Risk stratification based on immune signature and prediction of immune checkpoint inhibitor’s efficacy for bladder urothelial carcinoma

doi:10.21203/rs.3.rs-3785010/v1

Download PDF

Research Article

Risk stratification based on immune signature and prediction of immune checkpoint inhibitor’s efficacy for bladder urothelial carcinoma

https://doi.org/10.21203/rs.3.rs-3785010/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose：To investigate the impact of the tumor immune microenvironment on survival and response to immune checkpoint inhibitors (ICIs) in patients with bladder urothelial carcinoma (BLCA).

Materials and Methods: We constructed a prognostic immune signature based on the TCGA dataset and verified its predictive accuracy on both internal and total data sets. Firstly, we screened immune genes associated with prognosis based on univariate Cox risk regression analysis. Secondly, LASSO model is used for variable selection and regularization to avoid over-fitting. Thirdly, we also carried out immune status or immunophenoscore (IPS) analysis. Immune genes used in constructing a immune-related multivariate Cox risk regression analysis, and univariate and multivariate Cox regression analysis is used to evaluate the performance of high-risk and low-risk groups in tumor mutation load, patients' response to ICI treatment, tumor infiltrating lymphocytes and other aspects.

Results: This immunomarker consists of 6 immune-related genes, which we confirmed to be an independent prognostic element for BLCA patients. We found that patients in the high-risk group had poorer overall survival, while patients in the low-risk group had better overall survival. Although the difference was not significant, high-risk score patients had fewer mast cell rest and neutrophils, and more memory activation of b cell infants and t cell CD4, which was closely associated with clinical outcome. In addition, the tumor mutational burden (TMB) was significantly reduced in the low-risk group. When using IPS as a substitute for ICI response, we found that patients with low-risk scores had significantly higher IPS.

Conclusion: Our study concluded that 6 immune-related genes signature may be related to prognosis and prediction of ICI treatment, but further verification is needed.

the tumor immune microenvironment

bladder urothelial carcinoma

tumor mutation burden

immune checkpoint inhibitor

Bladder cancer (BC) includes urothelial carcinoma, squamous cell carcinoma and adenocarcinoma, Among them, BLCA is the most common, accounting for more than 90% of BC^[1]. BLCA is one of the most common malignant tumors in the urinary system, which is characterized by rapid tumor growth and high postoperative recurrence rate^[2]. The traditional treatment of bladder tumor mainly includes radical cystectomy and local bladder chemotherapy, but the traditional treatment has high recurrence rate and low survival rate^[3]. At present, the clinical treatment strategy for urothelial carcinoma is gradually transiting from advanced systemic treatment to early disease diagnosis, from single drug treatment to multifactor treatment^[4-7].

Tumor cells can escape the attack of immune system through immune escape, so as to obtain a good growth environment. Therefore, it is of great significance to study the BLCA immune microenvironment on survival and ICI response of BLCA patients.

BLCA often has high TMB and strong immune heredity, so BLCA is an ideal tumor immunotherapy^{[8, 9]}. Some studies have shown that immune-related genes (IRGs) such as TAP1, RBP7, STAT1, PDGFRA, AHNAK, OLR1, and RAC3 are not only potential targets for immunotherapy, but also their relationship and interaction with immune cells are important for further immune-related research have significant value^[10].Many studies have shown that immune checkpoint blockade is a promising approach and has been approved for the treatment of metastatic bladder tumors^[11]. So far, many studies have shown that there are several immunotherapy drugs that block immune checkpoints, Such as cytotoxic T lymphocyt-associated protein 4 (CTLA-4) and programmed cell death 1 receptor (PD-1), Pd-ligand-1 (PD-L1), has been shown to be useful for a variety of malignancies, including BLCA. Studies have shown that the interaction between tumor cells and immune cells inhibits the immune response to some extent, but on the other hand, immune cells promote the progression of tumors. In fact, the immune microenvironment affects the growth and spread of tumors, indicating that the effectiveness of our immunotherapy is closely related to patient survival. Many previous studies have shown that some IRG are biomarkers for the prognosis of breast cancer, colon cancer, and lung cancer ^[10-12]. But few scholars have studied the decisive value of IRGs and immunoassays in BLCA. We build an IRG-based immune characteristics and sought the real relationship between this immune signal and the clinicopathological characteristics of BLCA patients as clinical results in our research. In addition, we not only analyzed the TMB and mapped the immune profile of BLCA patients, but more importantly, predicted the treatment response to ICI in high-risk and low-risk patients.

