Cell free DNA as a diagnostic and prognostic marker for cardiovascular diseases

Abstract

Release of cell free DNA (cfDNA) from damaged or dead cells routinely occurs in normal physiology. Recently, cfDNA has emerged as an essential biomarker in cardiovascular disease (CVD) of potential prognostic and diagnostic significance. Within the last decade, significant research efforts have been devoted to uncovering the mechanisms mediating cfDNA release and its outcome-predicting ability. The current review focuses on the pathways for cfDNA release in myocardial infarction, heart failure and hypertension, and discusses implementation of cfDNA monitoring to assess the overall development of these disease states and predict future complications.

1. Introduction: cfDNA and its origins

Extracellular DNA, or cell-free DNA (cfDNA), is highly fragmented double-stranded DNA, which freely circulates in body fluids (plasma/serum, urine, cerebrospinal fluid, etc) under normal physiological conditions. cfDNA is known to increase due to exercise, increased age, as well as in pathological conditions, such as cardiovascular diseases (CVD) including hypertension, myocardial infarction (MI) and heart failure (Figure 1).[1–7]

FIGURE 1. Schematic representation of cfDNA involvement in cardiovascular disease.

In a disease state, such as hypertension, MI or HF, cfDNA release can be triggered from the injured tissue. cfDNA can be generated as a result of cellular breakdown, and subsequent release of the content of the dead cells into the bloodstream. Dependent on the type of cell death - apoptosis or necrosis - smaller or larger cfDNA fragments will be detected, respectively. Active release from living cells via exosomes, NETosis or vesicular transport is associated with immune cell activation and increased levels of cytokines, chemokines and MMPs.

cfDNA was first considered a promising new biomarker when the concept of “liquid biopsy” was conceived as a result of seminal research in the field of oncology.[8–10] Since then, multiple studies have been devoted to the role of cfDNA in various pathological conditions. Despite these research efforts, many aspects of cfDNA-related processes remain mostly unclear. For example, the origin of cfDNA is still to be fully determined. Potentially, there are two major sources of cfDNA bloodstream release, ie, due to cellular breakdown predominantly caused by necrosis, apoptosis and active release of cfDNA from living cells (exogenous and endogenous DNA fragments resulting from vesicular transport, exosomes, release from hematopoietic cells, macromolecular complexes, neutrophil extracellular traps etc.). [11] The term extracellular DNA may also include circulating mitochondrial DNA (mtDNA), which plays a crucial role in cardiac cell damage - mtDNA is recognized as a part of circulating danger-associated molecular patterns (DAMPs).[12, 13] mtDNA has similarity with bacterial DNA, and may become an inflammogenic factor (if autophagy is impaired).[14, 15] In many ways, the structural features of cfDNA depend on the mode of release and its origin; the ability to detect specific cfDNA characteristics and subtypes may possess prognostic potential and provide important information for diagnosis.[1, 16]

Whether cfDNA was formed as a result of apoptosis or necrosis may be characterized by the size of DNA fragments. DNA released via apoptotic mechanisms is initially cleaved into large fragments of 50–300 kb with subsequent degradation into smaller fragments (180 – 200 bp) facilitating by endonuclease(s).[17–20] Unlike programmed cell death, necrosis in response to injury, trauma, or sepsis can be recognized by the presence of high molecular weight cfDNA (~10000 bp).[17, 19] Normally, apoptotic and necrotic bodies with fragmented cellular organelles and DNA would be cleared via phagocytosis, which is essential for tissue homeostasis.[21, 22] However, impaired clearance mechanisms would cause accumulation of cfDNA that might later lead to an auto-inflammatory response.[23–25] Our understanding of the role of cfDNA during hypertension and post-MI remodeling however, is incomplete. Since cfDNA has been implicated as a strong biomarker for hypertension and myocardial damage, this review is focused on the strength and feasibility and potential mechanisms.

