|
|
||||||||
J Thorac Cardiovasc Surg 2008;135:823-829
© 2008 The American Association for Thoracic Surgery
General Thoracic Surgery |
a Division of Thoracic Surgery, Toronto General Hospital, University Health Network, University of Toronto, Ontario, Canada
b Division of Applied Molecular Oncology, Ontario Cancer Institute, University Health Network, University of Toronto, Ontario, Canada
c Division of Cellular and Molecular Biology, Toronto General Research Institute, University Health Network, University of Toronto, Ontario, Canada
Received for publication April 22, 2006; revisions received July 2, 2007; accepted for publication October 26, 2007. * Address for reprints: Marc de Perrot, MD, MSc, Toronto General Hospital 9N-961, 200 Elizabeth Street, Toronto, Ontario M5G 2C4, Canada. (Email: marc.deperrot{at}uhn.on.ca).
| Abstract |
|---|
|
|
|---|
Methods: We performed an immunohistochemical analysis of 32 extrapleural pneumonectomy specimens to assess the distribution of T-cell subtypes (CD3+, CD4+, and CD8+), regulatory subtypes (CD25+ and FOXP3+), and memory subtype (CD45RO+) within the tumor.
Results: Patients with high levels of CD8+ tumor-infiltrating lymphocytes demonstrated better survival than those with low levels (3-year survival: 83% vs 28%; P = .06). Moreover, high levels of CD8+ tumor-infiltrating lymphocytes were associated with a lower incidence of mediastinal node disease (P = .004) and longer progression-free survival (P = .05). Higher levels of CD8+ tumor-infiltrating lymphocytes were observed in patients treated with cisplatin and pemetrexed than in those treated with cisplatin and vinorelbine (P = .02). Patients presenting high levels of CD4+ or CD25+ tumor-infiltrating lymphocytes or low levels of CD45RO+ also demonstrated a trend toward shorter survival. However, the presence of FOXP3+ tumor-infiltrating lymphocytes did not affect survival. After multivariate adjustment, high levels of CD8+ tumor-infiltrating lymphocytes remained an independent prognostic factor associated with delayed recurrence (hazard ratio = 0.38; confidence interval = 0.09–0.87; P = .02) and better survival (hazard ratio = 0.39; confidence interval = 0.09–0.89; P = .02).
Conclusion: The presence of high levels of CD8+ tumor-infiltrating lymphocytes is associated with better prognosis in patients undergoing extrapleural pneumonectomy for malignant pleural mesothelioma. The stimulation of CD8+ lymphocytes can be a potential therapeutic strategy to improve outcome.
| Introduction |
|---|
|
|
|---|
Clinical studies conducted during the last decade demonstrated that MPM is sensitive to immune-based therapies. Intrapleural administration of interleukin-2,9,10
interferon alpha,11,12
and interferon gamma13
have had some impacts on tumor regression and decreased pleural effusion when combined with chemotherapy. The underlying pathophysiologic mechanism is complex and still largely unknown; however, the encouraging results of these immunomodulating approaches suggest that MPM is potentially immunogenic.
Recent evidence suggests that the tumor microenvironment and the interaction among the tumor cells, immune cells, stromal cells, and extracellular matrix are crucial in tumor progression.14
In particular, tumor-infiltrating lymphocytes (TILs) have been recognized as principal effectors of local antitumor immune response.15
Recent studies also demonstrated the presence of regulatory lymphocytes in the tumor microenvironment that possess various functions in modulating local tumor immunity.16
The presence and role of TILs remain poorly defined in MPM. In this study, we analyzed the impact of both cytotoxic and regulatory phenotypes of TILs on survival in patients with MPM. Immunostaining with T-cell subtypes (CD3, CD4, CD8), regulatory subtypes (CD25, FOXP3), and memory subtype (CD45RO) were used to select the different phenotypes within the tumors.
| Materials and Methods |
|---|
|
|
|---|
Histopathologic Analysis and Construction of Tissue Arrays
All available histologic sections of tumor stained with hematoxylin-eosin were assessed by an investigator (M.A.) and one of our institution's pathologists (K.S.C.) for TILs. On the basis of this assessment, paraffin-embedded samples corresponding to the hematoxylin-eosin–stained slides were processed into tissue array blocks, which consisted of 4 cores per case with a core diameter of 1.5 mm using a tissue microarray instrument (Tissue Arrayer I, Beecher Instruments Inc, Sun Prairie, Wis). Tissue array blocks were then cut into 4-µm sections for hematoxylin-eosin and immunohistochemical staining.
