Gene expression signature-based prognostic risk score in patients with primary central nervous system lymphoma

Machine learning-based pathomics signature of histology slides as a novel prognostic indicator in primary central nervous system lymphoma

PURPOSE

To develop and validate a pathomics signature for predicting the outcomes of Primary Central Nervous System Lymphoma (PCNSL).

METHODS

In this study, 132 whole-slide images (WSIs) of 114 patients with PCNSL were enrolled. Quantitative features of hematoxylin and eosin (H&E) stained slides were extracted using CellProfiler. A pathomics signature was established and validated. Cox regression analysis, receiver operating characteristic (ROC) curves, Calibration, decision curve analysis (DCA), and net reclassification improvement (NRI) were performed to assess the significance and performance.

RESULTS

In total, 802 features were extracted using a fully automated pipeline. Six machine-learning classifiers demonstrated high accuracy in distinguishing malignant neoplasms. The pathomics signature remained a significant factor of overall survival (OS) and progression-free survival (PFS) in the training cohort (OS: HR 7.423, p < 0.001; PFS: HR 2.143, p = 0.022) and independent validation cohort (OS: HR 4.204, p = 0.017; PFS: HR 3.243, p = 0.005). A significantly lower response rate to initial treatment was found in high Path-score group (19/35, 54.29%) as compared to patients in the low Path-score group (16/70, 22.86%; p < 0.001). The DCA and NRI analyses confirmed that the nomogram showed incremental performance compared with existing models. The ROC curve demonstrated a relatively sensitive and specific profile for the nomogram (1-, 2-, and 3-year AUC = 0.862, 0.932, and 0.927, respectively).

CONCLUSION

As a novel, non-invasive, and convenient approach, the newly developed pathomics signature is a powerful predictor of OS and PFS in PCNSL and might be a potential predictive indicator for therapeutic response.

Relevance of different prognostic scores in primary CNS lymphoma in the era of intensified treatment regimens: A retrospective, multicenter analysis of 174 patients

OBJECTIVES

Treatment intensification (including consolidative high-dose chemotherapy with autologous stem cell transplantation [HDT-ASCT]) significantly improved outcome in primary central nervous system lymphoma (PCNSL) patients.

METHODS

We conducted a multicenter, retrospective analysis of newly diagnosed PCNSL patients, treated with intensified treatment regimens. The following scores were evaluated in terms of overall survival (OS) and progression-free survival (PFS): Memorial Sloan-Kettering Cancer Center (MSKCC), International Extranodal Lymphoma Study Group (IELSG), and three-factor (3F) prognostic score. Further, all scores were comparatively investigated for model quality and concordance.

RESULTS

Altogether, 174 PCNSL patients were included. One hundred and five patients (60.3%) underwent HDT-ASCT. Two-year OS and 2-year PFS for the entire population were 73.3% and 48.5%, respectively. The MSKCC (p = .003) and 3F score (p < .001), but not the IELSG score (p = .06), had the discriminatory power to identify different risk groups for OS. In regard to concordance, the 3F score (C-index [0.71]) outperformed both the MSKCC (C-index [0.64]) and IELSG (C-index [0.53]) score. Moreover, the superiority of the 3F score was shown for PFS, successfully stratifying patients in three risk groups, which also resulted in the highest C-index (0.66).

CONCLUSION

The comparative analysis of established PCNSL risk scores affirm the clinical utility of the 3F score stratifying the widest prognostic spectrum among PCNSL patients treated with intensified treatment approaches.

Identification of a novel autophagy-related prognostic signature and small molecule drugs for glioblastoma by bioinformatics

OBJECTIVE

To explore the autophagy-related prognostic signature (ARPs) via data mining in gene expression profiles for glioblastoma (GBM).

METHODS

Using the Cancer Genome Atlas (TCGA) database, we obtained 156 GBM samples and 5 adjacent normal samples, and denoted them as discovery cohort. Univariate Cox regression analysis was used to screen autophagy genes that related to GBM prognosis. Then, the least absolute shrinkage and selection operator Cox regression model was used to construct an autophagy-based ARPs, which was validated in an external cohort containing 80 GBM samples. The patients in the above-mentioned cohorts were divided into low-risk group and high-risk group according to the median prognostic risk score, and the diagnostic performance of the model was assessed by receiver operating characteristic curve analyses. The gene ontology and Kyoto encyclopedia of genes and genomes pathway enrichment analyses were performed between the high-risk and low-risk patients. Additionally, the genetic features of ARPs, such as genetic variation profiles, correlations with tumor-infiltrating lymphocytes (TILs), and potential drug sensitivity, were further assessed in the TCGA-GBM data set.

RESULTS

A signature of ARPs including NDUFB9, BAK1, SUPT3H, GAPDH, CDKN1B, CHMP6, and EGFR were detected and validated. We identified a autophagy-related prognosis 7-gene signature correlated survival prognosis, immune infiltration, level of cytokines, and cytokine receptor in tumor microenvironment. Furthermore, the signature was tested in several pathways related to disorders of tumor microenvironment, as well as cancer-related pathways. Additionally, a range of small molecular drugs, shown to have a potential therapeutic effect on GBM.

CONCLUSIONS

We constructed an autophagy-based 7-gene signature, which could serve as an independent prognostic indicator for cases of GBM and sheds light on the role of autophagy as a potential therapeutic target in GBM.

