Identification and Validation of a Novel Multiomics Signature for Prognosis and Immunotherapy Response of Endometrial Carcinoma

Purpose Cancer development and immune escape involve DNA methylation, copy number variation, and other molecular events. However, there are remarkably few studies integrating multiomics genetic profiles into endometrial cancer (EC). This study aimed to develop a multiomics signature for the prognosis and immunotherapy response of endometrial carcinoma. Methods The gene expression, somatic mutation, copy number alteration, and DNA methylation data of EC were analyzed from the UCSC Xena database. Then, a multiomics signature was constructed by a machine learning model, with the ROC curve comparing its prognostic power with traditional clinical features. Two computational strategies were utilized to estimate the signature's performance in predicting immunotherapy response in EC. Further validation focused on the most frequently mutant molecule, ARID1A, in the signature. The association of ARID1A with survival, MSI (Microsatellite-instability), immune checkpoints, TIL (tumor-infiltrating lymphocyte), and downstream immune pathways was explored. Results The signature consisted of 22 multiomics molecules, showing excellent prognostic performance in predicting the overall survival of patients with EC (AUC = 0.788). After stratifying patients into a high and low-risk group according to the signature's median value, low-risk patients displayed a greater possibility of respond to immunotherapy. Further validation on ARID1A suggested it could induce immune checkpoints upregulation, promote interferon response pathway, and interact with Treg (regulatory T cell) to facilitate immune activation in EC. Conclusion A novel multiomics prognostic signature of EC was identified and validated in this study, which could guide clinical management of EC and benefit personalized immunotherapy.


Introduction
As the most prevalent gynecologic malignancy, endometrial carcinoma (EC) is one of the leading causes of female mortality worldwide [1]. Endometrial cancer develops in about 142,000 women worldwide every year, and an estimated 42,000 women die from this cancer. e introduction of ICB (Immune Checkpoint Blockade) has achieved favorable clinical effects in patients with end-stage EC where the chemotherapy regimen has little progression [2,3]. However, more than 80% of patients are nonresponders, or NDB (no durable clinical benefit), to immunotherapy, and the underlying factors resulting in heterogeneous prognoses are poorly understood. In fact, cancer development and immune response are determined by multiple factors, including genomic mutation [4], DNA methylation [5], and copy number variance [6], et al. erefore, analysis incorporating multiomics data is urgently needed for EC management.
We utilized meta-dimensional strategies to seek genetically susceptible molecules from gene expression, somatic mutation, copy number alteration, and DNA methylation data of EC, aiming to develop a multiomics signature for prognosis and immunotherapy response of EC. e signature was built by machine learning model, and its efficiency was compared with traditional clinical features. Two computational approaches were also deployed to estimate the signature's performance in predicting immunotherapy response. Further validation focused on the most frequently mutant molecule in the signature: ARID1A. e association of ARID1A with survival, MSI (Microsatellite-instability), immune checkpoints, TIL (tumor-infiltrating lymphocyte), and downstream immune pathways were explored and potential mechanisms was given. e present study constructed a novel multiomics prognostic signature for prognosis and immunotherapy response of EC, which could guide clinical management of EC and benefit from personalized immunotherapy.

Differential Expression and Function Enrichment
Analysis. To reveal the molecules of real value for EC in these multiomics datasets, a series of R packages were used for screening, for example, the limma package [10] to seek out differentially expressed genes between 543 tumor and 35 normal samples with |log2 Fold Change (FC)| > 1.5 and P value < 0.05 as the threshold, as well as the ChAMP package [11] to identify differential methylation loci with |log2 Fold Change (FC)| > 0.5 and P value < 10 −15 .
A heatmap and volcano plot were used to display the 457 differentially expressed genes (DEG) and 746 CpG sites between tumor and normal samples, with GO (https://wego. genomics.org.cn) and KEGG (https://wego.genomics.org. cn) enrichment analysis to dissect their biological function and related signaling pathways. Meanwhile, oncoprintplot was employed to present the top 30 mutated and copy number varied genes in EC.

