grading score of 2+ (10 to <40%) or 1+ (10%) was used. (2) Intratumoral—direct infiltration of tumor cells by CD68+ cells was graded as 3+ positive if at least 50% of tumor cells were infiltrated by a diffuse network of CD68+ cells. If only rare or no tumor cells were infiltrated by CD68+ cells this was considered 1+ negative and if 5 to <50% of tumor cells showed any degree of infiltration this was considered 2+. Carcinomas demonstrating either 3+ CD68+ cells in either stroma or tumor cells were considered positive.

PD-L1 expression through immunohistochemistry was assessed for both TILs and tumor cells. Positive staining for PD-L1 of any intensity which was present on >1% of either tumor-infiltrating mononuclear immune cells or tumor epithelium was considered positive.14,15 Tumor cells typically showed membrane staining of variable intensity with or without cytoplasmic staining and tumor-infiltrating immune cells showed both membrane and cytoplasmic staining.

Estrogen receptor (ER) expression and HER2/neu expression were both evaluated in core biopsies before chemotherapy as part of routine pathologic assessment and these results were obtained from our pathology reports. The anti-ER antibody clone used was 1D5 (Dako, Carpinteria, CA) or SP1 (Cell marque, Rockin, CA) and any invasive carcinoma expressing 1% or more positive-staining nuclei was considered ER+ as per American Society of Clinical Oncology and College of American Pathologists guidelines.16 The HercepTest (Dako) was used for determination of immunohistochemical expression of HER2/neu as per American Society of Clinical Oncology and College of American Pathologists guidelines in place at the time of interpretation.17 When the HercepTest yielded a equivocal 2+ result a fluorescent in situ hybridization test was performed using Vysis probe (Abbott Molecular) as per manufacturer’s guidelines. In cases where both tests yielded an equivocal result, the case was considered negative for the purposes of this study.

Back to Top | Article Outline

Statistical Analysis

Associations of HER2/neu expression, ER expression, TILs, CD68 and PD-L1 with an excellent pathologic response to NAC were examined using the Fisher exact tests. An excellent pathologic response to NAC was defined as having an RCB score of 0 or I. Logistic regression models were used to examine associations of TILs, CD68 and PD-L1 with an excellent pathologic response to NAC adjusted for ER expression and HER2/neu expression. Sensitivity and specificity were calculated to assess performance of TILs, CD68, and PD-L1 on identifying an excellent pathologic response to NAC. All statistical analyses were performed using SAS version 9.4 (SAS Inc., Cary, NC).

Back to Top | Article Outline

RESULTS

Study Population

Our study included 76 female patients with invasive breast cancer treated with NAC. The mean (±SD) age was 50.9±11.5 years with a range of 26 to 77 years. The clinical and pathologic characteristics of our patients before chemotherapy are shown in Table 1. Twenty-five (32.9%) patients had a RCB score of 0 representing no residual-invasive carcinoma and 10 (13.2%) had a RCB score of I representing minimal residual disease. Twenty-four (31.6%) patients had a RCB score of II, representing moderate residual disease and 17 (22.4%) patients had a RCB score of III representing significant residual disease.

TABLE 1

TABLE 1

Back to Top | Article Outline

RCB Score by ER and HER2/neu Expression

The primary tumor present in prechemotherapy core biopsies was ER+ and HER2/Neu− in 19 (25.0%) cases, both ER+ and HER2/neu+ in 15 (19.7%) cases, negative for both ER and HER2/neu in 33 (43.4%) cases, and HER2/neu+and ER− in 9 (11.8%) cases (Table 2). The primary tumor was HER2/neu equivocal in 2 patients who are considered HER2/Neu− for the purposes of this study. Among the 19 ER+ Her2/Neu− cases 3 (15.8%) had an RCB score of 0 or I and 16 (84.2%) had an RCB score or II or III. Among the 15 ER+ Her2/Neu+ cases, 5 (33.3%) had an RCB score of 0 or I and 10 (66.6%) had an RCB score of II or III. Among the 33 both HER2/neu− and ER− cases, 19 (57.6%) had an RCB score of 0 or I and 14 (42.4%) had an RCB score of II or III. Among the 9 Her2/Neu+ and ER− cases, 8(88.9%) had an RCB score of 0 or I and 1 (11.1%) had an RCB score of II or III. HER2/neu expression both with and without ER expression was not significantly associated with an excellent pathologic response to NAC (P=0.335). ER expression both with and without HER2/neu was significantly associated with a poor pathologic response (an RCB score of II or III) to NAC, as expected [76.5% (26/34) vs. 35.7% (15/42); P=0.0004; Table 2].

