Prognostication and prediction using gene expression profiling in oesophageal cancer

Development and validation of a novel artificial intelligence algorithm for precise prediction the postoperative prognosis of esophageal squamous cell carcinoma

BACKGROUND

Esophageal squamous cell carcinoma (ESCC) is a highly aggressive malignancy, and current postoperative prognostic assessment methods remain unsatisfactory, underlining the urgent to develop a reliable approach for precision medicine. Given the similarities with gametogenesis, cancer/testis genes (CTGs) are acknowledged for regulation unrestrained multiplication and immune microenvironment during oncogenic processes. These processes are associated with advanced disease and poorer prognosis, indicating that CTGs could serve as ideal prognostic biomarkers in ESCC. The purpose of this study is to develop a novel clinically prognostic prediction system to facilitate the individualized postoperative care.

METHODS

We conducted LASSO regression analysis of protein-coding CTGs and clinical characteristics from 119 pathologically confirmed ESCC patients to recognize powerful predictive variables. We employed nine supervised machine learning classifiers and integrated best predictive machine learning classifiers by weighted voting method to construct an ensemble model called PPMESCC. Additionally, functional assay was conducted to examine the potential effect of top-ranking CTG HENMT1 in ESCC.

RESULTS

LASSO regression identified five CTGs and TNM stage as optimized prognostic features. Six machine learning classifiers were integrated to construct an ensemble model, PPMESCC, which exhibited outstanding performance in ESCC prediction. The AUC for PPMESCC was 0.9828 (95% confidence interval: 0.9608 to 0.9926), with an accuracy of 98.32% (95% CI: 96.64-99.16%) in the discovery cohort and 0.9057 (95% CI: 0.8897 to 0.9583) of AUC with an accuracy of 90% (95% CI: 89.08-93.28%) in validation cohort. In addition, the top-ranking CTG HENMT1 encodes 2'-O-methyltransferase of piRNAs that was confirmed positively correlated with the proliferation capacity of ESCC cells. Then we systematically screen piRNAs associated with esophageal carcinoma based on GWAS, eQTL-piRNA, and i2OM databases, and successfully discovered 8 piRNAs potentially regulated by HENMT1.

CONCLUSION

The study highlights the clinical utility of PPMESCC algorithm in prognostic prediction that may facilitate to establish the personalized screening and management strategies for postoperative ESCC patients.

The prognostic value of a seven-lncRNA signature in patients with esophageal squamous cell carcinoma: a lncRNA expression analysis

Background

Long non-coding RNAs (lncRNAs) have been reported to be prognostic biomarkers in many types of cancer. We aimed to identify a lncRNA signature that can predict the prognosis in patients with esophageal squamous cell carcinoma (ESCC).

Methods

Using a custom microarray, we retrospectively analyzed lncRNA expression profiles in 141 samples of ESCC and 81 paired non-cancer specimens from Sun Yat-Sen University Cancer Center (Guangzhou, China), which were used as a training cohort to identify a signature associated with clinical outcomes. Then we conducted quantitative RT-PCR in another 103 samples of ESCC from the same cancer center as an independent cohort to verify the signature.

Results

Microarray analysis showed that there were 338 lncRNAs significantly differentially expressed between ESCC and non-cancer esophagus tissues in the training cohort. From these differentially expressed lncRNAs, we found 16 lncRNAs associated with overall survival (OS) of ESCC patients using Cox regression analysis. Then a 7-lncRNA signature for predicting survival was identified from the 16 lncRNAs, which classified ESCC patients into high-risk and low-risk groups. Patients with high-risk have shorter OS (HR: 3.555, 95% CI 2.195–5.757, p < 0.001) and disease-free survival (DFS) (HR: 2.537, 95% CI 1.646–3.909, p < 0.001) when compared with patients with low-risk in the training cohort. In the independent cohort, the 7 lncRNAs were detected by qRT-PCR and used to compute risk score for the patients. The result indicates that patients with high risk also have significantly worse OS (HR = 2.662, 95% CI 1.588–4.464, p < 0.001) and DFS (HR 2.389, 95% CI 1.447–3.946, p < 0.001). The univariate and multivariate Cox regression analyses indicate that the signature is an independent factor for predicting survival of patients with ESCC. Combination of the signature and TNM staging was more powerful in predicting OS than TNM staging alone in both the training (AUC: 0.772 vs 0.681, p = 0.002) and independent cohorts (AUC: 0.772 vs 0.660, p = 0.003).

