Loading...

JOURNAL OF CANCER RESEARCH AND ONCOBIOLOGY (ISSN:2517-7370)

Prognostic Value of Measurable Th1 Immune Response in the Tumors and Lymph Nodes of Patients with Lymph Node Positive and Lymph Node Negative Breast Cancer

Archana Thakur1 *, Dana L. Schalk1, Tayson Lin2, Griffin Calme3, Johnson Ung1, PatcharinPramoonjago4, Paul Tranchida, Sudeshna Bandyopadhyay5, Lydia Choi6, Lawrence G. Lum1

1 Department of Medicine, Division of Hematology/Oncology, University of Virginia Cancer Center, Division of Hematology/Oncology 1335 Lee Street, West Complex 7191 Charlottesville, United States
2 Providence Hospital, Southfield, MI, United States
3 Eugene Applebaum College of Pharmacy and Health Sciences, MI, United States
4 Department of Pathology University of Virginia Cancer Center, United States
5 Department of Pathology, Wayne State University, School of Medicine and Karmanos Cancer Institute, Detroit, MI, United States
6 Department of Hematology and Oncology,  Wayne State University, School of Medicine and Karmanos Cancer Institute, Detroit, MI, United States

CitationCitation COPIED

Thakur A, Schalk DL, Lin T, Calme G, Ung J, et al. Prognostic Value of Measurable Th1 Immune Response in the Tumors and Lymph Nodes of Patients with Lymph Node Positive and Lymph Node Negative Breast Cancer. J Cancer Res Oncobiol. 2019 Nov;2(2):125

© 2019 Thakur A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 international License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Abstract

Background

The functional significance of immune cell subsets may provide important information for maximizing the prognostic value of immunoscoring.

Objective

We hypothesize that immune cell population and T cell activation status (Th1 type) in the resected breast tumors and sentinel Lymph Nodes (LN) may serve as immunological biomarkers of tumor aggressiveness and offer useful prognostic information to facilitate specific therapy related clinical decision.

Methods

Biopsies from two groups of patients were examined for T cell activation status, 1) lymph node negative Breast Cancer (BC) and; 2) lymph node positive BC. Tissue slides from sentinel node biopsies and tumor resections from twenty four women who underwent sentinel node biopsy with invasive breast cancer resection were dual stained for CD3/IL10, CD3/IFN-γ or triple stained for CD3/CD20/CD68. Slides were imaged using a Digital Scanner, the density and intensity of each cell type and cytokine was recorded as the number of positive cells per unit tissue surface area.

Results

Our data show that patients who had no evidence of disease after 5 years showed significantly higher IFN-γ expression in BC (p<0.02) and LN (p<0.03) as well as increased infiltration of T cells in both BC (p<0.03) and LN (p<0.02) biopsies while patients who progressed rapidly had no expression of IFN- γ in the BC or in the LN.

Conclusions 

Understanding of the immune cell environment within the tumors and lymph nodes may have implications for improving clinical outcome of cancer patients.

Keywords

Tumor Infiltrating Lymphocytes; Breast Cancer; Immunohistochemistry; Immunoscoring

Introduction

Although T cell infiltration is recognized to be of positive prognostic significance in a broad range of solid tumors [1-11]. Early analyses in colorectal carcinoma show that the majority of patients with stage I and stage II cancer who lack a T cell infiltrate develop disease recurrence within 5 years, whereas the presence of a T cell infiltrate in patients with stage III cancer can predict an unusually long disease-free interval [4]. A subset of patients with breast cancer, renal cell carcinoma, melanoma, ovarian cancer and gastrointestinal stromal tumors with spontaneous T cell infiltrate have similar positive prognostic value [12-16]. A high ratio of CD8+ T cells to Foxp3+ regulatory T cells (Treg cells) in the ovarian tumor microenvironment has been associated with a favorable clinical outcome [17]. Immune scoring of lymph nodes (LN) continues to evolve and is not clearly defined for sentinel lymph nodes in breast cancer (BC). Identifying and quantitating the number and Th1 (interferon-gamma, IFN- γ) and Th2 (interleukin-10, IL-10) phenotypes in the sentinel LN is likely to more clearly delineate the functional significance of T cells and various immune cell subsets, and may provide important information for maximizing the prognostic value of immunoscoring. There was a 20-30% relapse rate in early stage breast cancer patients with no detectable lymph node invasion, while one-third of LN positive BC patients remain free of distant metastasis [18]. In other words, the prognostic gap remains and the vast majority of the patients in both groups had no prognostic marker to assist in driving clinical management based on risk factors. Understanding of the immune cell environment within the tumors and lymph nodes may have implications for improving clinical outcome of cancer patients.

