Comparative meta-analysis of prognostic gene signatures for late-stage ovarian cancer

MARCKS contributes to stromal cancer-associated fibroblast activation and facilitates ovarian cancer metastasis

The Cancer Genome Atlas network has revealed that the 'mesenchymal' epithelial ovarian cancer (EOC) subtype represents the poorest outcome, indicating a crucial role of stromal cancer-associated fibroblasts (CAFs) in disease progression. The cooperative role of CAFs in EOC metastasis has long been recognized, but the mechanisms of stromal CAFs activation are still obscure. Therefore, we carried out an integrative analysis to identify the regulator genes that are responsible for CAFs activation in microdissected tumor stroma profiles. Here, we determined that myristoylated alanine-rich C-kinase substrate (MARCKS) was highly expressed in ovarian stroma, and was required for the differentiation and tumor promoting function of CAFs. Suppression of MARCKS resulted in the loss of CAF features, and diminished role of CAFs in supporting tumor cell growth in 3D organotypic cultures and in murine xenograft model. Mechanistically, we found that MARCKS maintained CAF activation through suppression of cellular senescence and activation of the AKT/Twist1 signaling. Moreover, high MARCKS expression was associated with poor patient survival in EOC. Collectively, our findings identify the potential of MARCKS inhibition as a novel stroma-oriented therapy in EOC.

The expression of miRNAs is associated with tumour genome instability and predicts the outcome of ovarian cancer patients treated with platinum agents

miRNAs, a class of short but stable noncoding RNA molecules, have been revealed to play important roles in the DNA damage response (DDR). However, their functions in cancer genome instability and the consequent clinical effect as the response to chemotherapy have not been fully elucidated. In this study, we utilized multidimensional TCGA data and the known miRNAs involved in DDR to identify a miRNA-regulatory network that responds to DNA damage. Additionally, based on the expression of ten miRNAs in this network, we developed a 10-miRNA-score that predicts defects in the homologous recombination (HR) pathway and genome instability in ovarian cancer. Importantly, consistent with the association between HR defects and improved response to chemotherapeutic agents, the 10-miRNA-score predicts the outcome of ovarian cancer patients treated with platinum agents, with a surprisingly better performance than the indexes of DNA damage. Therefore, our study demonstrates the implication of miRNA expression on cancer genome instability and provides an alternative method to identify DDR defects in patients who show the best effect with platinum drug treatment.

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.

SIRT1 deacetylates KLF4 to activate Claudin-5 transcription in ovarian cancer cells

Malignant cancers are distinguished from more benign forms of cancers by enhanced ability to disseminate. A number of factors aid the migration and invasion of malignant cancer cells. Epithelial-to-mesenchymal transition (EMT), which greatly facilitates the dissemination of cancer cells, is characterized by the loss of epithelial markers and the acquisition of mesenchymal markers thereby rendering the cells more migratory and invasive. We have previously shown that the class III lysine deacetylase SIRT1 plays a critical role curbing the metastasis of ovarian cancer cells partly by blocking EMT. Here we investigated the mechanism by which SIRT1 regulates the transcription of Claudin 5 (CLDN5), an epithelial marker gene, in ovarian cancer cells. SIRT1 activation or over-expression up-regulated CLDN5 expression while SIRT1 inhibition or depletion down-regulated CLDN5 expression. SIRT1 interacted with and deacetylated Kruppel-like factor 4 (KLF4), a known transcriptional activator for CLDN5. Deacetylation by SIRT1 promoted nuclear accumulation of KLF4 and enhanced the binding of KLF4 to the CLDN5 promoter in the nucleus. SIRT1-mediated up-regulation of CLDN5 was abrogated in the absence of KLF4. In accordance, KLF4 depletion by siRNA rendered ovarian cancer cells more migratory and invasive despite of SIRT1 activation or over-expression. In conclusion, our data suggest that SIRT1 activates CLDN5 transcription by deacetylating and potentiating KLF4.

The Prognostic 97 Chemoresponse Gene Signature in Ovarian Cancer

Patient diagnosis and care would be significantly improved by understanding the mechanisms underlying platinum and taxane resistance in ovarian cancer. Here, we aim to establish a gene signature that can identify molecular pathways/transcription factors involved in ovarian cancer progression, poor clinical outcome, and chemotherapy resistance. To validate the robustness of the gene signature, a meta-analysis approach was applied to 1,020 patients from 7 datasets. A 97-gene signature was identified as an independent predictor of patient survival in association with other clinicopathological factors in univariate [hazard ratio (HR): 3.0, 95% Confidence Interval (CI) 1.66–5.44, p = 2.7E-4] and multivariate [HR: 2.88, 95% CI 1.57–5.2, p = 0.001] analyses. Subset analyses demonstrated that the signature could predict patients who would attain complete or partial remission or no-response to first-line chemotherapy. Pathway analyses revealed that the signature was regulated by HIF1α and TP53 and included nine HIF1α-regulated genes, which were highly expressed in non-responders and partial remission patients than in complete remission patients. We present the 97-gene signature as an accurate prognostic predictor of overall survival and chemoresponse. Our signature also provides information on potential candidate target genes for future treatment efforts in ovarian cancer.

Next generation sequencing of pancreatic ductal adenocarcinoma: right or wrong?

Pancreatic ductal adenocarcinoma (PDAC) has the highest mortality rate of all epithelial malignancies and a paradoxically rising incidence rate. Clinical translation of next generation sequencing (NGS) of tumour and germline samples may ameliorate outcomes by identifying prognostic and predictive genomic and transcriptomic features in appreciable fractions of patients, facilitating enrolment in biomarker-matched trials. Areas covered: The literature on precision oncology is reviewed. It is found that outcomes may be improved across various malignancies, and it is suggested that current issues of adequate tissue acquisition, turnaround times, analytic expertise and clinical trial accessibility may lessen as experience accrues. Also reviewed are PDAC genomic and transcriptomic NGS studies, emphasizing discoveries of promising biomarkers, though these require validation, and the fraction of patients that will benefit from these outside of the research setting is currently unknown. Expert commentary: Clinical use of NGS with PDAC should be used in investigational contexts in centers with multidisciplinary expertise in cancer sequencing and pancreatic cancer management. Biomarker directed studies will improve our understanding of actionable genomic variation in PDAC, and improve outcomes for this challenging disease.