Survival prediction based on the gene expression associated with cancer morphology and microenvironment in primary central nervous system lymphoma

Dysregulation of cell morphology and cell-cell interaction results in cancer cell growth, migration, invasion, and metastasis. Besides, a balance between the extracellular matrix (ECM) and matrix metalloprotease (MMP) is required for cancer cell morphology and angiogenesis. Here, we determined gene signatures associated with the morphology and microenvironment of primary central nervous system lymphoma (PCNSL) to enable prognosis prediction. Next-generation sequencing (NGS) on 31 PCNSL samples revealed gene signatures as follows: ACTA2, ACTR10, CAPG, CORO1C, KRT17, and PALLD in cytoskeleton, CDH5, CLSTN1, ITGA10, ITGAX, ITGB7, ITGA8, FAT4, ITGAE, CDH10, ITGAM, ITGB6, and CDH18 in adhesion, COL8A2, FBN1, LAMB3, and LAMA2 in ECM, ADAM22, ADAM28, MMP11, and MMP24 in MMP. Prognosis prediction formulas with the gene expression values and the Cox regression model clearly divided survival curves of the subgroups in each status. Furthermore, collagen genes contributed to gene network formation in glasso, suggesting that the ECM balance controls the PCNSL microenvironment. Finally, the comprehensive balance of morphology and microenvironment enabled prognosis prediction by a combinatorial expression of 8 representative genes, including KRT17, CDH10, CDH18, COL8A2, ADAM22, ADAM28, MMP11, and MMP24. Besides, these genes could also diagnose PCNSL cell types with MTX resistances in vitro. These results would not only facilitate the understanding of biology of PCNSL but also consider targeting pathways for anti-cancer treatment in personalized precision medicine in PCNSL.

A 5-gene DNA methylation signature is a promising prognostic biomarker for early-stage cervical cancer

The demographic information and overall survival (OS) of patients with cervical cancer (CC) (pathological stage: IA-IIA) were extracted from the TCGA database. A univariate and multivariate Cox proportional hazard model was performed to identify methylation markers significantly associated with the OS of patients in the training dataset. Then such a prognostic classifier was tested on the validation set and all subgroups. The Kaplan-Meier analysis and ROC analysis were performed to detect the ability to discriminate between patients with different risks and different OS. A DNA methylation signature which contained five genes was found to be significantly associated with the OS of CC patients by the Cox regression analysis in the training dataset. Such a signature could efficiently distinguish the patients into two risk groups with significantly different OS in both datasets. The receiver operating characteristic (ROC) analysis showed it had high sensitivity and specificity. Moreover, such a prognostic model also could be effectively applied to different subgroups, including groups of different ages, tumour sizes, histologic types, etc. A 5-DNA methylation signature identified by this study may act as a novel prognostic indicator for early-stage CC, and it may be helpful for the timely diagnosis and intervention of CC at pathological stages IA-IIA.Impact StatementWhat is already known on this subject? Cervical cancer (CC) is one of the most common gynaecological malignant tumours.What the results of this study add? This study constructed a risk model based on a 5-DNA methylation signature for early-stage CC patients' survival prediction.What the implications are of these findings for clinical practice and/or further research? Methylated markers have the potential to discriminate patients of different risks and different OS. Our results may shed new light on the early diagnosis and intervention, and potential therapeutic targets for CC patients at pathological stages IA-IIA.

Prognostic models in primary central nervous system lymphoma patients: A systematic review

Over the last decade, several prognostic models have been proposed for primary central nervous system lymphoma (PCNSL), but consensus on the optimal model for these patients is absent or lacking. This study aims to review available prognostic models for PCNSL and discuss their prognostic features. A comprehensive literature search performed in Pubmed/Embase identified ten studies with a variable number of analysed patients (range 32-3453), which proposed 12 prognostic models. Age and performance status were the most important prognostic factors in PCNSL and an integral part of the majority of the proposed models. However, there is no universally accepted prognostic model for PCNSL owning to a number of limitations such as a small number of patients, limited samples obtained for genetic analysis, retrospective nature of studies, single centre studies, and lack of validation. Future multicentre studies are necessary to determine the optimal prognostic model for PCNSL by combining different prognostic markers of significance.

The immune contexture of primary central nervous system diffuse large B cell lymphoma associates with patient survival and specific cell signaling