Data acquiring and cleaning

Our research method is through TCGA data portal website (https://portal.gdc.cancer.gov/) to download BLCA transcriptome sequencing data and clinical information from. Mutation spectrum by R software (version: 4.0.4), was obtained from the TCGA data portal website (https://portal.gdc.cancer.gov/) and package TCGA biological link ^[13]. R software for data cleaning. Which list from the import database (https://www.immport.org/shared/genelists) to download^[14].

Differentially expressed genes

We looked at genes that differ in BLCA expression from normal tissues, we aim to screen for IRGs involved in tumorigenesis. Abnormal expression genes were screened by package limma, edgeR and ggplot2, as described above (filter condition adjusted p<0.05 and logFCfilter=3)^[15-17].

Pathway and function enrichment analysis

We performed pathway and functional enrichment analyses using R software (version 4.0.4) to explore the biological significance of these differentially expressed IRGs. The research methods were also enriched by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Significantly enriched objects are P < 0.05 paths and the term is considered to be. Visualization was done by the clusterProfiler, org.hs.egg.db, enrichment plot, and ggplot2 package.

Constructing an IRG‑related immune signature for BLCA

We randomly selected 2/3 of TCGA BLCA patients as training set to identify prognostic immune genes and construct a prognostic risk model, and the remaining 1/3 of patients were used as testing set for model testing to validate its predictive ability and robustness. First, we screened the immune genes associated with prognosis by univariate Cox regression analysis. The minimum absolute contraction and selection operator (LASSO) analysis was used to screen genes with p< 0.05 to avoid overfitting. Then, a multivariate Cox regression analysis model was used to construct an immune-related risk model for the genes selected in the structured LASSO analysis: risk score = gene a level * coefficient a+ gene b level * coefficient b+ gene c level * coefficient c+...... + Gene n level * coefficient n^[18].

The risk score of the model reflects the prognosis of BLCA patients. The higher the risk score, the worse the prognosis of BLCA patients, conversely, the better the prognosis of BLCA patients. We used the median of the risk score as the cut-off value, which is higher than this value as a high-risk group, and less than this value is a low-risk group. Kaplan-Meier survival curves were used to calculate predictive power, and to assess the predictive power of this immune signature we used time-dependent receiver operating characteristic curves (packed time ROC).

Calculating the ratio of tumor‑infiltrating immune cells

Based on TCGA BCLA transcription profile and R script v1.03 of CIBERSORT, we calculated the proportion of 22 types of tumor-infiltrating immune cells from tissue transcription profiles by deconvolution algorithm.

TMB analysis

We used the software package maftools (MAF file pipeline: muse) for the analysis and visualization of somatic variant data from BCLA patients^[19]. Then, based on the above analysis results, the TMB of each patient was calculated.

Immune status or IPS analysis

Antigenomic and immune atlases of 20 solid tumors were analyzed in The Cancer Immunome Atlas (https://tcia.at/). The quantitative score of tumor immunogenicity is 0 to 10 points, which is called IPS. Relevant studies have confirmed that IPS can predict patients' response to ICI treatment ^[20].We know that the value of IPS is positively correlated with tumor immunogenicity. Our study extracted IPS data from BLCA for analysis. Furthermore, mRNA levels of immune checkpoints and their ligands were compared between high-risk and low-risk groups.

Statistical analysis

Our statistical method is the difference between variables using Student's t test and Chi-square test. Statistical analysis was performed by R software (4.0.4) in this study. Univariate and multivariate cox regression analyses were used to evaluate the effects of immune characteristics and various clinicopathological parameters on patient survival. Heat maps are drawn by packet heat maps. P<0.05 was considered statistically significant.