TABLE 2

TABLE 2

Back to Top | Article Outline

Relationship Between Stromal TILs and RCB Score

Stromal TILs were >30% in prechemotherapy core biopsies of invasive breast carcinoma in 19 (25%) cases of which 14 (74.3%) demonstrated an RCB score of 0 or I representing an excellent pathologic response to NAC. TILs was ≤30% in the remaining 57 patients (75%) of whom 21 (36.8%) had an excellent pathologic response (RCB score of 0 or I). TILs>30% in prechemotherapy core biopsies was significantly associated with an excellent pathology response to NAC (P=0.0075; Table 3). ER− Her2/Neu− cancers had the greatest number of cases with TILs>30% (13/33) when compared with other subtypes (6/43) (Table 4). When adjusted for ER and Her2/Neu status, TILs>30% became not significantly associated with an excellent pathologic response to NAC (odds ratio=2.73, 95% confidence interval (CI), 0.76-9.83; P=0.124).

TABLE 3

TABLE 3

TABLE 4

TABLE 4

Back to Top | Article Outline

PD-L1 Expression Associated With RCB

PD-L1 was expressed on >1% of TILs and/or tumor cells in prechemotherapy core biopsies in 27 (35.5%) of the 76 patients of whom 20 (74.1%) had an RCB score of 0 or 1 and 7 (25.9%) had an RCB score of II or III (Fig. 1). PD-L1 was present on ≤1% of TILs and/or tumor cells in 49 (64.5%) patients of whom 15 (30.6%) had an RCB score of 0 or I and 34 (69.4%) had an RCB score of II or III. PD-L1 expression, as determined by immunohistochemistry, when present on >1% of TILs and/or tumor cells was significantly associated with an excellent pathologic response to NAC (P=0.0003;Table 5). ER− Her/2Neu− breast cancers had the highest expression of PD-L1 (20/33) when compared with all the other subtypes (7/43) (Table 4). In multivariate analysis adjusting for ER and Her/2Neu expression, PD-L1 expression was significantly associated with excellent pathologic response to NAC (odds ratio=6.47, 95% CI, 1.80-23.30; P=0.004).

FIGURE 1

FIGURE 1

TABLE 5

TABLE 5

Back to Top | Article Outline

Monocyte/Macrophage Marker CD68 Expression Associated With RCB

Prechemotherapy core biopsies showed 3+ positive expression of the monocyte/macrophage marker CD68 in 22 (28.9%) of 76 cases (Fig. 2). Among the 22 CD68 3+positive tumors, 18 (81.8%) showed an RCB score of 0 or I and 4 (18.2%) had an RCB score of II or III. In the 54 cases which were negative for CD68, 17 (31.5%) had an RCB score 0 or I and 37 (68.5%) had an RCB score of II or III. Positive expression of CD68 was significantly associated with an excellent pathologic response to NAC (P<0.0001; Table 6). Similar to PD-L1 and TILs, highest expression of CD68 was present in ER− HER/2Neu− tumors (16/33) when compared with the other subtypes (6/43) (Table 4). In multivariate analysis adjusting for ER and HER/2Neu expression, CD68 expression was significantly associated with an excellent pathologic response to NAC (R=12.20, 95% CI, 2.90-51.44; P=0.0007).

FIGURE 2

FIGURE 2

TABLE 6

TABLE 6

Back to Top | Article Outline

Combination of TILs and Expression of PD-L1 and CD68 and RCB

To maximize the sensitivity of identifying an excellent pathologic response to NAC, the 3 markers were combined to assess response. TILs>30%, PD-L1>1% positive and CD68 3+ positive were present in 19, 27, and 22, respectively of the possible 35 patients who demonstrated an excellent pathologic response to NAC as measured by RCB Score of 0 or I. Hence, the sensitivity for identifying patients with an excellent pathologic response was 40%, 57.1%, and 51.4% for TILs>30%, PD-L1>1%, and CD68 3+ positive, respectively. The specificity was 87.8%, 82.9%, and 90.2% for these 3 immue parameters respectively. PD-L1 expression was positively associated with TILs>30% in prechemotherapy core biopsies (P<0.0001) and CD68 expression in tumors (P<0.0001). TILs>30% in prechemotherapy core biopsies was positively associated with CD68 expression in tumors (P=0.0001). Using expression of at least 1 of these 3 (ie, PD-L1>1%, TILs>30%, or CD68 3+ positive) immunologic tumor parameters identified 26 of the 35 patients with an RCB score of 0 or 1 (P<0.0001; Table 7). The sensitivity and specificity were 74.3% and 75.6%, respectively.