Rationale: Primary central nervous system diffuse large B-cell lymphoma (PCNSL) is a rare and aggressive entity that resides in an immune-privileged site. The tumor microenvironment (TME) and the disruption of the immune surveillance influence lymphoma pathogenesis and immunotherapy resistance. Despite growing knowledge on heterogeneous therapeutic responses, no comprehensive description of the PCNSL TME is available. We hence investigated the immune subtypes of PCNSL and their association with molecular signaling and survival. Methods: Analysis of PCNSL transcriptomes (sequencing, n = 20; microarrays, n = 34). Integrated correlation analysis and signaling pathway topology enabled us to infer intercellular interactions. Immunohistopathology and digital imaging were used to validate bioinformatic results. Results: Transcriptomics revealed three immune subtypes: immune-rich, poor, and intermediate. The immune-rich subtype was associated to better survival and characterized by hyper-activation of STAT3 signaling and inflammatory signaling, e.g., IFNγ and TNF-α, resembling the hot subtype described in primary testicular lymphoma and solid cancer. WNT/β-catenin, HIPPO, and NOTCH signaling were hyper-activated in the immune-poor subtype. HLA down-modulation was clearly associated with a low or intermediate immune infiltration and the absence of T-cell activation. Moreover, HLA class I down-regulation was also correlated with worse survival with implications on immune-intermediate PCNSL that frequently feature reduced HLA expression. A ligand-receptor intercellular network revealed high expression of two immune checkpoints, i.e., CTLA-4/CD86 and TIM-3/LAGLS9. TIM-3 and galectin-9 proteins were clearly upregulated in PCNSL. Conclusion: Altogether, our study reveals that patient stratification according to immune subtypes, HLA status, and immune checkpoint molecule quantification should be considered prior to immune checkpoint inhibitor therapy. Moreover, TIM-3 protein should be considered an axis for future therapeutic development.

Development and validation of a five-lncRNA signature with prognostic value in colon cancer

Dysregulation of long noncoding RNAs (lncRNAs) has been found in a large number of human cancers, including colon cancer. Therefore, the implementation of potential lncRNAs biomarkers with prognostic prediction value are very much essential. GSE39582 data set was downloaded from database of Gene Expression Omnibus. Re-annotation analysis of lncRNA expression profiles was performed by NetAffx annotation files. Univariate and multivariate Cox proportional analyses helped select prognostic lncRNAs. Algorithm of random survival forest-variable hunting (RSF-VH) together with stepwise multivariate Cox proportional analysis were performed to establish lncRNA signature. The log-rank test was carried out to analyze and compare the Kaplan-Meier survival curves of patients' overall survival (OS). Receiver operating characteristic (ROC) analysis was used for comparing the survival prediction regarding its specificity and sensitivity based on lncRNA risk score, followed by calculating the values of area under the curve (AUC). The single-sample GSEA (ssGSEA) analysis was used to describe biological functions associated with this signature. Finally, to determine the robustness of this model, we used the validation sets including GSE17536 and The Cancer Genome Atlas data set. After re-annotation analysis of lncRNAs, a total of 14 lncRNA probes were obtained by univariate and multivariate Cox proportional analysis. Then, the RSF-VH algorithm and stepwise multivariate Cox analysis helped to build a five-lncRNA prognostic signature for colon cancer. The patients in group with high risk showed an obviously shorter survival time compared with patients in group with low risk with AUC of 0.75. In addition, the five-lncRNA signature can be used to independently predict the survival of patients with colon cancer. The ssGSEA analysis revealed that pathways such as extracellular matrix-receptor interaction was activated with an increase in risk score. These findings determined the strong power of prognostic prediction value of this five-lncRNA signature for colon cancer.

GSEA-assisted gene signatures valid for combinations of prognostic markers in PCNSL

Primary central nervous system lymphoma (PCNSL) is a brain malignant non-Hodgkin’s B-cell lymphoma. The standard treatments are high-dose methotrexate (MTX)-based chemotherapies and deferred whole brain radiotherapy. However, MTX resistance-dependent global expression and signaling pathway changes and their relationship with prognoses have not yet been elucidated. Here, we conducted a global expression analysis with next-generation sequencing and gene set enrichment analysis (GSEA) in MTX-resistant PCNSL cell lines (HKBML-MTX and TK-MTX) and PCNSL tissues. In rank scores, genes listed in HKBML-MTX and TK-MTX were enriched in PCNSL with poor prognoses. In fold changes, a part of differentially-expressed genes in PCNSL tissues were also detected in HKBML-MTX and TK-MTX cells; FOXD2-AS1 and MMP19 were commonly expressed in both HKBML-MTX and TK-MTX, FABP5 and CD70 were HKBML-MTX-specifically expressed, and CLCN2, HOXB9, INE1, and LRP5L were TK-MTX-specifically expressed, which may provide a combination of prognostic markers on MTX-sensitivities in PCNSL. Additionally, PCNSL subgroups, divided with hierarchical clustering and Kaplan-Meier methods, included twenty commonly expressed genes in both HKBML-MTX and TK-MTX, ten HKBML-MTX-specifically expressed genes, and two TK-MTX-specifically expressed genes. These results suggest that the GSEA-assisted gene signatures can provide a combination for prognostic markers in recurrent PCNSL with MTX resistances.

A 10-gene prognostic methylation signature for stage I-III cervical cancer

PURPOSE

Cervical cancer (CC) patients usually have poor prognosis. The present study aims to find a DNA methylation signature for predicting survival of CC patients.

METHODS

We selected CC patients at pathological stage I-III with corresponding information on radiotherapy and overall survival (OS) from TCGA. Differential expression and methylation analysis was done between patients with and without radiotherapy. We selected feature genes using recursive feature elimination algorithm to build a support vector machine classifier. DNA methylation biomarkers predictive of prognosis were identified using a LASSO Cox-Proportional Hazards model to construct a prognostic scoring model. The classifier and the prognostic model were tested on the training set and the validation set. Nomogram combining risk score and prognostic clinical factors were used.