The characteristics of patients

We downloaded RNA-sequencing profiles and clinicopathological results of 395 BLCA patients from the TCGA-BLCA dataset. BLCA patients were randomly divided into the experimental group (131 cases) and the training group (264 cases). For most clinicopathological parameters in the Table 1, we can see that there is no significant difference between the testing set and the training set.

Table1

Differentially expressed genes analysis

We are based on filter conditions p<0.05 and logFCfilter=3, 4669 differentially expressed genes were screened. The results showed that among 4669 differentially expressed genes, 2726 genes were up-regulated and 1943 genes were down-regulated. We selected 350 differentially expressed IRGs according to the list of differentially expressed genes.

Immune signature can predict the prognosis of BLCA patients

We performed a univariate Cox regression analysis to determine the prognostic value of IRGs. To avoid overfitting, we performed LASSO analysis, which revealed six genes in Figure 1a-b as predictors of patient prognosis. Then, we perform univariate Cox proportional risk regression analysis based on the training set and build a predictive model: risk score=(LRP1*1.388) + (OAS1*0.767) + (CGA*1.303) + (PPY*1.719) + (SCG2*1.299) + (ADCYAP1R1*1.644) in Figure 1c. OAS1 was protective factors while five genes were risk factors.

Figure 1. LASSO analysis and forest plot.

Cross validation.
Lasso coefficient spectrum.
Variable Cox model results.

Establish prediction model based on multivariate Cox regression analysis: risk score=(LRP1*0.215)+(OAS1*-0.192)+(CGA*0.193)+(PPY*0.491)+(SCG2*0.144). We used multivariate Cox regression analysis to obtain a risk score for each patient in the training set. We divided patients into high-risk and low-risk groups using a median risk score as a cut-off. In addition, we ranked the risk scores for all patients and outlined their distribution. The patient's risk score and survival time were represented by a dot plot. We used the heatmap to show the expression patterns of the five IRGs in patients with different risk scores.

In order to verify the accuracy of the training group, we verified the data obtained by the training group through the test set. We divided the test set into high-risk and low-risk groups based on patient risk scores. The results showed that, according to the survival analysis, patients with high-risk scores had a low survival rate and those with low-risk scores had a high survival rate. The prognosis was better than that of the high-risk group (p=0.006), as shown in Figure 2a. The AUC in Figure 2b is 0.639. Survival status, risk score, and survival time distribution of patients were represented by scatter plots, as shown in Figure 2c-d. The expression of the five resulting genes was shown by heat map, as shown in Figure 2e.

Figure 2. The six immune-related gene signatures in the test set were validated. A Survival analysis. B ROC curve analysis. C risk score. D survival status. E heat map

We also evaluated the robustness of 6-IRGs signature. In Figure 3a, risk score is a factor in poor prognosis in BLCA patients (p<0.001). The AUC is 0.689, as shown in Figure 3b. The life status, risk score distribution and expression of 17 IRGs of patients are shown in Figure 3c-e.

Figure 3. The markers of 6 immune-related genes were verified in all sets. A Survival analysis. B ROC curve analysis. C risk score. D survival status. E heat map

The 6‑genes immune signature is an independent prognostic factor for BLCA patients

We used univariate Cox regression analysis to evaluate the impact of patients' clinicopathological factors and immune characteristic risk scores on the overall survival time of patients, as shown in Table 2. It is concluded that age, grade, stage and risk score are the risk factors affecting survival time. Multivariate Cox regression analysis showed that 6-IRGs risk score was an independent prognostic factor (HR= 1.14, 95% CI 1.09-1.18;p<0.001).

Table2

The 6‑genes immune signature and relationship between clinical case parameters

This study analyzed the relationship between 6-IRGs features in Figure 4a-f and clinicopathological parameters of patients, including grade, TNM stage, age, lymph node status, tumor size, and distant metastasis. The results showed that 6-IRGs risk scores were significantly higher in patients with TNM stage, age ≥65 years, lymph node metastasis, large tumor size, and distant metastasis.

Figure 4. Relationship between clinicopathological parameters and immune characteristics of patients.

The 6‑IRGs signature and tumor immune microenvironment

According to the CIBERSORT algorithm, we obtained the proportion of 22 immune cells in each BLCA sample. We then compared 22 types of immune cells in the high-risk and low-risk groups. The proportions of T cell follicular helper and dendritic cells activated were significantly higher in the high-risk group in Figure 5a. In 22 types of immune cells, B cells naive, Mast cells resting, Neutrophils, T cells CD4 memory activated were related to better survival time in Figure 5b-e (on the verge of statistical significance, p=0.039, 0.021,0.031, 0.021, respectively).

Figure 5. Risk score, overall survival and tumor-infiltrating immune cells.

The immune signature and TMB

We analyzed the mutational spectrum of each BCLA patient. The results showed that from the mutant spectrum of the training group and the experimental group, the top 20 genes with the most significant mutations were TP53, TTN, KMT2D, etc(Figure 6AB）.Subsequently, we observed that Survival probability of TMB high-risk group is higher than low-risk group (p=0.006) in Figure 6C. As shown in Figure 6D, we also calculated the TMB of each sample and found that the TMB was lower in the high risk score group (p=0.033).

Figure 6. AB. BLCA patients with high-risk and low-risk population mutant spectrum.

C. Relationship between survival time and tumor mutation load.

D. Differences in tumor mutation burden between high-risk and low-risk populations.

The immune signature and patients response to ICI treatment

IPS predicts BLCA patients' response to ICIs on a computer through a machine learning-based evaluation scheme. Since we did not find information on ICI treatment in the TCGA BLCA dataset, we used two subtype IPS values (IPS-CTLA-4_pos and IPS_PD-1_pos) as proxies for BCLA patients' response to anti-CTLA-4 and PD-1 treatment. In Figure 7 AB, we can see that the relative probability of response to anti-CTLA-4 and PD-1 therapy in low-risk ratings was higher (p=0.0055). Studies have shown that patients with high immune signature scores are not suitable for ICI treatment, while low scores patients may be more suitable for ICI treatment. In addition, we compared the expression levels of immune checkpoints and their ligands in high-risk and low-risk groups, as shown in Figure 7C. It can be seen that PD-L1 was slightly increased in high-risk score patients, and PD-L1 was significantly increased (p=0.01).

Figure 7. PD-1/PD-L1 and CTLA4 immunotherapy analysis

One of the most common cancers of the urinary system is bladder tumors, and treatments for bladder tumors have stagnated over the past 30 years, with few new advances emerging. In recent years, the discovery of immune checkpoints has promoted the development of cancer immunotherapy to a new level, achieving specific blocking of immunosuppressive effects and enhancing the immune response of anti-tumor effects^[21]. The purpose of our study is to construct immune characteristics according to IRGs, explore the real relationship between these immune signals, analyze TMB, draw immune landscape, and use regression analysis to build a prediction model to predict the treatment response of patients with different risk scores to ICI.

In our study, the TCGA-BLCA dataset was used to construct prognostic immune profiles, including 6 IRGs. A previous study used univariate Cox proportional hazard regression analysis and multivariate Cox proportional hazard regression analysis to construct a risk scoring model. 27 immune genes such as OAS1 are related to the prognosis of patients with bladder urothelial cancer^[22]. It is worth noting that most selected IRGs can simultaneously regulate the occurrence and development of cancer and effectively regulate the tumor immune microenvironment and malignant biological characteristics of tumor cells^[23]. We compared the differences in proportions of immune cells between the high-risk score group and low-risk score group. The results demonstrated that the proportion of T follicular helper cells and activated dendritic cells was significantly higher in the high-risk group compared with the low-risk group. Among the 22 types of immune cells, Mast cells resting, B cells naive, T cells CD4 memory activated，Neutrophils were related to better survival time. The microenvironment of tumor tissue is similar to the chronic inflammatory state, which contains many different mediators and inflammatory cells^[24]. Tumor infiltrating neutrophils and lymphocytes (TILs) are the main components of inflammatory cells in the tumor microenvironment. When inflammation and infection occur, neutrophils are often considered the first line of defense^[25]. Regarding the mechanism, reactive oxygen species come from activated neutrophils, which cause tumor cell lysis through oxidation, and damage and stimulate T cell-mediated immune CD8+ T lymphocytes by producing costimulatory molecules that promote CD4+ proliferation. However, it has been found that Neutrophils participate in malignant tumor signals, matrix degrading proteases and angiogenesis mediators by releasing substances that stimulate growth in some cases^[26]. Related studies have found that many TILs are found in the lymphocytes surrounding tumor cells, and they have different functions in different subgroups. CD4+, CD3+, FoxP3+ T lymphocytes and CD8+ are the most common subgroups of TIL. Studies have shown that TILs play an important role in the dual regulation of tumor-related immune response. For example, they can not only inhibit tumor growth, but also help tumors by creating microenvironments that stimulate tumor growth or selectively protect tumor cells in immunocompetent hosts^[27].

