Esophageal cancer is the sixth frequent cause of death among malignant tumors. Due to late-stage diagnosis, about 70% of patients die within 1 year after diagnosis. There are two main subtypes of esophageal cancer described: esophageal squamous cell carcinoma (ESCC) comprising 90% esophageal cancer cases and esophageal adenocarcinoma (EAC).
Available data indicate that in order to understand the pathogenesis of esophageal cancer, it is necessary to understand not only the molecular repertoire of the tumor cells but also the properties of the cells of the tumor microenvironment, which contains various cells of the immune system that support the development of the tumor at all its stages. The escape of the tumor from immunological control is crucial for the survival, progression, and metastasis of the tumor. Tumor cells can suppress the antitumor immune response through the production of various soluble factors, which, in turn, attract and direct the differentiation and activation of stromal cells in the direction necessary for the tumor. In this work, we examined cell populations associated with the tumor immune escape, namely, macrophages and T cells.
TAMs show a number of protumorigenic features. It is widely accepted that macrophages may display a broad spectrum of phenotypes where type 1 (M1) and type 2 (M2) macrophages represent its extremes. M1 stimulate inflammation, produce proinflammatory cytokines, and show antitumor cytotoxic activity; M2 produce anti-inflammatory cytokines, extracellular matrix components, and remodeling enzymes and show high phagocytic and low cytotoxic activities [
Tumor-infiltrating lymphocytes (TILs) represent another important part of tumor stromal cells. They are found in different tumors, and their population is mainly comprised of CD3+ and CD8+ T cells. CD3+ T cells have antitumor activity [
Recent advances in cancer immunological therapeutics have revealed the importance of programmed death-1- (PD-1-) activated signaling. The combination of PD-1 and its ligand PD-L1 is the key immune checkpoint for inhibition of T cell activation. Recently developed PD-L1 inhibitor antibodies are now used for treatment of various cancers including esophageal cancer. However, in contrast to many other tumors, the association of PD-L1 expression with the clinicopathological relationship in ESCC remains controversial. Some studies demonstrated that PD-L1 expression correlates with poor prognosis [
In this study, we have examined the prognostic impact of different components of tumor stroma basing on immunohistochemical analysis of macrophage and T cell markers in a group of 48 curatively resected esophageal squamous cancers. We established that out of all macrophage markers studied, only CD206 correlates with the clinicopathological features of the tumor. Analysis of survival revealed that the number of CD163+ TAMs and FOXP3+ TILs correlates with prolonged survival of the patients. We also tested PU.1 as a potential general marker for macrophages and demonstrated its high correlation with CD68, which confirms our hypothesis of the possible use of nuclear PU.1 staining for labelling TAMs.
A total of 48 surgically resected and formalin-fixed paraffin-embedded (FFPE) human ESCC tissues were collected from the Clinical Oncology Department of N. N. Blokhin Russian Cancer Research Centre (Moscow, Russia) (collected from 2005 to 2012). The patients consisted of 36 men and 12 women with an age range of 43–79 years old and mean age of 61 years old; all had been diagnosed with ESCC. All specimens were sectioned into 5
Study population.
Category | All cases |
---|---|
Age | |
≤60 | 26 (54%) |
>60 | 22 (46%) |
Gender | |
Male | 36 (75%) |
Female | 12 (25%) |
Stage | |
I-II | 25 (52%) |
III-IV | 23 (48%) |
Nodal Status | |
N- | 25 (52%) |
N+ | 23 (48%) |
Histologic grade | |
G1/2 | 38 (79%) |
G3 | 10 (21%) |
The Institutional Review Board of N.N. Blokhin Russian Cancer Research Center approved the project (approval date 09/2018), and all patients, who were involved in the study, gave written informed consents that their samples could be used for research purposes. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki. Data were analyzed anonymously. All potential participants who declined to participate or otherwise did not participate were eligible for treatment (if applicable) and were not disadvantaged in any other way by not participating in the study.
