Development of a prognostic model for breast cancer survival in an open challenge environment

A Regularized Cox Hierarchical Model for Incorporating Annotation Information in Predictive Omic Studies

Background

Associated with high-dimensional omics data there are often "meta-features" such as biological pathways and functional annotations, summary statistics from similar studies that can be informative for predicting an outcome of interest. We introduce a regularized hierarchical framework for integrating meta-features, with the goal of improving prediction and feature selection performance with time-to-event outcomes.

Methods

A hierarchical framework is deployed to incorporate meta-features. Regularization is applied to the omic features as well as the meta-features so that high-dimensional data can be handled at both levels. The proposed hierarchical Cox model can be efficiently fitted by a combination of iterative reweighted least squares and cyclic coordinate descent.

Results

In a simulation study we show that when the external meta-features are informative, the regularized hierarchical model can substantially improve prediction performance over standard regularized Cox regression. We illustrate the proposed model with applications to breast cancer and melanoma survival based on gene expression profiles, which show the improvement in prediction performance by applying meta-features, as well as the discovery of important omic feature sets with sparse regularization at meta-feature level.

Conclusions

The proposed hierarchical regularized regression model enables integration of external meta-feature information directly into the modeling process for time-to-event outcomes, improves prediction performance when the external meta-feature data is informative. Importantly, when the external meta-features are uninformative, the prediction performance based on the regularized hierarchical model is on par with standard regularized Cox regression, indicating robustness of the framework. In addition to developing predictive signatures, the model can also be deployed in discovery applications where the main goal is to identify important features associated with the outcome rather than developing a predictive model.

A primer on the use of machine learning to distil knowledge from data in biological psychiatry

Applications of machine learning in the biomedical sciences are growing rapidly. This growth has been spurred by diverse cross-institutional and interdisciplinary collaborations, public availability of large datasets, an increase in the accessibility of analytic routines, and the availability of powerful computing resources. With this increased access and exposure to machine learning comes a responsibility for education and a deeper understanding of its bases and bounds, borne equally by data scientists seeking to ply their analytic wares in medical research and by biomedical scientists seeking to harness such methods to glean knowledge from data. This article provides an accessible and critical review of machine learning for a biomedically informed audience, as well as its applications in psychiatry. The review covers definitions and expositions of commonly used machine learning methods, and historical trends of their use in psychiatry. We also provide a set of standards, namely Guidelines for REporting Machine Learning Investigations in Neuropsychiatry (GREMLIN), for designing and reporting studies that use machine learning as a primary data-analysis approach. Lastly, we propose the establishment of the Machine Learning in Psychiatry (MLPsych) Consortium, enumerate its objectives, and identify areas of opportunity for future applications of machine learning in biological psychiatry. This review serves as a cautiously optimistic primer on machine learning for those on the precipice as they prepare to dive into the field, either as methodological practitioners or well-informed consumers.

Calibration and Uncertainty in Neural Time-to-Event Modeling

Models for predicting the time of a future event are crucial for risk assessment, across a diverse range of applications. Existing time-to-event (survival) models have focused primarily on preserving pairwise ordering of estimated event times (i.e., relative risk). We propose neural time-to-event models that account for calibration and uncertainty while predicting accurate absolute event times. Specifically, an adversarial nonparametric model is introduced for estimating matched time-to-event distributions for probabilistically concentrated and accurate predictions. We also consider replacing the discriminator of the adversarial nonparametric model with a survival-function matching estimator that accounts for model calibration. The proposed estimator can be used as a means of estimating and comparing conditional survival distributions while accounting for the predictive uncertainty of probabilistic models. Extensive experiments show that the distribution matching methods outperform existing approaches in terms of both calibration and concentration of time-to-event distributions.

Survival prediction for patients with glioblastoma multiforme using a Cox proportional hazards denoising autoencoder network

Objectives

This study aimed to establish and validate a prognostic model based on magnetic resonance imaging and clinical features to predict the survival time of patients with glioblastoma multiforme (GBM).

Methods

In this study, a convolutional denoising autoencoder (DAE) network combined with the loss function of the Cox proportional hazard regression model was used to extract features for survival prediction. In addition, the Kaplan-Meier curve, the Schoenfeld residual analysis, the time-dependent receiver operating characteristic curve, the nomogram, and the calibration curve were performed to assess the survival prediction ability.

Results

The concordance index (C-index) of the survival prediction model, which combines the DAE and the Cox proportional hazard regression model, reached 0.78 in the training set, 0.75 in the validation set, and 0.74 in the test set. Patients were divided into high- and low-risk groups based on the median prognostic index (PI). Kaplan-Meier curve was used for survival analysis (p = < 2e-16 in the training set, p = 3e-04 in the validation set, and p = 0.007 in the test set), which showed that the survival probability of different groups was significantly different, and the PI of the network played an influential role in the prediction of survival probability. In the residual verification of the PI, the fitting curve of the scatter plot was roughly parallel to the x-axis, and the p-value of the test was 0.11, proving that the PI and survival time were independent of each other and the survival prediction ability of the PI was less affected than survival time. The areas under the curve of the training set were 0.843, 0.871, 0.903, and 0.941; those of the validation set were 0.687, 0.895, 1.000, and 0.967; and those of the test set were 0.757, 0.852, 0.683, and 0.898.

Conclusion

The survival prediction model, which combines the DAE and the Cox proportional hazard regression model, can effectively predict the prognosis of patients with GBM.

Landscape of the intratumroal microenvironment in bladder cancer: Implications for prognosis and immunotherapy

Introduction

This study aims to present the landscape of the intratumoral microenvironment and by which establish a classification system that can be used to predict the prognosis of bladder cancer patients and their response to anti-PD-L1 immunotherapy.

Methods

The expression profiles of 1554 bladder cancer cases were downloaded from seven public datasets. Single-sample gene set enrichment analysis (ssGSEA), univariate Cox regression analysis, and meta-analysis were employed to establish the bladder cancer immune prognostic index (BCIPI). Extensive analyses were executed to investigate the association between BCIPI and overall survival, tumor-infiltrated immunocytes, immunotherapeutic response, mutation load, etc.

Results

The results obtained from seven independent cohorts and meta-analyses suggested that the BCIPI is an effective classification system for estimating bladder cancer patients' overall survival. Patients in the BCIPI-High subgroup revealed different immunophenotypic outcomes from those in the BCIPI-Low subgroup regarding tumor-infiltrated immunocytes and mutated genes. Subsequent analysis suggested that patients in the BCIPI-High subgroup were more sensitive to anti-PD-L1 immunotherapy than those in the BCIPI-Low subgroup.

Conclusions

The newly established BCIPI is a valuable tool for predicting overall survival outcomes and immunotherapeutic responses in patients with bladder cancer.

MPVNN: Mutated Pathway Visible Neural Network architecture for interpretable prediction of cancer-specific survival risk

MOTIVATION

Survival risk prediction using gene expression data is important in making treatment decisions in cancer. Standard neural network (NN) survival analysis models are black boxes with a lack of interpretability. More interpretable visible neural network architectures are designed using biological pathway knowledge. But they do not model how pathway structures can change for particular cancer types.

RESULTS

We propose a novel Mutated Pathway Visible Neural Network (MPVNN) architecture, designed using prior signaling pathway knowledge and random replacement of known pathway edges using gene mutation data simulating signal flow disruption. As a case study, we use the PI3K-Akt pathway and demonstrate overall improved cancer-specific survival risk prediction of MPVNN over other similar-sized NN and standard survival analysis methods. We show that trained MPVNN architecture interpretation, which points to smaller sets of genes connected by signal flow within the PI3K-Akt pathway that is important in risk prediction for particular cancer types, is reliable.

AVAILABILITY AND IMPLEMENTATION

The data and code are available at https://github.com/gourabghoshroy/MPVNN.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

CCDC69 is a prognostic marker of breast cancer and correlates with tumor immune cell infiltration

Purpose

Breast cancer (BC) is the most common malignancy and the leading cause of cancer-related death among women worldwide. Early detection, treatment, and metastasis monitoring are very important for the prognosis of BC patients. Therefore, effective biomarkers need to be explored to help monitor the prognosis of BC patients and guide treatment decisions.

