|Year : 2020 | Volume
| Issue : 2 | Page : 320-326
Comprehensive analysis of the association between tumor-infiltrating immune cells and the prognosis of lung adenocarcinoma
Yitong Pan1, Yeqin Sha2, Hongye Wang2, Hao Zhuang2, Xiaohan Ren3, Xianji Zhu4, Xiyi Wei3
1 Department of Respiratory Medicine, Shanghai Pudong Hospital, Fudan University Pudong Medical Center, 2800 Gongwei Road, Huinan Town, Pudong, Shanghai 201399; Department of Bioinformatics, School of Biomedical Engineering and Informatics, Nanjing Medical University, Nanjing, China
2 Department of Bioinformatics, First Clinical Medical College of Nanjing Medical University, Nanjing, China
3 Department of Urology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
4 Department of Respiratory Medicine, Shanghai Pudong Hospital, Fudan University Pudong Medical Center, 2800 Gongwei Road, Huinan Town, Pudong, Shanghai 201399, China
|Date of Submission||06-Nov-2019|
|Date of Decision||22-Jan-2020|
|Date of Acceptance||11-Mar-2020|
|Date of Web Publication||28-May-2020|
Department of Respiratory Medicine, Shanghai Pudong Hospital, Fudan University Pudong Medical Center, 2800 Gongwei Road, Huinan Town, Pudong, Shanghai 201399
Department of Urology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, 210 029
Source of Support: None, Conflict of Interest: None
Context: Increasing evidence has indicated an association between immune cell infiltration in lung adenocarcinoma (LUAD) and clinical outcomes.
Aims: This study aimed to investigate the effect of 22 tumor-infiltrating immune cells (TIICs) on the prognosis of patients with LUAD.
Settings and Design: This was a case–control study.
Materials and Methods: The CIBERSORT algorithm calculated the proportion of cases from the Cancer Genome Atlas (TCGA) cohort. Cox regression analysis evaluated the effect of TIICs on the prognosis of LUAD. The immune risk score model was constructed based on a statistical correlation. Multivariate cox regression analysis investigated independent factors. P < 0.05 was considered to be statistically significant.
Results: Certain immune cells had differential infiltration between normal tissues and LUAD. Univariate Cox regression analysis revealed that four immune cell types were statistically correlated with LUAD-related survival risk, and an immune risk scoring model was constructed. The results indicated that patients in the high-risk group were associated with poor outcomes. In addition, the multivariate cox analysis revealed that the immune risk scoring model was an independent factor for LUAD prognosis prediction. Ultimately, a nomogram was established to comprehensively predict the survival of LUAD patients.
Conclusions: TIICs played an essential role in the prognosis of LUAD. Furthermore, the immune risk score was a poor predictive factor of LUAD, and the established model was reliable in predicting the prognosis of LUAD.
Keywords: Lung adenocarcinoma, nomogram, prognosis, tumor-infiltrating immune cells
|How to cite this article:|
Pan Y, Sha Y, Wang H, Zhuang H, Ren X, Zhu X, Wei X. Comprehensive analysis of the association between tumor-infiltrating immune cells and the prognosis of lung adenocarcinoma. J Can Res Ther 2020;16:320-6
|How to cite this URL:|
Pan Y, Sha Y, Wang H, Zhuang H, Ren X, Zhu X, Wei X. Comprehensive analysis of the association between tumor-infiltrating immune cells and the prognosis of lung adenocarcinoma. J Can Res Ther [serial online] 2020 [cited 2020 Jul 5];16:320-6. Available from: http://www.cancerjournal.net/text.asp?2020/16/2/320/285204
FNx01Yitong Pan, Yeqin Sha and Hongye Wang contributed equally in this work.
| > Introduction|| |
Lung cancer, also known as lung carcinoma, is a malignant lung tumor characterized by uncontrolled cell growth in tissues of the lung. Recent worldwide statistics have shown that lung cancer occurred in 1.8 million people and resulted in 1.6 million deaths, making it the most common cause of cancer-related death in men and the second most common in women after breast cancer. Moreover, the most common age at diagnosis is 70 years. Non-small-cell lung carcinoma (NSCLC) is the most prominent subtype of lung cancer, with lung adenocarcinoma (LUAD) being one of the three most common histologic types of NSCLC. LUADs account for approximately 50% of lung cancer cases each year. Currently, the evaluation of the prognosis of LUAD primarily depends on tumor–node–metastasis (TNM) staging, and surgical resection was regarded as the most effective treatment for patients with early LUAD. However, nearly a quarter of patients with LUAD relapse after surgical resection., Therefore, it is critical to accurately assess a patient's survival risk and predict the prognosis.