MKL1 links epigenetic activation of MMP2 to ovarian cancer cell migration and invasion

Responding to pro-metastatic cues such as low oxygen tension, cancer cells develop several different strategies to facilitate migration and invasion. During this process, expression levels of matrix metalloproteinases (MMPs) are up-regulated so that cancer cells can more easily enter or exit the circulation. In this report we show that message levels of the transcriptional modulator MKL1 were elevated in malignant forms of ovarian cancer tissues in humans when compared to more benign forms accompanying a similar change in MMP2 expression. MKL1 silencing blocked hypoxia-induced migration and invasion of ovarian cancer cells (SKOV-3) in vitro. Over-expression of MKL1 activated while MKL1 depletion repressed MMP2 transcription in SKOV-3 cells. MKL1 was recruited to the MMP2 promoter by NF-κB in response to hypoxia. Mechanistically, MKL1 recruited a histone methyltransferase, SET1, and a chromatin remodeling protein, BRG1, and coordinated their interaction to alter the chromatin structure surrounding the MMP2 promoter leading to transcriptional activation. Both BRG1 and SET1 were essential for hypoxia-induced MMP2 trans-activation. Finally, expression levels of SET1 and BRG1 were positively correlated with ovarian cancer malignancies in humans. Together, our data suggest that MKL1 promotes ovarian cancer cell migration and invasion by epigenetically activating MMP2 transcription.

Distance in cancer gene expression from stem cells predicts patient survival

The degree of histologic cellular differentiation of a cancer has been associated with prognosis but is subjectively assessed. We hypothesized that information about tumor differentiation of individual cancers could be derived objectively from cancer gene expression data, and would allow creation of a cancer phylogenetic framework that would correlate with clinical, histologic and molecular characteristics of the cancers, as well as predict prognosis. Here we utilized mRNA expression data from 4,413 patient samples with 7 diverse cancer histologies to explore the utility of ordering samples by their distance in gene expression from that of stem cells. A differentiation baseline was obtained by including expression data of human embryonic stem cells (hESC) and human mesenchymal stem cells (hMSC) for solid tumors, and of hESC and CD34+ cells for liquid tumors. We found that the correlation distance (the degree of similarity) between the gene expression profile of a tumor sample and that of stem cells orients cancers in a clinically coherent fashion. For all histologies analyzed (including carcinomas, sarcomas, and hematologic malignancies), patients with cancers with gene expression patterns most similar to that of stem cells had poorer overall survival. We also found that the genes in all undifferentiated cancers of diverse histologies that were most differentially expressed were associated with up-regulation of specific oncogenes and down-regulation of specific tumor suppressor genes. Thus, a stem cell-oriented phylogeny of cancers allows for the derivation of a novel cancer gene expression signature found in all undifferentiated forms of diverse cancer histologies, that is competitive in predicting overall survival in cancer patients compared to previously published prediction models, and is coherent in that gene expression was associated with up-regulation of specific oncogenes and down-regulation of specific tumor suppressor genes associated with regulation of the multicellular state.

Preoperative red cell distribution width and neutrophil-to-lymphocyte ratio predict survival in patients with epithelial ovarian cancer

Several parameters of preoperative complete blood count (CBC) and inflammation-associated blood cell markers derived from them have been reported to correlate with prognosis in patients with epithelial ovarian cancer (EOC), but their prognostic importance and optimal cutoffs are still needed be elucidated. Clinic/pathological parameters, 5-year follow-up data and preoperative CBC parameters were obtained retrospectively in 654 EOC patients underwent primary surgery at Mayo Clinic. Cutoffs for neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and monocyte-to-lymphocyte ratio (MLR) were optimized by receiver operating characteristic (ROC) curve. Prognostic significance for overall survival (OS) and recurrence free survival (RFS) were determined by Cox proportional hazards models and Kaplan-Meier method. Associations of RDW and NLR with clinic/pathological parameters were analyzed using non-parametric tests. RDW with cutoff 14.5 and NLR with cutoff 5.25 had independent prognostic significance for OS, while combined RDW and NLR scores stratified patients into low (RDW-low and NLR-low), intermediate (RDW-high or NLR-high) and high risk (RDW-high and NLR-high) groups, especially in patients with high-grade serous ovarian cancer (HGSOC). Moreover, high NLR was associated with poor RFS as well. Elevated RDW was strongly associated with age, whereas high NLR was strongly associated with stage, preoperative CA125 level and ascites at surgery.

Personalized dynamic prediction of death according to tumour progression and high-dimensional genetic factors: Meta-analysis with a joint model

Developing a personalized risk prediction model of death is fundamental for improving patient care and touches on the realm of personalized medicine. The increasing availability of genomic information and large-scale meta-analytic data sets for clinicians has motivated the extension of traditional survival prediction based on the Cox proportional hazards model. The aim of our paper is to develop a personalized risk prediction formula for death according to genetic factors and dynamic tumour progression status based on meta-analytic data. To this end, we extend the existing joint frailty-copula model to a model allowing for high-dimensional genetic factors. In addition, we propose a dynamic prediction formula to predict death given tumour progression events possibly occurring after treatment or surgery. For clinical use, we implement the computation software of the prediction formula in the joint.Cox R package. We also develop a tool to validate the performance of the prediction formula by assessing the prediction error. We illustrate the method with the meta-analysis of individual patient data on ovarian cancer patients.

MicroRNA-30c inhibits metastasis of ovarian cancer by targeting metastasis-associated gene 1

BACKGROUND

It is important to find reliable molecular markers or biological targets that associate with ovarian cancer (OC) metastasis for diagnosis and treatment. In this study, researchers investigated the regulated chain of microRNA-30c (miR-30c) and metastasis-associated gene 1 (MTA1) in OC tissues and cells.