In this report we asked a question if there is a quantifiable immune response in the tumors and sentinel lymph nodes of breast cancer patients and weather immunoscoring and immune profiling is suggestive of clinical outcomes of cancer patients. Tissue sections from breast cancer biopsies along with auxiliary lymph nodes were stained for cancer-infiltrating immune cells (T and B lymphocytes and macrophages) as well as IFN- γ and IL-10, which can stimulate or suppress anti-tumor immune mechanisms, respectively

Since the nodal immune environment is influenced in part by tumor-derived factors [19], we tested a small cohort for cytoplasmic expression of IFN- γ and IL-10 to explore the association between Th1 and Th2 status of the CD3+ T cells along with the CD20+ B cells and CD68+ macrophages in tumor and lymph node biopsies with clinical outcomes of breast cancer patients. Our exploratory studies showed that patients who had no evidence of disease after 5 years showed IFN- γ expression while patients who progressed rapidly had no expression of IFN- γ by T cells in the tumors or in the lymph nodes.

Material and Methods

Patients and tumor tissues

Breast cancer patients included in this study (n = 24) were seen at the Karmanos Cancer Institute, patients who underwent surgery signed the Institutional Review Board–approved informed consent releasing the use of specimens (tissues) for research purposes. The tissues were processed and banked at the Wayne State University Tissue Procurement Facility, tumor and lymph node slides were obtained from 24 patients undergoing surgery for treatment of primary disease. None of the patients received neoadjuvant chemotherapy

Immunohistochemical staining

All tissue blocks were sectioned at 4 μm and mounted on positively charged glass slides. Tissue sections were stained with monoclonal antibodies recognizing CD3 (T cells), CD20 (B-cells), and CD68 (Macrophages) for immune cell population scoring. For cytokine detection, antibodies recognizing IFN- γ and IL-10 were used. Standardization of the dilution, incubation, and development times appropriate for each antibody allowed an accurate comparison of expression levels in all cases. In the entire antibody staining studies conducted, positive and negative control slides were analyzed in parallel. Before use, all mAbs were titrated using breast biopsy and lymph node sections. Staining was carried out on paraffin embedded resected breast tissue sections by dual staining for Th1 using anti-CD3/anti-IFN-γ and Th2 using anti-CD3/anti-IL-10 expression. For double staining, HRP Polymer anti-Mouse secondary antibody and AP Polymer anti-Rabbit secondary antibody was used (GBI Labs Polink DS-MR-Hu A1 Kit). Color was developed using DAB Chromogen and GBI Permanent Red Chromogen followed by counter stain with hematoxylin. Isotype-matched mouse monoclonal antibodies were used as negative controls. Second set of slides were triple stained withanti-CD3/anti-CD19/anti-CD68 antibody cocktail, using automated protocol on a Ventana Discovery Ultra instrument (Roche Diagnostics, Ventana Medical Systems, Tucson, AZ). Sections of human tumor known to contain abundant CD8+ lymphocytes were used as positive controls. Slides were imaged using a Panoramic Digital Scanner and Viewer (Caliper Life Sciences) for semiquantitative evaluation of the density and intensity of each cell type and cytokine. A breast pathologist at the KCI reviewed H&E and all stained slides from tumor and LN biopsies for scoring.