Methods

In this study, the relationship between CCDC69 expression levels and tumor clinical characteristics were analyzed using RNA-seq information in BC samples from the TCGA database. Kaplan-Meier survival analysis was performed to analyze the prognostic value of CCDC69 in BC patients. Besides, gene enrichment analysis in BC samples was used to confirm the main function of CCDC69 in BC. The correlation between the expression of CCDC69 and the number of tumor-infiltrating lymphocytes was confirmed by interaction analysis of TIMER and GEPIA.

Results

The results showed that CCDC69 expression was significantly lower in cancer samples than in normal tissues, and was significantly lower in highly invasive BC than in carcinoma in situ. Meanwhile, low levels of CCDC69 were associated with a further poor prognosis. CDCC69 expression was positively correlated with the amount of different tumor-infiltrating lymphocytes. Mechanically, it could be presumed that the low expression of CCDC69 in BC might be caused by hypermethylation of the promoter region.

Conclusions

Summarily, CDCC69 could be used as a potential biomarker to predict the prognosis of BC and the sensitivity to immunotherapy such as PD-1/PD-L1 checkpoint inhibitors.

Exploration of the prognostic signature reflecting tumor microenvironment of lung adenocarcinoma based on immunologically relevant genes

Lung adenocarcinoma (LUAD) represents the major histological type of lung cancer with high mortality globally. Due to the heterogeneous nature, the same treatment strategy to various patients may result in different therapeutic responses. Hence, we aimed to elaborate an effective signature for predicting patient survival outcomes. The TCGA-LUAD cohort from the TCGA portal was used as a training dataset. The GSE26939 and GSE68465 cohorts from the GEO database were taken as validation datasets. All immunologically relevant genes were extracted from the ImmPort. The ESTIMATE algorithm was employed to explore LUAD microenvironment in the training dataset. Further, the DEGs were picked out based on the immune-associated genes reflecting different statuses in the immune context of TME. Univariate/multivariate Cox regression was performed to determine six prognosis- specific genes (PIK3CG, BTK, VEGFD, INHA, INSL4, and PTPRC) and established a risk predictive signature. The time-dependent ROC indicated that AUC values were all greater than 0.70 at 1-, 3-, and 5- year intervals. Corresponding RiskScore of each LUAD patient was calculated from the signature, and they were stratified into the high- and low-risk groups by the median value of RiskScore. K-M curves and Log-rank test demonstrated significant survival differences between the two groups (P < 0.05). Similar results were exhibited in the validation datasets. The RiskScore was incredibly relevant to clinicopathological factors like gender, AJCC stage, and T stage. Also, it can mirror the distribution state of 15 kinds of TIICs and have some predictive value for the sensitivity of therapeutic drugs.

A framework for validating AI in precision medicine: considerations from the European ITFoC consortium

BACKGROUND

Artificial intelligence (AI) has the potential to transform our healthcare systems significantly. New AI technologies based on machine learning approaches should play a key role in clinical decision-making in the future. However, their implementation in health care settings remains limited, mostly due to a lack of robust validation procedures. There is a need to develop reliable assessment frameworks for the clinical validation of AI. We present here an approach for assessing AI for predicting treatment response in triple-negative breast cancer (TNBC), using real-world data and molecular -omics data from clinical data warehouses and biobanks.

METHODS

The European "ITFoC (Information Technology for the Future Of Cancer)" consortium designed a framework for the clinical validation of AI technologies for predicting treatment response in oncology.

RESULTS

This framework is based on seven key steps specifying: (1) the intended use of AI, (2) the target population, (3) the timing of AI evaluation, (4) the datasets used for evaluation, (5) the procedures used for ensuring data safety (including data quality, privacy and security), (6) the metrics used for measuring performance, and (7) the procedures used to ensure that the AI is explainable. This framework forms the basis of a validation platform that we are building for the "ITFoC Challenge". This community-wide competition will make it possible to assess and compare AI algorithms for predicting the response to TNBC treatments with external real-world datasets.

CONCLUSIONS

The predictive performance and safety of AI technologies must be assessed in a robust, unbiased and transparent manner before their implementation in healthcare settings. We believe that the consideration of the ITFoC consortium will contribute to the safe transfer and implementation of AI in clinical settings, in the context of precision oncology and personalized care.

Immunohistochemical Evaluation of FGD3 Expression: A New Strong Prognostic Factor in Invasive Breast Cancer

Among new prognostic factors for breast cancer, the most promising one seems to be FGD3 (Facio-Genital Dysplasia 3) gene, whose expression improves outcome by inhibiting cell migration. The aim of the study was to evaluate the prognostic role of FGD3 in invasive breast cancer in a series of 401 women, treated at our unit, by evaluating the expression of this gene by immunohistochemistry. Patients with high FGD3 expression showed a significantly better disease-free survival (DFS) (p < 0.001) and overall survival (OS) (p < 0.001). The prognostic value of FGD3 expression was stronger than that of classical pathologic parameters such as histological grade of differentiation, Ki-67 index and molecular subtype. By multivariate Cox analysis, FGD3 expression was confirmed as significant and independent prognostic factor, ranking second after age at diagnosis (≤40 years) for DFS (p = 0.003) and the second strongest predictor of OS, after AJCC Stage (p < 0.001). Our data suggest that inclusion of FGD3 evaluation in the routine workup of breast cancer patients may result in a more accurate stratification of the individual risk. The possibility to assess FGD3 expression by a simple and cheap technique such as immunohistochemistry may enhance the spread of its use in the clinical practice.

Single-cell analysis reveals the pan-cancer invasiveness-associated transition of adipose-derived stromal cells into COL11A1-expressing cancer-associated fibroblasts

During the last ten years, many research results have been referring to a particular type of cancer-associated fibroblasts associated with poor prognosis, invasiveness, metastasis and resistance to therapy in multiple cancer types, characterized by a gene expression signature with prominent presence of genes COL11A1, THBS2 and INHBA. Identifying the underlying biological mechanisms responsible for their creation may facilitate the discovery of targets for potential pan-cancer therapeutics. Using a novel computational approach for single-cell gene expression data analysis identifying the dominant cell populations in a sequence of samples from patients at various stages, we conclude that these fibroblasts are produced by a pan-cancer cellular transition originating from a particular type of adipose-derived stromal cells naturally present in the stromal vascular fraction of normal adipose tissue, having a characteristic gene expression signature. Focusing on a rich pancreatic cancer dataset, we provide a detailed description of the continuous modification of the gene expression profiles of cells as they transition from APOD-expressing adipose-derived stromal cells to COL11A1-expressing cancer-associated fibroblasts, identifying the key genes that participate in this transition. These results also provide an explanation to the well-known fact that the adipose microenvironment contributes to cancer progression.

Predicting survival time of Korean hepatocellular carcinoma patients using the Cox proportional hazards model: a retrospective study based on big data analysis

AIM

To predict survival time of Korean hepatocellular carcinoma (HCC) patients by analyzing big data using Cox proportional hazards model.

METHODS

Big data of the patients who underwent treatment for HCC from 2008 to 2015, provided by Korea Central Cancer Registry, National Cancer Center, and Ministry of Health and Welfare, were analyzed. A total of 10 742 patients with HCC were divided into two groups, with Group I (3021 patients) confirmed on biopsy and Group II (5563 patients) diagnosed as HCC according to HCC diagnostic criteria as outlined in Korean Liver Cancer Association guidelines. Univariate and multivariate Cox regression analyses were performed to identify independent risk factors of recurrence after treatment and survival status.

RESULTS

A total of 3021 patients in Group I and 5563 patients in Group II were included in the study and the difference in survival time between the two groups was statistically significant (P < 0.05). Recurrence was only included in intrahepatic cases, and the rates were 21.2 and 19.8% while the periods from the first treatment to recurrence were 15.57 and 14.19 months, respectively. Age, diabetes, BMI, platelet, alpha-fetoprotein, histologic tumor maximum size, imaging T stage, presence of recurrence, and duration of recurrence were included in multivariate analysis.

CONCLUSION

By using nationwide, multicenter big data, it is possible to predict recurrence rate and survival time which can provide the basis for treatment response to develop a predictive program.