Immunohistochemistry Analysis
The antibodies used were as follows: anti-human CD3 (polyclonal; Dako, Carpinteria, Calif), anti-human CD4 (clone BC/1F6; Abcam, Cambridge, UK), anti-human CD8 (clone 4B11; Novocastra Laboratories, Ltd. Newcastle upon Tyne, UK), anti-human CD45RO (clone OPD4, Dako), anti-human CD25 (clone 4C9; Vector Laboratories, Burlingame, Calif), and anti-human FOXP3 (mAbcam22510; Abcam). Sections from the tissue array blocks were deparaffinized in xylene, washed in phosphate-buffered saline (pH 7.4), and rehydrated through a graded ethanol series. For anti-human CD3 antibody, the enzyme digestion method was applied with 1% pepsin in 0.01 N HCl (pH 2.0) at 37°C for 15 minutes. For anti-human CD8, CD25, CD45RO, and FOXP3 antibodies, deparaffinized sections were immersed into 10 mmol/L of preheated citrate buffer (pH 6.0), incubated at 95°C for 20 minutes, and allowed to cool to room temperature. For anti-human CD4 antibody, preheated Tris-EDTA buffer (pH 9.0) was used for antigen retrieval at 95°C for 20 minutes. Endogenous peroxidase and biotin activities were blocked respectively using 3% hydrogen peroxide and an avidin/biotin blocking kit (Vector Laboratories). After blocking for 15 minutes with 10% normal goat serum, sections were incubated accordingly at room temperature for either 1 hour (anti-human CD3, CD8, and CD45RO) or overnight (anti-human CD4, CD25, and FOXP3) with each primary antibody. NovaRed solution (Vector Laboratories) was applied for color development followed by counterstaining with hematoxylin. Negative control slides omitting primary antibodies were included for all experiments.
Assessment of Tumor Cell Apoptosis
To analyze the interrelationship between the presence of TILs and tumor-cell apoptosis, we performed terminal deoxyuride-5'-triphosphate biotin nick end labeling (TUNEL) on the tissue array slides. The In Situ Cell Death Detection Kit (Roche, Nutley, NJ) was used according to the manufacturer's protocol. To distinguish tumor cells from other types of cells, a double immunofluorescent staining with antihuman mesothelin antibody (clone SPM143; Abcam) was performed after TUNEL. Mesothelin is a maker for mesothelium-derived cells, including MPM.18
Slides were incubated with antihuman mesothelin antibody at room temperature for 2 hours followed by conjugation with Alexa Fluor 488 (Invitrogen, Carlsbad, Calif) as a secondary antibody. Nuclei were stained with Hoechst 33258 (Sigma, St Louis, Mo). Only cells positive for both TUNEL and antihuman mesothelin were considered as apoptotic tumor cells.
Positive Cell Quantification
Antibody stained slides were digitally scanned using Aperio ScanScope CS (Aperio Technologies, Vista, Calif), and the scanned images were analyzed with ImageScope (Aperio Technologies) for the positive cell quantification. All 4 spots of each case were evaluated for TILs, and a minimum of 4 independent areas (up to 6 areas) with the most abundant immunohistochemically positive TILs were selected at a size of 0.0625 mm2. Positive cells in the selected areas for each antibody were counted independently by 2 investigators (M.A. and Z.Y.) without knowledge of clinical information. The averages of positive TILs for each antibody were used for statistical analyses. For the apoptotic tumor-cell quantification, both TUNEL and mesothelin-positive cells were counted for all 4 spots per case on the tissue array slides, and the average was taken for the following analyses.
Statistical Analysis
Kaplan–Meier survival curves were used to estimate survivals on the basis of the levels of positive TILs and were compared with the log-rank test. Multivariate Cox proportional hazards models were used to calculate adjusted hazard ratios. Covariates included in the multivariate analyses were the level of CD8+ T cells (high vs low level), tumor stage (stage II vs stage III/IV), and cell type (epithelioid vs others). Cutoff values for the positive TIL counts to define subgroups were the 25th or 50th percentile, that is, the top 25th or 50th percentile was defined as a high level and all others were defined as low levels. Correlation between the numbers of positive TILs and the patients' clinicopathologic factors were tested with either the t test or the Wilcoxon–Mann–Whitney test, depending on the distribution of positive cell counts. The normality of the distribution of positive cell counts for each antibody was determined with the Shapiro–Wilk test (the distribution of TIL counts was judged as a normal distribution when the P value was more than .05). Apoptotic tumor cell counts were analyzed in association with TIL counts and clinicopathologic factors. The chi-square test was used when categoric variables were analyzed. All analyses were performed with JMP 5.0 software (SAS Institute Inc, Cary, NC).
| Results |
|---|
|
|
|---|
|
|
|
Tumor Apoptosis and Tumor-infiltrating Lymphocytes
Mesothelin was positive for all tumors in the current study subjects (
Figure 3). The median number of both TUNEL and mesothelin double-positive cells (ie, apoptotic tumor cells) was 11.4 (range, 0–85). Significantly higher numbers of apoptotic tumor cells were observed in the tumors with high levels of CD8+ TILs compared with those with low levels (P = .02). No correlations were observed between the presence of apoptotic tumor cells and other types of TILs. There was no significant difference in the numbers of apoptotic tumor cells in the tumors treated with cisplatin and pemetrexed when compared with those treated with cisplatin and vinorelbine (P = .7).
