Abstract
Background/Aim: Immunoscore (IS) is an important evaluation method for the tumor immune microenvironment (TIME); however, formal IS analysis requires designated reagents and a specific digital pathology software and image data analysis. This study aimed to investigate whether simplified IS (s-IS) can substitute formal IS upon modifying the location of the assessment of the numbers of immune cells and verify that the addition of T cell subset markers to s-IS can enhance the prognostic impact in patients with colorectal cancer (CRC). Patients and Methods: A total of 82 CRC cases were included in this study. Immunohistochemical analysis was performed using CD3/CD8/CD45RO/FOXP3 on tissue specimens; the expression levels were calculated in the center and perimeter of the tumors using digital pathology. The clinical prognostic significance of the expression of these markers was investigated by concordance index comparison according to their location of assessment and combinations. Results: In the univariate analysis, the CD3, CD8, and FOXP3 levels were significant prognostic factors. Moreover, for each T cell subset marker, the assessment of each T cell subset marker at the tumor perimeter had a stronger prognostic power than that in the tumor center. The modified s-IS (s-IS plus FOXP3 evaluation) was an independent prognostic factor for recurrence-free survival and overall survival through multivariate analysis and demonstrated the best prognostic power compared to other T subset marker combinations. Conclusion: In CRC, TIME evaluation could be simplified by assessing CD3- and CD8-positive T cells in the perimeter of the tumor, and additional FOXP3 evaluation would empower the ability of s-IS evaluation in prognostic assessment.
The tumor-node-metastasis (TNM) classification is a powerful tool for determining the patient’s treatment strategy. However, immunocompetence is not considered for disease stage determination. Galon et al. proposed an immunoscore (IS) to evaluate anti-tumor immunocompetence. The IS is based on the quantification of lymphocyte populations, particularly CD3- and CD8-positive T cells, both at the center of tumor (CT) and invasive margin (IM). Thus, IS is a better prognostic predictor than the conventional TNM classification for colorectal cancer (CRC) (1).
Based on the IS concept, the cytotoxic T cells (CD8) play a central role in anti-tumor immunity; however, subsets of T cells with various immunological functions, such as memory T cells, helper T cells, and regulatory T cells, exist and act in a complex manner in the tumor immune microenvironment (TIME). Thus, the analysis including T cell subsets would provide a deeper understanding of the TIME and beneficial information for accurate stratification of patients with CRC for immunotherapy.
Immune cells are also expected to have different effects based on their spatial localization in the TIME (2). Therefore, the evaluation of lymphocyte spatial location (center vs. marginal) is considered one of the key factors for TIME evaluation.
The proposed IS analysis was incorporated into digital pathology (3). However, their recommendation presents a considerable challenge; therefore, more simplified and easy-to-perform alternatives should be explored. Our study aimed to investigate the simplified evaluation of TIME, including T cell subsets, and assess their spatial localization, which can confer the prognostic value in patients with CRC.
Patients and Methods
Patients and samples. From 2009 to 2010, 165 patients with CRC were surgically treated at the Department of Surgery, Kurume University Hospital. Among these patients, 82 with Stage II and III CRC were included in the study. Double cancer, multiple primary cancer, and preoperative chemotherapy/radiation treated cases were excluded from the study.
Postoperative disease staging was performed according to the 8th edition of the UICC TNM Classification. Clinical and pathological data were collected and retrospectively analyzed. The study was conducted in accordance with the provisions of the Declaration of Helsinki and approved by the Institutional Review Board of Kurume University Hospital (no. 300).
Immunohistochemistry. Paraffin-embedded tissue samples were sliced at 4-μm thickness, fixed on a coated glass slide, and labeled with the following antibodies using a Bond-III autostainer (Leica Microsystems, Newcastle, UK). Primary antibodies (diluted) were as follows: CD3 (×300, clone LN10, Leica Microsystems), CD8 (×200, clone 4B11, Leica Microsystems), FOXP3 (×100, clone 236A/E7, Abcam, Cambridge, MA, USA), and CD45RO (×5000, clone UCH-L1, Abcam).