SurvCNN: A Discrete Time-to-Event Cancer Survival Estimation Framework Using Image Representations of Omics Data

The utility of multi-omics in personalized therapy and cancer survival analysis has been debated and demonstrated extensively in the recent past. Most of the current methods still suffer from data constraints such as high-dimensionality, unexplained interdependence, and subpar integration methods. Here, we propose SurvCNN, an alternative approach to process multi-omics data with robust computer vision architectures, to predict cancer prognosis for Lung Adenocarcinoma patients. Numerical multi-omics data were transformed into their image representations and fed into a Convolutional Neural network with a discrete-time model to predict survival probabilities. The framework also dichotomized patients into risk subgroups based on their survival probabilities over time. SurvCNN was evaluated on multiple performance metrics and outperformed existing methods with a high degree of confidence. Moreover, comprehensive insights into the relative performance of various combinations of omics datasets were probed. Critical biological processes, pathways and cell types identified from downstream processing of differentially expressed genes suggested that the framework could elucidate elements detrimental to a patient's survival. Such integrative models with high predictive power would have a significant impact and utility in precision oncology.

Accurate cancer phenotype prediction with AKLIMATE, a stacked kernel learner integrating multimodal genomic data and pathway knowledge

Advancements in sequencing have led to the proliferation of multi-omic profiles of human cells under different conditions and perturbations. In addition, many databases have amassed information about pathways and gene "signatures"-patterns of gene expression associated with specific cellular and phenotypic contexts. An important current challenge in systems biology is to leverage such knowledge about gene coordination to maximize the predictive power and generalization of models applied to high-throughput datasets. However, few such integrative approaches exist that also provide interpretable results quantifying the importance of individual genes and pathways to model accuracy. We introduce AKLIMATE, a first kernel-based stacked learner that seamlessly incorporates multi-omics feature data with prior information in the form of pathways for either regression or classification tasks. AKLIMATE uses a novel multiple-kernel learning framework where individual kernels capture the prediction propensities recorded in random forests, each built from a specific pathway gene set that integrates all omics data for its member genes. AKLIMATE has comparable or improved performance relative to state-of-the-art methods on diverse phenotype learning tasks, including predicting microsatellite instability in endometrial and colorectal cancer, survival in breast cancer, and cell line response to gene knockdowns. We show how AKLIMATE is able to connect feature data across data platforms through their common pathways to identify examples of several known and novel contributors of cancer and synthetic lethality.

Phenomapping of Patients with Primary Breast Cancer Using Machine Learning-Based Unsupervised Cluster Analysis

Primary breast cancer (PBC) is a heterogeneous disease at the clinical, histopathological, and molecular levels. The improved classification of PBC might be important to identify subgroups of the disease, relevant to patient management. Machine learning algorithms may allow a better understanding of the relationships within heterogeneous clinical syndromes. This work aims to show the potential of unsupervised learning techniques for improving classification in PBC. A dataset of 712 women with PBC is used as a motivating example. A set of variables containing biological prognostic parameters is considered to define groups of individuals. Four different clustering methods are used: K-means, self-organising maps, hierarchical agglomerative (HAC), and Gaussian mixture models clustering. HAC outperforms the other clustering methods. With an optimal partitioning parameter, the methods identify two clusters with different clinical profiles. Patients in the first cluster are younger and have lower values of the oestrogen receptor (ER) and progesterone receptor (PgR) than patients in the second cluster. Moreover, cathepsin D values are lower in the first cluster. The three most important variables identified by the HAC are: age, ER, and PgR. Unsupervised learning seems a suitable alternative for the analysis of PBC data, opening up new perspectives in the particularly active domain of dissecting clinical heterogeneity.

Germline genetic contribution to the immune landscape of cancer

Understanding the contribution of the host's genetic background to cancer immunity may lead to improved stratification for immunotherapy and to the identification of novel therapeutic targets. We investigated the effect of common and rare germline variants on 139 well-defined immune traits in ∼9000 cancer patients enrolled in TCGA. High heritability was observed for estimates of NK cell and T cell subset infiltration and for interferon signaling. Common variants of IFIH1, TMEM173 (STING1), and TMEM108 were associated with differential interferon signaling and variants mapping to RBL1 correlated with T cell subset abundance. Pathogenic or likely pathogenic variants in BRCA1 and in genes involved in telomere stabilization and Wnt-β-catenin also acted as immune modulators. Our findings provide evidence for the impact of germline genetics on the composition and functional orientation of the tumor immune microenvironment. The curated datasets, variants, and genes identified provide a resource toward further understanding of tumor-immune interactions.

Pan-cancer driver copy number alterations identified by joint expression/CNA data analysis

Analysis of large gene expression datasets from biopsies of cancer patients can identify co-expression signatures representing particular biomolecular events in cancer. Some of these signatures involve genomically co-localized genes resulting from the presence of copy number alterations (CNAs), for which analysis of the expression of the underlying genes provides valuable information about their combined role as oncogenes or tumor suppressor genes. Here we focus on the discovery and interpretation of such signatures that are present in multiple cancer types due to driver amplifications and deletions in particular regions of the genome after doing a comprehensive analysis combining both gene expression and CNA data from The Cancer Genome Atlas.

Deep learning for risk prediction in patients with nasopharyngeal carcinoma using multi-parametric MRIs

BACKGROUND

Magnetic resonance images (MRI) is the main diagnostic tool for risk stratification and treatment decision in nasopharyngeal carcinoma (NPC). However, the holistic feature information of multi-parametric MRIs has not been fully exploited by clinicians to accurately evaluate patients.

OBJECTIVE

To help clinicians fully utilize the missed information to regroup patients, we built an end-to-end deep learning model to extract feature information from multi-parametric MRIs for predicting and stratifying the risk scores of NPC patients.

METHODS

In this paper, we proposed an end-to-end multi-modality deep survival network (MDSN) to precisely predict the risk of disease progression of NPC patients. Extending from 3D dense net, this proposed MDSN extracted deep representation from multi-parametric MRIs (T1w, T2w, and T1c). Moreover, deep features and clinical stages were integrated through MDSN to more accurately predict the overall risk score (ORS) of individual NPC patient.

RESULT

A total of 1,417 individuals treated between January 2012 and December 2014 were included for training and validating the end-to-end MDSN. Results were then tested in a retrospective cohort of 429 patients included in the same institution. The C-index of the proposed method with or without clinical stages was 0.672 and 0.651 on the test set, respectively, which was higher than the that of the stage grouping (0.610).

CONCLUSIONS

The C-index of the model which integrated clinical stages with deep features is 0.062 higher than that of stage grouping alone (0.672 vs 0.610). We conclude that features extracted from multi-parametric MRIs based on MDSN can well assist the clinical stages in regrouping patients.

Nicotinamide N-Methyltransferase Is a Prognostic Biomarker and Correlated With Immune Infiltrates in Gastric Cancer

Gastric cancer (GC) is the third most common cause of cancer-related death in the word. Immunotherapy is a promising treatment of cancer. However, it is unclear which GC subpopulation would benefit most from immunotherapy and it is necessary to develop effective biomarkers for predicting immunotherapy response. Nicotinamide N-methyltransferase (NNMT) is a metabolic regulator of cancer-associated fibroblast (CAF) differentiation and cancer progression. In this study, we explored the correlations of NNMT to tumor-infiltrating immune cells (TIICs) and immune marker sets in The Cancer Genome Atlas Stomach Adenocarcinoma STAD (TCGA-STAD). Subsequently, we screened the NNMT correlated genes and performed the enrichment analysis of these genes. We eventually predicted the 19 most potential small-molecule drugs using the connectivity map (CMap) and Comparative Toxicogenomics Database (CTD). Also, nadolol, tranexamic acid, felbinac and dapsone were considered the four most promising drugs for GC. In summary, NNMT can be used as a prognostic biomarker that reflect immune infiltration level and a novel therapeutic target in GC.