Standard immunohistochemical procedure was used for staining FFPE sections with antibodies against stromal cell markers. We used the following antibodies: mouse anti-CD163 (Clone 10D6; BIOCARE, USA, 1 : 100 dilution), rabbit anti-CD206 (HPA004114; Sigma, USA, 1 : 2000 dilution), rabbit anti-iNOS (SAB5500152; Sigma, USA, 1 : 150 dilution), rabbit anti-FOXP3 (Clone D2W8E; Cell Signaling Technology, USA, 1 : 200 dilution), rabbit PU.1 (Clone 9G7; Cell Signaling Technology, USA, 1 : 200 dilution), rabbit anti-PD-L1 (E1L3N; Cell Signaling Technology, USA, 1 : 200 dilution), rabbit anti-CD68 (Clone GR021, 61-0184 Genemed, USA, 1 : 100 dilution), mouse anti-CD8 (Clone CD8/144B, 61-0124 Genemed, USA, 1 : 100 dilution), and rabbit anti-CD3 (61-0011 Genemed, USA, 1 : 100 dilution). We used UltraVision Quanto Detection System HRP DAB (Thermo Fisher Scientific, USA).
IHC scoring was done as described [
For iNOS, the sample was considered to have low expression if less than 1% of tumor cells showed positive staining. Samples having more than 1% tumor cells expressing iNOS were considered high expressing.
For evaluation of PD-L1 expression, we used Combined Positive Score (CPS) (PD-L1 IHC 22C3 pharmDx Interpretation Manual—Esophageal Squamous Cell Carcinoma), which is the number of PD-L1 staining cells (tumor cells, lymphocytes, and macrophages) divided by the total number of viable tumor cells, multiplied by 100. According to obtained results, the samples were divided into low expression (less than 1%) and high expression (more than 1%) groups.
The statistical analysis was performed with GraphPad Prism, Version 8.3, software (San Diego, CA, USA).
We used CD68 as a common macrophage marker, CD163 and CD206 as M2 markers, and iNOS as an M1 marker. As an additional general macrophage marker, we used PU.1. We selected PU.1 due to its relatively specific macrophage expression and nuclear pattern of staining which generally simplifies the scoring; in contrast, CD68, CD163, and CD206 demonstrate diffuse membrane and cytoplasmic staining that may lead to uncertainties in quantification.
We found CD68+, CD163+, and CD206+ TAMs distributed in both tumor stroma and tumor islets. By immunohistochemical analysis, in tumor tissue, the median level of CD68+ cells/HPF was 49 (range, 25–87), the median level of CD163+ cells/HPF was 45 (range, 9–104), the median level of CD206+ cells/HPF was 7 (range, 3–37), and the median level of PU.1+ cells/HPF was 58 (range, 12–115). Analyzed cases were divided into groups with high and low number of M2 macrophages as described above. These groups were used to analyze the association with clinicopathological characteristics. For CD206, a tendency for correlation with the histologic grade was observed, though it was not statistically significant (
Clinicopathological characteristics and TAM markers in ESCC.
CD68 | CD163 | CD206 | PU.1 | iNOS | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
High | Low | High | Low | High | Low | High | Low | High | Low | ||||||
Stage | |||||||||||||||
I-II | 12 | 13 | >0.999 | 13 | 12 | >0.999 | 13 | 12 | >0.999 | 13 | 12 | >0.999 | 1 | 24 | 0.044 |
III-IV | 12 | 11 | 11 | 12 | 11 | 12 | 11 | 12 | 6 | 17 | |||||
Nodal status | |||||||||||||||
N- | 11 | 14 | 0.564 | 13 | 12 | >0.999 | 13 | 12 | >0.999 | 12 | 13 | >0.999 | 2 | 23 | 0.237 |
N+ | 13 | 10 | 11 | 12 | 11 | 12 | 12 | 11 | 5 | 18 | |||||