Conclusions

The 7-lncRNA signature is a potential prognostic biomarker in patients with ESCC and may help in treatment decision when combined with the TNM staging system.

Integrated analysis of genomic and transcriptomic profiles identified a prognostic immunohistochemistry panel for esophageal squamous cell cancer

Background

The poor outcome of patients with esophageal squamous cell carcinoma (ESCC) highlights the importance of the identification of novel effective prognostic biomarkers. We aimed to identify a clinically applicable prognostic immunohistochemistry (IHC) panel for ESCC.

Methods

An integrated analysis was performed to screen and establish a prognostic panel using exome sequencing profile from 81 pairs of ESCC samples and RNA expression microarray data from 119 ESCC subjects. Two independent ESCC cohorts were recruited as training and validation groups to test the prognostic value.

Results

Three genes were selected, namely, ANO1, GAL, and MMP3, which were aberrantly expressed in ESCC tumor tissues (P < .001). Among them, ANO1 and MMP3 were reserved for the construction of the prognostic panel due to their significant association with the prognosis of ESCC patients (P = .015 and P < .001). Patients with both ANO1+ and MMP3+ had a poorer prognosis than that with ANO1−/MMP3+, ANO1+/MMP3−, or ANO1−/MMP3 − in both the training set and validation set (P < .001). Receiver operating characteristic analysis showed that the combination of IHC panel and eighth American Joint Commission on Cancer staging yielded a better prognostic predictive efficacy compared with the two indexes alone (P < .001, area under curve: 0.752). Finally, a nomogram was created by integrating the IHC markers and clinicopathological risk factors to predict prognosis with a C‐index of 0.695 (95% confidence interval: 0.657‐0.734).

Conclusion

Using an integrated multistage screening strategy, we identified and validated a valuable prognostic IHC panel for ESCC.

Elevated Pretreatment CEA and CA19-9 Levels are Related to Early Treatment Failure in Esophageal Adenocarcinoma

INTRODUCTION

Chemoradiotherapy and surgery are the basis of the potentially curative treatment for esophageal cancer. Approximately 1 in 5 patients, however, do not benefit from this intensive treatment due to early treatment failure. The aim of this study was to evaluate levels of carcinoembryonic antigen (CEA) and carbohydrate antigen (CA) 19-9 at diagnosis, in relation to survival and early treatment failure (disease recurrence or death within 1 year after surgery).

METHODS

Patients with esophageal adenocarcinoma scheduled for chemoradiotherapy followed by surgery between 1998 and 2014 were selected from a retrospectively collected database if both CEA and CA19-9 levels were measured before the start of treatment.

RESULTS

Pretreatment CEA and CA19-9 levels were known in 102 patients. Median overall survival differed (P<0.001) between patients with normal levels of both CEA and CA19-9 (n=59; 51 mo), patients with elevated CEA only (n=13; 43 mo), patients with elevated CA19-9 only (n=19; 24 mo), and those with elevated levels of both CEA and CA19-9 (n=11; 11 mo). Elevation of both CEA and CA19-9 was associated with early treatment failure (odds ratio: 10.4; 95% confidence interval: 2.4-45.5, P=0.002). Median time to tumor recurrence was 34 months in patients with normal CEA and CA19-9 levels, and 7 months in those with elevated levels of both (P=0.003).

CONCLUSIONS

Pretreatment elevated CEA and CA19-9 levels were significantly associated with early treatment failure and decreased overall survival in this esophageal adenocarcinoma patient cohort treated with curative intent. Until prospective validation, CEA and CA19-9 might play a role in identifying high-risk patients before the start of intensive locoregional therapy.