Manual scoring criteria for Intensities and Cell Densities of Immunostained Samples

Immunostaining for cytokines and different cell populations were scored semi-quantitatively for both the intensity and the density of immune cell populations by 2 trained personnel for all sample sets in a blinded fashion. Scoring for staining intensity and cell densities are shown Table 1. Semi-quantitative analysis of IHC stained slides for tumor infiltrating immune cells (TIIC) was performed on paired tumor and lymph node sections. The TIIC were evaluated for overall impression tlow microscopic magnification (40x) followed by a higher magnification (100x– 200x) fields for tumor margin, tumor stained stroma and tumor center. Cell densities for triple stained slides were scored as 0 (-) for the absence of specific immune cells, 1 (+) for sparse 2% - 10% positive staining, 2 (++) for moderately dense = 11% - 30% positive staining, 3 (+++) for dense= 31% - 60% positive staining and 4 (++++) for very dense = 61% - 100% positive staining of CD3+ T cells [Purple or brown in dual stained samples], CD20+ B cells [Brown] and CD68+ macrophages [Yellow]. Intensities were scored as 0 (-) for absence of IL-10 and IFN- γ, 1 (+) for weak staining, 2 (++) for moderate staining, 3 (+++) for strong and 4 (++++) for very strong expression of extracellular and intracellular staining of IFN- γ and IL-10.

Algorithm to analyze staining densities and Intensities of stained sections

The Python programming language (version 3.5) was chosen for this project due to its open source, high readability, quick to develop and well-supported by computing packages. The third-party packages we used include Num Py for efficient multidimensional array and linear algebra operations, Matplotlib for visualizations, and ScikitImage for image operations and several computer vision algorithm implementations. The source code and documentation are accessible through a repository on GitHub (search “Micro Deconvolution”). The digital images were imported into Micro Deconvolution software and analyzed using the program’s ability to differentiate immunostained cells and hematoxylin-stained nuclei as well as extracellular staining. Software based scoring was called Digital Scoring (DS).

Validation

To validate DS results, we first performed an acceptability check of the DS data; high-level visual assessment was performed at high resolution for each slide to identify major errors in the DS, such as incorrect exclusion of histological artifacts etc. In parallel stained slides were validated by a pathologist and two scientists blindly manually for Scoring (MS).


Table 1: Shows the manual scoring criteria for intensity

Results

Manual scoring of cell densities and expression intensities of immunostained sections

In Representative IHC results for several paired samples are presented in Figures. 1A-1D showing negative (-) or + to ++++ scores for BC and LN specimens. In general patients with progressive disease (PD) show low infiltration of T cells (Figure 1A) and complete lack of IFN- γ compared to patients with no evidence of disease (NED) (Figure 1B).One progressive disease case (P-636) showed a large infiltration of CD3+ T cells (+++) and macrophages (++) but complete lack of IFN- γ (Figure 1A) suggesting that Th1 response may be of better prognostic significance (Figure 1A and 1B). Figure 1C shows various intensities of CD3+ T cells, IFN-γ and IL-10 scored as + to +++ for BC and LN specimens. Figure 1D shows densities of CD+ T cells, B cells and macrophages (Top panel) and intensities of CD3+ T cells, IFN-γ and IL-10 scored as + to +++ in LN specimens (Lower panel).

Digital scoring (DS) and validation of cd3+ t cells in tumor and lymph node sections

We used automated computer vision algorithms that provided a unique ability to analyze IHC dual or multicolor stained specimens fast and accurately. Using color deconvolution and random walker segmentation, when combined in series, showed consistent segmentation ability. The numbers of CD3+ lymphocytes determined by algorithm based DS were compared with those scored semiquantitatively by the 2 trained personnel for all sample sets in a blinded fashion. We found that the majority of CD3+ lymphocytes that were evident microscopically were detected by the software (Figure 2). The overall numbers of CD3+ cells counted by our software and microscopic evaluation based scoring (manual scoring) showed good agreement (Table 3).