Briefly, sections were heat-treated using epitope retrieval solution 2 (pH 9.0) for 20 min and incubated with FOXP3 antibody for 20 min and CD3 and CD8 antibodies for 15 min. For CD45RO antigen retrieval sections were heat-treated using epitope retrieval solution 2 (pH 9.0) for 5 min and incubated with CD45RO antibody for 15 min. Immunostaining was fully automated. This automated system used a Refine polymer detection system (Leica Microsystems) with horseradish peroxidase-polymer as the secondary antibody and 3,3′-diaminobenzidine (DAB) as the chromogen.
Image analysis and evaluation of expression. All the stained slides were scanned and digitized using NanoZoomer2.0-HT: C9600-13 (Hamamatsu Photonics KK, Shizuoka, Japan). The scanned images were observed using NDP.view2: U12388-01 software (Hamamatsu Photonics K.K.), and five randomly selected points at the CT and IM were captured and stored as JPEG images at a 200×field view. The stored images were processed and evaluated using ImageJ software; for evaluation, deconvolution of the images was performed, and a red image was selected, following which a binary image was created. The color density threshold was set to a constant value, and positive cells were counted and quantified using particle count.
Evaluation of CD3, CD8, CD45RO, FOXP3. The number of CD3-, CD8-, CD45RO, and FOX-P3 positive lymphocytes was quantified in five randomly selected locations at the CT and IM. The median value was calculated, and the cutoff values were determined using the receiver-operating characteristic curve analysis.
Simplified IS (s-IS) evaluation. s-IS was quantified based on the particle counts of CD3+ and CD8+ lymphocytes in the CT and IM. Specifically, we quantified TIME from 0 to 4, according to the procedure described in a previous report (4). We denoted our TIME scoring as s-IS because it differs from the method currently recommended by the Immunoscore Consortium using digital pathology (Developer XD digital pathology software, Definiens, Munich, Germany).
Statistical analysis. Correlations between CD3, CD8, CD45RO, FOXP3 expression, s-IS, and the clinicopathological characteristics of patients were analyzed using the chi-square test. Survival curves were estimated using the Kaplan–Meier method, and statistical significance was evaluated using the long-rank test. Overall survival (OS) and relapse-free survival (RFS) were defined as the time from surgery to death and disease recurrence, respectively. Univariate and multivariate analyses of RFS and OS were performed using the Cox hazard model. The association between CD3, CD8, CD45RO, FOXP3, and OS and RFS was examined using the Pearson correlation coefficient test. The accuracy of the prediction of OS and RFS was compared with the CD3, CD8, CD45RO, and FOXP3 values, and the performance of each model was evaluated using Harrell’s C-statistics. All the statistical analyses were conducted using JMP software (version 16.0; SAS Institute Inc., Cary, NC, USA), and a p-value <0.05 was considered statistically significant.
Results
Staining result of each marker. The median value of each lymphocyte particle count in the CT and IM groups was as follows: CD3: CT-151, IM-453; CD8: CT-23, IM-181; CD45RO: CT-10, IM-119; and FOXP3: CT-10, IM-40. The expression values of all the T subset markers were significantly higher at the IM than at the CT.
s-IS and clinicopathological variables. s-IS was classified into five levels (s-IS0 to s-IS4) according to the CD3 and CD8 tumor-infiltrating lymphocyte values at the CT and IM. The cutoff value was determined using the median value of the total particle counts at the CT and IM. The s-IS was categorized into two groups: s-IS High (s-IS3 and s-IS4) and s-IS Low (s-IS0 to s-IS2).
In the s-IS high and s-IS low groups, regarding the clinicopathological factors, only positive lymphatic invasion was significantly more common in the s-IS-L group (p=0.044). However, regarding other factors [sex, mean age, tumor location, pT stage, pN stage, TNM stage (UICC 8th edition), tumor differentiation, venous invasion, microsatellite instability (MSI) status, and prognostic nutritional index (PNI) status], there were no significant differences between the two groups.
s-IS and survival analysis. Univariate and multivariate analyses were performed to compare the effects of the clinicopathological variables and s-IS on the RFS and OS. In the univariate analysis for RFS, the hazard ratio (HR) was significantly lower in the s-IS high group and significantly higher in the lymph node metastasis group. On multivariate analysis, s-IS [high vs. low: HR=0.15, 95% confidence interval (CI)=0.03-0.66] and lymph node metastasis (N0 vs. N1-2: HR=3.53, 95%CI=1.14-10.9) were independent prognostic factors for RFS. However, neither of these factors was statistically significant for OS in this analysis (data not presented in this article).