2. cfDNA in Cardiovascular Disease

2.1. Hypertension.

Uncontrolled hypertension has been shown to be an independent determinant for elevated cfDNA.[2, 26] Hypertension is a chronic complex condition associated with a high mortality risk. A growing body of literature suggests a strong link between inflammation and hypertension, however, whether inflammation is a source or outcome of blood pressure increase still remains a matter for debate.[27–30] Augmented cfDNA levels have been linked to multiple cardiometabolic risk factors including increased systemic inflammation, elevated low-density lipoprotein cholesterol and triglyceride levels, and higher systolic blood pressure and pulse pressure in both sexes.[2] During hypertension, tissues and organs undergo prolonged exposure to damaging factors such as oxidative stress, inflammation, and increased Ang II levels.[31] An important outcome of the inflammation induced by Ang II is increased vascular reactive oxygen species (ROS) formation.[32–34] As a consequence of oxidative stress, damage of genomic DNA occurs, which may contribute to the total pool of cfDNA circulating in the bloodstream, and result in activation of toll-like receptor (TLR) signaling pathway followed by vascular dysfunction and increased blood pressure.[12, 35]

Patients with elevated blood pressure and diabetes were demonstrated to have elevated levels of cfDNA indicting a possible association with decreased arterial elasticity.[2, 26] Plasma cfDNA levels are elevated in patients with undergoing hemodialysis compared with healthy controls and were linked to vascular injury not inflammatory status.[26] It was shown that in post-menopausal women not using hormone replacement therapy, an increase in cfDNA levels indicated arterial stiffness (higher stiffness index and Young’s elastic modulus and lower carotid artery compliance), systemic inflammation (elevated C-reactive protein (CRP), IL-6 and TNF-α levels), impaired glucose metabolism and elevated blood pressure.[2] The association of cfDNA levels with increased inflammation was attenuated in women on estrogen-based hormone replacement therapy. These studies suggests that the elevated cfDNA may be indicative of ongoing vascular endothelial injury and support the notion that cfDNA may represent a novel way to monitor hypertension progression at the molecular level.

2.2. Myocardial Infarction and Ischemic Heart Failure.

The use of cfDNA as a biomarker for MI and development of heart failure is an attractive alternative to existing tests. After MI, cfDNA increases up to 50-fold compared with healthy controls.[1, 5, 36–38] When assessing time of release, cfDNA is significantly elevated in MI patients prior to intervention (0–2 h after onset of chest pain) compared to healthy individuals (AUC =0.8117, p-value <0.001).[1] After percutaneous coronary intervention, cfDNA accumulates at a faster rate, and then returns to baseline after 1–2 days, while the more commonly used biomarker, troponin, remains elevated. [1]

In vitro studies have shown that genetic content in cfDNA may contribute to the immune response.[39] An increase in circulating myeloperoxidase–DNA complexes was found to be strictly specific to neutrophil activation in patients with severe coronary atherosclerosis.[40] Significant positive associations were also observed between neutrophil activation marker elastase–α1-PI complexes and cfDNA levels (Spearman correlation value =0.257; p-value <0.001).[40] Neutrophils are a major source of MMPs, specifically MMP-8 and −9, and play a critical role in the cardiac wound healing process after an MI.[41–43] Whether cfDNA is acting on inflammatory cells to initiate changes in activation status post-MI is unknown, and requires further studies.

In plasma from MI patients, cfDNA correlates with well-established markers of necrosis such as troponin, creatine kinase (CK), CRP, as well as a decrease in ejection fraction.[5, 44–48] Such a tight correlation between cfDNA levels and these necrotic markers may indicate that the amount of released cfDNA depends on the severity of the myocardial injury. The CK test is often used as a means to test injury or stress to the heart.[1, 5, 44] Interestingly, cardiac cfDNA levels were increased in MI patients with normal (<200 μg/L) CK levels suggesting greater sensitivity.[1] A comparison of troponin levels to cardiac cfDNA in 57 samples from MI patients demonstrated a strong relationship between the two biological variables (Spearman correlation value=0.79; p-value < 0.0001). Majority (79%) of the MI samples were positive for both troponin and cardiac cfDNA, 7% were negative for both, 11% were positive only for troponin, and 4% were weakly positive only for cardiac cfDNA.[1] These studies seem to indicate that cfDNA may be a stronger biomarker than some of the more traditional ones currently used in the clinic.