RESULTS

We obtained 497 differentially expressed genes (DEGs) and 865 differentially methylated genes (DMGs). Fifteen feature genes were selected from the 292 common genes between the DEGs and the DMGs to construct a classification model for radiotherapy. A DNA methylation signature including 10 genes was identified and used to establish a prognostic scoring model. The 10-gene methylation signature could effectively separate patients into two risk groups with markedly different OS time. Predictive capability of the methylation signature was successfully confirmed on the validation set. A nomogram comprised of risk score, radiotherapy, and recurrence was applied, with calibration plots displaying good concordance between predicted and actual OS. The DEGs were involved in 12 KEGG pathways most of which were correlated with metastasis and proliferation of various cancers, such as pathways in cancer, basal cell carcinoma, transcriptional misregulation in cancer and ECM-receptor interaction.

CONCLUSION

We Identified a 10-gene methylation signature for risk stratification of CC patients at pathological stages I-III, and ten methylation biomarkers might be novel therapeutic targets for CC.

miR-101, miR-548b, miR-554, and miR-1202 are reliable prognosis predictors of the miRNAs associated with cancer immunity in primary central nervous system lymphoma

MicroRNAs (miRNAs) inhibit protein function by silencing the translation of target mRNAs. However, in primary central nervous system lymphoma (PCNSL), the expression and functions of miRNAs are inadequately known. Here, we examined the expression of 847 miRNAs in 40 PCNSL patients with a microarray and investigated for the miRNA predictors associated with cancer immunity-related genes such as T helper cell type 1/2 (Th-1/Th-2) and regulatory T cell (T-reg) status, and stimulatory and inhibitory checkpoint genes, for prognosis prediction in PCNSL. The aim of this study is to find promising prognosis markers based on the miRNA expression in PCNSL. We detected 334 miRNAs related to 66 cancer immunity-related genes in the microarray profiling. Variable importance measured by the random survival forest analysis and Cox proportional hazards regression model elucidated that 11 miRNAs successfully constitute the survival formulae dividing the Kaplan-Meier curve of the respective PCNSL subgroups. On the other hand, univariate analysis shortlisted 23 miRNAs for overall survival times, with four miRNAs clearly dividing the survival curves-miR-101/548b/554/1202. These miRNAs regulated Th-1/Th-2 status, T-reg cell status, and immune checkpoints. The miRNAs were also associated with gene ontology terms as Ras/MAP-kinase, ubiquitin ligase, PRC2 and acetylation, CDK, and phosphorylation, and several diseases including acquired immunodeficiency syndrome, glioma, and those related to blood and hippocampus with statistical significance. In conclusion, the results demonstrated that the four miRNAs comprising miR-101/548b/554/1202 associated with cancer immunity can be a useful prognostic marker in PCNSL and would help us understand target pathways for PCNSL treatments.

Promising Prognosis Marker Candidates on the Status of Epithelial-Mesenchymal Transition and Glioma Stem Cells in Glioblastoma

Multivariable analyses of global expression profiling are valid indicators of the prognosis of various diseases including brain cancers. To identify the candidates for markers of prognosis of glioblastoma, we performed multivariable analyses on the status of epithelial (EPI)-mesenchymal (MES) transition (EMT), glioma (GLI) stem cells (GSCs), molecular target therapy (MTT), and potential glioma biomarkers (PGBs) using the expression data and clinical information from patients. Random forest survival and Cox proportional hazards regression analyses indicated significant variable values for DSG3, CLDN1, CDH11, FN1, HDAC3/7, PTEN, L1CAM, OLIG2, TIMP4, IGFBP2, and GFAP. The analyses also comprised prognosis prediction formulae that could distinguish between the survival curves of the glioblastoma patients. In addition to the genes mentioned above, HDAC1, FLT1, EGFR, MGMT, PGF, STAT3, SIRT1, and GADD45A constituted complex genetic interaction networks. The calculated status scores obtained by principal component analysis indicated that GLI genes covered the status of EPI, GSC, and MTT-related genes. Moreover, survival tree analyses indicated that MES^high, MES^highGLI^low, GSC^highGLI^low, MES^highMTT^low, and PGB^high showed poor prognoses and MES^middle, GSC^low, and PGB^low showed good prognoses, suggesting that enhanced EMT and GSC are associated with poor survival and that lower expression of EPI markers and the pre-stages of EMT are relatively less malignant in glioblastoma. These results demonstrate that the assessment of EMT and GSC enables the prediction of the prognosis of glioblastoma that would help develop novel therapeutics and de novo marker candidates for the prognoses of glioblastoma.