Comparison of the T cell subset marker localization between the CT and IM. A correlation analysis between the localization of T cell subset markers at the CT and IM was performed to investigate the differences in the T cell subsets and their spatial location (Figure 1). For each subset marker, significant positive correlations were observed for the CT and IM (CD3: r=0.556, CD8: r=0.606, CD45RO: r=0.541, and FOXP3: r=0.567).
Correlation analysis of each T cell subset marker at the center of the tumor (CT) and invasive margin (IM).
Prognostic impact of T cell subset marker localization. The univariate analysis of the effect of T cell localization (CT or IM) on RFS and OS is shown in Table I. High CD3 expression at both the CT and IM significantly lowered the HR for RFS (CT: HR=0.30, 95%CI=0.10-0.93; IM: HR=0.21, 95%CI=0.07-0.65). High CD8 and FOXP3 expression at the IM significantly lowered the HR for RFS (CD8, HR=0.30, 95%CI=0.10-0.91; FOXP3, HR=0.32, 95%CI=0.11-0.90). Furthermore, high CD3, CD8, and FOXP3 expression at the IM significantly reduced the HR for OS (CD3, HR=0.19, 95%CI=0.04-0.90; CD8, HR=0.11; 95%CI=0.01-0.89; FOXP3, HR=0.21, 95%CI=0.04-0.98). In contrast, the CD45RO expression at the CT and IM showed no significant effect on the HR for RFS and OS.
Univariate analysis of relapse-free survival (RFS) and overall survival (OS) in each T cell subset.
The prognostic ability of the different observed localizations of CD3 was compared using the concordance index (c-index) analysis. IM had a higher c-index value than CT for both RFS and OS (RFS, 0.631 vs. 0.738; OS, 0.590 vs. 0.765).
Development of a modified s-IS. We developed a modified s-IS (ms-IS), in which FOXP3 was added to CD3 and CD8, and only the IM was evaluated.
Based on the above evaluation, the cases were reclassified as ms-IS0 to ms-IS3 [mIS0: 21 (25.6%), ms-IS1: 18 (22.0%), ms-IS2: 18 (22.0%), and ms-IS3: 25 (30.5%)]. The ms-IS was categorized into two groups: a high score group with scores of 2-3 (ms-IS High) and a low score group with scores of 0-1 (ms-IS Low).
ms-IS and survival analysis. Univariate and multivariate analyses were performed to investigate the influence of the clinicopathological variables and ms-IS scores on the RFS and OS (Table II).
Univariate and multivariate analysis among ms-Immunoscore and clinicopathological variables for relapse-free survival (RFS) and overall survival (OS).
In the univariate analysis of RFS, a significant difference was observed between the ms-IS and lymph node metastasis (HR=0.17, 95%CI=0.05-0.59; HR=3.45, 95%CI=1.12-10.6, respectively); the multivariate analysis revealed that the ms-IS and lymph node metastasis were independent prognostic factors (HR=0.18, 95%CI=0.05-0.64; HR=3.14, 95%CI=1.02-9.69). Univariate analysis of OS demonstrated a significant difference only for ms-IS (HR=0.10, 95%CI=0.01-0.78), and multivariate analysis also revealed only ms-IS as an independent prognostic factor (HR=0.11, 95%CI=0.01-0.87).
Comparison of the Significance of ms-IS and conventional s-IS. To compare the usefulness of s-IS and ms-IS as prognostic factors, we compared the c-index for RFS and OS of the following four groups: CD3-IM+CD8-IM, s-IS, ms-IS, and CD3-CT+IM+CD8-IM+FOXP3-IM [all univariate significant factor IS (ausIS)].