|
Overall estimated 3- and 5-year survivals were 35% and 28%, respectively. The median survival for all patients was 12.2 months. Tumor recurrence was found in 17 patients (53%) at either local (n = 6) or distant (n = 11) sites. There were 3 postoperative deaths, and these were excluded from the analyses to assess the correlation between survival and the levels of TILs. CD3+ (pan-T-cell marker) TILs were not associated with either prolonged or shortened survival (P = .8). There was a clear correlation between high levels of CD8+ TILs and overall survival (
Figure 4, A), as well as progression-free survival (Figure 4, B). High levels of helper TILs (CD4+) and regulatory/activated TILs (CD25+) indicated poor prognosis (3-year survival: 15% vs 53%; P = .08, and 17% vs 45%; P = .09, respectively;
Figure 5, A and B). No difference in survival according to the levels of FOXP3+ TILs was observed (P = .8; Figure 5, C). High levels of CD45RO+ (memory T-cell marker) TILs adversely affected survival, although the difference between the groups did not reach statistical significance (P = .08; Figure 5, D). After multivariate adjustments, the presence of CD8+ TILs remained a significant prognostic factor for both prolonged progression-free survival (hazard ratio = 0.38; 95% confidence interval = 0.09–0.87; P = .02) and overall survival (hazard ratio = 0.39; 95% confidence interval = 0.09–0.89; P = .02) (
Table 2).
|
|
|
| Discussion |
|---|
|
|
|---|
Chemotherapy may alter the tumor microenvironment and increase the susceptibility to an antitumor immune response. It has been shown that chemotherapy-induced apoptotic tumor cells can potentially augment antitumor cytotoxic T-cell responses.23
In our study, we observed that the presence of TUNEL-positive apoptotic tumor cells correlated with high levels of CD8+ TILs. In addition, a significant correlation between induction cisplatin/pemetrexed chemotherapy and high levels of CD8+ TILs was also observed in our study, suggesting that this drug combination may induce greater antitumor immune responses. Because cisplatin/pemetrexed chemotherapy is the only regimen that has been shown to have a significant impact on survival in a randomized clinical trial for MPM,24
the mechanism by which the chemotherapy-induced apoptosis augments CD8+ T-cell–mediated antitumor responses is of great interest.
In our study, we found that CD25+ TILs were negatively correlated with prognosis; however, this was not statistically significant. Regulatory T cells, which are characterized by the expression of both CD4 and CD25, can induce a "tumor tolerance" state by suppressing effector or helper T cells.19
Regulatory T cells can also impair the function of dendritic cells that are indispensable in antigen presentations to both CD4+ helper and CD8+ cytotoxic T cells. On the contrary, FOXP3, a specific marker for naturally occurring regulatory T cells and a crucial transcription factor that mediates the regulatory T-cell development and function,25
did not have any impact on the posttreatment prognosis of MPM. This may be partly because the numbers of lymphocytes positive for FOXP3 were small in our study. Although there are a few studies exploring the role of regulatory T cells in MPM, it is often not well defined because of their small sample size.26,27
Hegmans and colleagues26
demonstrated a massive influx of CD4+CD25+FOXP3+ T cells in 4 biopsy specimens of MPM, and DeLong and colleagues27
reported fewer CD4+CD25+ T cells in pleural effusions from 7 patients with MPM when compared with those from patients with non–small cell lung cancer or breast cancer. Extended research including functional studies of TILs is required to determine the role of regulatory T cells in MPM, especially their interaction with CD8+ cytotoxic T cells.
| Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
D. Roulois, V. Vignard, F. Gueugnon, N. Labarriere, M. Gregoire, and J.-F. Fonteneau Recognition of pleural mesothelioma by mucin-1(950-958)/human leukocyte antigen A*0201-specific CD8+ T-cells Eur. Respir. J., November 1, 2011; 38(5): 1117 - 1126. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. K. Broussard and M. L. Disis TNM Staging in Colorectal Cancer: T Is for T Cell and M Is for Memory J. Clin. Oncol., February 20, 2011; 29(6): 601 - 603. [Full Text] [PDF] |
||||
![]() |
M. Anraku, T. Tagawa, L. Wu, Z. Yun, S. Keshavjee, L. Zhang, M. R. Johnston, and M. de Perrot Synergistic Antitumor Effects of Regulatory T Cell Blockade Combined with Pemetrexed in Murine Malignant Mesothelioma J. Immunol., July 15, 2010; 185(2): 956 - 966. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Lucchi, A. Picchi, G. Ali, A. Chella, G. Guglielmi, A. Cristaudo, G. Fontanini, and A. Mussi Multimodality treatment of malignant pleural mesothelioma with or without immunotherapy: does it change anything? Interact CardioVasc Thorac Surg, April 1, 2010; 10(4): 572 - 576. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Jackaman, S. Cornwall, A. M. Lew, Y. Zhan, B. W. S. Robinson, and D. J. Nelson Local effector failure in mesothelioma is not mediated by CD4+ CD25+ T-regulator cells Eur. Respir. J., July 1, 2009; 34(1): 162 - 175. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |