Cell-free DNA levels of twins and sibling pairs indicate individuality and possible use as a personalized biomarker

Abstract

Cell-free DNA (cfDNA) in the human blood circulation has been under investigation since its initial observation in 1948. Plasma cfDNA is known to be significantly elevated in diseased people. Due to possible variation in the population, evaluating cfDNA as a non-invasive biomarker at disease onset alone may not be sensitive enough to accurately diagnose diseases, particularly early stage cancers on a personal level. To understand the factors that define the cfDNA levels on the personal level and for better use as a non-invasive biomarker, we isolated cfDNA from the plasma of healthy individuals with varying degrees of genetic and/or environmental similarities (monozygotic twins, dizygotic twins, sibling pairs, and unrelated individuals) as well as from patients with varying stages of breast and ovarian cancer undergoing treatment. Cell-free DNA levels were quantified by a fluorometer (ng/ml) and/or real-time PCR (copies/ml). The associations between individuals with various degrees of genetic and/or environmental similarities and their plasma cfDNA levels were evaluated. The ACE model (A = additive genetic, C = common environment, and E = specific environmental factors) was used to determine the proportion of each factor on the cfDNA levels. We found a high correlation (r = 0.77; p < 0.0001) in plasma cfDNA levels between monozygotic twins (n = 39). However, the correlation was gradually reduced to moderate (r = 0.47; p = 0.016) between dizygotic twins (n = 13) and low correlation (r = 0.28; p = 0.043) between sibling pairs (n = 26). The ACE model analysis showed that the plasma cfDNA level of a given healthy individual is influenced both by genetic and the environmental components in similar proportions (53% and 47%, respectively; A = 53%, C = 22.5%, E = 24.5%). Moreover, while age had no effect, gender significantly influenced the individual’s plasma cfDNA level. As expected, cfDNA levels were significantly higher in both breast (n = 26) (p<0.0001) and ovarian (n = 64) (p<0.0001) cancer patients compared to the healthy individuals. Our study demonstrated that both genome and environmental factors modulate the individual’s cfDNA level suggesting that its diagnostic sensitivity may be improved only if the person’s cfDNA level is known prior to disease presentation.