MATERIALS AND METHODS

Expression of miR-30c and MTA1 was detected with quantitative real-time polymerase chain reaction and immunohistochemistry in 33 OC and matched adjacent tissues. MiR-30c mimics were synthetized and transfected into SKOV3 cells to target MTA1. The wound healing and transwell assays were detected to observe migration and invasion of transfected OC cells.

RESULTS

Compared with matching normal ovarian tissues, the MTA1 expression was upregulated and localized in the cytoplasm, and the expression of miR-30c was significantly reduced. The expression intensity of MTA1 was correlated with the Federation of Gynecology and Obstetrics stage, tumor grade, and metastasis of OC. Transfecting miR-30c mimics could significantly reduce the expression of MTA1 in SKOV3 cells and obviously inhibit the migration and invasion of SKOV3 cells.

CONCLUSION

MiR-30c and MTA1 abnormally expressed in OC, which may be related to metastasis of OC. In MiR-30c as a tumor suppressor gene, its expression in OC could lead to reduced expression of MTA1, which may be one of the mechanisms of metastasis of OC cells.

Prediction of Possible Biomarkers and Novel Pathways Conferring Risk to Post-Traumatic Stress Disorder

Post-traumatic stress disorder is one of the common mental ailments that is triggered by exposure to traumatic events. Till date, the molecular factors conferring risk to the development of PTSD is not well understood. In this study, we have conducted a meta-analysis followed by hierarchical clustering and functional enrichment, to uncover the potential molecular networks and critical genes which play an important role in PTSD. Two datasets of expression profiles from Peripheral Blood Mononuclear Cells from 62 control samples and 63 PTSD samples were included in our study. In PTSD samples of GSE860 dataset, we identified 26 genes informative when compared with Post-deploy PTSD condition and 58 genes informative when compared with Pre-deploy and Post-deploy PTSD of GSE63878 dataset. We conducted the meta-analysis using Fisher, roP, Stouffer, AW, SR, PR and RP methods in MetaDE package. Results from the rOP method of MetaDE package showed that among these genes, the following showed significant changes including, OR2B6, SOX21, MOBP, IL15, PTPRK, PPBPP2 and SEC14L5. Gene ontology analysis revealed enrichment of these significant PTSD-related genes for cell proliferation, DNA damage and repair (p-value ≤ 0.05). Furthermore, interaction network analysis was performed on these 7 significant genes. This analysis revealed highly connected functional interaction networks with two candidate genes, IL15 and SEC14L5 highly enriched in networks. Overall, from these results, we concluded that these genes can be recommended as some of the potential targets for PTSD.

NUAK1 (ARK5) Is Associated with Poor Prognosis in Ovarian Cancer

Background and objective

Nua kinase 1 (NUAK1) was identified in multigene signatures of survival and suboptimal debulking in high-grade serous ovarian cancer (HGSOC). This study investigates the individual clinical and biologic contributions of NUAK1 in HGSOC patients and cell lines.

Methods

Public transcript expression, clinical, and outcome data were used to interrogate the relationship between NUAK1 and clinicopathologic factors and patient outcomes including progression-free survival (PFS) and molecular subtypes using logistic and Cox modeling. Analysis of NUAK1 transcript expression was performed in primary tumors from 34 HGSOC patients with < or ≥2 years PFS. The impact of silencing NUAK1 by RNA interference (RNAi) on the migratory potential and chemosensitivity of SOC cells was assessed in vitro.

Results

Elevated NUAK1 transcript expression was associated with worse PFS (hazard ratio = 1.134), advanced stage (odds ratio, OR = 1.7), any residual disease (OR = 1.58), and mesenchymal disease subtype (OR = 7.79 ± 5.89). Elevated NUAK1 transcript expression was observed in HGSOC patients with < vs. ≥2 years PFS (p < 0.045). RNAi-mediated silencing of NUAK1 expression attenuated migration of OV90 and E3 HGSOC cells in vitro, but did not modulate sensitivity to cisplatin or paclitaxel.

Conclusion

Elevated NUAK1 was associated with poor survival as well as advanced stage, residual disease after cytoreductive surgery and mesenchymal molecular subtype. NUAK1 impacted migration, but not chemosensitivity, in vitro. Additional studies are needed to further develop the concept of NUAK1 as a clinically deployable biomarker and therapeutic target in HGSOC.

Genetic and molecular changes in ovarian cancer

Epithelial ovarian cancer represents the most lethal gynecological malignancy in the developed world, and can be divided into five main histological subtypes: high grade serous, endometrioid, clear cell, mucinous and low grade serous. These subtypes represent distinct disease entities, both clinically and at the molecular level. Molecular analysis has revealed significant genetic heterogeneity in ovarian cancer, particularly within the high grade serous subtype. As such, this subtype has been the focus of much research effort to date, revealing molecular subgroups at both the genomic and transcriptomic level that have clinical implications. However, stratification of ovarian cancer patients based on the underlying biology of their disease remains in its infancy. Here, we summarize the molecular changes that characterize the five main ovarian cancer subtypes, highlight potential opportunities for targeted therapeutic intervention and outline priorities for future research.

Machine Learning Meta-analysis of Large Metagenomic Datasets: Tools and Biological Insights

Shotgun metagenomic analysis of the human associated microbiome provides a rich set of microbial features for prediction and biomarker discovery in the context of human diseases and health conditions. However, the use of such high-resolution microbial features presents new challenges, and validated computational tools for learning tasks are lacking. Moreover, classification rules have scarcely been validated in independent studies, posing questions about the generality and generalization of disease-predictive models across cohorts. In this paper, we comprehensively assess approaches to metagenomics-based prediction tasks and for quantitative assessment of the strength of potential microbiome-phenotype associations. We develop a computational framework for prediction tasks using quantitative microbiome profiles, including species-level relative abundances and presence of strain-specific markers. A comprehensive meta-analysis, with particular emphasis on generalization across cohorts, was performed in a collection of 2424 publicly available metagenomic samples from eight large-scale studies. Cross-validation revealed good disease-prediction capabilities, which were in general improved by feature selection and use of strain-specific markers instead of species-level taxonomic abundance. In cross-study analysis, models transferred between studies were in some cases less accurate than models tested by within-study cross-validation. Interestingly, the addition of healthy (control) samples from other studies to training sets improved disease prediction capabilities. Some microbial species (most notably Streptococcus anginosus) seem to characterize general dysbiotic states of the microbiome rather than connections with a specific disease. Our results in modelling features of the “healthy” microbiome can be considered a first step toward defining general microbial dysbiosis. The software framework, microbiome profiles, and metadata for thousands of samples are publicly available at http://segatalab.cibio.unitn.it/tools/metaml.