Construction of the Multiomics Prognostic Signature for
EC. Subsequent filtration of the 457 significant DEGs, 746 differential methylation loci, 482 mutated, and 380 copy number varied genes was completed by LASSO penalized Cox regression with overall survival as the dependent variable. Finally, 22 molecules were adopted for modeling. Next, Kaplan-Meier curves were depicted to show the prognostic power of the 22-gene-signature where the risk score of each patient was calculated with the following formula: n i Coef i * X i (Coef i : cox regression coefficient, X i : expression value of corresponding molecule, n � 22). Following that, patients were stratified into a high-and low-risk group according to the median risk score. A ROC (receiver operator characteristic) curve and multivariate Cox regression were also used to evaluate its prognostic performance and independent prognostic efficiency.

Relationship of the Prognostic Signature with Immunotherapy Response in EC.
To assess the relationship of the signature with immunotherapy, algorithms of TIDE [12] (tumor immune dysfunction and exclusion) and Immune Cell AI [13] were applied to predict patients' responses to ICB (immune checkpoint blockade) treatment. A hundredpercent bar-chart and a heatmap were used to display the response difference to ICB between the high and low-risk groups.

Validation on ARID1A for Its Prognostic Ability and
Association with Immunotherapy. Further validation focused on the most frequently mutant molecule in the signature: ARID1A. e association of ARID1A mutation with patients' survival, MSI (microsatellite-instability), immune checkpoints or T cell exhaustion markers (LAG3, SIGLEC15, CTLA4, HAVCR2 (TIM3), PDCD1LG2 (PD-L2), CD274 (PD-L1), PDCD1 (PD1), and TIGIT) and downstream immune pathways were explored. In addition, the impact of the ARID1A mutation on the abundances of 22 tumorinfiltrating immune cells was assessed by the CIBERSORT algorithm.

Underlying Mechanism from ARID1A Mutation to Cancer
Immune Activation. To identify the underlying mechanism from ARID1A mutation to cancer immune activation, a ternary interaction network was constructed. First, differential expression analysis was carried out between 235 ARID1A-mut samples and 291 ARID1A-wild tumor samples of the UCEC cohort, with 25 upregulated and 46 downregulated DEGs being obtained. By performing correlation analyses between the71 DEGs, abundances of 22 immune cells computed by the CIBERSORT, and enrichment scores of 29 cancer specialized immune pathways [14] quantified by GSVA [15], the interaction pairs of DEG-Immune Cell and DEG-Immune Pathway with a correlation coefficient >0.3 were screened out. A further regulating network of 71 DEG, 22 immune cells, and 29 immune pathways was completed by Cytoscape software (https:// cytoscape.org/).

Statistical Analysis.
Data processing and all analyses were accomplished by R 4.0.4. (Package: limma, ggplot2, survminer, ChAMP, ggcorrplot, GSVA, CIBERSORT, and so on). A chi-square test was used for counting data. Wilcoxon or Kruskal-Wallis tests were applied for comparisons between groups, while the Pearson and Spearman's rank correlation were adopted to estimate the statistical correlation of parametric or nonparametric variables. Two-sided P < 0.05 was considered a significant threshold for all statistical tests.

Differential Expression Analysis between Tumor and Normal Samples.
e study protocol was illustrated in Figure 1 and Table 1 summarized the demographic features of the TCGA-UCEC cohort. 457 differentially expressed genes (DEG) and 746 differential CpG sites are shown in the heatmap and volcano-plot (Figure 2

Construction of the Multiomics Prognostic Signature.
22 molecules stood out in LASSO-Cox analysis after shrinking most factors' coefficient towards zero (Figure 3(a)-3(b)), including 9 genes with somatic mutation, 4 with copy number variance, 3 with differential CpG sites, and 6 DEGs, their regression coefficients are shown in Table 2. e risk score of each patient was illustrated which well-stratified patients into two groups, according to the median value, with a huge discrepancy in survival probability (Figures 3(c)-3(d)). Patients were illustrated from a database. ROC curve showed a better prognostic performance of the signature than traditional clinical features, such as pathological stage and tumor grade (Figure 3(e)). Subsequent univariate and multivariate Cox analyses proved the signature can be an independent factor for the prognosis of EC (Figures 3(f)-3(g)). Response. In light of immunotherapy, no matter TIDE or Immune Cell AI algorithm, more patients were seen to be responders to ICB treatment (anti-PD-1 or anti-CTLA4) in the low-risk group than people in the high-risk group (71 vs 46 and 130 vs 74, respectively, P < 0.001) with statistically significant difference (Figure 4(a) and 4(b)).

