Early and less tedious cancer diagnosis can prevent future complications and improve patient outcome, however past studies have revealed that the markers of most cancers are as unique as snowflakes. An enticing solution for this heterogeneity would be to screen for any type of cancer using one detection system to ‘rule them all’. Differential DNA methylation is very common in many cancers and has been utilized as a biomarker in different cancer types.
Armed with this vital evidence, researchers at King’s College in the UK and their collaborators decided to identify the whole blood DNA methylation commonality across cancer types including breast, cervical, colon, endometrial, thyroid, melanoma, ovarian, and pancreatic cancers with the intention of finding a pan-cancer biomarker. To achieve their goal, the team analyzed genome-wide DNA methylation profile of 41 female cancer discordant monozygotic twin-pairs with diagnosis at different body sites. The lab of Jordana Bell used the 450K array to analyze whole blood DNA samples. Additionally, the differential DNA methylation data was integrated with differential gene expression in independent control samples.
Here’s what they found:
- No significant global differences in whole blood DNA methylation profile across cancers.
- A novel pan-cancer differentially methylated region (DMR) 70kb upstream of the tumor suppressor gene, SASHI and three other putative pan-cancer biomarker sites at COL11A2, AXYL, and LINC0000340, with consistent replicability in independent samples.
- DMRs upstream of SASHI, COL11A2, and LINC0000340 were present in samples up to 5 years prior to cancer diagnosis, which hints at the possibility of a pan-cancer surveillance system.
- Gene expression analysis from lymphoblastoid cell lines (LCL), skin, and adipose tissues found nominally significant correlation with DNA methylation with LCL gene expression of the closest gene, prolactin (PRL) and at with skin tissue expression of AXL receptor tyrosine kinase gene.
- 500 top-ranked pan-cancer differentially methylated positions (DMPs) were assessed for two risk factors, age and smoking, and no significant enrichment was found.
- No significant association between cancer status and first five principal components were found, although a significant correlation with CD8+ T cells, CD4+ T cells, natural killer cells, granulocytes, monocytes, batch, and array was detected.
- There was a significant enrichment for differential methylation at enhancers.
Given the heterogenous nature of the samples and the relatively low statistical power of the research, more work needs to be done on the putative biomarkers before they are ready for the clinic. Pending further verification and optimization, the identified putative pan-cancer biomarkers hold great promises for the future of cancer screening and surveillance.
To find out more about this pan-cancer masterpiece, jump ahead to Clinical Epigenetics, January 2016.