It is well known that the immune system plays an essential role in tumor progression. The immune response involves the interaction between several specific cells, which can affect the clinical outcome of LUAD. It has been reported that the function and composition of tumor-infiltrating immune cells (TIICs) vary among differing host immune status that can be effectively targeted by medications. Nevertheless, these therapeutics may have beneficial or adverse effects on the clinical outcome of patients. The CIBERSORT method is a gene expression-based deconvolution algorithm that uses a set of barcode gene expression values to characterize immune cell composition. Compared with traditional immunohistochemistry and flow cytometry methods, the CIBERSORT algorithm can comprehensively, quickly, and accurately infer the relative proportion of 22 types of invasive immune cells in tumors. Given this method's several benefits, a large number of studies have recently used this method to investigate the effect of 22 tumor immune cell subtypes on prognosis.,, In this study, we used the CIBERSORT algorithm to calculate the infiltration ratio of 22 immune cells in LUAD and to investigate the effects of 22 TIICs on the prognosis of patients with LUAD. Furthermore, an immune–risk-scoring model and a nomogram model were constructed to predict the survival rate of LUAD.
| > Materials and Methods|| |
The transcriptome data of LUAD were derived from the public database, the Cancer Genome Atlas (TCGA). We identified and downloaded the transcriptome data of 594 patients from the TCGA database through the R-package “TCGA-Assembler (Bell Laboratories, New Jersey, USA),” including 59 cases of normal patients and 535 cases of patients with LUAD. Furthermore, relevant clinical information of the 535 patients with LUAD was obtained, such as age, gender, tumor stage, survival status, and survival duration. Finally, the “limma” package in the R software was utilized to correct the downloaded transcriptome data. All the data were utilized under the ethics committee permission of Shanghai Pudong Hospital, Fudan University Pudong Medical Center.
Assessment of immune infiltration
CIBERSORT is a deconvolution algorithm that uses 547 tag gene expression values to characterize the composition of immune cells in tissues. In this study, we used this algorithm to estimate the relative proportion of 22 infiltrating immune cells in the transcriptome data of the corrected LUAD. We uploaded the corrected transcriptome data to the CIBERSORT website (http://cibersort.stanford.edu/) and set the algorithm to 1000 rows. P < 0.05 was considered to be statistically significant.
All analyses were performed using R 3.6.1 (Bell Laboratories, New Jersey, USA). All statistical tests were two sided, and P < 0.05 was considered statistically significant. Continuous variables that conformed to the normal distribution were compared with the use of independent t-test for comparison between groups, whereas continuous variables with skewed distribution were compared with the Mann–Whitney U-test. The relationship between immune cell infiltration and overall survival was analyzed through the Kaplan–Meier (KM) curve which was evaluated by the log-rank test. Time-dependent receiver operating characteristic (ROC) curves were used to analyze the sensitivity and specificity of the recurrence prediction model. The one-factor Cox regression model was used to analyze the effects of individual variables on survival, and the multivariate Cox regression model was used to analyze independent impact factors associated with survival. The nomogram was constructed with the regression coefficients in the Cox analysis.
| > Results|| |
Distribution of tumor-infiltrating immune cells
In this study, the dataset included 535 cases of LUAD samples and 59 cases of normal tissues. The clinicopathological characteristics of these samples are summarized in [Table 1]. The composition of TIICs in LUAD and normal samples is shown in [Figure 1]a and [Figure 1]b, respectively. The correlation between immune cells is depicted in [Figure 1]c. The violin diagram revealed that there were large differences in the composition of TIICs between normal tissue and LUAD tissue [Figure 1]d. Excluding CD8+ T cells, activated natural killer (NK) cells, and activated mast cells, the remaining immune cells were differentially infiltrated between normal tissues and LUAD tissues. According to [Figure 2], macrophages M2 and B memory cells were most significantly associated with tumor progression.