We found that immune signature can reflect the state of the tumor immune microenvironment and predict the survival of tumor patients based on RNA-sequencing data. Immune profiles not only predict tumor survival, but also predict patient response to ICI treatment. As mentioned above, we used IPS as a surrogate indicator of ICI treatment efficacy^[20]. In numerous current study, this immune signature is also often used as a predictor of response to ICI therapy in tumor patients. In contrast, our study showed that TMB was significantly lower in the low-risk rating group. Because IPS is a complex model involving multiple factors, we believe that certain variables, such as up-regulation of immune checkpoint signals, may enhance the efficacy of low-risk group therapy.

Our results show that the high-risk scores of patients have lower immune expression checkpoint molecules and conversely, the lower risk scores, the higher immune expression checkpoint molecules. If we increase the level of immune checkpoints, such as CTLA-4 and PD-L1, it indirectly indicates that existing T cells in the low-risk group are activated. Therefore, the high-risk scores of patients are not sensitive to ICI, while the low risk scores of patients may be more sensitive to ICI efficacy.

Despite some positive results, a number of problems remain. First, the construction of this immune feature is based on a public data set. Its predictive power needs to be further validated in prospective trials. In addition, we used IPS to simulate patient responses to ICI efficacy. we used IPS to simulate the response of patient to ICI efficacy. Although previous studies have demonstrated the correlation between IPS and ICI treatment response, IPS cannot completely replace the true treatment response.

Abbreviation	Full name
ICI	immune checkpoint inhibitor
BLCA	Bladder urothelial carcinoma
IPS	immunophenoscore
TMB	Tumor mutation burden
BC	Bladder cancer
IRG	immune-related gene
PD-1	programmed cell death 1 receptor
PD-L1	programmed cell death-Ligand 1
CTLA-4	cytotoxic T lymphocyteassociated protein 4
GO	Gene ontology
KEGG	Kyoto Encyclopedia of Genes and Genomes
LASSO	least absolute shrinkage and selection operator
AUC	areas under the curves

Ethics approval and consent to participate: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee. The Ethics Committee of Tongren Hospital, Shanghai Jiao Tong University School of Medicine approved the study.

Consent for publication: yes

Availability of data and material: All the original data of this study can be made public and can be provided at any time as needed.

Funding: This study was not funded.

Authors' contributions: (I) Conception and design: Lu Ma; (II) Administrative support: Haining Qian; (III) Provision of study materials or patients: Lu Ma, Shutian Zhao,An Xu ,Zhou Wang, Haining Qian; (IV) Data analysis and interpretation: Lu Ma, Shutian Zhao,An Xu ,Zhou Wang, Haining Qian; (V) Manuscript writing: All authors ;(VI) Final approval of manuscript: All authors.

Conflicts of interest: All authors have completed the ICMJE uniform disclosure form. The authors have no conflicts of interest to declare.

Code availability: All the codes involved in this study are submitted with the article and can be made public.

Ethics approval: The author is responsible for all aspects of the work to ensure that issues related to the accuracy or completeness of any part of the work are properly investigated and resolved. The study is based on the Helsinki Declaration (JAMA 2000; 284:3043–3049). The study was approved by the Ethics Committee of Tongren Hospital, Shanghai Jiao Tong University School of Medicine and obtained the informed consent of all participants.