Histological grade | |||||||||||||||
G1/2 | 19 | 19 | >0.999 | 19 | 19 | >0.999 | 22 | 16 | 0.072 | 19 | 19 | >0.999 | 6 | 32 | >0.999 |
G3 | 5 | 5 | 5 | 5 | 2 | 8 | 5 | 5 | 1 | 9 | |||||
Age | |||||||||||||||
≤60 | 13 | 13 | >0.999 | 11 | 15 | 0.385 | 12 | 14 | 0.773 | 12 | 14 | 0.773 | 6 | 20 | 0.106 |
>60 | 11 | 11 | 13 | 9 | 12 | 10 | 12 | 10 | 1 | 21 | |||||
Gender | |||||||||||||||
Male | 18 | 18 | >0.999 | 18 | 18 | >0.999 | 16 | 20 | 0.318 | 18 | 18 | >0.999 | 6 | 30 | 0.662 |
Female | 6 | 6 | 6 | 6 | 8 | 4 | 6 | 6 | 1 | 11 |
Further, we demonstrate that none of the samples contained iNOS+ type 1 macrophages. Though iNOS expression in tumor cells was detected in 12 out of 48 samples in 5 cases, this expression was in less than 1% of cells. Expression of iNOS correlated with the disease stage (
For tumor-infiltrating T cells, the situation was similar to that with macrophage. In nearly all samples (98%), CD3+ cells were detected within tumor stroma and tumor islets, and CD8+ cells were detected in 96% of cases. By immunohistochemical analysis, in tumor tissue, the median level of CD3+ cells/HPF was 40 (range, 16–126), CD8+ cells/HPF was 22 (range 5-65), and FOXP3+ cells/HPF was 5 (range 0-46). For the analysis of correlation with clinical data, the same approach as for macrophages was used. Samples were divided into high- and low-density TIL groups according to positive cell count and evaluated possible correlations with clinicopathological parameters, including age, gender, histological grade, nodal status, and clinical stages (Table
Clinicopathological characteristics and TIL markers in ESCC.
CD3 | CD8 | FOXP3 | |||||||
---|---|---|---|---|---|---|---|---|---|
High | Low | High | Low | High | Low | ||||
Stage | |||||||||
I-II | 15 | 10 | 0.248 | 15 | 10 | 0.248 | 12 | 13 | >0.999 |
III-IV | 9 | 14 | 9 | 14 | 12 | 11 | |||
Nodal status | |||||||||
N- | 15 | 10 | 0.248 | 15 | 10 | 0.248 | 14 | 11 | 0.564 |
N+ | 9 | 14 | 9 | 14 | 10 | 13 | |||
Histological grade | |||||||||
G1/2 | 19 | 19 | >0.999 | 20 | 18 | 0.724 | 19 | 19 | >0.999 |
G3 | 5 | 5 | 4 | 6 | 5 | 5 | |||
Age | |||||||||
≤60 | 12 | 14 | 0.776 | 11 | 15 | 0.385 | 9 | 17 | 0.042 |
>60 | 12 | 10 | 13 | 9 | 15 | 7 | |||
Gender | |||||||||
Male | 18 | 18 | >0.999 | 17 | 19 | 0.740 | 15 | 21 | 0.093 |
Female | 6 | 6 | 7 | 5 | 9 | 3 |
Statistically significant correlation was found solely for FOXP3+ that correlated with the age of patients (
Programmed death ligand 1 (PD-L1) is a ligand for the inhibitory programmed cell death protein 1 (PD-1), which is targeted by several anti-PD-1 and PD-L1 drugs for a variety of human cancers including metastatic squamous cell carcinoma of the esophagus. In our study group, 27% of samples were PD-L1-negative and 62.5% with Combined Positive Score lower than 1% (including PD-L1 negative samples). No statistically significant correlations of PD-L1 expression and clinical parameters were found (data not shown).
To identify markers of potential prognostic significance in the patients with ESCC, the impacts of TAMs and TIL subgroup and other clinicopathological parameters on the prognosis were explored. To establish the prognostic effect of these clinicopathologic characteristics and markers of immune cells, univariate analysis was used (results are presented in Table
Statistical analysis of the prognostic value of immune cells of tumor stroma.
