August 4, 2015

expert reaction to study reporting potential clinical biomarker associated with breast cancer risk

It is becoming increasingly clear that genes can be turned on and off by having different molecular tags added or removed; these epigenetic changes allow fine tuning of the DNA code. In a paper published in the journal Clinical Epigenetics, a group of scientists has studied one such epigenetic change in several hundred women in relation to breast cancer and reported that high levels of it were related to lower levels of breast cancer.

Prof. Shirley Hodgson, Professor of Cancer Genetics, St George’s, University of London, said:

“This is interesting preliminary data but there are some potential factors that could influence the results including the methods used for processing and storing blood samples, and differences in the demographics of the different cohorts, so the study should be interpreted with caution. However there does seem to be a small difference in methylation in cases and controls in three of the cohorts studied, and the fact that gene body methylation shows a difference whereas promoter methylation does not is an argument in favour of this being a real predictor of risk, but the numbers of cases and controls are not very large and the fact that one study does not show this effect is worrying. There could also be problems with methodology, as the authors themselves say, as different cell types may have different methylation profiles, and the patient cohorts are also rather different, with fewer years of follow up in the Norwegian cohort. The observation that there was no correlation with the hypomethylation and length of time to diagnosis was interesting.

“Overall, this is a preliminary finding of interest and worthy of follow up, but much larger cohorts will be needed to confirm this and explain the differences between the studies.”

Prof. Paul Pharoah, Professor of Cancer Epidemiology, University of Cambridge, said:

“This is an intriguing study reporting on a possible association between a blood-based biomarker (called DNA hypomethylation) and future breast cancer risk.  However, the findings, as emphasised by the authors, are somewhat preliminary and much more work needs to be done to confirm or refute these findings.

“While the authors report the results from four cohort studies – studies in which healthy people are followed up and subsequent cancer diagnoses determined – the methods for determining DNA hypomethylation were the same in two cohorts and different in each of the other two. In the two cohorts in which the same method was used, the results were notably different. In one cohort (EPIC Italy) there was a modest association between hypomethylation and reduced risk and in the other cohort (NOWAC) no association was observed. This difference is hard to explain and emphasises the need for further research. In a third cohort (BGS – Breakthrough Generations Study) the authors report an association of hypomethylation with risk, but no statistical P-value is reported to enable the robustness of this association to be judged. Finally, the fourth cohort (MCCS) reports data that have previously been published and provides support for the association observed in EPIC.

“I think these results are too preliminary to consider estimating what level of risk, if any, is associated with blood DNA hypomethylation.”

‘Epigenome-wide association study reveals decreased average methylation levels years before breast cancer diagnosis’ by Karin van Veldhoven et al. published in Clinical Epigenetics on Tuesday 4 August 2015. 

Declared interests

Prof. Shirley Hodgson declares no interests.

Prof. Paul Pharoah: “I have no conflicts of interest to report other than knowing many of the authors of this study as collaborators in other projects.”