The human microbiome–the entire set of microbial organisms associated with the human host–interacts closely with host immune and metabolic functions and is crucial for human health. Significant advances in the characterization of the microbiome associated with healthy and diseased individuals have been obtained through next-generation DNA sequencing technologies, which permit accurate estimation of microbial communities directly from uncultured human-associated samples (e.g., stool). In particular, shotgun metagenomics provide data at unprecedented species- and strain- levels of resolution. Several large-scale metagenomic disease-associated datasets are also becoming available, and disease-predictive models built on metagenomic signatures have been proposed. However, the generalization of resulting prediction models on different cohorts and diseases has not been validated. In this paper, we comprehensively assess approaches to metagenomics-based prediction tasks and for quantitative assessment of microbiome-phenotype associations. We consider 2424 samples from eight studies and six different diseases to assess the independent prediction accuracy of models built on shotgun metagenomic data and to compare strategies for practical use of the microbiome as a prediction tool.

The Doppelgänger Effect: Hidden Duplicates in Databases of Transcriptome Profiles

Whole-genome analysis of cancer specimens is commonplace, and investigators frequently share or re-use specimens in later studies. Duplicate expression profiles in public databases will impact re-analysis if left undetected, a so-called "doppelgänger" effect. We propose a method that should be routine practice to accurately match duplicate cancer transcriptomes when nucleotide-level sequence data are unavailable, even for samples profiled by different microarray technologies or by both microarray and RNA sequencing. We demonstrate the effectiveness of the method in databases containing dozens of datasets and thousands of ovarian, breast, bladder, and colorectal cancer microarray profiles and of matching microarray and RNA sequencing expression profiles from The Cancer Genome Atlas (TCGA). We identified probable duplicates among more than 50% of studies, originating in different continents, using different technologies, published years apart, and even within the TCGA itself. Finally, we provide the doppelgangR Bioconductor package for screening transcriptome databases for duplicates. Given the potential for unrecognized duplication to falsely inflate prediction accuracy and confidence in differential expression, doppelgänger-checking should be a part of standard procedure for combining multiple genomic datasets.

The Doppelgänger Effect: Hidden Duplicates in Databases of Transcriptome Profiles

Whole-genome analysis of cancer specimens is commonplace, and investigators frequently share or re-use specimens in later studies. Duplicate expression profiles in public databases will impact re-analysis if left undetected, a so-called "doppelgänger" effect. We propose a method that should be routine practice to accurately match duplicate cancer transcriptomes when nucleotide-level sequence data are unavailable, even for samples profiled by different microarray technologies or by both microarray and RNA sequencing. We demonstrate the effectiveness of the method in databases containing dozens of datasets and thousands of ovarian, breast, bladder, and colorectal cancer microarray profiles and of matching microarray and RNA sequencing expression profiles from The Cancer Genome Atlas (TCGA). We identified probable duplicates among more than 50% of studies, originating in different continents, using different technologies, published years apart, and even within the TCGA itself. Finally, we provide the doppelgangR Bioconductor package for screening transcriptome databases for duplicates. Given the potential for unrecognized duplication to falsely inflate prediction accuracy and confidence in differential expression, doppelgänger-checking should be a part of standard procedure for combining multiple genomic datasets.

Characteristics of 10-year survivors of high-grade serous ovarian carcinoma

OBJECTIVE

High-grade serous carcinoma (HGSC) generally presents at an advanced stage with poor long-term (LT) survival. Here we describe clinical features found in women surviving HGSC for ten or more years.

METHODS

A multi-center research consortium was established between five participating academic centers. Patient selection criteria included high-grade serous ovarian, fallopian tube, or peritoneal carcinoma with at least ten years of follow up. Non-serous, borderline tumors and low-grade serous subtypes were excluded.

RESULTS

The 203 identified LT ten-year survivors with HGSC were diagnosed at a median age of 57years (range 37-84years). The majority of patients had stage IIIC (72.4%) disease at presentation. Of those who underwent primary cytoreductive surgery, optimal cytoreduction was achieved in 143 (85.6%) patients. After a median follow up of 144months, 88 (46.8%) patients did not develop recurrent disease after initial treatment. Unexpected findings from this survey of LT survivors includes 14% of patients having had suboptimal cytoreduction, 11% of patients having an initial platinum free interval of <12months, and nearly 53% of patients having recurrent disease, yet still surviving more than ten years after diagnosis.

CONCLUSIONS

LT survivors of HGSC of the ovary generally have favorable clinical features including optimal surgical cytoreduction and primary platinum sensitive disease. The majority of patients will develop recurrent disease, however many remained disease free for more than 10years. Future work will compare the clinical features of this unusual cohort of LT survivors with the characteristics of HGSC patients having less favorable outcomes.

Characteristics of 10-year survivors of high-grade serous ovarian carcinoma

OBJECTIVE

High-grade serous carcinoma (HGSC) generally presents at an advanced stage with poor long-term (LT) survival. Here we describe clinical features found in women surviving HGSC for ten or more years.

METHODS

A multi-center research consortium was established between five participating academic centers. Patient selection criteria included high-grade serous ovarian, fallopian tube, or peritoneal carcinoma with at least ten years of follow up. Non-serous, borderline tumors and low-grade serous subtypes were excluded.