Univariate analysis | |||
---|---|---|---|
HR | 95% CI | ||
CD3 (high/low) | 0.6930 | (0.2996-1.603) | 0.3623 |
CD8 (high/low) | 0.8064 | (0.3478-1.869) | 0.5979 |
FOXP3 (high/low) | 0.4420 | (0.1985-0.9842) | 0.0325 |
CD68 (high/low) | 0.8953 | (0.4038-1.985) | 0.7781 |
CD163 (high/low) | 0.4447 | (0.1957-1.010) | 0.0456 |
CD206 (high/low) | 0.9158 | (0.3994-2.100) | 0.8292 |
iNOS (high/low) | 0.6928 | (0.2635-1.821) | 0.4929 |
PU.1 (high/low) | 0.6414 | (0.2796-1.472) | 0.2700 |
PD-L1 (high/low) | 0.7251 | (0.3171-1.658) | 0.4504 |
We established that increased CD163+ macrophages and FOXP3+ lymphocytes were significantly associated with prolonged overall survival (OS) in ESCC (
Kaplan–Meier curves of overall survival (OS) in esophageal squamous cell carcinoma (ESCC) based on TILs and TAMs.
There is an urgent need for new macrophage markers suitable for immunohistochemical analysis showing nuclear staining. As such a marker, we used PU.1 in this study. It has a nuclear pattern of expression, which makes it easier to evaluate the data and also allow for multiplex analysis together with other macrophage markers.
We performed a correlation analysis of various macrophage markers in esophageal tumor and demonstrated that PU.1 expression strongly correlates with that of CD68 (
Spearman’s rank correlation coefficient for macrophage markers.
IHC analysis of macrophage markers on serial tumor sections also demonstrates highly overlapping staining patterns for PU.1 and other macrophage markers (Figure
Immunohistochemical analysis of CD68, PU.1, and CD163 on serial tissue sections of 2 different tissue samples (magnification 100x).
Tumor immune escape is an important aspect of tumor development that ensures tumor progression. Tumor cells produce soluble factors that modify microenvironment, attract various immune cells, and drive their differentiation to immunosuppressive phenotype. In this study, using various markers of tumor stroma cells, we investigated the immunosuppressive phenotype of esophageal squamous cell carcinoma (ESCC) (Figure
Immunosuppressive cells of esophageal squamous cell carcinoma stroma.
The main cell population we have studied is composed of TAMs. Like other immune effector and regulatory cells, macrophages demonstrate high degree of functional versatility and express different surface markers and secretable factors [
M2 are usually considered to be able to suppress antitumor properties of M1 TAMs and modulate tissue remodeling by producing matrix metalloproteinases, transglutaminases, and extracellular matrix components [
There are contradictory literature data regarding the prognostic value of M2 number in the tumor. In most of the cases, high number of M2 TAMs correlates with poor prognosis, since these macrophages promote vascularization, invasion, and metastasis in many cancer types [
Since the total amount of macrophages is a highly important criterion, there is an urgent need for a macrophage marker that allows clear identification of the cell. We selected PU.1 as such a marker. PU.1 is a transcription factor regulating hematopoietic differentiation pathways [
In the present study, we also explored the impact of TILs on the clinical significance in ESCC. It was demonstrated that high numbers of TILs are a marker of good prognosis and longer survival in ESCC. Particularly, the presence of T cells (CD3+) and T cell subpopulations (e.g., CD4+, CD8+, and CD103+) was established to be markers of a good prognosis [
Another T cell type that has diagnostic and prognostic value in different types of cancer is regulatory T cells, expressing FOXP3. FOXP3 is a member of the forkhead/winged-helix family of transcription factors that is critically involved in the development and function of Tregs [
In conclusion, data we obtained are in a good agreement with a number of studies, indicating that TAMs and TILs may provide important diagnostic and prognostic information for esophageal squamous cell carcinoma. However, discrepancies found suggest that there is a need for a general agreement on the methodology of stromal cell evaluation and specifically macrophage counting. Also, usage of a nuclear marker for macrophage identification can be recommended which will facilitate stromal cell identification.
All data that support conclusions made in the manuscript are included in the manuscript. Raw data, i.e., stained slides, are available for review in our laboratory.
The authors declare that there is no conflict of interests.
O. Kovaleva analyzed data and wrote the original draft preparation. P. Podlesnaya analyzed data. D. Samoilova, M. Rashidova, and V. Mochalnikova performed experiments. A. Gratchev was responsible for supervision, study conceptualization, writing, reviewing, and editing.
We would like to acknowledge the contribution of our colleague Dr. Irina Zborovskaya, who contributed to the design of this study. This work was supported by the Russian Foundation for Basic Research, grant number 18-29-09069.