Meta-Analyses of Microarray Datasets Identifies ANO1 and FADD as Prognostic Markers of Head and Neck Cancer

The head and neck squamous cell carcinoma (HNSCC) transcriptome has been profiled extensively, nevertheless, identifying biomarkers that are clinically relevant and thereby with translational benefit, has been a major challenge. The objective of this study was to use a meta-analysis based approach to catalog candidate biomarkers with high potential for clinical application in HNSCC. Data from publically available microarray series (N = 20) profiled using Agilent (4X44K G4112F) and Affymetrix (HGU133A, U133A_2, U133Plus 2) platforms was downloaded and analyzed in a platform/chip-specific manner (GeneSpring software v12.5, Agilent, USA). Principal Component Analysis (PCA) and clustering analysis was carried out iteratively for segregating outliers; 140 normal and 277 tumor samples from 15 series were included in the final analysis. The analyses identified 181 differentially expressed, concordant and statistically significant genes; STRING analysis revealed interactions between 122 of them, with two major gene clusters connected by multiple nodes (MYC, FOS and HSPA4). Validation in the HNSCC-specific database (N = 528) in The Cancer Genome Atlas (TCGA) identified a panel (ECT2, ANO1, TP63, FADD, EXT1, NCBP2) that was altered in 30% of the samples. Validation in treatment naïve (Group I; N = 12) and post treatment (Group II; N = 12) patients identified 8 genes significantly associated with the disease (Area under curve>0.6). Correlation with recurrence/re-recurrence showed ANO1 had highest efficacy (sensitivity: 0.8, specificity: 0.6) to predict failure in Group I. UBE2V2, PLAC8, FADD and TTK showed high sensitivity (1.00) in Group I while UBE2V2 and CRYM were highly sensitive (>0.8) in predicting re-recurrence in Group II. Further, TCGA analysis showed that ANO1 and FADD, located at 11q13, were co-expressed at transcript level and significantly associated with overall and disease-free survival (p<0.05). The meta-analysis approach adopted in this study has identified candidate markers correlated with disease outcome in HNSCC; further validation in a larger cohort of patients will establish their clinical relevance.

How do patients with uveal melanoma experience and manage uncertainty? A qualitative study

OBJECTIVES

Cancer survivors experience uncertainty about the future, which can be distressing. A prognostication tool is available for uveal melanoma survivors, which can provide accurate estimates of life expectancy - a key source of uncertainty. Accurate prognostic information has not previously been available for healthy cancer survivors. The aims of this study were to identify how patients experience prognostic information and how it affects their experience of uncertainty.

METHODS

Semi-structured interviews were conducted with 25 healthy survivors of uveal melanoma 6-60 months after treatment (approximately 8-62 months after receiving prognostic information). Data were analysed qualitatively.

RESULTS

Patients did not feel that the prognostic information relieved uncertainty, which still overshadowed their lives. Different prognoses engendered different experiences of uncertainty. Those receiving poor life expectancy estimates reported uncertainties regarding the timing and form of metastases that they were likely to experience, but they also used uncertainty to justify feeling hopeful. Those receiving good prognoses were often unable wholly to accept these. Patients whose test results failed or were intermediate retained their original uncertainties. Patients managed their uncertainties by suppressing thoughts about them and by trusting in the care of clinicians and the health-care system.

CONCLUSIONS

Uncertainty in the context of uveal melanoma is a complex and multifaceted experience that is not easily resolved by prognostication. Additional approaches are needed to help patients with the uncertainty that persists despite prognostication.