TABLE 7

TABLE 7

Back to Top | Article Outline

DISCUSSION

NAC is commonly used in the treatment of localized invasive breast cancer. Response to NAC can serve as an in vivo test of tumor chemosensititvity as well as facilitate breast conservation by decreasing tumor size.18 Complete pathologic response or only minimal residual disease after NAC, as defined by RCB 0 or I is highly correlated with improved disease free survival.13,19 Pathologic features which predict for a greater likelihood of pCR include high tumor grade, HER/2neu expression and triple-negative immunophenotype.20 However, even among high grade and triple-negative tumors the majority do not show pCR and some ER+ tumors may respond as well.20 Thus the development of alternative biomarkers which could be used to predict response to NAC are important. The immune system is an important intrinsic mechanism for tumor elimination or containment and tumors exhibiting features of high levels of immune activation have a better prognosis.1 TILs have emerged as a marker for improved prognosis as well as improved response to NAC in invasive breast cancer.9 However, there are other effectors of the immune system which could potentially be used as predictive markers.

The aim of our study was to develop an immune biomarker profile predictive of response to NAC. We evaluated TILs in combination with CD68+ monocyte/macrophage cells and PD-L1 by immunohistochemistry, and correlated them with response to NAC. We found that high TILs, presence of CD68 expressing immune cells or PD-L1 positivity were predictive of excellent response to NAC. Furthermore, using a combination of at least 1 of these 3 immunologic tumor parameters showed a significantly higher level of sensitivity than TILs alone for response prediction (sensitivity, 40% vs. 74.3%), with a specificity level of 75.6%.

In our study, there is a significant association between positive presence of CD68-expressing cells, which include monocytes, and TAMS, and excellent response to NAC. Similar studies looking at CD68+ cells in whole-tissue core sections of human breast cancers before NAC and tumor response are few if any. A significant body of research exists which demonstrates that the presence of a high infiltration of TAMS in the breast cancer microenvironment, as demonstrated by CD68 positivity, is associated with an unfavorable outcome in breast cancer.21 Specifically, high TAMS has been linked to reduced survival, high tumor grade, larger tumor size, and triple-negative phenotype.12,21 TAMS have been shown to enhance tumor progression through promotion of tumor cell proliferation, stimulating angiogenesis, motility and extravasation of tumor cells and to suppress T-cell function.22

In addition to their protumor effects, CD68+ TAMs can also exhibit tumoricidal properties. It has been demonstrated that some chemotherapeutic agents exert their anticancer effects through the tumor killing actions of TAMS. TAMS mediate the antibody dependent cellular cytotoxicity which is the mechanism of anticancer action of monoclonal therapies such as anti HER2/neu therapy.23 In addition, anthracylines have been demonstrated to result in immunogenic cell death which activates the CD68+ dendritic cells. These activated dendritic cells than go on to stimulate tumor specific cytotoxic T cells which mediate the antitumor effects of the drug.7 CD68 is part of the 21 gene oncotype profile where its presence predicts for a greater benefit from chemotherapy.24 Recent research looking at immunohistochemical tumor-immune cell profiles in tissue microarray core sections of breast cancer and response to NAC also demonstrated that a pretreatment profile associated with a high rate of pCR included a high CD68 population of cells.25 Thus a significant body of research exists which supports the role of CD68+ cells in enhancing the effects of chemotherapy. Conversly, TAMs have also been associated with promoting chemoresistance through a variety of mechanisms including a misdirected tissue repair response.26 It is clear that TAMS and other CD68+ cells of the monocyte/macrophage lineage interact with chemotherapies in complex ways. Further studies, possibly to include looking at subsets of TAMS or focusing on their interaction with specific drugs are necessary.