PtdIns(3,4,5)P3-dependent Rac exchanger 1 (P-Rex1) promotes mammary tumor initiation and metastasis

The Rac-GEF, P-Rex1, activates Rac1 signaling downstream of G protein-coupled receptors and PI3K. Increased P-Rex1 expression promotes melanoma progression; however, its role in breast cancer is complex, with differing reports of the effect of its expression on disease outcome. To address this we analyzed human databases, undertook gene array expression analysis, and generated unique murine models of P-Rex1 gain or loss of function. Analysis of PREX1 mRNA expression in breast cancer cDNA arrays and a METABRIC cohort revealed that higher PREX1 mRNA in ER^+ve/luminal tumors was associated with poor outcome in luminal B cancers. Prex1 deletion in MMTV-neu or MMTV-PyMT mice reduced Rac1 activation in vivo and improved survival. High level MMTV-driven transgenic PREX1 expression resulted in apicobasal polarity defects and increased mammary epithelial cell proliferation associated with hyperplasia and development of de novo mammary tumors. MMTV-PREX1 expression in MMTV-neu mice increased tumor initiation and enhanced metastasis in vivo, but had no effect on primary tumor growth. Pharmacological inhibition of Rac1 or MEK1/2 reduced P-Rex1-driven tumoroid formation and cell invasion. Therefore, P-Rex1 can act as an oncogene and cooperate with HER2/neu to enhance breast cancer initiation and metastasis, despite having no effect on primary tumor growth.

TILRR (FREM1 isoform 2) is a prognostic biomarker correlated with immune infiltration in breast cancer

In atherosclerosis, upregulated TILRR (FREM1 isoform 2) expression increases immune cell infiltration. We hypothesized that TILRR expression is also correlated with cancer progression. By analyzing data from Oncomine and the Tumor Immune Estimation Resource, we found that TILRR mRNA expression was significantly lower in breast cancer tissue than adjacent normal tissue. Kaplan-Meier survival analysis and immunohistochemical staining revealed shortened overall survival and disease-free survival in patients with low TILRR expression. TILRR transcript expression was positively correlated with immune score, immune cell biomarkers and the expression of CXCL10 and CXCL11. TILRR expression was also positively correlated with CD8+ and CD4+ T-cell infiltration. These correlations were verified using the ESTIMATE algorithm, gene set enrichment analysis and Q-PCR. We concluded that impaired TILRR expression is correlated with breast cancer prognosis and immune cell infiltration.

Chloride intracellular channel protein 2: prognostic marker and correlation with PD-1/PD-L1 in breast cancer

Immune checkpoint inhibition has emerged as an effective treatment for multiple solid tumors, including advanced-stage breast cancer (BC). During the past decade, the US Food and Drug Administration has approved a number of agents for immune checkpoint blockade (ICB). However, the limited data on monotherapy anti-tumor activity in BC underscores the need for robust predictive biomarker development. Here, we used weighted gene coexpression network analysis of genes differentially expressed between BC and normal tissue to identify genes coexpressed with programmed death-1 (PD-1) and its ligand (PD-L1). Tumor Immune Estimation Resource and Gene Expression Profiling Interaction Analysis were used to assess the relationship between gene expression and the abundance of tumor-infiltrating lymphocytes (TILs). We found that chloride intracellular channel protein 2 (CLIC2) was not only coexpressed with PD-1 and PD-L1, but its increased expression was associated with a favorable prognosis and enrichment of multiple TIL types, particularly CD8+ T cells. These results suggest that CLIC2 is a potentially useful biomarker for identifying BC patients who could benefit from ICB.

Prognostication of a 13-immune-related-gene signature in patients with early triple-negative breast cancer

PURPOSE

We investigated the expression profiles of immune genes in patients with triple-negative breast cancer (TNBC) to identify the prognostic value of immune genes and their clinical implications.

METHODS

NanoString nCounter Analysis of 770 immune-related genes was used to measure immune gene expression in patients with TNBC who underwent curative surgery followed by adjuvant chemotherapy at Samsung Medical Center between 2000 and 2004. Statistical analyses were conducted to identify the associations between gene expression and distant recurrence-free survival (DRFS).

RESULTS

Of 1189 patients who underwent curative BC surgery, 200 TNBC patients were included and stage was the only clinical factor predictive of DRFS. In terms of immune genes, 155 of 770 genes were associated with DRFS (p < 0.01). Further multivariate analysis revealed that 13 genes, CD1B, CD53, CT45A1, GTF3C1, IL11RA, IL1RN, LRRN3, MAPK1, NEFL, PRKCE, PTPRC, SPACA3 and TNFSF11, were associated with patient prognosis (p < 0.05). The prognostic value of stage and expression levels of 13 immune genes was analyzed and the area under the receiver operating characteristic curve (AUC) was 0.923. Based on the AUC, we divided patients into three genetic risk groups and DRFS rate was significantly different according to genetic risk groups, even in the same stage (p < 0.001).

CONCLUSIONS

In this study, a 13-gene expression profile in combination with stage precisely predicted distant recurrence of early TNBC. Therefore, this 13-immune-gene signature could help predict TNBC prognosis and provide guidance for treatment as well as the opportunity to develop new targets for immunotherapy in TNBC patients.

An integrated prognosis model of pharmacogenomic gene signature and clinical information for diffuse large B-cell lymphoma patients following CHOP-like chemotherapy

Background

As the most common form of lymphoma, diffuse large B-cell lymphoma (DLBCL) is a clinical highly heterogeneous disease with variability in therapeutic outcomes and biological features. It is a challenge to identify of clinically meaningful tools for outcome prediction. In this study, we developed a prognosis model fused clinical characteristics with drug resistance pharmacogenomic signature to identify DLBCL prognostic subgroups for CHOP-based treatment.

Methods

The expression microarray data and clinical characteristics of 791 DLBCL patients from two Gene Expression Omnibus (GEO) databases were used to establish and validate this model. By using univariate Cox regression, eight clinical or genetic signatures were analyzed. The elastic net-regulated Cox regression analysis was used to select the best prognosis related factors into the predictive model. To estimate the prognostic capability of the model, Kaplan–Meier curve and the area under receiver operating characteristic (ROC) curve (AUC) were performed.

Results

A predictive model comprising 4 clinical factors and 2 pharmacogenomic gene signatures was established after 1000 times cross validation in the training dataset. The AUC of the comprehensive risk model was 0.78, whereas AUC value was lower for the clinical only model (0.68) or the gene only model (0.67). Compared with low-risk patients, the overall survival (OS) of DLBCL patients with high-risk scores was significantly decreased (HR = 4.55, 95% CI 3.14–6.59, log-rank p value = 1.06 × 10⁻¹⁵). The signature also enables to predict prognosis within different molecular subtypes of DLBCL. The reliability of the integrated model was confirmed by independent validation dataset (HR = 3.47, 95% CI 2.42–4.97, log rank p value = 1.53 × 10⁻¹¹).

Conclusions

This integrated model has a better predictive capability to ascertain the prognosis of DLBCL patients prior to CHOP-like treatment, which may improve the clinical management of DLBCL patients and provide theoretical basis for individualized treatment.

Immune cell infiltration-based signature for prognosis and immunogenomic analysis in breast cancer

Breast cancer is one of the most human malignant diseases and the leading cause of cancer-related death in the world. However, the prognostic and therapeutic benefits of breast cancer patients cannot be predicted accurately by the current stratifying system. In this study, an immune-related prognostic score was established in 22 breast cancer cohorts with a total of 6415 samples. An extensive immunogenomic analysis was conducted to explore the relationships between immune score, prognostic significance, infiltrating immune cells, cancer genotypes and potential immune escape mechanisms. Our analysis revealed that this immune score was a promising biomarker for estimating overall survival in breast cancer. This immune score was associated with important immunophenotypic factors, such as immune escape and mutation load. Further analysis revealed that patients with high immune scores exhibited therapeutic benefits from chemotherapy and immunotherapy. Based on these results, we can conclude that this immune score may be a useful tool for overall survival prediction and treatment guidance for patients with breast cancer.

The role of the CDCA gene family in ovarian cancer

Background

Ovarian cancer is a frequently-occurring reproductive system malignancy in females, which leads to an annual of over 100 thousand deaths worldwide.

Methods

The electronic databases, including GEPIA, ONCOMINE, Metascape, and Kaplan-Meier Plotter, were used to examine both survival and transcriptional data regarding the cell division cycle associated (CDCA) gene family among ovarian cancer patients.