|Table 1: Clinical characteristics of 482 lung adenocarcinoma patients derived from the Cancer Genome Atlas database|
Click here to view
|Figure 1: The composition of tumor-infiltrating immune cells in lung adenocarcinoma and normal samples. (a and b) Expression of 22 tumor-infiltrating immune cells in lung adenocarcinoma and normal samples. (c) The correlation between 22 immune cells. (d) Differentially expressed immune cells between lung adenocarcinoma and normal tissues|
Click here to view
|Figure 2: Correlation analysis between tumor–node–metastasis and stage and 22 tumor-infiltrating immune cells in 535 lung adenocarcinoma cases|
Click here to view
Univariate and multivariate Cox regression analysis to establish an immune–risk score
To identify a subset that is statistically associated with the risk of LUAD survival, we used a univariate Cox regression method to analyze the 22 immune cells. P < 0.1 was set as the screening criterion. The results indicated that activated NK cells, resting dendritic cells, activated dendritic cells, and M1 macrophages were associated with LUAD survival risk [Table 2]. We hypothesized that the four screened immune cells had an important impact on the survival risk of patients with LUAD, and a multivariate Cox regression method [Figure 3] was used to construct an immune–risk score model based on these four immune cells. Each patient received a risk score according to the risk model. Based on the median value of the risk score, patients were divided into high-risk and low-risk groups. The KM curve suggested that patients in the high-risk group had a poor prognosis [Figure 4]a. The ROC curve revealed that the risk model had a good sensitivity and specificity in predicting survival risk [Figure 4]b. In addition, [Figure 4]c, [Figure 4]d, [Figure 4]e shows the risk score, survival status, and infiltration patterns of the four immune cells in patients with LUAD, respectively. The correlation between risk scores and clinical indicators is displayed in [Figure 4]f.
|Table 2: Results of univariate Cox regression analysis for 22 immune cells|
Click here to view
|Figure 3: Multivariate regression analysis of four immune cells to establish a risk score model|
Click here to view
|Figure 4: Relevant characteristics of immune–risk-scoring model. (a) The survival curve of immune–risk score. (b) Area under the curve value of the immune–risk-scoring model. (c) The distribution of patients' risk score. (d) The distribution of patients' survival state. (e) Correlation between four immune cells and survival risk (f)The correlation between 4 immune cells and clinical features|
Click here to view
The immune–risk-scoring model is an independent influence factor for predicting the prognosis of lung adenocarcinoma
To explore whether the constructed immune–risk-scoring model was independent of the patient's age, gender, stage, and other clinicopathological parameters, we performed univariate and multivariate Cox regression analyses for age, gender, stage, TNM, and risk score [Figure 5]a and [Figure 5]b. Multivariate regression analysis showed that both stage and risk scores were independent prognostic predictors for LUAD. On the basis of the coefficients of the multivariate Cox regression analysis, we constructed a nomogram to predict the patient survival rate [Figure 5]c.
|Figure 5: Prognostic value of immune–risk score. (a) Univariate regression analysis of immune–risk scores and other clinical features. (b) Multivariate regression analysis of immune–risk scores and other clinical features. (c) Establishment of a nomogram to predict survival|
Click here to view
| > Discussion|| |
In addition to malignant tumor cells, cancer tissues include immune cells, fibroblasts, endothelial cells, and a large number of cytokines, chemokines, and growth factors. These components and their interactions constitute the tumor microenvironment (TME). TME has an inhibitory effect on malignant cells. However, tumor cells will avoid these inhibitory signals during tumor progression and use immune cells and other conditions conducive to their own environment to promote their own growth, invasion, and metastasis. Previous studies have shown that prognosis and therapeutic response of cancer was closely related to TME, especially tumor-infiltrating immune cells (TIICs).,, Moreover, different types of cancers have different TIICs subpopulations. Even with the same pathological type, the TIIC's subpopulation could be different among patients. For example, in colorectal cancer, macrophages M1 and activated dendritic cells are associated with a relatively better prognosis. In contrast, M2 macrophages, eosinophils, and neutrophils are associated with a relatively poor prognosis. Therefore, it is crucial to investigate the TIICs for an accurate evaluation of the risk and tumor prognosis.