The significant association of CD68+ monocytic cells/TAMs with TILs and PD-L1 expression, as demonstrated in our study, supports that it is part a tumor profile associated with a generalized level of immune activation within the tumor. It is noteworthy that although TILs were assessed only in the stromal component, as recommended by the TILs working group, it has been demonstrated that high-intratumoral TILs are also associated with a favorable response to NAC.4,27 CD68+ macrophages and dendritic cells are part of the mononuclear cells broadly characterized as TILs.4 As we assessed both intratumoral and stromal CD68+ cells, this may have allowed for recognition of additional cases with high-intratumoral TILs only and in part explaining the higher sensitivity of CD68 over TILs alone in response prediction.

A significant body of research exists demonstrating the association between TILs in breast cancer and favorable response to NAC and the results of our study support these findings. However, although we found TILs to be highly specific for RCB 0 or I (specificity of 87.8%), TILs identified less than half of the excellent responders in our study (sensitivity 40%). The addition of either CD68 and/or PD-L1 to the immune profile increased the sensitivity to 74.3% with only a minor loss of specificity (specificity, 75.6%). A recent meta-analysis, which reviewed 13 published studies on the value of TILs in predicting NAC response in breast cancer, concluded that although TILs predict higher pCR in triple negative and HER2/neu+ breast cancer, they were not predictive for pCR in ER+ disease.9 In our study, none of the 3 ER+ cancers which showed RCB 0/1 showed TILs>30%; however, 2 of them were identified by either a positive CD68 score or positive PD-L1. When adjusted for ER expression and HER2/neu expression both PD-L1 and CD68 remained significantly associated with excellent response to NAC, but TILs did not. Thus our results suggest that immune related markers other than TILs may be necessary to signal immune activation.

Previous studies have shown that TILs are prognostic in breast cancer using either a categorical cut point as well as in a linear scale.27,28 However, using a categorical cutoff would be helpful for pathologists to define good prognostic and bad prognostic populations similar to interpretations of HER2 or ER/PR status. Although the TILs cut point for predicting response to NAC varies considerably among studies, there is support in the literature for utilizing the TILs cut point of >30% to predict response as was utilized in our study. Very recent work demonstrated by multivariate analysis that TILs>30% significantly contributed to predicting pCR.29 In the neoadjuvant NeoALTTO trial, which looked at NAC for HER2/neu+ patients only, although any levels of TILs>5% were associated with higher pCR, they noted that only the subgroup of patients with both pCR and TILs>40% defined a group with 100% event-free survival at 3 years, whereas for those with pCR and TILs<40% event-free survival was only 72%.27 Other studies have used higher TILs cut points, including 1 study which used a TILs cut point of ≥60% in the neoadjuvant setting in Her2/neu− breast cancer to define a group with a high pCR.30 Further studies will be necessary to define a TILs cut point for prediction of response to NAC.

In our study PD-L1 expression, as measured by the monoclonal antibody SP142, was significantly correlated with response to NAC and correlated with the other 2 immune parameters of TILs and CD68.31 PD-L1 has been previously demonstrated to be present on tumor cells, lymphocytes and CD68+ macrophages.32 Research has shown that breast carcinoma with PD-L1+ tumor cells were more likely to harbor PD-L1+ TILs and that high number of TILs, which are inclusive of all mononuclear cells, were significantly associated with PD-L1 expression.33 These observations support our findings and suggest these 3 immune parameters are interrelated and are evidence of an active immunogenic tumor response.

PD-L1 is involved in immune evasion.34 Drugs which block the PD-L1 pathways, thereby effectively lifting the immune blockade, have shown response rates in breast cancer.6 Sabatier et al35 have shown that PD-L1 upregulation in breast cancer, as demonstrated by mRNA measurements, is associated with over expression of genes involved in the immune response and more specifically upregulating of cytotoxic T cells which are known to be important mediators of response to chemotherapy. They have also shown PD-L1 upregulation, as measured by mRNA, is associated with pCR after NAC. Rimm and colleagues evaluated TILs and PD-L1 expression by immunohistochemistry in 80 cases of invasive breast cancer treated with NAC and found the PD-L1 expression predicts pCR and correlates with high TILs.11 Thus despite the immune-suppressive function of PDL-1, overall it appears to indicate the presence of an active antitumor immunogenic response and therefore a favorable environment for chemotherapeutic-mediated tumor death.

