Author(s): Felipe Vaca-Paniagua aff1 aff2 aff3 , Javier Oliver aff1 aff4 , Andre Nogueira da Costa aff1 aff5 , Philippe Merle aff6 aff7 , James McKay aff1 , Zdenko Herceg aff1 , Reetta Holmila [*] aff1
cfDNA; early cancer detection; hepatocellular carcinoma; methylation; next-generation sequencing
With the recent development of high-throughput sequencing technologies there has been a growing interest in circulating cell-free DNA (cfDNA), which consists of small double-stranded fragments of DNA found in plasma in variable amounts between different individuals [1 ]. Of particular interest is a fraction of cfDNA that appears to originate from tumor cells (circulating tumor DNA, ctDNA), and therefore has the potential to be used in the context of noninvasive biomarkers that would support the development of early detection, early diagnosis and prognostic tools [2-12 ]. The release of ctDNA in the bloodstream is thought to be related to apoptotic, necrotic and secretion events from tumor cells [13 ] with the abundance of cfDNA, as well as the fraction of ctDNA, increasing with the severity of the disease [14,15 ]. For example, Newman et al . [15 ] found that the levels of ctDNA highly correlated with tumor volume in non-small-cell lung cancer and the fractions of the ctDNA ranged from 0.02 to 3.2%. Nevertheless, as ctDNA may only represent a very small fraction of cfDNA, there are technical challenges that limit the sensitivity of ctDNA detection [16 ].
Massive parallel sequencing comprises a new set of technologies with higher sensitivity that have potential to underpin the studies involving ctDNA in clinical and molecular epidemiology settings. Several studies have already shown that these technologies can be applied to the detection of tumor somatic mutations in ctDNA and suggested that they could be gradually translated to the clinical setting [15-17 ]. However, with the exception of three recent studies that focused on analyzing plasma methylation by massively parallel sequencing and one using genome-wide amplification of bisulfite-treated DNA [18-21 ], studies aimed at the detection of methylation in ctDNA have used methylation-specific PCR or similar methods [21-25 ], which typically provide information on the presence of a restricted number of the methylated sites. Therefore, although current evidence suggests that DNA methylation changes in plasma samples have potential for diagnostic and monitoring in various cancers including hepatocellular carcinoma (HCC), the application of deep-sequencing approaches for methylation analysis of cfDNA is still scarce.
In the present study, we applied massively parallel semiconductor sequencing to analyze the methylation profiles of specific targets in plasma cfDNA obtained from HCC patients and controls. The aim of this work was to conduct a proof-of-principle study demonstrating the applicability of massively parallel semiconductor sequencing as a noninvasive, cost-effective and time-efficient approach to identify, develop and validate novel biomarkers potentially translatable into a clinical setting. This approach may support the development of strategies for early detection, diagnosis and prognosis of cancer pathologies.
Materials & methods
Patient characteristics, samples, DNA extraction & bisulfite treatment
Plasma and buffy coat specimens were collected and processed as previously described [26 ]. Blood specimens were obtained from controls...