Characterizing dynamic regulatory programs in mouse lung development and their potential association with tumourigenesis via miRNA-TF-mRNA circuits

Background

In dynamic biological processes, genes, transcription factors(TF) and microRNAs(miRNAs) play vital regulation roles. Many researchers have focused on the transcription factors or miRNAs in transcriptional or post transcriptional stage, respectively. However, the transcriptional regulation and post transcriptional regulation is not isolated in the whole dynamic biological processes, there are few reserchers who have tried to consider the network composed by genes, miRNAs and TFs in this dynamic biological processes, especially in the mouse lung development. Moreover, it is widely acknowledged that cancer is a kind of developmental disorders, and some of pathways involved in tissue development might be also implicated in causing cancer. Although it has been found that many genes differentially expressed during mouse lung development are also differentially expressed in lung cancer, very little work has been reported to elucidate the combinational regulatory programs of such kind of associations.

Results

In order to investigate the association of transcriptional and post-transcriptional regulating activities in the mouse lung development, we define the significant triple relations among miRNAs, TFs and mRNAs as circuits. From the lung development time course data GSE21053, we mine 142610 circuit candidates including 96 TFs, 129 miRNAs and 13403 genes. After removing genes with little variation along different time points, we finally find 64760 circuit candidates, containing 8299 genes, 50 TFs, and 118 miRNAs in total. Further analysis on the circuits shows that the circuits vary in different stages of the lung development and play different roles. By investigating the circuits in the context of lung specific genes, we identify out the regulatory combinations for lung specific genes, as well as for those lung non-specific genes. Moreover, we show that the lung non-specific genes involved circuits are functionally related to the lung development. Noticing that some tissue developmental systems may be involved in tumourigenesis, we also check the cancer genes involved circuits, trying to find out their regulatory program, which would be useful for the research of lung cancer.

Conclusions

The relevant transcriptional or post-transcriptional factors and their roles involved in the mouse lung development are both changed greatly in different stages. By investigating the cancer genes involved circuits, we can find miRNAs/TFs playing important roles in tumour progression. Therefore, the miRNA-TF-mRNA circuits can be used in wide translational biomedicine studies, and can provide potential drug targets towards the treatment of lung cancer.

Hypoxia-regulated gene expression and prognosis in loco-regional gastroesophageal cancer

Gastroesophageal cancers are heterogeneous diseases with a poor outcome. Prognostic and predictive factors are needed to improve patient survival. Hypoxia is an adverse prognostic factor and is associated with resistance to chemo- and radiotherapy in various cancers. However, knowledge on the impact of hypoxia in gastroesophageal cancer is limited. The aim of this study was to evaluate potential prognostic factors in terms of a subset of hypoxia-responsive genes and clinicopathological parameters in patients with gastroesophageal cancer.

MATERIAL AND METHODS

Ninety-five patients with loco-regional gastroesophageal cancer treated with curative intent were retrospectively analyzed. Based on formalin-fixed paraffin-embedded diagnostic biopsies gene expressions of 15 hypoxia-induced and pH-independent genes from a previously described hypoxia gene expression classifier was quantified. The prognostic impact was evaluated for overall survival (OS) and disease-specific survival (DSS). Uni- and multivariate Cox proportional hazards model was used to identify hypoxia-responsive gene expression and clinicopathological parameters as prognostic markers.

RESULTS

An unsupervised hierarchical clustering of hypoxia regulated genes showed two well-differentiated patient clusters: One cluster of tumors with high gene expression and another with low gene expression, indicating a more hypoxic genotype versus a less hypoxic genotype respectively. As the group of esophageal squamous cell carcinomas (ESCC) alone showed intra-group heterogeneity this group was ranked according to the gene expression of the 15 genes. The most hypoxic third showed a trend towards a poorer outcome in terms of OS [HR = 0.48 (CI 0.21-1.07), p = 0.07] and DSS [HR = 0.48 (CI 0.18-1.24), p = 0.13]. Treatment response was identified as an independent prognostic factor for DSS in the group of ESCC [HR = 0.21 (CI 0.05-0.95), p = 0.04].

CONCLUSION

Gene expression analysis of 15 hypoxia-responsive genes was identified as a promising prognostic marker in patients with ESCC. Further studies confirming these results in larger patient cohorts are needed.