A multimarker test that includes cfDNA may complement CK and troponin testing. Due to what we do know about cfDNA and inflammation, adding cfDNA to a multimarker panel may also give us insight into the inflammatory status of the patient. The use of a multimarker panel will also capitalize on the differences in the rates of release and clearance of these biomarkers (Table 1). Cardiac troponin has been shown to be detectable as early as 6 hours post-MI; however, troponin does not reflect current cardiovascular status as levels can remain elevated up to 14 hours after hospital admission.[49] The half-life of cfDNA is not completely clear, although multiple studies suggest dynamic changes in the levels – from 4 – 30 min.[9, 10, 17, 50] Rapid changes in cfDNA should be considered an advantage as it might serve as a powerful tool to monitor the response immediately after treatment.

Table 1.

Kinetics of known biomarkers for myocardial infarction

Marker	Kinetics		References
	Detection limits (from baseline to peak)	Clearance
cfDNA	Increases within 0 – 2 hours after onset of a chest pain	Remains elevated at 24 hours	[1, 5, 37, 38, 44, 45]
Cardiac troponin	Increases within 6 hours up to 24 – 48 hours of post-MI (the degree of elevation correlates with infarct zone size)	Elevated up to 14 days	[1, 5, 44, 49]
CK	Increases within 6 hours peaking around 24 hours of onset of MI	Decreases within 2–3 days	[1, 5, 44]
CRP	Peaks around days 1–2	Elevated for several weeks	[1, 5, 46, 48]
Inflammatory cytokines (i.e. IL-6 and Tnfα)	Increases as early as 45 min and peaks within 1–2 days post-MI	Above normal levels after 12 weeks	[47, 48, 77]
MMPs	increase during the first hours post-MI	Remains elevated up to 14 days	[42, 43, 77]

CK – creatine kinase; MMPs – matrix metalloproteinases; CRP - C-reactive protein

There is a clear need for the identification of biomarkers that have mechanistic implications, and can accurately and reliably predict the development of heart failure post-MI. The potential for cfDNA as a biomarker for heart failure post-MI is still under evaluation. To ensure the translation of cfDNA into the clinic, the clinical and research community will need to define the practical considerations of this approach.

3. Future Directions

As a relatively new biomarker, cfDNA has shown great promise in medical practice, including during prenatal testing and as a liquid biopsy for cancer; however, not much is understood in regards to the mechanism of action in CVD. Further exploration of the association of CVD and cfDNA will enhance its clinical utility to better address patients’ needs and treatment (Table 2). To detect cfDNA with high specificity and reactivity, a variety of approaches have been developed including droplet digital PCR and molecular index-based next generation sequencing technologies.[51, 52] All of them have benefits and limitations associated with increased sample variation during sample preparation steps, such as extraction and purification.[53, 54]

Table 2.

Strengths and weaknesses of cell-free DNA as a biomarker for hypertension and myocardial infarction.

Disease	Strengths	Limitations	References
Hypertension	May be an early predictor of hypertension induced complications due to its link to endothelial dysfunction	Short half-life/detection window Low concentration/poor sensitivity	[2, 26]
Myocardial infarction	Early detection Sensitive; more responsive to disease changes Robust and reproducible	Short half-life/detection window No data with long-term monitoring of patients and the links to disease complications.	[1, 5, 37, 38, 44, 45]

Depending on the methodology used, a cutoff level of cfDNA >0.20  mg/ml has been suggested to distinguish between experimental group and controls with a sensitivity of 69–79% and a specificity of 83–89%.[55] Amplification-based target enrichment techniques such as multiplex-PCR next-generation sequencing improves sensitivity by eliminating genomic DNA regions that are not of interest and usually do not require a large amount of DNA.[56] In lung cancer patients, multiplex-PCR next-generation sequencing demonstrated a sensitivity >99% for single-nucleotide variants and a specificity of 99.6% with as less as 20  ng of cfDNA as input material.[57] Optimizing the workflow is another way to increase the reliability of the cfDNA analysis. Adding quality control steps such as measuring cfDNA concentration and integrity and cellular DNA contamination and potential PCR inhibitors (e.g. heparin, hormones, immunoglobulin G and lactoferrin) during cfDNA extraction can facilitate the implementation of cfDNA analysis into clinical routine.[58]