Upon comparing the prognostic value of the ms-IS group to the other groups, the ms-IS group showed similar or slightly better results for RFS, and the ms-IS was the strongest factor for OS (ms-IS: RFS=0.699, OS=0.728; CD3-IM+CD8-IM: RFS=0.680, OS=0.671; s-IS: RFS=0.691, OS=0.656; ausIS: RFS=0.699, OS=0.681).
Discussion
In this study, we analyzed the degree of CD3, CD8, FOXP3, and CD45RO infiltration in the TIME of CRC and found that the expression of each subset in CT and IM was positively correlated. We also identified that the IM than the CT was more important as the evaluation. Finally, we proposed an s-IS, a simplified measurement of TIME, and further suggested an ms-IS for the prediction of a better prognosis in patients with CRC.
Significance of tumor-infiltrating lymphocyte (TIL) evaluation at the IM. TIME contains a variety of cells, including T lymphocytes, B lymphocytes, macrophages, neutrophils, and dendritic cells. The interaction of these immune cells results in an anti-tumor effect and favorable prognosis for CRC (5). In particular, tumor-infiltrating lymphocytes (TILs) in the TIME are known to play an important role in the anti-tumor effect (6). The localization of TILs in the TIME may reportedly influence their anti-tumor effects (7).
In our analysis, TIL localization at the IM had a strong influence on the OS and RFS according to the univariate and c-index analyses. Spatially and temporally, tumor cells logically infiltrate or metastasize to distant sites more frequently from the IM than the CT (8). The presence of immune cells at the IM may have had a significant impact on the prognosis because it is a factor that hinders the migration of metastatic cells. Several studies have reported that the degree of TIL infiltration at the CT compared to that of the IM is not associated with the prognosis, which is consistent with the results of our analysis (9-11).
Significance of FOXP3 expressing cells. FOXP3 expressing regulatory T cells (Tregs) are usually considered suppressive of immunity, so that the inclusion of FOXP3 seems to be contradictory. However, the incorporation of FOXP3 enhanced the prognostic effect in this study.
Effective anti-tumor immunity requires well-organized TIME, and Tregs play an important role in it (12, 13). FOXP3+ cells reportedly regulate excessive inflammation caused by Fusobacterium nucleatum and suppress the growth of CRC (14). FOXP3+ cells in TILs also reportedly influence the prognosis of CRC (15). Our results demonstrated a more accurate cancer microenvironment response by adding FOXP3 to s-IS, which may possess better prognostic potential than IS.
Usefulness and significance of ms-IS compared to IS. The original Immunoscore requires the use of designated antibodies for IHC and a digital pathology software for international standardization (1). However, considering the beneficial use of IS, it is difficult to meet all the requests owing to complex circumstances, such as lack of money and availability of the designated materials. Thus, the IS needs to be more simple and flexible.
IS secures the prognostic power by evaluating immune cells at the CT and IM. We propose that rather than increasing the number of sites of observation, the addition of a factor, FOXP3, would further improve prognostic ability.
In conclusion, in CRC, TIME evaluation could be simplified by examining CD3 and CD8 expression (CD3 and CD8 positive T cells) in the perimeter of the tumor, and additional FOXP3 evaluation would empower the ability of s-IS evaluation in prognostic assessment.
Acknowledgements
The Authors would like to thank Editage (www.editage.jp) for English language editing.
Footnotes
Authors’ Contributions
Sudo T, Takatou Y, Murotani K, and Akagi Y: Study concept and design, analysis, and interpretation of data, drafting and revision of the article. Kawahara A: Immunohistochemical staining and pathology evaluation. Fujita F, Shigaki T, Fujiyoshi K, Ogata T: Important suggestions on the study and revision of the manuscript. All the Authors have read and approved the final manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
- Received May 13, 2023.
- Revision received June 1, 2023.
- Accepted June 7, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- Contribution of Immunoscore to Survival Prediction in Pancreatic Ductal Adenocarcinoma
- Identification of DNA Repair-related Genes as Biomarkers Reflecting MSI, TMB, and TIL in Colorectal Cancer
- Identification and Quantification of Radiotherapy-related Protein Expression in Cancer Tissues Using the Qupath Software and Prediction of Treatment Response