Validation of digital scoring of dual stained CD3/IFN- γ and CD3/IL-10 tumor and lymph node sections

We compared the DAB global threshold with intensity conservation at T = 0.5 (50%) and, adaptive thresholds to correct the regional variations in lighting by utilizing localized intensity information to determine the threshold value for each pixel based on the defined regional block size [20,21]. We concluded that our DS could accurately measure CD3+ TILs and IFN- γ or IL-10 as part of the dual stain and was comparable to manual scoring. These findings support the use of DS to analyze cell densities, intracellular and extracellular intensities in the TME. The DS algorithm we developed should, however, provide a useful analytical tool for multi-color staining analysis (Figures 3 and 4).

Comparison of tumor and LN biopsies in patients with no evidence of disease versus progressive disease

We scored IHC stained slides for cellular infiltrates and cytokines in paired tumor (n=24) and LN (n=24) biopsies for CD3, CD20, CD68, IL-10 and IFN-γ. Tumor biopsies. Tumor (BC) biopsies of patients with no evidence of disease had significantly increased infiltration of CD3+ T cells (p<0.03) compared to patients with progressive disease (median score for T cells, 2.5 vs 1.0) and significantly lower macrophage infiltration (p<0.03) compared to patients with progressive disease (median score for macrophages, 1.0 vs 2.0). Median score of B cells was 1.5 fold higher in patients with no evidence of disease. Striking difference was seen for Th1 cytokine IFN-γ (p<0.02), all eight patients with progressive diseases showed no expression of IFN- γ in tumor biopsies while scores ranged between 0.5-1.0 in 8/16 patients with no evidence of disease (median score of 0.25). No changes were observed for IL-10 expression between two groups. LN biopsies. In LN biopsies, significantly higher relative score for T cells (p<0.02) and increased B-cell infiltration in patients with no evidence of disease compared to patients with progressive disease. No changes were observed for macrophages between two groups. In order to determine the functional basis of immune cell infiltration profiles we analyzed the IFN- γ levels, a marker of Th1 response, and IL-10, a Th2 mediator, by IHC in node positive and node negative breast and LN biopsies. IFN- γ / IL-10 expression was scored negative or + to ++++. Similar to tumor (BC) biopsies, IFN-γ expression was significantly higher (p<0.03) in patients with no evidence of disease compared to patients with progressive disease (median score of 1 vs 0.2). IL-10 was also higher in no evidence of disease patients compared to patients with progressive disease (1.5 vs 1). Figure 5 show the IHC scores for paired tumor (BC)/LN samples. These data suggest that patients with no evidence of disease showed significantly higher IFN- γ expression, consistent with a dominant Th1 response profile, compared to patients with progressive disease (Table 3).


Figure 1 A: Sections of tumor specimens and nodes from 24 patients (12 node+ and 12 node-) were dual immunostained for CD3/IFN-γ or CD3/IL-10. CD3+ TILs were immunolabeled brown and IFN-γ/IL-10 in permanent red in the dual stained slides. Second set of serial sections of tumor specimens and nodes were triple stained for anti-CD3 (purple), anti-CD20 (brown) and antiCD68 (yellow) color indicates distribution of immune cells. Neg (-) to ++ indicates cell densities or intensities of IFN-γ and IL10. Upper panel shows the whole biopsy images showing CD3+/ CD20+/CD68+ tumor infiltrating T cells, B cells and macrophages in resected paired tumor and lymph node biopsies in 3 patients with progressive disease. In general breast tumor biopsies show reduced cellular infiltrates and lack of IFN-γ. Lower panel shows a representative PD case with of large numbers of TILs but unusually high infiltration of macrophages (stained yellow) in breast biopsy may have generated immune suppressive tumor promoting environment.


Figure 1 B: Image analysis show CD3+/CD20+/CD68+ TILs and macrophages in resected tumor and lymph node biopsies in patients with no evidence of disease (NED). Representative whole scanned slide images show the paired tumor and LN sections stained for CD3 (purple), CD20 (brown) and CD68 (yellow). Compared to patients with PD, biopsy sections from patients with NED show increased infiltration of T and B cells, increased number of IFN-γ secreting cells with much reduced infiltration of macrophages. Lower panel shows the corresponding IL-10 and IFN γ from two patients shown in the upper panel, T cells are stained brown and IFN-γ is stained red.