Results

All CDCA genes expression levels were up-regulated in ovarian cancer tissues relative to those in non-carcinoma ovarian counterparts. Besides, CDCA5/7 expression levels were related to the late tumor stage. In addition, the Kaplan-Meier Plotter database was employed to carry out survival analysis, which suggested that ovarian cancer patients with increased CDCA2/3/5/7 expression levels had poor overall survival (OS) (P<0.05). Moreover, ovarian cancer patients that had up-regulated mRNA expression levels of CDCA2/5/8 had markedly reduced progression-free survival (PFS) (P<0.05); and up-regulated CDCA4 expression showed remarkable association with reduced post-progression survival (PPS) (P<0.05). Additionally, the following processes were affected by CDCA genes alterations, including R-HAS-2500257: resolution of sister chromatid cohesion; GO:0051301: cell division; CORUM: 1118: Chromosomal passenger complex (CPC, including CDCA8, INCENP, AURKB and BIRC5); CORUM: 127: NDC80 kinetochore complex; M129: PID PLK1 pathway; and GO: 0007080: mitotic metaphase plate congression, all of which were subjected to marked regulation since the alterations affected CDCA genes.

Conclusions

Up-regulated CDCA gene expression in ovarian cancer tissues probably played a crucial part in the occurrence of ovarian cancer. The up-regulated CDCA2/3/5/7 expression levels were used as the potential prognostic markers to improve the poor ovarian cancer survival and prognostic accuracy. Moreover, CDCA genes probably exerted their functions in tumorigenesis through the PLK1 pathway.

Analysis of immune-related signatures of lung adenocarcinoma identified two distinct subtypes: implications for immune checkpoint blockade therapy

Immune checkpoint blockade (ICB) therapies have revolutionized the treatment of human cancers including lung adenocarcinoma (LUAD). However, our understanding of the immune subtyping of LUAD and its association with clinical response of immune checkpoint inhibitor remains incomplete. Here we performed molecular subtyping and association analysis of LUAD from the Cancer Genome Atlas (TCGA) and validated findings from TCGA cohort in 9 independent validation cohorts. We conducted consensus molecular subtyping with nonnegative matrix factorization (NMF). Potential response of ICB therapy was estimated with Tumor Immune Dysfunction and Exclusion (TIDE) algorithm. We identified 2 distinct subtypes of LUAD in TCGA cohort that were characterized by significantly different survival outcomes (i.e., high- and low-risk subtypes). The high-risk subtype was featured by lower TIDE score, upregulation of programmed death-ligand 1 (PD-L1) expression, and higher tumor mutation burden (TMB). The high-risk subtype also harbored significantly elevated cell cycle modulators CDK4/CDK6 and TP53 mutation. These observations were validated in 9 independent LUAD cohorts. Our findings suggest that immune checkpoint blockade therapy may be efficacious for high-risk subtype of LUAD patients.

Comprehensive Analysis of Aberrantly Expressed Profiles of lncRNAs and miRNAs with Associated ceRNA Network in Lung Adenocarcinoma and Lung Squamous Cell Carcinoma

Lung cancer (LC) continues to be the leading cause of cancer-related deaths worldwide and the prognosis remains poor worldwide. At present, the long non-coding RNAs (lncRNAs) was considered as a part of competing endogenous RNA (ceRNA) network act as natural microRNA (miRNA) sponges to regulate protein-coding gene expression. However, functional roles of lncRNA-mediated ceRNAs in LC are insufficiently understood. To classify the specific mechanism of lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), we comprehensively compared the expression profiles of mRNAs, lncRNAs and miRNAs obtained from 509 LUAD, 473 LUSC tissues and 49 adjacent non-cancerous lung tissues, based on The Cancer Genome Atlas (TCGA). After screening for differently expressed (DE) mRNAs, DEmiRNAs, DElncRNAs and weighted gene co-expression network analysis (WGCNA) (|log2FC| > 2.0 and an adjusted p value <0.05), a total of 4478 DEmRNAs, 526 DElncRNAs and 75 DEmiRNAs in LUAD, while 6237 DEmRNAs, 843 DElncRNAs and 117 DEmiRNAs in LUSC were discovered. Interaction (PPI) network analysis was performed to identify 656 nodes and 2987 edges (minimum required interaction score > 0.9), as well as 8 different protein-protein interactions. Gene ontology (GO) analysis mainly associated with cell proliferation. KEGG pathway enrichment analyses most partly associated with metabolism pathway and cytokine-cytokine receptor interaction. Finally, the dysregulated lncRNA-miRNA-ceRNA network was constructed based on correlation analyses and a total of 62 dysregulated lncRNAs, 28 DEmRNAs and 18 DEmiRNAs were involved. The most significant lncRNAs included DElncRNAs, LINC00641 and AC004947.2, miRNAs included miR-6860, miR-1285-3p, miR-767-3p and miR-7974, mRNAs included MAP3K3, FGD3 and ATP1B2. Then we analyzed and described the potential characteristics of biological function and pathological roles of the LUAD and LUSC ceRNA co-regulatory network. Our findings revealed ceRNA network will be beneficial for promoting the understanding of lncRNA-mediated ceRNA regulatory mechanisms in the pathogenesis of LUAD and LUSC.

JAK1 as a prognostic marker and its correlation with immune infiltrates in breast cancer

Clinical trials testing Janus kinase-1 (JAK1) inhibitors in cancers are under way. Whether the JAK1 mRNA levels in breast tumors correlates with outcome has not been evaluated. JAK1 expression was analyzed via the Oncomine database and Tumor IMmune Estimation Resource site. Tumor tissues from 57 breast cancer patients were used for qRT-PCR and immune infiltration assessment. JAK1 expression was significantly lower in breast invasive carcinoma compared with adjacent normal tissues. Public databases (Kaplan-Meier plotter and PrognoScan) showed that low JAK1 expression was associated with poorer survival. Data from The Cancer Genome Atlas (TCGA) showed that high JAK1 expression was associated with increased survival in both TNM I-II and TNM III-IV patients. JAK1 was inversely correlated with tumor size status, lymph node status, and TNM of breast cancer patients. JAK1 levels were correlated with the T cell transcript-enriched LYM metagene signature and was significantly lower in the low tumor infiltrating lymphocytes (TILs) group. JAK1 expression levels had significant positive correlations with infiltrating levels of CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells in breast cancer and not with other B cells. In conclusion, JAK1 mRNA levels were correlated with prognosis and immune infiltrating levels in breast cancer.

Gene expression based survival prediction for cancer patients-A topic modeling approach

Cancer is one of the leading cause of death, worldwide. Many believe that genomic data will enable us to better predict the survival time of these patients, which will lead to better, more personalized treatment options and patient care. As standard survival prediction models have a hard time coping with the high-dimensionality of such gene expression data, many projects use some dimensionality reduction techniques to overcome this hurdle. We introduce a novel methodology, inspired by topic modeling from the natural language domain, to derive expressive features from the high-dimensional gene expression data. There, a document is represented as a mixture over a relatively small number of topics, where each topic corresponds to a distribution over the words; here, to accommodate the heterogeneity of a patient's cancer, we represent each patient (≈ document) as a mixture over cancer-topics, where each cancer-topic is a mixture over gene expression values (≈ words). This required some extensions to the standard LDA model-e.g., to accommodate the real-valued expression values-leading to our novel discretized Latent Dirichlet Allocation (dLDA) procedure. After using this dLDA to learn these cancer-topics, we can then express each patient as a distribution over a small number of cancer-topics, then use this low-dimensional "distribution vector" as input to a learning algorithm-here, we ran the recent survival prediction algorithm, MTLR, on this representation of the cancer dataset. We initially focus on the METABRIC dataset, which describes each of n = 1,981 breast cancer patients using the r = 49,576 gene expression values, from microarrays. Our results show that our approach (dLDA followed by MTLR) provides survival estimates that are more accurate than standard models, in terms of the standard Concordance measure. We then validate this "dLDA+MTLR" approach by running it on the n = 883 Pan-kidney (KIPAN) dataset, over r = 15,529 gene expression values-here using the mRNAseq modality-and find that it again achieves excellent results. In both cases, we also show that the resulting model is calibrated, using the recent "D-calibrated" measure. These successes, in two different cancer types and expression modalities, demonstrates the generality, and the effectiveness, of this approach. The dLDA+MTLR source code is available at https://github.com/nitsanluke/GE-LDA-Survival.