RESULTS

The 203 identified LT ten-year survivors with HGSC were diagnosed at a median age of 57years (range 37-84years). The majority of patients had stage IIIC (72.4%) disease at presentation. Of those who underwent primary cytoreductive surgery, optimal cytoreduction was achieved in 143 (85.6%) patients. After a median follow up of 144months, 88 (46.8%) patients did not develop recurrent disease after initial treatment. Unexpected findings from this survey of LT survivors includes 14% of patients having had suboptimal cytoreduction, 11% of patients having an initial platinum free interval of <12months, and nearly 53% of patients having recurrent disease, yet still surviving more than ten years after diagnosis.

CONCLUSIONS

LT survivors of HGSC of the ovary generally have favorable clinical features including optimal surgical cytoreduction and primary platinum sensitive disease. The majority of patients will develop recurrent disease, however many remained disease free for more than 10years. Future work will compare the clinical features of this unusual cohort of LT survivors with the characteristics of HGSC patients having less favorable outcomes.

Single Gene Prognostic Biomarkers in Ovarian Cancer: A Meta-Analysis

PURPOSE

To discover novel prognostic biomarkers in ovarian serous carcinomas.

METHODS

A meta-analysis of all single genes probes in the TCGA and HAS ovarian cohorts was performed to identify possible biomarkers using Cox regression as a continuous variable for overall survival. Genes were ranked by p-value using Stouffer's method and selected for statistical significance with a false discovery rate (FDR) <.05 using the Benjamini-Hochberg method.

RESULTS

Twelve genes with high mRNA expression were prognostic of poor outcome with an FDR <.05 (AXL, APC, RAB11FIP5, C19orf2, CYBRD1, PINK1, LRRN3, AQP1, DES, XRCC4, BCHE, and ASAP3). Twenty genes with low mRNA expression were prognostic of poor outcome with an FDR <.05 (LRIG1, SLC33A1, NUCB2, POLD3, ESR2, GOLPH3, XBP1, PAXIP1, CYB561, POLA2, CDH1, GMNN, SLC37A4, FAM174B, AGR2, SDR39U1, MAGT1, GJB1, SDF2L1, and C9orf82).

CONCLUSION

A meta-analysis of all single genes identified thirty-two candidate biomarkers for their possible role in ovarian serous carcinoma. These genes can provide insight into the drivers or regulators of ovarian cancer and should be evaluated in future studies. Genes with high expression indicating poor outcome are possible therapeutic targets with known antagonists or inhibitors. Additionally, the genes could be combined into a prognostic multi-gene signature and tested in future ovarian cohorts.

Integration of genomic, transcriptomic and proteomic data identifies two biologically distinct subtypes of invasive lobular breast cancer

Invasive lobular carcinoma (ILC) is the second most frequently occurring histological breast cancer subtype after invasive ductal carcinoma (IDC), accounting for around 10% of all breast cancers. The molecular processes that drive the development of ILC are still largely unknown. We have performed a comprehensive genomic, transcriptomic and proteomic analysis of a large ILC patient cohort and present here an integrated molecular portrait of ILC. Mutations in CDH1 and in the PI3K pathway are the most frequent molecular alterations in ILC. We identified two main subtypes of ILCs: (i) an immune related subtype with mRNA up-regulation of PD-L1, PD-1 and CTLA-4 and greater sensitivity to DNA-damaging agents in representative cell line models; (ii) a hormone related subtype, associated with Epithelial to Mesenchymal Transition (EMT), and gain of chromosomes 1q and 8q and loss of chromosome 11q. Using the somatic mutation rate and eIF4B protein level, we identified three groups with different clinical outcomes, including a group with extremely good prognosis. We provide a comprehensive overview of the molecular alterations driving ILC and have explored links with therapy response. This molecular characterization may help to tailor treatment of ILC through the application of specific targeted, chemo- and/or immune-therapies.

Meta-Analysis in Gene Expression Studies

This chapter introduces methods to synthesize experimental results from independent high-throughput genomic experiments, with a focus on adaptation of traditional methods from systematic review of clinical trials and epidemiological studies. First, it reviews methods for identifying, acquiring, and preparing individual patient data for meta-analysis. It then reviews methodology for synthesizing results across studies and assessing heterogeneity, first through outlining of methods and then through a step-by-step case study in identifying genes associated with survival in high-grade serous ovarian cancer.

Leveraging global gene expression patterns to predict expression of unmeasured genes

Background

Large collections of paraffin-embedded tissue represent a rich resource to test hypotheses based on gene expression patterns; however, measurement of genome-wide expression is cost-prohibitive on a large scale. Using the known expression correlation structure within a given disease type (in this case, high grade serous ovarian cancer; HGSC), we sought to identify reduced sets of directly measured (DM) genes which could accurately predict the expression of a maximized number of unmeasured genes.

Results

We developed a greedy gene set selection (GGS) algorithm which returns a DM set of user specified size based on a specific correlation threshold (|r_P|) and minimum number of DM genes that must be correlated to an unmeasured gene in order to infer the value of the unmeasured gene (redundancy). We evaluated GGS in the Cancer Genome Atlas (TCGA) HGSC data across 144 combinations of DM size, redundancy (1–3), and |r_P| (0.60, 0.65, 0.70). Across the parameter sweep, GGS allows on average 9 times more gene expression information to be captured compared to the DM set alone. GGS successfully augments prognostic HGSC gene sets; the addition of 20 GGS selected genes more than doubles the number of genes whose expression is predictable. Moreover, the expression prediction is highly accurate. After training regression models for the predictable gene set using 2/3 of the TCGA data, the average accuracy (ranked correlation of true and predicted values) in the 1/3 testing partition and four independent populations is above 0.65 and approaches 0.8 for conservative parameter sets. We observe similar accuracies in the TCGA HGSC RNA-sequencing data. Specifically, the prediction accuracy increases with increasing redundancy and increasing |r_P|.

Conclusions

GGS-selected genes, which maximize expression information about unmeasured genes, can be combined with candidate gene sets as a cost effective way to increase the amount of gene expression information obtained in large studies. This method can be applied to any organism, model system, disease, or tissue type for which whole genome gene expression data exists.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-2250-5) contains supplementary material, which is available to authorized users.

