Enhanced evaluation of RNA expression in Esophageal Cancer tissues: Combining Microarray and RNA-Sequence approaches

Esophageal1 cancer (EC) is a serious global health challenge, associated with high mortality rates and limited advances in early detection. The two predominant malignant subtypes, esophageal1 squamous cell carcinoma (ESCC) and esophageal1 adenocarcinoma (EAC), arise through distinct biological and epidemiological pathways, yet both are associated with poor prognoses. Barrett’s esophagus1 (BE), a metaplastic transformation of the normal oesophageal epithelium driven primarily by chronic gastroesophageal reflux, is a recognised precursor lesion for EAC. Together, BE, EAC, and ESCC represent a complex disease spectrum characterised by extensive molecular and transcriptomic dysregulation. Many transcriptomic studies have investigated these conditions, but is often limited by small sample sizes, different methodologies and population-specific biases. This leads to limited reproducibility and biological interpretability, especially in areas like Africa where genomic research is still lacking. The objective of this study was to conduct a thorough integrative analysis of publicly available transcriptome data to discern molecular pathways linked to BE, EAC, and ESCC. To do this, microarray and bulk RNA-sequencing (RNA-seq) datasets were chosen from many independent studies and evaluated utilising, platform-specific techniques. For the microarray component, Affymetrix-based datasets were harmonised through rigorous preprocessing and subjected to a rank-based meta-analysis using a modified RankProd framework to accommodate large-scale data integration. Ranked gene lists were subsequently analysed using SetRank, a pathway enrichment approach designed to reduce bias arising from overlapping gene sets, with additional subsampling applied to identify biologically stable and reproducible pathways. In parallel, RNA-seq data were processed using the nf-core/rnaseq pipeline to ensure standardised and reproducible preprocessing, alignment, and quantification. Batch effects deriving from study-specific variation were handled using ComBat-seq prior to differential expression analysis with DESeq2. Gene set enrichment analysis was then performed using FGSEA against curated pathway databases to uncover significantly enriched biological processes across subtypes. Integration of results across platforms revealed both shared and subtype-specific transcriptomic signatures. BE was characterised by enrichment of immune signalling, extracellular matrix organisation, metabolic processes, and developmental pathways, reflecting its role as an adaptive and inflammatory precursor state. EAC showed enrichment of receptor tyrosine kinase signalling, cell cycle control, and metabolic reprogramming pathways, consistent with oncogenic transformation and proliferation. ESCC showed clear dysregulation of pathways linked to chronic inflammation, immunological response, and signal transduction, which fits with its cause of toxins and injuries. The integrative framework strengthened confidence in pathways recognised across platforms, while also emphasising constraints, such as data heterogeneity and the restricted availability of African transcriptome datasets. Overall, this study reveals that integrated transcriptome meta-analysis combining both microarray and RNA-seq data provides a comprehensive technique for revealing molecular pathways across the oesophageal illness range. These findings expand understanding of EC biology and identify biologically relevant pathways that may inform future biomarker development and treatment efforts.

Grantor dc:publisher

Stellenbosch : Stellenbosch University

Year dc:date.issued

2026

Author dc:creator

• Otto, Nicola Miriam Elizabeth

Advisors dc:contributor.advisor

• Tromp, Gerardus
• Simba, Hannah

Language dc:language.iso

en

Repository record dc:identifier.uri

https://scholar.sun.ac.za/handle/10019.1/135967

OAI identifier oai:identifier

oai:scholar.sun.ac.za:10019.1/135967