Validation on ARID1A for Its Prognostic Ability and
Association with Immunotherapy. As the most frequently mutant gene in EC (Figures 4(c)-4(d), ARID1A can well stratify patients into two groups with noticeable survival differences in the UCEC cohort (4E-4 F), but did not affect their mRNA transcription. ARID1A mutation was also associated with MSI-H status, higher level of immune checkpoints expression, and TIL (tumor-infiltrating lymphocyte) ( Figure 5(a)-5(c)).

Discussion
e present study constructed a novel multiomics prognostic signature for prognosis and immunotherapy response of EC, which could guide clinical management of EC and benefit personalized immunotherapy. Following validation, it indicated the ARID1A mutation may interact with Treg and promote Type-I-IFN-Response pathway to facilitate tumor immune response and better survival outcomes for EC patients.
ARID1A (BAF250a), though connected with a superior outcome of ICB treatment in several cancer types, has rarely been reported for its prognostic and predictive ability in the immunotherapy cohort of EC [16][17][18]. As a subunit of the SWI/SNF chromatin-remodeling complex, it harbors an Nterminal DNA binding ARID (∼110 residues) and a Cterminal folded region (∼250 residues) [19], which are essential to increasing chromatin accessibility, binding to the promoter regions and facilitating transcription of multiple genes [20]. Inconsistently, the majority of DEGs were found to be downregulated in the ARID1A-mut group in our study (46 vs 25), partly accounting for the tumor suppression effect of ARID1A deficiency in a wide range of cancer types [21][22][23]. ese results were in line with the advantageous role of ARID1A mutation for patients' survival outcomes in the TCGA-UCEC in this study.
In fact, association between ARID1A mutation and favorable ICB treatment outcome in other cancer types is not scarce. Shen J et al. have reported a greater proportion of ICB responses in the ARID1A-deficient group than in the ARID1A-wild group in ovarian cancer mouse models [24]. A similar result was also observed in two melanoma cohorts [25][26][27] (42.86% responders versus 25.81% nonresponders and 100% responders versus 51.43% nonresponders, respectively). In addition, favorable survival outcomes in ARID1A mutant patients when receiving ICB treatment were also revealed in a pan-cancer study [16], but merely 10 EC samples with the ARID1A mutation were included, not sufficient to demonstrate the survival difference.
Elsewhere, ARID1A mutation was seen to be involved in Type-I-IFN-Response pathway and regulatory T cell to interact with EC development, partly accounting for its      Journal of Oncology advantageous role in many kinds of cancer. e previous study has already linked IFN I [28] and IFN II [29] pathway to ICB therapy outcome in multiple cancers and there was data also connecting the ARID1A mutation with IFN I and II Response pathway activity [17]. Apart from IFN pathways, in agreement with our findings, ARID1A mutation could also result in a higher level of PD-1, MSI, and T cell infiltration [30][31][32] to promote cancer immunity and potentiating favorable ICB treatment response.
Given the inherent fault of bioinformatics analysislacking of convincing data from reality. e conclusion of this study may be constrained. Furthermore, multicentric clinical studies and experiments at the cell and animal levels are warranted to validate the results under different circumstances. Following validation, it indicated that ARID1A mutation may interact with Treg and promote Type-I-IFN-Response pathway to facilitate tumor immune response and better survival outcomes for EC patients.

Conclusion
e present study constructed a novel multiomics prognostic signature for prognosis and immunotherapy response of EC, which could guide clinical management of EC and benefit from personalized immunotherapy.

Ethical Approval
As all datasets involved in this study were from public databases, ethics approval is not required.

Consent
All authors approved the submission and the International Committee of Medical Journal Editors (ICMJE) criteria for authorship were met.

Conflicts of Interest
Author Jiantong Zheng was employed by the company Shenzhen Dymind Biotechnology Company Limited. e remaining authors declare that the research conducted have no conflicts of interest.

Authors' Contributions
XFL and LJ conceived of the idea and wrote the manuscript. ZCW, QW, XQL, and JTZ prepared the data and analyzed the results. XFL and LJ supervised this work. Zhicheng Wu, Qiu Wang and Xiuqing Liu contributed equally to this study.