Stroma-associated master regulators of molecular subtypes predict patient prognosis in ovarian cancer

High-grade serous ovarian carcinoma (HGS-OvCa) has the lowest survival rate among all gynecologic cancers and is hallmarked by a high degree of heterogeneity. The Cancer Genome Atlas network has described a gene expression-based molecular classification of HGS-OvCa into Differentiated, Mesenchymal, Immunoreactive and Proliferative subtypes. However, the biological underpinnings and regulatory mechanisms underlying the distinct molecular subtypes are largely unknown. Here we showed that tumor-infiltrating stromal cells significantly contributed to the assignments of Mesenchymal and Immunoreactive clusters. Using reverse engineering and an unbiased interrogation of subtype regulatory networks, we identified the transcriptional modules containing master regulators that drive gene expression of Mesenchymal and Immunoreactive HGS-OvCa. Mesenchymal master regulators were associated with poor prognosis, while Immunoreactive master regulators positively correlated with overall survival. Meta-analysis of 749 HGS-OvCa expression profiles confirmed that master regulators as a prognostic signature were able to predict patient outcome. Our data unraveled master regulatory programs of HGS-OvCa subtypes with prognostic and potentially therapeutic relevance, and suggested that the unique transcriptional and clinical characteristics of ovarian Mesenchymal and Immunoreactive subtypes could be, at least partially, ascribed to tumor microenvironment.

E2F4 Program Is Predictive of Progression and Intravesical Immunotherapy Efficacy in Bladder Cancer

Bladder cancer is a common malignant disease, with non-muscle-invasive bladder cancer (NMIBC) representing the majority of tumors. This cancer subtype is typically treated by transurethral resection. In spite of treatment, up to 70% of patients show local recurrences. Intravesical BCG (Bacillus Calmette-Guerin) immunotherapy has been widely used to treat NMIBC, but it fails to suppress recurrence of bladder tumors in up to 40% of patients. Therefore, the development of prognostic markers is needed to predict the progression of bladder cancer and the efficacy of intravesical BCG treatment. This study demonstrates the effectiveness of an E2F4 signature for prognostic prediction of bladder cancer. E2F4 scores for each sample in a bladder cancer expression dataset were calculated by summarizing the relative expression levels of E2F4 target genes identified by ChIP-seq, and then the scores were used to stratify patients into good- and poor-outcome groups. The molecular signature was investigated in a single bladder cancer dataset and then its effectiveness was confirmed in two meta-bladder datasets consisting of specimens from multiple independent studies. These results were consistent in different datasets and demonstrate that the E2F4 score is predictive of clinical outcomes in bladder cancer, with patients whose tumors exhibit an E2F4 score >0 having significantly shorter survival times than those with an E2F4 score <0, in both non-muscle-invasive, and muscle-invasive bladder cancer. Furthermore, although intravesical BCG immunotherapy can significantly improve the clinical outcome of NMIBC patients with positive E2F4 scores (E2F4>0 group), it does not show significant treatment effect for those with negative scores (E2F4<0 group).

IMPLICATIONS

The E2F4 signature can be applied to predict the progression/recurrence and the responsiveness of patients to intravesical BCG immunotherapy in bladder cancer.

Characterization of twenty-five ovarian tumour cell lines that phenocopy primary tumours

Currently available human tumour cell line panels consist of a small number of lines in each lineage that generally fail to retain the phenotype of the original patient tumour. Here we develop a cell culture medium that enables us to routinely establish cell lines from diverse subtypes of human ovarian cancers with >95% efficiency. Importantly, the 25 new ovarian tumour cell lines described here retain the genomic landscape, histopathology and molecular features of the original tumours. Furthermore, the molecular profile and drug response of these cell lines correlate with distinct groups of primary tumours with different outcomes. Thus, tumour cell lines derived using this methodology represent a significantly improved platform to study human tumour pathophysiology and response to therapy.

Orchestrating high-throughput genomic analysis with Bioconductor

Bioconductor is an open-source, open-development software project for the analysis and comprehension of high-throughput data in genomics and molecular biology. The project aims to enable interdisciplinary research, collaboration and rapid development of scientific software. Based on the statistical programming language R, Bioconductor comprises 934 interoperable packages contributed by a large, diverse community of scientists. Packages cover a range of bioinformatic and statistical applications. They undergo formal initial review and continuous automated testing. We present an overview for prospective users and contributors.

A reproducible approach to high-throughput biological data acquisition and integration

Modern biological research requires rapid, complex, and reproducible integration of multiple experimental results generated both internally and externally (e.g., from public repositories). Although large systematic meta-analyses are among the most effective approaches both for clinical biomarker discovery and for computational inference of biomolecular mechanisms, identifying, acquiring, and integrating relevant experimental results from multiple sources for a given study can be time-consuming and error-prone. To enable efficient and reproducible integration of diverse experimental results, we developed a novel approach for standardized acquisition and analysis of high-throughput and heterogeneous biological data. This allowed, first, novel biomolecular network reconstruction in human prostate cancer, which correctly recovered and extended the NFκB signaling pathway. Next, we investigated host-microbiome interactions. In less than an hour of analysis time, the system retrieved data and integrated six germ-free murine intestinal gene expression datasets to identify the genes most influenced by the gut microbiota, which comprised a set of immune-response and carbohydrate metabolism processes. Finally, we constructed integrated functional interaction networks to compare connectivity of peptide secretion pathways in the model organisms Escherichia coli, Bacillus subtilis, and Pseudomonas aeruginosa.

Open access to large scale datasets is needed to translate knowledge of cancer heterogeneity into better patient outcomes

In this guest editorial, Andrew Beck discusses the importance of open access to big data for translating knowledge of cancer heterogeneity into better outcomes for cancer patients.