Differential expression of individual transcript variants of PD-1 and PD-L2 genes on Th-1/Th-2 status is guaranteed for prognosis prediction in PCNSL

In current molecular medicine, next-generation sequencing (NGS) for transcript variant detection and multivariable analyses are valid methods for evaluating gene expression, cancer mechanisms, and prognoses of patients. We conducted RNA-sequencing on samples from patients with primary central nervous system lymphoma (PCNSL) using NGS and performed multivariable analysis on gene expression data and correlations focused on Th-1/Th-2 helper T cell balance and immune checkpoint to identify diagnosis/prognosis markers and cancer immune pathways in PCNSL. We selected 84 transcript variants to limit the analysis range for Th-1/Th-2 balance and stimulatory and inhibitory checkpoints in 31 PCNSLs. Of these, 21 highly-expressed transcript variants were composed of the formulas for prognoses based on Th-1/Th-2 status and checkpoint activities. Using formulas, Th-1^low, Th-2^high, and stimulatory checkpoint^high resulted in poor prognoses. Further, Th-1^highTh-2^low was associated with good prognoses. On the other hand, CD40-001^high and CD70-001^high as stimulatory genes, and LAG3-001^high, PDCD1 (PD-1)-001/002/003^high, and PDCD1LG2 (PD-L2)-201^low as inhibitory genes were associated with poor prognoses. Interestingly, Th-1^highTh-2^low and Th-1^lowTh-2^high were correlated with stimulatory checkpoint^low as CD70-001^low and inhibitory checkpoint^low as HAVCR2 (TIM-3)-001^low and PDCD1LG2-001/201^low, respectively. Focused on the inhibitory checkpoint, specific variants of CD274 (PD-L1)-001 and PDCD1-002 served severe hazard ratios. In particular, PDCD1-002^high by a cut off score was associated with poor prognoses, in addition to PDCD1-001/003^high, PDCD1LG2-201^low, and LAG3-001^high. These results mainly suggest that expression of transcript variants of PDCD1 and PDCD1LG2 on the Th-1/Th-2 balance enable prognostic prediction in PCNSL. This study provides insights for development of molecular target therapies and identification of diagnosis/prognosis markers in PCNSL.

CD276 and the gene signature composed of GATA3 and LGALS3 enable prognosis prediction of glioblastoma multiforme

Glioma is the most common type of primary brain tumor, accounting for 40% of malignant brain tumors. Although a single gene may not be a marker, an expression profiling and multivariate analyses for cancer immunotherapy must estimate survival of patients. In this study, we conducted expression profiling of immunotherapy-related genes, including those in Th1/2 helper T and regulatory T cells, and stimulatory and inhibitory checkpoint molecules associated with survival prediction in 571 patients with malignant and aggressive form of gliomas, glioblastoma multiforme (GBM). Expression profiling and Random forests analysis of 21 immunosuppressive genes and Kaplan-Meier analysis in 158 patients in the training data set suggested that CD276, also known as B7-H3, could be a single gene marker candidate. Furthermore, prognosis prediction formulas, composed of Th2 cell-related GATA transcription factor 3 (GATA3) and immunosuppressive galactose-specific lectin 3 (LGALS3), based on 67 immunotherapy-related genes showed poor survival with high scores in training data set, which was also validated in another 413 patients in the test data set. The CD276 expression helped distinguish survival curves in the test data set. In addition, inhibitory checkpoint genes, including T cell immunoreceptor with Ig and ITIM domains, V-set domain containing T cell activation inhibitor 1, T-cell immunoglobulin and mucin-domain containing 3, and tumor necrosis factor receptor superfamily 14, showed potential as secondary marker candidates. These results suggest that CD276 expression and the gene signature composed of GATA3 and LGALS3 are effective for prognosis in GBM and will help us understanding target pathways for immunotherapy in GBM.

MicroRNA signature constituted of miR-30d, miR-93, and miR-181b is a promising prognostic marker in primary central nervous system lymphoma

MicroRNAs (miRNAs) are small RNA molecules that inhibit gene function by suppressing translation of target genes. However, in primary central nervous system lymphoma (PCNSL), the biological significance of miRNAs is largely unknown, although some miRNAs are known to be prognosis markers. Here, we analyzed 847 miRNAs expressed in 27 PCNSL specimens using microarray profiling and surveyed miRNA signature for prognostic prediction. Of these, 16 miRNAs were expressed in 27 PCNSL specimens at a frequency of 48%. Their variable importance measured by Random forest model revealed miR-192, miR-486, miR-28, miR-52, miR-181b, miR-194, miR-197, miR-93, miR-708, and let-7g as having positive effects; miR-29b-2*, miR-126, and miR-182 as having negative effects; and miR-18a*, miR-425, and miR-30d as neutral. After principal component analysis, the prediction formula for prognosis, consisting of the expression values of the above-mentioned miRNAs, clearly divided Kaplan-Meier survival curves by the calculated Z-score (HR = 6.4566, P = 0.0067). The 16 miRNAs were enriched by gene ontology terms including angiogenesis, cell migration and proliferation, and apoptosis, in addition to signaling pathways including TGF-β/SMAD, Notch, TNF, and MAPKinase. Their target genes included BCL2-related genes, HMGA2 oncogene, and LIN28B cancer stem cell marker. Furthermore, three miRNAs including miR-181b, miR-30d, and miR-93, selected from the 16 miRNAs, also showed comparable results for survival (HR = 8.9342, P = 0.0007), suggestive of a miRNA signature for prognostic prediction in PCNSL. These results indicate that this miRNA signature is useful for prognostic prediction in PCNSL and would help us understand target pathways for therapies in PCNSL.

Development of a Novel Autophagy-related Prognostic Signature for Serous Ovarian Cancer

Purpose: Considerable evidence suggests that autophagy plays a crucial role in the biological processes of ovarian cancer. The aim of this study was to develop a novel autophagy-related prognostic signature for serous ovarian cancer.