The main limitations of this study are the small sample size, the retrospective nature and the variable tumor hormone receptor and HER/2neu expression phenotypes which did not allow us to make any specific conclusions about the relationships of CD68, TILs, or PD-L1 for the various tumor receptor subtypes. A strength of this study was the use of core biopsies rather than tissue microarrays, which allow us to evaluate heterogeneity in tumor expression.

Back to Top | Article Outline

CONCLUSIONS

In summary, in our study we have shown that in addition to TILs, immunohistochemical assessment of both PD-L1 expression and CD68 expression by cells of the macrophage/monocyte lineage both show a statistically significant correlation with response to NAC. In addition, an immune biomarker profile which includes assessment of CD68 and PD-L1 in addition to the now widely accepted TILs, is more sensitive (40% vs. 74.3%) for response prediction to NAC in invasive breast cancer. To our knowledge, this is the first study looking at this particular combination of an immune biomarker profile for response. It appears likely that both immune-targeted therapies as well as many conventional cytotoxic chemotherapies are most effective in tumors showing some evidence of immune activation. Therefore, assessment for activation of multiple immune pathways may become clinically important for therapeutic decision making. The methodologies we describe are simple and easily accessible to all pathologists in clinical practice. Validation of the findings in this exploratory study in other cohorts is needed.