Expression of FGD3 gene as prognostic factor in young breast cancer patients

The FGD3 gene works as a cell migration inhibitor and seems to be a promising indicator of outcome in some human cancers including breast. In this study, we analysed for the first time the prognostic role of FGD3 in young breast cancer patients. We studied the relationship between traditional prognostic factors, FGD3 expression and outcome in ≤40 years breast cancer patients. We found that lower FGD3 expression decreased the probability of disease-free survival (p = 0.042) and overall survival (p = 0.007). In a multivariate analysis for overall survival AJCC stage (p = 0.005) and FGD3 expression (p = 0.03) resulted independent prognostic factors. Low FGD3 expression increased the risk of death from disease (HR 5.73, p = 0.03). Moreover, low FGD3 expression was associated with more widespread lymph node involvement (p = 0.04) and a lower FGD3 staining intensity was found in positive-lymph-node patients vs negative (p = 0.003) and in patients with ≥10 involved lymph nodes vs <10 (p = 0.05). Our results suggest FGD3 to be a significant independent prognostic factor in young breast cancer patients in terms of disease-free survival and overall survival. A lower expression increased the risk of recurrence and death from disease and was associated with widespread lymph node metastases.

PREX1 drives spontaneous bone dissemination of ER+ breast cancer cells

A significant proportion of breast cancer patients develop bone metastases, but the mechanisms regulating tumor cell dissemination from the primary site to the skeleton remain largely unknown. Using a novel model of spontaneous bone metastasis derived from human ER+ MCF7 cells, molecular profiling revealed increased PREX1 expression in a cell line established from bone-disseminated MCF7 cells (MCF7b), which were more migratory, invasive, and adhesive in vitro compared to parental MCF7 cells, and this phenotype was mediated by PREX1. MCF7b cells grew poorly in the primary tumor site when re-inoculated in vivo, suggesting these cells are primed to grow in the bone, and were enriched in skeletal sites of metastasis over soft tissue sites. Skeletal dissemination from the primary tumor was reversed with PREX1 knockdown, indicating that PREX1 is a key driver of spontaneous dissemination of tumor cells from the primary site to the bone marrow. In breast cancer patients, PREX1 levels are significantly increased in ER+ tumors and associated with invasive disease and distant metastasis. Together, these findings implicate PREX1 in spontaneous bone dissemination and provide a significant advance to the molecular mechanisms by which breast cancer cells disseminate from the primary tumor site to bone.

Application of machine learning in rheumatic disease research

Over the past decade, there has been a paradigm shift in how clinical data are collected, processed and utilized. Machine learning and artificial intelligence, fueled by breakthroughs in high-performance computing, data availability and algorithmic innovations, are paving the way to effective analyses of large, multi-dimensional collections of patient histories, laboratory results, treatments, and outcomes. In the new era of machine learning and predictive analytics, the impact on clinical decision-making in all clinical areas, including rheumatology, will be unprecedented. Here we provide a critical review of the machine-learning methods currently used in the analysis of clinical data, the advantages and limitations of these methods, and how they can be leveraged within the field of rheumatology.

A deep survival analysis method based on ranking

Survival analyses of populations and the establishment of prognoses for individual patients are important activities in the practice of medicine. Standard survival models, such as the Cox proportional hazards model, require extensive feature engineering or prior knowledge to model at an individual level. Some survival analysis models can avoid these problems by using machine learning extended the CPH model, and higher performance has been reported. In this paper, we propose an innovative loss function that is defined as the sum of an extended mean squared error loss and a pairwise ranking loss based on ranking information on survival data. We apply this loss function to optimize a deep feed-forward neural network (RankDeepSurv), which can be used to model survival data. We demonstrate that the performance of our model, RankDeepSurv, is superior to that of other state-of-the-art survival models based on an analysis of 4 public medical clinical datasets. When modelling the prognosis of nasopharyngeal carcinoma (NPC), RankDeepSurv achieved better prognostic accuracy than the CPH established by clinical experts. The difference between high and low risk groups in the RankDeepSurv model is greater than the difference in the CPH. The results show that our method has considerable potential to model survival data in medical settings.

Systems biology: perspectives on multiscale modeling in research on endocrine-related cancers

Drawing on concepts from experimental biology, computer science, informatics, mathematics and statistics, systems biologists integrate data across diverse platforms and scales of time and space to create computational and mathematical models of the integrative, holistic functions of living systems. Endocrine-related cancers are well suited to study from a systems perspective because of the signaling complexities arising from the roles of growth factors, hormones and their receptors as critical regulators of cancer cell biology and from the interactions among cancer cells, normal cells and signaling molecules in the tumor microenvironment. Moreover, growth factors, hormones and their receptors are often effective targets for therapeutic intervention, such as estrogen biosynthesis, estrogen receptors or HER2 in breast cancer and androgen receptors in prostate cancer. Given the complexity underlying the molecular control networks in these cancers, a simple, intuitive understanding of how endocrine-related cancers respond to therapeutic protocols has proved incomplete and unsatisfactory. Systems biology offers an alternative paradigm for understanding these cancers and their treatment. To correctly interpret the results of systems-based studies requires some knowledge of how in silico models are built, and how they are used to describe a system and to predict the effects of perturbations on system function. In this review, we provide a general perspective on the field of cancer systems biology, and we explore some of the advantages, limitations and pitfalls associated with using predictive multiscale modeling to study endocrine-related cancers.

Prognostic Impact of Src, CDKN1B, and JAK2 Expression in Metastatic Breast Cancer Patients Treated with Trastuzumab

BACKGROUND: Src, CDKN1B, and JAK2 play a crucial role in the coordination of cell signaling pathways. In the present study, we aim to investigate the prognostic significance of these biomarkers in HER2-positive metastatic breast cancer (MBC) patients treated with trastuzumab (T). METHODS: Formalin-fixed paraffin-embedded tumor tissue samples from 197 patients with HER2-positive MBC treated with T were retrospectively collected. All tissue samples were centrally assessed for ER, PgR, Ki67, HER2, and PTEN protein expression; EGFR gene amplification; PI3KCA mutational status; and tumor-infiltrating lympocytes density. Src, CDKN1B, and JAK2 mRNA expression was evaluated using quantitative reverse transcription-polymerase chain reaction. RESULTS: Only 133 of the 197 patients (67.5%) were found to be HER2-positive by central assessment. CDKN1B mRNA expression was strongly correlated with Src (rho = 0.71) and JAK2 (rho = 0.54). In HER2-positive patients, low CDKN1B conferred higher risk for progression [hazard ratio (HR) = 1.58, 95% confidence interval (CI) 1.08-2.32, P = .018]. In HER2-negative patients, low Src was associated with longer survival (HR = 0.56, 95% CI 0.32-0.99, P = .045). Upon multivariate analyses, only low CDKN1B and JAK2 mRNA expression remained unfavorable factors for PFS in de novo and relapsed (R)-MBC patients, respectively (HR = 2.36, 95% CI 1.01-5.48, P = .046 and HR = 1.76, 95% CI 1.01-3.06, P = .047, respectively). CONCLUSIONS: Low CDKN1B and JAK2 mRNA expressions were unfavorable prognosticators in a cohort of T-treated MBC patients. Our results suggest that CDKN1B and JAK2, if validated, may serve as prognostic factors potentially implicated in T resistance, which seems to be associated with distinct pathways in de novo and R-MBC.

Gene expression profiles analysis identifies a novel two-gene signature to predict overall survival in diffuse large B-cell lymphoma

Diffuse large B-cell lymphoma (DLBCL) is the most common hematologic malignancy, however, specific tumor-associated genes and signaling pathways are yet to be deciphered. Differentially expressed genes (DEGs) were computed based on gene expression profiles from GSE32018, GSE56315, and The Cancer Genome Atlas (TCGA) DLBC. Overlapping DEGs were then evaluated for gene ontology (GO), pathways enrichment, DNA methylation, protein–protein interaction (PPI) network analysis as well as survival analysis. Seventy-four up-regulated and 79 down-regulated DEGs were identified. From PPI network analysis, majority of the DEGs were involved in cell cycle, oocyte meiosis, and epithelial-to-mesenchymal transition (EMT) pathways. Six hub genes including CDC20, MELK, PBK, prostaglandin D2 synthase (PTGDS), PCNA, and CDK1 were selected using the Molecular Complex Detection (MCODE). CDC20 and PTGDS were able to predict overall survival (OS) in TCGA DLBC and in an additional independent cohort GSE31312. Furthermore, CDC20 DNA methylation negatively regulated CDC20 expression and was able to predict OS in DLBCL. A two-gene panel consisting of CDC20 and PTGDS had a better prognostic value compared with CDC20 or PTGDS alone in the TCGA cohort (P=0.026 and 0.039). Overall, the present study identified a set of novel genes and pathways that may play a significant role in the initiation and progression of DLBCL. In addition, CDC20 and PTGDS will provide useful guidance for therapeutic applications.