Reporting Checklist: The authors have completed the REMARK reporting checklist.

LOPEZ-BELTRAN A. Bladder cancer: clinical and pathological profile [J]. Scand J Urol Nephrol Suppl, 2008, 218): 95-109.
PENG M, HUANG Y, TAO T, et al. Metformin and gefitinib cooperate to inhibit bladder cancer growth via both AMPK and EGFR pathways joining at Akt and Erk [J]. Sci Rep, 2016, 6(28611.
FARLING K B. Bladder cancer: Risk factors, diagnosis, and management [J]. Nurse Pract, 2017, 42(3): 26-33.
BELLMUNT J, POWLES T, VOGELZANG N J. A review on the evolution of PD-1/PD-L1 immunotherapy for bladder cancer: The future is now [J]. Cancer Treat Rev, 2017, 54(58-67.
ZIBELMAN M, RAMAMURTHY C, PLIMACK E R. Emerging role of immunotherapy in urothelial carcinoma-Advanced disease [J]. Urol Oncol, 2016, 34(12): 538-547.
BOEGEMANN M, AYDIN A M, BAGRODIA A, et al. Prospects and progress of immunotherapy for bladder cancer [J]. Expert Opin Biol Ther, 2017, 17(11): 1417-1431.
BELLMUNT J, DE WIT R, VAUGHN D J, et al. Pembrolizumab as Second-Line Therapy for Advanced Urothelial Carcinoma [J]. N Engl J Med, 2017, 376(11): 1015-1026.
SONG Y, JIN D, CHEN J, et al. Identification of an immune-related long non-coding RNA signature and nomogram as prognostic target for muscle-invasive bladder cancer [J]. Aging (Albany NY), 2020, 12(12): 12051-12073.
ROUANNE M, BAJORIN D F, HANNAN R, et al. Rationale and Outcomes for Neoadjuvant Immunotherapy in Urothelial Carcinoma of the Bladder [J]. Eur Urol Oncol, 2020, 3(6): 728-738.
ZHANG X, ZHAO H, SHI X, et al. Identification and validation of an immune-related gene signature predictive of overall survival in colon cancer [J]. Aging (Albany NY), 2020, 12(24): 26095-26120.
PRAT A, NAVARRO A, PARÉ L, et al. Immune-Related Gene Expression Profiling After PD-1 Blockade in Non-Small Cell Lung Carcinoma, Head and Neck Squamous Cell Carcinoma, and Melanoma [J]. Cancer Res, 2017, 77(13): 3540-3550.
MA Y F, CHEN Y, FANG D, et al. The immune-related gene CD52 is a favorable biomarker for breast cancer prognosis [J]. Gland Surg, 2021, 10(2): 780-798.
COLAPRICO A, SILVA T C, OLSEN C, et al. TCGAbiolinks: an R/Bioconductor package for integrative analysis of TCGA data [J]. Nucleic Acids Res, 2016, 44(8): e71.
BHATTACHARYA S, ANDORF S, GOMES L, et al. ImmPort: disseminating data to the public for the future of immunology [J]. Immunol Res, 2014, 58(2-3): 234-239.
RITCHIE M E, PHIPSON B, WU D, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies [J]. Nucleic Acids Res, 2015, 43(7): e47.
ROBINSON M D, MCCARTHY D J, SMYTH G K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data [J]. Bioinformatics, 2010, 26(1): 139-140.
MAAG J L V. gganatogram: An R package for modular visualisation of anatograms and tissues based on ggplot2 [J]. F1000Res, 2018, 7(1576.
YANG S, WU Y, DENG Y, et al. Identification of a prognostic immune signature for cervical cancer to predict survival and response to immune checkpoint inhibitors [J]. Oncoimmunology, 2019, 8(12): e1659094.
MAYAKONDA A, LIN D C, ASSENOV Y, et al. Maftools: efficient and comprehensive analysis of somatic variants in cancer [J]. Genome Res, 2018, 28(11): 1747-1756.
CHAROENTONG P, FINOTELLO F, ANGELOVA M, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade [J]. Cell Rep, 2017, 18(1): 248-262.
CAO J, YANG X, LI J, et al. Screening and Identifying Immune-Related Cells and Genes in the Tumor Microenvironment of Bladder Urothelial Carcinoma: Based on TCGA Database and Bioinformatics [J]. Front Oncol, 2019, 9(1533.
NA L, BAI Y, SUN Y, et al. Identification of 9-Core Immune-Related Genes in Bladder Urothelial Carcinoma Prognosis [J]. Front Oncol, 2020, 10(1142.
TAN B, SHI X, ZHANG J, et al. Inhibition of Rspo-Lgr4 Facilitates Checkpoint Blockade Therapy by Switching Macrophage Polarization [J]. Cancer Res, 2018, 78(17): 4929-4942.
BALKWILL F, MANTOVANI A. Inflammation and cancer: back to Virchow? [J]. Lancet, 2001, 357(9255): 539-545.
AMULIC B, CAZALET C, HAYES G L, et al. Neutrophil function: from mechanisms to disease [J]. Annu Rev Immunol, 2012, 30(459-489.
PICCARD H, MUSCHEL R J, OPDENAKKER G. On the dual roles and polarized phenotypes of neutrophils in tumor development and progression [J]. Crit Rev Oncol Hematol, 2012, 82(3): 296-309.
RAKAEE M, BUSUND L T, PAULSEN E E, et al. Prognostic effect of intratumoral neutrophils across histological subtypes of non-small cell lung cancer [J]. Oncotarget, 2016, 7(44): 72184-72196.