In order to comprehensively investigate the prognostic value of TIICs in patients with LUAD, we used the CIBERSORT algorithm to estimate the relative proportion of 22 infiltrating immune cells in LUAD tissue. The results showed that except for CD8+ T cells, activated NK cells, and activated mast cells, the remaining 19 kinds of immune cells all displayed differential infiltration in normal tissues and LUAD (P < 0.05). An additional univariate Cox regression analysis showed that activated NK cells, resting dendritic cells, activated dendritic cells, and macrophages M1 were associated with the LUAD survival risk. To date, several studies have shown that NK cells display protumor activity in certain subtypes., Bruno et al. observed that CD56 (+) CD16 (− ) NK cells in lung cancer were associated with vascular endothelial growth factor (VEGF), placental growth factor (PGF), and interleukin-8 (IL-8)/CXCL8 production, which may act as pro-angiogenic cells to promote lung cancer cell proliferation. In addition, Kerdidani et al. recently reported that Wnt1-exposed dendritic cells reduced T-cell-attracting chemokine secretion in a CC- and CXC-dependent manner, suggesting that dendritic cells induce an adaptive immune resistance in LUAD. Concerning macrophages, Yuan et al. revealed that the activation of the PI3K/AKT signaling pathway in macrophages promoted the proliferation, invasion, and migration of LUAD. Their results demonstrated that NK cells, dendritic cells, and macrophages could promote LUAD cell growth, which is consistent with our findings that these subsets of immune cells were associated with a poor prognosis of LUAD.
Furthermore, we used a multivariate Cox regression analysis to construct an immune–risk-scoring model based on these four immune cells. The KM curve indicated that patients in the high-risk group were associated with poor outcomes. The ROC curve indicated that the model was reliable in predicting the survival risk of LUAD. We conducted univariate and multivariate Cox regression analyses on risk score, age, gender, stage, and TNM stages. The results showed that the immune–risk-scoring model was an independent factor for predicting the prognosis of colon cancer. In 2019, based on the systematic evaluation of the immune status of 751 LUAD patients, Yang et al. established an immune infiltrating risk score model for LUAD, which was similar to this study. Through testing, the immune–risk model was demonstrated to be an independent prognostic factor and had a better prognostic value than TNM staging. The developed immune–risk-scoring model represented promising novel signatures for the prognosis prediction of LUAD. At present, the assessment of a prognosis still relies on clinical staging and histopathological classification. However, even under the same stage, patients with LUAD may have a different prognosis. Therefore, exploring effective and specific prognostic TIICs is of important clinical value for improving the prognosis of LUAD. Recently, a large number of studies provided their prognostic models for LUAD. All models showed a higher prediction accuracy and stability than the traditional classification methods.,, Given the rapid development of high-throughput technologies, it is reasonable to believe that our immune–risk-scoring model has great potential for clinical practice. Ultimately, a nomogram was established to comprehensively predict the survival of patients with LUAD, according to the results from the multivariate Cox regression analysis.
In addition, we also found that naive B-cells, memory B-cells, plasma cells, CD4+ T cells, helper T cells, regulatory T cells, monocytes, and other cells had little significance in providing a prognosis. Nevertheless, they were differentially expressed in normal lung tissues and LUAD tissues, indicating that these cells might be closely related to the occurrence and development of LUAD.
Our study inevitably had certain limitations. First, the amount of data published in the public dataset was limited. Thus, the clinical pathology parameters used for analysis in this study were not comprehensive and may lead to potential errors or biases. Second, we did not consider the heterogeneity of the immune microenvironment associated with the location of the immune infiltration. Third, all data series downloaded to construct an immune–risk-scoring model were from Western countries. Thus, the study results may not be applicable to patients in Asian countries. Further research is needed for verification.
In summary, our study demonstrated the utility of immune cells in the prognosis of LUAD. The constructed immune–risk-scoring model was reliable in predicting the prognosis of LUAD, and this risk-scoring model was an independent influencing factor for the prognosis of LUAD.
Financial support and sponsorship
This research was funded by Key Specialty Construction Project of Pudong Health and Family Planning Commission of Shanghai (Grant No. PWZzk2017-30).
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2017. CA Cancer J Clin 2017;67:7-30.
Detterbeck FC, Boffa DJ, Tanoue LT. The new lung cancer staging system. Chest 2009;136:260-71.
Baltayiannis N, Chandrinos M, Anagnostopoulos D, Zarogoulidis P, Tsakiridis K, Mpakas A, et al
. Lung cancer surgery: An up to date. J Thorac Dis 2013;5 Suppl 4:S425-39.
Fedor D, Johnson WR, Singhal S. Local recurrence following lung cancer surgery: Incidence, risk factors, and outcomes. Surg Oncol 2013;22:156-61.
Remark R, Becker C, Gomez JE, Damotte D, Dieu-Nosjean MC, Sautès-Fridman C, et al
. The non-small cell lung cancer immune contexture. A major determinant of tumor characteristics and patient outcome. Am J Respir Crit Care Med 2015;191:377-90.
Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al
. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 2015;12:453-7.
Zhang S, Zhang E, Long J, Hu Z, Peng J, Liu L, et al
. Immune infiltration in renal cell carcinoma. Cancer Sci 2019;110:1564-72.
Zhou R, Zhang J, Zeng D, Sun H, Rong X, Shi M, et al
. Immune cell infiltration as a biomarker for the diagnosis and prognosis of stage I-III colon cancer. Cancer Immunol Immunother 2019;68:433-42.
McDonald KA, Kawaguchi T, Qi Q, Peng X, Asaoka M, Young J, et al
. Tumor heterogeneity correlates with less immune response and worse survival in breast cancer patients. Ann Surg Oncol 2019;26:2191-9.
Sato Y, Yoshizato T, Shiraishi Y, Maekawa S, Okuno Y, Kamura T, et al
. Integrated molecular analysis of clear-cell renal cell carcinoma. Nat Genet 2013;45:860-7.
Bremnes RM, Al-Shibli K, Donnem T, Sirera R, Al-Saad S, Andersen S, et al
. The role of tumor-infiltrating immune cells and chronic inflammation at the tumor site on cancer development, progression, and prognosis: Emphasis on non-small cell lung cancer. J Thorac Oncol 2011;6:824-33.
Wang M, Zhao J, Zhang L, Wei F, Lian Y, Wu Y, et al
. Role of tumor microenvironment in tumorigenesis. J Cancer 2017;8:761-73.
Wang H, Liu Y, Shen K, Dong Y, Sun J, Shu Y, et al
. A comparison between radiofrequency ablation combined with transarterial chemoembolization and surgical resection in hepatic carcinoma: A meta-analysis. J Cancer Res Ther 2019;15:1617-23.
Cui F, Ji J, Lv H, Qu D, Yu C, Yang Y, et al
. Immune responsiveness in a mouse model of combined adoptive immunotherapy with NK and dendritic cells. J Cancer Res Ther 2013;9 Suppl: S162-8.
Xiong Y, Wang K, Zhou H, Peng L, You W, Fu Z. Profiles of immune infiltration in colorectal cancer and their clinical significant: A gene expression-based study. Cancer Med 2018;7:4496-508.
Stabile H, Fionda C, Gismondi A, Santoni A. Role of distinct natural killer cell subsets in anticancer response. Front Immunol 2017;8:293.
Platonova S, Cherfils-Vicini J, Damotte D, Crozet L, Vieillard V, Validire P, et al
. Profound coordinated alterations of intratumoral NK cell phenotype and function in lung carcinoma. Cancer Res 2011;71:5412-22.
Bruno A, Focaccetti C, Pagani A, Imperatori AS, Spagnoletti M, Rotolo N, et al
. The proangiogenic phenotype of natural killer cells in patients with non-small cell lung cancer. Neoplasia 2013;15:133-42.
Kerdidani D, Chouvardas P, Arjo AR, Giopanou I, Ntaliarda G, Guo YA, et al
. Wnt1 silences chemokine genes in dendritic cells and induces adaptive immune resistance in lung adenocarcinoma. Nat Commun 2019;10:1405.
Yuan S, Dong Y, Peng L, Yang M, Niu L, Liu Z, et al
. Tumor-associated macrophages affect the biological behavior of lung adenocarcinoma A549 cells through the PI3K/AKT signaling pathway. Oncol Lett 2019;18:1840-6.
Yang X, Shi Y, Li M, Lu T, Xi J, Lin Z, et al
. Identification and validation of an immune cell infiltrating score predicting survival in patients with lung adenocarcinoma. J Transl Med 2019;17:217.
Luo C, Lei M, Zhang Y, Zhang Q, Li L, Lian J, et al
. Systematic construction and validation of an immune prognostic model for lung adenocarcinoma. J Cell Mol Med 2020;24:1233-44.
Chen G, Dong Z, Wu D, Chen Y. Profiles of immune infiltration in lung adenocarcinoma and their clinical significant: A gene-expression-based retrospective study. J Cell Biochem 2020.
Zhang M, Zhu K, Pu H, Wang Z, Zhao H, Zhang J, et al
. An Immune-Related Signature Predicts Survival in Patients With Lung Adenocarcinoma. Front Oncol 2019;9:1314.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2]