The Immune Landscape of Cancer

We performed an extensive immunogenomic analysis of more than 10,000 tumors comprising 33 diverse cancer types by utilizing data compiled by TCGA. Across cancer types, we identified six immune subtypes-wound healing, IFN-γ dominant, inflammatory, lymphocyte depleted, immunologically quiet, and TGF-β dominant-characterized by differences in macrophage or lymphocyte signatures, Th1:Th2 cell ratio, extent of intratumoral heterogeneity, aneuploidy, extent of neoantigen load, overall cell proliferation, expression of immunomodulatory genes, and prognosis. Specific driver mutations correlated with lower (CTNNB1, NRAS, or IDH1) or higher (BRAF, TP53, or CASP8) leukocyte levels across all cancers. Multiple control modalities of the intracellular and extracellular networks (transcription, microRNAs, copy number, and epigenetic processes) were involved in tumor-immune cell interactions, both across and within immune subtypes. Our immunogenomics pipeline to characterize these heterogeneous tumors and the resulting data are intended to serve as a resource for future targeted studies to further advance the field.

Health-Related Quality of Life in a Predictive Model for Mortality in Older Breast Cancer Survivors

OBJECTIVES

To develop a predictive model and risk score for 10-year mortality using health-related quality of life (HRQOL) in a cohort of older women with early-stage breast cancer.

DESIGN

Prospective cohort.

SETTING

Community.

PARTICIPANTS

U.S. women aged 65 and older diagnosed with Stage I to IIIA primary breast cancer (N=660).

MEASUREMENTS

We used medical variables (age, comorbidity), HRQOL measures (10-item Physical Function Index and 5-item Mental Health Index from the Medical Outcomes Study (MOS) 36-item Short-Form Survey; 8-item Modified MOS Social Support Survey), and breast cancer variables (stage, surgery, chemotherapy, endocrine therapy) to develop a 10-year mortality risk score using penalized logistic regression models. We assessed model discriminative performance using the area under the receiver operating characteristic curve (AUC), calibration performance using the Hosmer-Lemeshow test, and overall model performance using Nagelkerke R² (NR).

RESULTS

Compared to a model including only age, comorbidity, and cancer stage and treatment variables, adding HRQOL variables improved discrimination (AUC 0.742 from 0.715) and overall performance (NR 0.221 from 0.190) with good calibration (p=0.96 from HL test).

CONCLUSION

In a cohort of older women with early-stage breast cancer, HRQOL measures predict 10-year mortality independently of traditional breast cancer prognostic variables. These findings suggest that interventions aimed at improving physical function, mental health, and social support might improve both HRQOL and survival.

DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network

Background

Medical practitioners use survival models to explore and understand the relationships between patients’ covariates (e.g. clinical and genetic features) and the effectiveness of various treatment options. Standard survival models like the linear Cox proportional hazards model require extensive feature engineering or prior medical knowledge to model treatment interaction at an individual level. While nonlinear survival methods, such as neural networks and survival forests, can inherently model these high-level interaction terms, they have yet to be shown as effective treatment recommender systems.

Methods

We introduce DeepSurv, a Cox proportional hazards deep neural network and state-of-the-art survival method for modeling interactions between a patient’s covariates and treatment effectiveness in order to provide personalized treatment recommendations.

Results

We perform a number of experiments training DeepSurv on simulated and real survival data. We demonstrate that DeepSurv performs as well as or better than other state-of-the-art survival models and validate that DeepSurv successfully models increasingly complex relationships between a patient’s covariates and their risk of failure. We then show how DeepSurv models the relationship between a patient’s features and effectiveness of different treatment options to show how DeepSurv can be used to provide individual treatment recommendations. Finally, we train DeepSurv on real clinical studies to demonstrate how it’s personalized treatment recommendations would increase the survival time of a set of patients.

Conclusions

The predictive and modeling capabilities of DeepSurv will enable medical researchers to use deep neural networks as a tool in their exploration, understanding, and prediction of the effects of a patient’s characteristics on their risk of failure.

Identification of a novel gene pairs signature in the prognosis of gastric cancer

Current prognostic signatures need to be improved in identifying high-risk patients of gastric cancer (GC). Thus, we aimed to develop a reliable prognostic signature that could assess the prognosis risk in GC patients. Two microarray datasets of GSE662254 (n = 300, training set) and GSE15459 (n = 192, test set) were included into analysis. Prognostic genes were screened to construct prognosis-related gene pairs (PRGPs). Then, a penalized Cox proportional hazards regression model identified seven PRGPs, which constructed a prognostic signature and divided patients into high- and low-risk groups according to the signature score. High-risk patients showed a poorer prognosis than low-risk patients in both the training set (hazard ratios [HR]: 6.086, 95% confidence interval [CI]: 4.341-8.533) and test set (1.773 [1.107-2.840]). The PRGPs signature also achieved a higher predictive accuracy (concordance index [C-index]: 0.872, 95% CI: 0.846-0.897) than two existing molecular signatures (0.706 [0.667-0.744] for a 11-gene signature and 0.684 [0.642-0.726] for a 24-lncRNA signature) and TNM stage (0.764 [0.715-0.814]). In conclusion, our study identified a novel gene pairs signature in the prognosis of GC.

Consensus of gene expression phenotypes and prognostic risk predictors in primary lung adenocarcinoma

Transcriptional profiling of lung adenocarcinomas has identified numerous gene expression phenotype (GEP) and risk prediction (RP) signatures associated with patient outcome. However, classification agreement between signatures, underlying transcriptional programs, and independent signature validation are less studied. We classified 2395 transcriptional adenocarcinoma profiles, assembled from 17 public cohorts, using 11 GEP and seven RP signatures, finding that 16 signatures were associated with patient survival in the total cohort and in multiple individual cohorts. For significant signatures, total cohort hazard ratios were ~2 in univariate analyses (mean=1.95, range=1.4-2.6). Strong classification agreement between signatures was observed, especially for predicted low-risk patients by adenocarcinoma-derived signatures. Expression of proliferation-related genes correlated strongly with GEP subtype classifications and RP scores, driving the gene signature association with prognosis. A three-group consensus definition of samples across 10 GEP classifiers demonstrated aggregation of samples with specific smoking patterns, gender, and EGFR/KRAS mutations, while survival differences were only significant when patients were divided into low- or high-risk. In summary, our study demonstrates a consensus between GEPs and RPs in lung adenocarcinoma through a common underlying transcriptional program. This consensus generalizes reported problems with current signatures in a clinical context, stressing development of new adenocarcinoma-specific single sample predictors for clinical use.

Distinct expression of CDCA3, CDCA5, and CDCA8 leads to shorter relapse free survival in breast cancer patient

Breast cancer is a dangerous disease that results in high mortality rates for cancer patients. Many methods have been developed for the treatment and prevention of this disease. Determining the expression patterns of certain target genes in specific subtypes of breast cancer is important for developing new therapies for breast cancer. In the present study, we performed a holistic approach to screening the mRNA expression of six members of the cell division cycle-associated gene family (CDCA) with a focus on breast cancer using the Oncomine and The Cancer Cell Line Encyclopedia (CCLE) databases. Furthermore, Gene Expression-Based Outcome for Breast Cancer Online (GOBO) was also used to deeply mine the expression of each CDCA gene in clinical breast cancer tissue and breast cancer cell lines. Finally, the mRNA expression of the CDCA genes as related to breast cancer patient survival were analyzed using a Kaplan-Meier plot. CDCA3, CDCA5, and CDCA8 mRNA expression levels were significantly higher than the control sample in both clinical tumor sample and cancer cell lines. These highly expressed genes in the tumors of breast cancer patients dramatically reduced patient survival. The interaction network of CDCA3, CDCA5, and CDCA8 with their co-expressed genes also revealed that CDCA3 expression was highly correlated with cell cycle related genes such as CCNB2, CDC20, CDKN3, and CCNB1. CDCA5 expression was correlated with BUB1 and TRIP13, while CDCA8 expression was correlated with BUB1 and CCNB1. Altogether, these findings suggested CDCA3, CDCA5, and CDCA8 could have a high potency as targeted breast cancer therapies.