Table1The clinical characteristics of BLCA patients

Covariates	Type	Total	Test	Train	P value
Age	<=65	159(40.25%)	52(39.69%)	107(40.53%)	0.9597
Age	>65	236(59.75%)	79(60.31%)	157(59.47%)
Gender	FEMALE	104(26.33%)	32(24.43%)	72(27.27%)	0.629
Gender	MALE	291(73.67%)	99(75.57%)	192(72.73%)
Grade	High Grade	374(94.68%)	123(93.89%)	251(95.08%)	1
	Low Grade	18(4.56%)	6(4.58%)	12(4.55%)
	unknow	3(0.76%)	2(1.53%)	1(0.38%)
Stage	Stage I-II	125(31.65%)	43(32.82%)	82(31.06%)	0.7347
	Stage III-IV	268(67.85%)	86(65.65%)	182(68.94%)
	unknow	2(0.51%)	2(1.53%)	0(0%)
T	T1-2	116(29.37%)	40(30.53%)	76(28.79%)	0.8423
	T3-4	247(62.53%)	81(61.83%)	166(62.88%)
	unknow	32(8.1%)	10(7.63%)	22(8.33%)
M	M0	189(47.85%)	66(50.38%)	123(46.59%)	1
	M1	10(2.53%)	4(3.05%)	6(2.27%)
	unknow	196(49.62%)	61(46.56%)	135(51.14%)
N	N0	228(57.72%)	73(55.73%)	155(58.71%)	0.5561
	N1-3	126(31.9%)	45(34.35%)	81(30.68%)
	unknow	41(10.38%)	13(9.92%)	28(10.61%)

Table2 The 6‑gene immune signature univariate and multivariate Cox regression analysis

Covariates	Univariate analysis		Multivariate analysis
Covariates	HR (95%CI)	P value	HR (95%CI)	P value
Age	1.03(1.02-1.05)	＜0.001	1.03(1.01-1.05)	＜0.001
Gender	0.87(0.63-1.21)	0.42	-	-
Grade	2.80 (0.69-11.33)	0.15	-	-
Stage	1.74 (1.44-2.12)	＜0.001	1.70(1.40-2.08)	＜0.001
Risk score	1.14 (1.10-1.18)	＜0.001	1.14(1.09-1.18)	＜0.001

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Risk stratification based on immune signature and prediction of immune checkpoint inhibitor’s efficacy for bladder urothelial carcinoma

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Results

Discussion

Abbreviations

Declarations

References

Tables

Additional Declarations

Status:

Version 1