Back to Top | Article Outline

REFERENCES

1. Andre F, Dieci M, Dubsky P, et al. Molecular pathways: Involvement of the immune pathways in the therapeutic response and outcome in breast cancer. Clin Cancer Res. 2012;19:28–33.
2. Sugita Y, Wada H, Fujitia S, et al. NY-ESO-1 expression and immunogenicity in malignant and benign breast tumors. Cancer Res. 2004;64:2199–2204.
3. Dushyanthem S, Beavis P, Savas P, et al. Relevance of tumor- infiltrating lymphocytes in breast cancer. BMC Med. 2015;13:202–214.
4. Salgado R, Denkert C, Demaria S, et al. Harmonization of the evaluation of tumor infiltrating lymphocytes (TILs in breast cancer: Recommendation by an international TILs—Working Group 2014. Ann Oncol. 2015;26:259–271.
5. Dirix L, Takacs I, Nikolinakos P, et al. Avelumab (MSB001018C), and anti-PD-L1 antibody, in patients with locally advanced metastatic breast cancer: a phase 1b JAVELIN Solid Tumor Trial. 2015 San Antonio Breast Cancer Symposium, San Antonio, TX, 2015, December 6-10, 2015.
6. Hugo R, Delord J, Im S, et al. Preliminary efficacy and safety of Penbrolizumab in patients with PD-L1- positive estrogen receptor-positive/HER2-Negative advanced breast cancer enrolled in KEYNOTE-028. 2015 San Antonio Breast Cancer Symposium, San Antonio, TX 2015, December 6-10 2015.
7. Apetoh L, Ghiringhelli F, Tesniere A, et al. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat Med. 2007;13:1050–1059.
8. Denkert C, Darb-esfahani S, Loibl S, et al. Anti-cancer immune response mechanisms in neoadjuvant and targeted therapy. Semin Immunopathol. 2011;33:341–351.
9. Mao Y, Qu Q, Zhang Y, et al. The value of tumor infiltrating lymphocytes (TILs) for predicting response to neoadjuvant chemotherapy in breast cancer: a systematic review and meta- analysis. PLOS one. 2014;9:e115103.
10. Muenst S, Schaerli A, Gao F, et al. Expression of programmed death ligand I (PD-L1) is associated with poor prognosis in human breast cancer. Cancer Res Treat. 2014;146:15–24.
11. Wimberly H, Brown J, Schalper K, et al. PD-L1 expression correlates with tumor- Infiltrating lymphocytes and response to neadjuvant chemotherapy in breast cancer. Cancer Immunol Res. 2015;3:326–332.
12. Medrek C, Ponten F, Jirstrom K, et al. The presence of tumor associated macrophages in tumor stroma as a prognostic marker for breast cancer patients. BMC Cancer. 2012;12:306.
13. Symmans S, Peintinger F, Hatzis C, et al. Measurement of residual breast cancer burden to predict survival after neoadjuvant chemotherapy. J Clin Oncol. 2007;25:4414–4422.
14. Solinas C, Buisseret L, Garaud S, et al. Evaluation of PDL1 expression in breast cancer by immunohistochemistry. Ann Oncol. 2015;26:25–26.
15. Herbst R, Soria J, Kowantetz M, et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL328A in cancer patients. Nature. 2014;515:563.
16. Hammond E, Hayes D, Dowsett M, et al. American society of clinical oncology/college of American pathologists guideline recommendation for immunohistochemical testing of estrogen and progesterone receptors in breast cancer. Arch Pathol Lab Med. 2010;134:907–922.
17. Wolff A, Hammond E, Hicks D, et al. Recommendations for Human epidermal Growth factor Receptor 2 for testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update. J Clin Oncol. 2013;31:3997–4013.
18. Kaufman M, Von Minckwitz G, Mamounas EP, et al. Recommendations from an international consensus conference on the current status and the future of neoadjuvant systemic therapy in primary invasive breast cancer. Ann Surg Oncol. 2012;19:1508–1516.
19. Von Minckwitz G, Untch M, Blohmer J. Definition and impact of pathologic complete response on prognosis after neoadjuvant chemotherapy in various intrinsic breast cancer subtypes. J Clin Oncol. 2012;30:1796–1804.
20. Guarneri V, Broglio K, Kau SW, et al. Prognostic value of pathologic complete response after primary chemotherapy in relation to hormone receptor status and other factors. J Clin Oncol. 2006;24:1037–1044.
21. Zhang Y, Cheng S, Zhang M, et al. High infiltration of tumor-associated macrophages predicts unfavorable clinical outcome for node- negative breast cancer. PLOS one. 2013;8:e76146.
22. Noy R, Pollard J. Tumor associated macrophage: from mechanisms to therapy. Immunity. 2014;41:49–61.
23. Sliwkowski MX, Mellman I. Antibody therapeutics in cancer. Science. 2013;341:1192–1198.
24. Paik S, Tung G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative estrogen receptor-positive chemotherapy breast cancer. J Clin Oncol. 2006;24:3726–3734.
25. Maritnwz E, Gil G, Benito A, et al. Tumor-infiltrating immune cell profiles and their change after neoadjuvant chemotherapy predict response and prognosis of breast cancer. Breast Cancer Res. 2014;16:488.
26. Mantovani A, Allavena P. The interaction of anticancer therapies with tumor associated macrophages. J Exp Med. 2015;212:435.
27. Denkert C, Von Minckwitz G, Brase J, et al. Tumor infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without carboplatin in human epidermal growth factor receptor-2 positive and triple negative primary breast cancers. J Clin Oncol. 2015;33:983–991.
28. Salgado R, Denkert C, Campbell C, et al. Tumor infiltrating lymphocytes and associations with pathologic complete response and event-free survival in HER2-positive early-stage breast cancer treated with Lapatinib and Trastuzamab: a secondary analysis of the Neo ALLTO Trial. JAMA Oncol. 2015;1:1–9.
29. Ohtani H, Kazuko M, Nakajima M, et al. Defining lymphocyte-predominant breast cancer by the proportion of lymphocyte-rich stroma and its signify cancer in routine histopathological diagnosis. Path Int. 2015;65:644–651.
30. Issa-nummer Y, Esfahani S, Loibi S, et al. Prospective validation of immunologic infiltrate for prediction of response to neoadjuvant chemotherapy in HER2-Negative breast cancer: a substudy of the neoadjuvant GeparQuinto trial. PLOS one. 2014;8:e79775.
31. Herbst R, Soria J, Kowanetz M, et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature. 2014;515:563–567.
32. Li X, Li M, Zhu H, et al. Prognostic role of Programmed Death Ligand-1 expression in breast cancer: a systemic review and meta-analysis. Targ Oncol. 2016. DOI 10.1007/s11523-016-0451-8.
33. Cimino-Mathews A, Thompson E, Taube J, et al. PD-L1 (B&-H1) expression and the immune tumor microenvironment in primary and metastatic breast carcinomas. Hum Pathol. 2016;47:62–63.
34. Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 2102;12:252–264.
35. Sabatier R, Finetti P, Mamessier E, et al. Prognostic and predictive value of ODL1 expression in breast cancer. Oncotarget. 2014;6:5449–5464.
Keywords:

breast cancer; neoadjuvant chemotherapy; tumor-infiltrating lymphocytes; CD68; PD-L1

Copyright 2018 Wolters Kluwer Health, Inc. All rights reserved.