Methods: A univariate Cox proportional regression model was used to analyze mRNA microarray and clinical data in The Cancer Genome Atlas (TCGA) for the purpose of selecting autophagy-related prognostic genes. A multivariate Cox proportional regression model and the survival analysis were used to develop an eight-gene prognostic signature. The multivariate Cox and stratification analysis suggested that this signature was an independent prognostic factor for serous ovarian cancer patients. Bioinformatics functions were investigated by a principal components analysis and gene set enrichment analysis (GSEA). Finally, the correlation between the prognostic signature and gene mutation status was further analyzed in serous ovarian cancer, and especially with regard to the mutation status of BRCA1 and BRCA2 (BRCA1/2) genes.

Results: Distinctly different autophagy-related gene expression profiles were identified in normal ovarian tissues and serous ovarian cancer tissues. We profiled an autophagy-related gene set and identified eight genes with significant prognostic values for serous ovarian cancer. Subsequently, an autophagy-related ovarian cancer risk signature was constructed, and patients at a high-risk or low-risk for poor prognosis were identified based on their signature. High-risk patients had significantly shorter overall survival (OS) and disease-free survival (DFS) times than low-risk patients. GSEA results suggested an enhanced intensity of autophagy regulation in high-risk patients when compared with low-risk patients. When studied as an independent prognostic factor for serous ovarian cancer, the significant prognostic value of this signature could be seen in the stratified cohorts. For clinical use, we developed a nomogram that included the prognostic classifier and seven clinical risk factors. Additionally, we identified the 10 most frequently mutated genes found in serous ovarian cancer patients, and analyzed them for their differences in high-risk and low-risk patients. Among 293 patients, 62 had BRCA1/2 gene mutations, and this result was significantly correlated with the autophagy-related prognostic signature.

Conclusions: Our findings suggest that the eight-gene autophagy-related signature could serve as an independent prognostic indicator for cases of serous ovarian cancer.

Cancer stem cell associated eight gene-based signature predicts clinical outcomes of colorectal cancer

Previous studies have suggested that cancer stem cells serve crucial functions in tumorigenesis, metastasis and therapy failure. Stem cell signaling transduction pathways are frequently dysregulated in cancer and associated with tumorigenesis, metastasis and the cell cycle, which are necessary for cancer proliferation. However, cancer stem cell-associated gene signatures have not been established for predicting patient outcomes in colorectal cancer. Using a gene-mining approach, the present study performed mRNA expression profiling in large colorectal cancer cohorts from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database, including a TCGA colorectal cancer cohort (n=383) and three independent validation series GSE39582 (n=582), GSE17536 (n=177) and GSE17537 (n=55). The present study identified that an eight-gene signature in cancer stem cell signaling was associated with the overall survival and disease/recurrence-free survival of patients with colorectal. On the basis of this signature, patients in the TCGA training sets were divided into high-risk and low-risk subgroups with a significantly different overall survival rate (hazard ratio, 2.38; P=0.0005). The prognostic value of this signature was confirmed using three independent GEO colorectal cancer sets. Identifying this prognostic stem cell signaling signature may provide an efficient classification tool for clinical prognosis evaluation, and facilitate cancer stem cell-targeted therapy.

The expression of PD-1 ligands and IDO1 by macrophage/microglia in primary central nervous system lymphoma

Recent progress in anti-tumor immunotherapy has focused on the significance of the tumor microenvironment in tumor progression and resistance to chemo/radio-therapy. Myeloid cells such as macrophages are predominant stromal components in hematological malignancies. In the present study, we investigated the regulation of programmed death-1 (PD-1) ligand expression in primary central nervous system lymphoma (PCNSL) using PCNSL cell lines and human monocyte-derived macrophages. TK PCNSL cell line-derived soluble factors induced overexpression of PD-1 ligands, indoleamine 2,3-dioxygenase (IDO1), and several other cytokines in macrophages. The expression of PD-1 ligands was dependent on the activation of signal transducer and activator of transcription 3. PD-L1 and IDO1 were overexpressed by macrophage/microglia in PCNSL tissues, and gene expression profiling indicated that IDO1 expression was positively correlated with the expression of macrophage and lymphocyte markers. Macrophage-derived factors did not influence the proliferation or chemo-sensitivity of cell lines. These data suggest that the expression of immunosuppressive molecules, including PD-1 ligands and IDO1, by macrophage/microglia may be involved in immune evasion of lymphoma cells.

Advances in Pathobiology of Primary Central Nervous System Lymphoma

Objective:

Primary central nervous system lymphoma (PCNSL) is a specific type of non-Hodgkin lymphoma with poor prognosis. The rare incidence of this disease and difficulty to obtain sufficient tissue material impede deep research into PCNSL. However, application of modern molecular techniques makes it possible to find biological characteristics exclusive to PCNSL. Therefore, we systematically reviewed the latest research progress on biological characteristics and pathogenesis of PCNSL.

Data Sources:

The data analyzed in this review were from the articles listed in PubMed database.

Study Selection:

Articles focusing on the biology of PCNSL at the cytogenetic or molecular level were reviewed, including clinical, basic research, and review articles.