Integrating Clinical and Multiple Omics Data for Prognostic Assessment across Human Cancers

Multiple omic profiles have been generated for many cancer types; however, comprehensive assessment of their prognostic values across cancers is limited. We conducted a pan-cancer prognostic assessment and presented a multi-omic kernel machine learning method to systematically quantify the prognostic values of high-throughput genomic, epigenomic, and transcriptomic profiles individually, integratively, and in combination with clinical factors for 3,382 samples across 14 cancer types. We found that the prognostic performance varied substantially across cancer types. mRNA and miRNA expression profile frequently performed the best, followed by DNA methylation profile. Germline susceptibility variants displayed low prognostic performance consistently across cancer types. The integration of omic profiles with clinical variables can lead to substantially improved prognostic performance over the use of clinical variables alone in half of cancer types examined. Moreover, we showed that the kernel machine learning method consistently outperformed existing prognostic signatures, suggesting that including a large number of omic biomarkers may provide substantial improvement in prognostic assessment. Our study provides a comprehensive portrait of omic architecture for tumor prognosis across cancers, and highlights the prognostic value of genome-wide omic biomarker aggregation, which may facilitate refined prognostic assessment in the era of precision oncology.

Development and Validation of an Individualized Immune Prognostic Signature in Early-Stage Nonsquamous Non-Small Cell Lung Cancer

Importance

The prevalence of early-stage non-small cell lung cancer (NSCLC) is expected to increase with recent implementation of annual screening programs. Reliable prognostic biomarkers are needed to identify patients at a high risk for recurrence to guide adjuvant therapy.

Objective

To develop a robust, individualized immune signature that can estimate prognosis in patients with early-stage nonsquamous NSCLC.

Design, Setting, and Participants

This retrospective study analyzed the gene expression profiles of frozen tumor tissue samples from 19 public NSCLC cohorts, including 18 microarray data sets and 1 RNA-Seq data set for The Cancer Genome Atlas (TCGA) lung adenocarcinoma cohort. Only patients with nonsquamous NSCLC with clinical annotation were included. Samples were from 2414 patients with nonsquamous NSCLC, divided into a meta-training cohort (729 patients), meta-testing cohort (716 patients), and 3 independent validation cohorts (439, 323, and 207 patients). All patients underwent surgery with a negative surgical margin, received no adjuvant or neoadjuvant therapy, and had publicly available gene expression data and survival information. Data were collected from July 22 through September 8, 2016.

Main Outcomes and Measures

Overall survival.

Results

Of 2414 patients (1205 men [50%], 1111 women [46%], and 98 of unknown sex [4%]; median age [range], 64 [15-90] years), a prognostic immune signature of 25 gene pairs consisting of 40 unique genes was constructed using the meta-training data set. In the meta-testing and validation cohorts, the immune signature significantly stratified patients into high- vs low-risk groups in terms of overall survival across and within subpopulations with stage I, IA, IB, or II disease and remained as an independent prognostic factor in multivariate analyses (hazard ratio range, 1.72 [95% CI, 1.26-2.33; P < .001] to 2.36 [95% CI, 1.47-3.79; P < .001]) after adjusting for clinical and pathologic factors. Several biological processes, including chemotaxis, were enriched among genes in the immune signature. The percentage of neutrophil infiltration (5.6% vs 1.8%) and necrosis (4.6% vs 1.5%) was significantly higher in the high-risk immune group compared with the low-risk groups in TCGA data set (P < .003). The immune signature achieved a higher accuracy (mean concordance index [C-index], 0.64) than 2 commercialized multigene signatures (mean C-index, 0.53 and 0.61) for estimation of survival in comparable validation cohorts. When integrated with clinical characteristics such as age and stage, the composite clinical and immune signature showed improved prognostic accuracy in all validation data sets relative to molecular signatures alone (mean C-index, 0.70 vs 0.63) and another commercialized clinical-molecular signature (mean C-index, 0.68 vs 0.65).

Conclusions and Relevance

The proposed clinical-immune signature is a promising biomarker for estimating overall survival in nonsquamous NSCLC, including early-stage disease. Prospective studies are needed to test the clinical utility of the biomarker in individualized management of nonsquamous NSCLC.

Decentralized Learning Framework of Meta-Survival Analysis for Developing Robust Prognostic Signatures

PURPOSE

A significant hurdle in developing reliable gene expression-based prognostic models has been the limited sample size, which can cause overfitting and false discovery. Combining data from multiple studies can enhance statistical power and reduce spurious findings, but how to address the biologic heterogeneity across different datasets remains a major challenge. Better meta-survival analysis approaches are needed.

MATERIAL AND METHODS

We presented a decentralized learning framework for meta-survival analysis without the need for data aggregation. Our method consisted of a series of proposals that together alleviated the influence of data heterogeneity and improved the performance of survival prediction. First, we transformed the gene expression profile of every sample into normalized percentile ranks to obtain platform-agnostic features. Second, we used Stouffer's meta-z approach in combination with Harrell's concordance index to prioritize and select genes to be included in the model. Third, we used survival discordance as a scale-independent model loss function. Instead of generating a merged dataset and training the model therein, we avoided comparing patients across datasets and individually evaluated the loss function on each dataset. Finally, we optimized the model by minimizing the joint loss function.

RESULTS

Through comprehensive evaluation on 31 public microarray datasets containing 6,724 samples of several cancer types, we demonstrated that the proposed method has outperformed (1) single prognostic genes identified using conventional meta-analysis, (2) multigene signatures trained on single datasets, (3) multigene signatures trained on merged datasets as well as by other existing meta-analysis methods, and (4) clinically applicable, established multigene signatures.

CONCLUSION

The decentralized learning approach can be used to effectively perform meta-analysis of gene expression data and to develop robust multigene prognostic signatures.

High Expression of FGD3, a Putative Regulator of Cell Morphology and Motility, Is Prognostic of Favorable Outcome in Multiple Cancers

Purpose

Identification of single-gene biomarkers that are prognostic of outcome can shed new insights on the molecular mechanisms that drive breast cancer and other cancers.

Methods

Exploratory analysis of 20,464 single-gene messenger RNAs (mRNAs) in the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) discovery cohort indicates that low expression of FGD3 mRNA is prognostic for poor outcome. Prognostic significance of faciogenital dysplasia 3 (FGD3), SUSD3, and other single-gene proliferation markers was evaluated in breast cancer and The Cancer Genome Atlas (TCGA) cohorts.

Results

A meta-analysis of Cox regression of FGD3 mRNA as a continuous variable for overall survival of estrogen receptor (ER)–positive samples in METABRIC discovery, METABRIC validation, TCGA breast cancer, and Combination Chemotherapy in Treating Women With Breast Cancer (E2197) cohorts resulted in a combined hazard ratio (HR) of 0.69 (95% CI, 0.63 to 0.75), indicating better outcome with high expression. In the ER-negative samples, the combined meta-analysis HR was 0.72 (95% CI, 0.63 to 0.82), suggesting that FGD3 is prognostic regardless of ER status. The potential of FGD3 as a biomarker for freedom from recurrence was evaluated in the Breast International Group 1-98 (BIG 1-98; Letrozole or Tamoxifen in Treating Postmenopausal Women With Breast Cancer) study (HR, 0.85; 95% CI, 0.76 to 0.93) for breast cancer–free interval. In the Hungarian Academy of Science (HAS) breast cancer cohort, splitting on the median had an HR of 0.49 (95% CI, 0.42 to 0.58) for recurrence-free survival. A comparison of the Stouffer P value in five ER-positive cohorts showed that FGD3 (P = 3.8^E-14) outperformed MKI67 (P = 1.06^E-8) and AURKA (P = 2.61^E-5). A comparison of the Stouffer P value in four ER-negative cohorts showed that FGD3 (P = 3.88^E-5) outperformed MKI67 (P = .477) and AURKA (P = .820).

Conclusion

FGD3 was previously shown to inhibit cell migration. FGD3 mRNA is regulated by ESR1 and is associated with favorable outcome in six distinct breast cancer cohorts and four TCGA cancer cohorts. This suggests that FGD3 is an important clinical biomarker.