Patient-specific driver gene prediction and risk assessment through integrated network analysis of cancer omics profiles

Extensive and multi-dimensional data sets generated from recent cancer omics profiling projects have presented new challenges and opportunities for unraveling the complexity of cancer genome landscapes. In particular, distinguishing the unique complement of genes that drive tumorigenesis in each patient from a sea of passenger mutations is necessary for translating the full benefit of cancer genome sequencing into the clinic. We address this need by presenting a data integration framework (OncoIMPACT) to nominate patient-specific driver genes based on their phenotypic impact. Extensive in silico and in vitro validation helped establish OncoIMPACT's robustness, improved precision over competing approaches and verifiable patient and cell line specific predictions (2/2 and 6/7 true positives and negatives, respectively). In particular, we computationally predicted and experimentally validated the gene TRIM24 as a putative novel amplified driver in a melanoma patient. Applying OncoIMPACT to more than 1000 tumor samples, we generated patient-specific driver gene lists in five different cancer types to identify modes of synergistic action. We also provide the first demonstration that computationally derived driver mutation signatures can be overall superior to single gene and gene expression based signatures in enabling patient stratification and prognostication. Source code and executables for OncoIMPACT are freely available from http://sourceforge.net/projects/oncoimpact.

Optimized Prediction of Extreme Treatment Outcomes in Ovarian Cancer

Ovarian cancer is the fifth leading cause of death among female cancers. Front-line therapy for ovarian cancer is platinum-based chemotherapy. However, the response of patients is highly nonuniform. The TCGA database of serous ovarian carcinomas shows that ~10% of patients respond poorly to platinum-based chemotherapy, with tumors relapsing in seven months or less. Another 10% or so enjoy disease-free survival of three years or more. The objective of the present research is to identify a small number of highly predictive biomarkers that can distinguish between the two extreme responders and then extrapolate to all patients. This is achieved using the lone star algorithm that is specifically developed for biological applications. Using this algorithm, we are able to identify biomarker panels of 25 genes (of 12,000 genes) that can be used to classify patients into one of the three groups: super responders, medium responders, and nonresponders. We are also able to determine a discriminant function that can divide the entire patient population into two classes, such that one group has a clear survival advantage over the other. These biomarkers are developed using the TCGA Agilent platform data and cross-validated on the TCGA Affymetrix platform data, as well as entirely independent data from Tothill et al. The P-values on the training data are extremely small, sometimes below machine zero, while the P-values on cross-validation are well below the widely accepted threshold of 0.05.

OvMark: a user-friendly system for the identification of prognostic biomarkers in publically available ovarian cancer gene expression datasets

BACKGROUND

Ovarian cancer has the lowest survival rate of all gynaecologic cancers and is characterised by a lack of early symptoms and frequent late stage diagnosis. There is a paucity of robust molecular markers that are independent of and complementary to clinical parameters such as disease stage and tumour grade.

METHODS

We have developed a user-friendly, web-based system to evaluate the association of genes/miRNAs with outcome in ovarian cancer. The OvMark algorithm combines data from multiple microarray platforms (including probesets targeting miRNAs) and correlates them with clinical parameters (e.g. tumour grade, stage) and outcomes (disease free survival (DFS), overall survival). In total, OvMark combines 14 datasets from 7 different array platforms measuring the expression of ~17,000 genes and 341 miRNAs across 2,129 ovarian cancer samples.

RESULTS

To demonstrate the utility of the system we confirmed the prognostic ability of 14 genes and 2 miRNAs known to play a role in ovarian cancer. Of these genes, CXCL12 was the most significant predictor of DFS (HR = 1.42, p-value = 2.42x10-6). Surprisingly, those genes found to have the greatest correlation with outcome have not been heavily studied in ovarian cancer, or in some cases in any cancer. For instance, the three genes with the greatest association with survival are SNAI3, VWA3A and DNAH12.

CONCLUSIONS/IMPACT

OvMark is a powerful tool for examining putative gene/miRNA prognostic biomarkers in ovarian cancer (available at http://glados.ucd.ie/OvMark/index.html). The impact of this tool will be in the preliminary assessment of putative biomarkers in ovarian cancer, particularly for research groups with limited bioinformatics facilities.

Cross-study validation for the assessment of prediction algorithms

Motivation: Numerous competing algorithms for prediction in high-dimensional settings have been developed in the statistical and machine-learning literature. Learning algorithms and the prediction models they generate are typically evaluated on the basis of cross-validation error estimates in a few exemplary datasets. However, in most applications, the ultimate goal of prediction modeling is to provide accurate predictions for independent samples obtained in different settings. Cross-validation within exemplary datasets may not adequately reflect performance in the broader application context.

Methods: We develop and implement a systematic approach to ‘cross-study validation’, to replace or supplement conventional cross-validation when evaluating high-dimensional prediction models in independent datasets. We illustrate it via simulations and in a collection of eight estrogen-receptor positive breast cancer microarray gene-expression datasets, where the objective is predicting distant metastasis-free survival (DMFS). We computed the C-index for all pairwise combinations of training and validation datasets. We evaluate several alternatives for summarizing the pairwise validation statistics, and compare these to conventional cross-validation.

Results: Our data-driven simulations and our application to survival prediction with eight breast cancer microarray datasets, suggest that standard cross-validation produces inflated discrimination accuracy for all algorithms considered, when compared to cross-study validation. Furthermore, the ranking of learning algorithms differs, suggesting that algorithms performing best in cross-validation may be suboptimal when evaluated through independent validation.

Availability: The survHD: Survival in High Dimensions package (http://www.bitbucket.org/lwaldron/survhd) will be made available through Bioconductor.

Contact:levi.waldron@hunter.cuny.edu

Supplementary information:Supplementary data are available at Bioinformatics online.

The future of laboratory medicine - a 2014 perspective

Predicting the future is a difficult task. Not surprisingly, there are many examples and assumptions that have proved to be wrong. This review surveys the many predictions, beginning in 1887, about the future of laboratory medicine and its sub-specialties such as clinical chemistry and molecular pathology. It provides a commentary on the accuracy of the predictions and offers opinions on emerging technologies, economic factors and social developments that may play a role in shaping the future of laboratory medicine.

Genome-driven integrated classification of breast cancer validated in over 7,500 samples

BACKGROUND

IntClust is a classification of breast cancer comprising 10 subtypes based on molecular drivers identified through the integration of genomic and transcriptomic data from 1,000 breast tumors and validated in a further 1,000. We present a reliable method for subtyping breast tumors into the IntClust subtypes based on gene expression and demonstrate the clinical and biological validity of the IntClust classification.