Results:

With respect to histopathology, perivascular growth pattern and reactive perivascular T-cell infiltration are regarded as typical histopathological manifestations of tumor cells in PCNSL. Moreover, tumor cells of PCNSL predominantly express an activated B-cell-like phenotype, including CD10⁻ BCL-6⁺ MUM1⁺, CD10⁻ BCL-6⁻ MUM1⁺, and CD10⁻ BCL-6⁻ MUM1⁻. On the molecular level, some molecular and genetic alterations may contribute to malignant transformation, including mutations of proto-oncogenes and tumor suppressor genes, gains and losses of genetic material, as well as aberrant activation of some important signaling pathways, such as nuclear factor-κB and JAK/STAT pathway.

Conclusions:

The integrated molecular mechanisms involved in pathogenesis of PCNSL are not well understood. The important biomarkers indicating prognosis are not identified. Multicenter studies should be carried out to elucidate pathogenesis of PCNSL to find novel and effective therapeutic strategies.

Correlation between lower balance of Th2 helper T-cells and expression of PD-L1/PD-1 axis genes enables prognostic prediction in patients with glioblastoma

Common cancer treatments include radiation therapy, chemotherapy including molecular targeted drugs and anticancer drugs, and surgical treatment. Recent studies have focused on investigating the mechanisms by which immune cells attack cancer cells and produce immune tolerance-suppressing cytokines, as well as on their potential application in cancer immunotherapy. We conducted expression profiling of CD274 (PD-L1), GATA3, IFNG, IL12R, IL12RB2, IL4, PDCD1 (PD-1), PDCD1LG2 (PD-L2), and TBX21 (T-bet) using data of 158 glioblastoma multiforme (GBM) patients with clinical information available at The Cancer Genome Atlas. Principal component analysis of the expression profiling data was used to derive an equation for evaluating the status of Th1 and Th2 cells. GBM specimens were divided based on the median of the Th scores. The results revealed that Th1^HighTh2^Low and Th1^LowTh2^Low statuses indicated better prognosis than Th1^HighTh2^High, and were evaluated based on the downregulation of PD-L1, PD-L2, and PD-1. Furthermore, Th2^Low divided based on the threshold, as well as CD274^Low and PDCD1^Low, were associated with good prognosis. In the Th2^Low subgroup, 14 genes were identified as potential prognostic markers. Of these, SLC11A1^Low, TNFRSF1B^Low, and LTBR^Low also indicated good prognosis. These results suggest that low Th2 balance and low activity of the PD-L1/PD-1 axis predict good prognosis in GBM. The set of genes identified in the present study could reliably predict survival in GBM patients and serve as useful molecular markers. Furthermore, this set of genes could prove to be novel targets for cancer immunotherapy.

Panel of seven long noncoding RNA as a candidate prognostic biomarker for ovarian cancer

Ovarian cancer is one of the most common and lethal gynecological malignancies. The diagnosis of ovarian cancer is often at an advanced stage. Accumulated evidence suggests that long noncoding RNAs (lncRNAs) play important roles during ovarian tumorigenesis. In this study, using the lncRNA-mining approach, we analyzed lncRNA expression profiles of 493 ovarian cancer patients from Gene Expression Omnibus datasets, and identified a signature group of seven lncRNAs (BC037530, AK021924, AK094536, AK094536, BC062365, BC004123 and BC007937) associated with patient survival in the training dataset GSE9891. We also formulated a risk score model to divide patients into low-risk and high-risk groups based on the expression of these seven lncRNAs. We further validated the predictive power of our risk score model in two other datasets, GSE26193 and GSE63885. Our analysis showed that the seven-lncRNA signature can serve as an independent predictor apart from Federation of Gynecology and Obstetrics (FIGO) stage and patient age. Further investigation revealed the seven-lncRNA signature correlated with few critical signaling pathways involved in cancer. Combined, all these findings strongly support that the seven-lncRNA signature can serve as a strong prognosis biomarker.

Development of a novel prognostic signature of long non-coding RNAs in lung adenocarcinoma

OBJECTIVES

Increasing evidence suggests that long non-coding RNAs (lncRNAs) may play a crucial role in many biological processes in a variety of cancers and serve as the basis for many clinical applications including prognostic biomarkers and potential therapeutic targets. The aim of this study is to develop a prognostic lncRNA signature with RNA-seq data in lung adenocarcinomas.

METHODS

LncRNA expression profiles and clinical data of lung adenocarcinoma patients from The Cancer Genome Atlas (TCGA) were analyzed. Univariate Cox proportional regression model was used to identify prognostic lncRNAs, and then multivariate Cox proportional regression model was used to develop a prognostic signature. Survivals were compared using log-rank test, and the biological implications of prognostic lncRNAs were analyzed using the KEGG pathway functional enrichment analysis.

RESULTS

We identified eight lncRNAs which had prognostic association with p value <0.01 in a TCGA lung adenocarcinoma cohort of 478 patients. Then a novel prognostic signature with the eight lncRNAs was developed using Cox regression model. Signature high-risk cases had worse overall survival (OS, median 85.97 vs. 38.34 months, p < 0.001) and disease-free survival (DFS, median 44.02 vs. 26.58 months, p = 0.007) than low-risk cases. Multivariate Cox regression analysis suggested that the eight-lncRNA signature was independent of clinical and pathological factors. KEGG pathway functional enrichment analysis indicated potential functional roles of the eight prognostic lncRNAs in tumorigenesis.