Moss et al. developed a method focused on identifying and locating tissue damage using the specific methylation pattern of cfDNA.[1, 16] By deconvolution of the genome-wide methylation profiles, they were able to determine cellular origins of cfDNA in healthy and diseased patients.[16] Moss et al. suggested was that refinement of the methylome atlas of individual cell types (purified from fresh tissue), rather than whole tissues is needed to enhance cfDNA interpretation. The analysis of the amount and methylation status of cfDNA at early stages after MI may be a good biomarker for evaluation of severity of cardiac myocytes injury, and further prognosis of disease.

Liquid biopsy studies have also suggested that cfDNA molecules originating from different tissues have distinct base-pair sizes.[59–62] Pregnant women have been shown to have two populations of circulating DNA: one from the fetus and one from the mother.[59, 60] Interestingly, the DNA molecules from the fetus were shorter than the maternal DNA. This was also demonstrated after transplant in solid organ transplantation patients and mouse xenograft models where cfDNA molecules that had originated from the transplanted organ/donor were shorter than the recipient cfDNA.[61, 62] The detection and performance of plasma DNA as a biomarker for cancer progression improved when size-capture sequencing was used to enrich for tumor DNA.[63, 64]. Considering the successful applications in pre-natal and cancer screening, it may be of clinical significance to incorporate the size characteristic in future CVD studies.

Using these techniques may facilitate in a better understanding of the cell-specific role cfDNA plays in CVD. In a mouse model of rheumatoid arthritis, blocking cfDNA actions decreased immune cell activation and MMP secretion.[65] In post-MI patients, cfDNA within the coronary artery positively correlated with macrophage accumulation in coronary plaques.[66] Neutrophil extracellular traps also contain cfDNA (nuclear and/or mitochondrial) and are believed to initiate an inflammatory response through the activation of TLR signaling pathways.[12, 14, 67, 68] In a similar fashion, mtDNA has been shown to activate Ly6C+ monocytes through TLR-9 resulting in prostaglandin E2 production.[69] Further evidence for the involvement of TLRs in cfDNA-evoked response was provided by Nishimoto et al. whereby elevated TLR9 expression and cfDNA were found to increase capillary density and promote blood flow recovery after ischemic injury in skeletal muscle.[70] Additional studies evaluating the mechanism behind cfDNA, inflammation, and MMP production (and how these processes are associated with CVD risk) is needed.

One of the main limitations of the current literature is the lack of long-term data. Because the mean half-life of cfDNA is relatively short (degraded within 4 – 30 minutes), understanding the kinetic parameters of cfDNA is important in order to develop a longitudinal diagnostic tool. [6, 9, 10, 17, 50, 71, 72] How cfDNA correlates with chronic hypertension or later MI-induced complications such as development of heart failure is unknown. Follow-up studies that link the initial rise in cfDNA to chronic disease states and late-onset complications would allow us to better understand the strength of cfDNA as a CVD biomarker.

4. Conclusions

Detection of circulating cfDNA fragments is a promising, non-invasive, inexpensive and readily available prognostic biomarker. It has already found widespread application in prenatal care, where the fetal cfDNA in maternal plasma is used for early diagnosis of genetic disease, and preeclampsia; in addition, cfDNA is an emerging biomarker in oncology, rheumatology, transplant medicine and some kidney diseases.[17, 23, 73–76] Based on numerous research and clinical studies, cfDNA analysis may be implemented into diagnosis and prognosis of CVD.

HIGHLIGHTS.

• cfDNA is a biomarker for diagnosis and prognosis of cardiovascular disease

• cfDNA is released from living cells via exosomes, NETosis, or vesicular transport

• cfDNA activates immune cells and increases cytokines, chemokines, and MMPs

• Future studies should be focused on optimizing sample preparation to increase reliability of cfDNA analysis.