RESULTS

We developed a gene expression-based approach for classifying breast tumors into the ten IntClust subtypes by using the ensemble profile of the index discovery dataset. We evaluate this approach in 983 independent samples for which the combined copy-number and gene expression IntClust classification was available. Only 24 samples are discordantly classified. Next, we compile a consolidated external dataset composed of a further 7,544 breast tumors. We use our approach to classify all samples into the IntClust subtypes. All ten subtypes are observable in most studies at comparable frequencies. The IntClust subtypes are significantly associated with relapse-free survival and recapitulate patterns of survival observed previously. In studies of neo-adjuvant chemotherapy, IntClust reveals distinct patterns of chemosensitivity. Finally, patterns of expression of genomic drivers reported by TCGA (The Cancer Genome Atlas) are better explained by IntClust as compared to the PAM50 classifier.

CONCLUSIONS

IntClust subtypes are reproducible in a large meta-analysis, show clinical validity and best capture variation in genomic drivers. IntClust is a driver-based breast cancer classification and is likely to become increasingly relevant as more targeted biological therapies become available.

Más-o-menos: a simple sign averaging method for discrimination in genomic data analysis

MOTIVATION

The successful translation of genomic signatures into clinical settings relies on good discrimination between patient subgroups. Many sophisticated algorithms have been proposed in the statistics and machine learning literature, but in practice simpler algorithms are often used. However, few simple algorithms have been formally described or systematically investigated.

RESULTS

We give a precise definition of a popular simple method we refer to as más-o-menos, which calculates prognostic scores for discrimination by summing standardized predictors, weighted by the signs of their marginal associations with the outcome. We study its behavior theoretically, in simulations and in an extensive analysis of 27 independent gene expression studies of bladder, breast and ovarian cancer, altogether totaling 3833 patients with survival outcomes. We find that despite its simplicity, más-o-menos can achieve good discrimination performance. It performs no worse, and sometimes better, than popular and much more CPU-intensive methods for discrimination, including lasso and ridge regression.

AVAILABILITY AND IMPLEMENTATION

Más-o-menos is implemented for survival analysis as an option in the survHD package, available from http://www.bitbucket.org/lwaldron/survhd and submitted to Bioconductor.

Risk prediction for late-stage ovarian cancer by meta-analysis of 1525 patient samples

BACKGROUND

Ovarian cancer causes more than 15000 deaths per year in the United States. The survival of patients is quite heterogeneous, and accurate prognostic tools would help with the clinical management of these patients.

METHODS

We developed and validated two gene expression signatures, the first for predicting survival in advanced-stage, serous ovarian cancer and the second for predicting debulking status. We integrated 13 publicly available datasets totaling 1525 subjects. We trained prediction models using a meta-analysis variation on the compound covariable method, tested models by a "leave-one-dataset-out" procedure, and validated models in additional independent datasets. Selected genes from the debulking signature were validated by immunohistochemistry and quantitative reverse-transcription polymerase chain reaction (qRT-PCR) in two further independent cohorts of 179 and 78 patients, respectively. All statistical tests were two-sided.

RESULTS

The survival signature stratified patients into high- and low-risk groups (hazard ratio = 2.19; 95% confidence interval [CI] = 1.84 to 2.61) statistically significantly better than the TCGA signature (P = .04). POSTN, CXCL14, FAP, NUAK1, PTCH1, and TGFBR2 were validated by qRT-PCR (P < .05) and POSTN, CXCL14, and phosphorylated Smad2/3 were validated by immunohistochemistry (P < .001) as independent predictors of debulking status. The sum of immunohistochemistry intensities for these three proteins provided a tool that classified 92.8% of samples correctly in high- and low-risk groups for suboptimal debulking (area under the curve = 0.89; 95% CI = 0.84 to 0.93).

CONCLUSIONS

Our survival signature provides the most accurate and validated prognostic model for early- and advanced-stage high-grade, serous ovarian cancer. The debulking signature accurately predicts the outcome of cytoreductive surgery, potentially allowing for stratification of patients for primary vs secondary cytoreduction.

Risk prediction for late-stage ovarian cancer by meta-analysis of 1525 patient samples

BACKGROUND

Ovarian cancer causes more than 15000 deaths per year in the United States. The survival of patients is quite heterogeneous, and accurate prognostic tools would help with the clinical management of these patients.

METHODS

We developed and validated two gene expression signatures, the first for predicting survival in advanced-stage, serous ovarian cancer and the second for predicting debulking status. We integrated 13 publicly available datasets totaling 1525 subjects. We trained prediction models using a meta-analysis variation on the compound covariable method, tested models by a "leave-one-dataset-out" procedure, and validated models in additional independent datasets. Selected genes from the debulking signature were validated by immunohistochemistry and quantitative reverse-transcription polymerase chain reaction (qRT-PCR) in two further independent cohorts of 179 and 78 patients, respectively. All statistical tests were two-sided.

RESULTS

The survival signature stratified patients into high- and low-risk groups (hazard ratio = 2.19; 95% confidence interval [CI] = 1.84 to 2.61) statistically significantly better than the TCGA signature (P = .04). POSTN, CXCL14, FAP, NUAK1, PTCH1, and TGFBR2 were validated by qRT-PCR (P < .05) and POSTN, CXCL14, and phosphorylated Smad2/3 were validated by immunohistochemistry (P < .001) as independent predictors of debulking status. The sum of immunohistochemistry intensities for these three proteins provided a tool that classified 92.8% of samples correctly in high- and low-risk groups for suboptimal debulking (area under the curve = 0.89; 95% CI = 0.84 to 0.93).

CONCLUSIONS

Our survival signature provides the most accurate and validated prognostic model for early- and advanced-stage high-grade, serous ovarian cancer. The debulking signature accurately predicts the outcome of cytoreductive surgery, potentially allowing for stratification of patients for primary vs secondary cytoreduction.