CONCLUSIONS

Our findings suggest that the eight-lncRNA signature might provide an effective independent prognostic model for the prediction of lung adenocarcinoma patients.

Expression of Tim-1 in primary CNS lymphoma

Primary central nervous system lymphoma (PCNSL) is a distinct subtype of extranodal lymphoma with aggressive clinical course and poor outcome. As increased IL‐10/IL‐6 ratio is recognized in the cerebrospinal fluid (CSF) of PCNSL patients, we hypothesized that PCNSL might originate from a population of B cells with high IL‐10‐producing capacity, an equivalent of “regulatory B cells” in mice. We intended in this study to clarify whether Tim‐1, a molecule known as a marker for regulatory B cells in mice, is expressed in PCNSL. By immunohistochemical analysis, Tim‐1 was shown to be positive in as high as 54.2% of PCNSL (26 of 58 samples), while it was positive in 19.1% of systemic diffuse large B‐cell lymphoma (DLBCL) samples (17 of 89 samples; P < 0.001). Tim‐1 expression positively correlated with IL‐10 expression in PCNSL (Cramer's V = 0.55, P < 0.001), and forced expression of Tim‐1 in a PCNSL cell line resulted in increased IL‐10 secretion, suggesting that Tim‐1 is functionally linked with IL‐10 production in PCNSL. Moreover, soluble Tim‐1 was detectable in the CSF of PCNSL patients, and was suggested to parallel disease activity. In summary, PCNSL is characterized by frequent Tim‐1 expression, and its soluble form in CSF may become a useful biomarker for PCNSL.

A long non-coding RNA signature to improve prognosis prediction of gastric cancer

BACKGROUND

Increasing evidence suggests long non-coding RNAs (lncRNAs) are frequently aberrantly expressed in cancers, however, few related lncRNA signatures have been established for prediction of cancer prognosis. We aimed at developing alncRNA signature to improve prognosis prediction of gastric cancer (GC).

METHODS

Using a lncRNA-mining approach, we performed lncRNA expression profiling in large GC cohorts from Gene Expression Ominus (GEO), including GSE62254 data set (N = 300) and GSE15459 data set (N = 192). We established a set of 24-lncRNAs that were significantly associated with the disease free survival (DFS) in the test series.

RESULTS

Based on this 24-lncRNA signature, the test series patients could be classified into high-risk or low-risk subgroup with significantly different DFS (HR = 1.19, 95 % CI = 1.13-1.25, P < 0.0001). The prognostic value of this 24-lncRNA signature was confirmed in the internal validation series and another external validation series, respectively. Further analysis revealed that the prognostic value of this signature was independent of lymph node ratio (LNR) and postoperative chemotherapy. Gene set enrichment analysis (GSEA) indicated that high risk score group was associated with several cancer recurrence and metastasis associated pathways.

CONCLUSIONS

The identification of the prognostic lncRNAs indicates the potential roles of lncRNAs in GC biogenesis. Our results may provide an efficient classification tool for clinical prognosis evaluation of GC.

A four-long non-coding RNA signature in predicting breast cancer survival

Background

Many long non-coding RNAs(lncRNAs) have been found to be a good marker for several tumors. Using lncRNA-mining approach, we aimed to identify lncRNA expression signature that can predict breast cancer patient survival.

Methods

We performed LncRNA expression profiling in 887 breast cancer patients from Gene Expression Omnibus (GEO) datasets. The association between lncRNA signature and clinical survival was analyzed using the training set(n = 327, from GSE 20685). The validation for the association was performed in another three independent testing sets(252 from GSE21653, 204 from GSE12276, and 104 from GSE42568).

Results

A set of four lncRNA genes (U79277, AK024118, BC040204, AK000974) have been identified by the random survival forest algorithm. Using a risk score based on the expression signature of these lncRNAs, we separated the patients into low-risk and high-risk groups with significantly different survival times in the training set. This signature was validated in the other three cohorts. Further study revealed that the four-lncRNA expression signature was independent of age and subtype. Gene Set Enrichment Analysis (GSEA) suggested that gene sets were involved in several cancer metastasis related pathways.

Conclusions

These findings indicate that lncRNAs may be implicated in breast cancer pathogenesis. The four-lncRNA signature may have clinical implications in the selection of high-risk patients for adjuvant therapy.

Electronic supplementary material

The online version of this article (doi:10.1186/s13046-014-0084-7) contains supplementary material, which is available to authorized users.

Gene expression signature-based prognostic risk score in patients with glioblastoma

The present study aimed to identify genes associated with patient survival to improve our understanding of the underlying biology of gliomas. We investigated whether the expression of genes selected using random survival forests models could be used to define glioma subgroups more objectively than standard pathology. The RNA from 32 non-treated grade 4 gliomas were analyzed using the GeneChip Human Genome U133 Plus 2.0 Expression array (which contains approximately 47 000 genes). Twenty-five genes whose expressions were strongly and consistently related to patient survival were identified. The prognosis prediction score of these genes was most significant among several variables and survival analyses. The prognosis prediction score of three genes and age classifiers also revealed a strong prognostic value among grade 4 gliomas. These results were validated in an independent samples set (n = 488). Our method was effective for objectively classifying grade 4 gliomas and was a more accurate prognosis predictor than histological grading.