Currently, with the development of new technologies for imaging and pathology, we have further insight into the mechanism of myocardial infarction. Myocardial infarction can be divided into different plaque types according to OCT examination, and it is important to distinguish plaque rupture from erosion because patients with different plaque types have different risks and prognoses. Patients with erosion may benefit from pharmacological therapy rather than mechanical revascularization.13 It is more convenient to determine the types of plaque by biomarkers rather than OCT, which is more expensive for patients. Our previous study revealed that plasma trimethylamine N-oxide (TMAO) can be a useful biomarker to predict plaque rupture.14 However, although the mechanism of plaque erosion has been partly studied, there is still no acknowledged biomarker for predicting plaque erosion. In this study, we first reported the distinction of plasma HA and CD44 between patients with STEMI and healthy people. Additionally, we found that plasma HA levels in erosion patients were significantly lower than in rupture, SCAD patients and healthy subjects, which meant that plasma HA levels can be used as a potential biomarker to predict plaque erosion.
HA plays a crucial role in the progression of cardiovascular disease and is involved in several important phases of coronary artery disease (CAD), such as inflammation and angiogenesis.15 HA exists in plasma in two forms: high molecular weight (HMW)-HA and low molecular weight (LMW)-HA. At homeostasis, HA is predominantly in its high molecular mass form of over 1,000 kDa. However, in tissue injury or inflammation, HYAL2 and HYAL1 are upregulated to break down HA to 20 kDa, which binds to the receptors of immune cells.16 It has been reported that HYAL2-deficient mice display a significant increase in plasma HA levels.17 HYAL2 and CD44 exist in many kinds of components in blood, such as platelets and monocytes, all of which contribute to HMW-HA degradation when inflammation occurs. Furthermore, the metabolism of HA is influenced by not only inflammation but also disturbed flow, which is related to plaque erosion according to the current study.18 Several in vivo and in vitro mechanistic studies indicated that disturbed flow might be the initial factor inducing erosion, which is related to an alteration of HA metabolism. Additionally, an increasing number of clinical investigations have found that HA is correlated with the pathogenesis of plaque erosion, and pathological observations have shown that eroded plaques have few inflammatory cells but abundant proteoglycan and glycosaminoglycans, including HA and its receptor CD44.19 A recent clinical study revealed that the expression of HYAL2 and CD44 in peripheral blood mononuclear cells (PBMCs) increased in acute coronary syndrome (ACS), especially in patients with plaque erosion compared with stable CAD and healthy people.7
However, most previous studies focused on HA inside the plaque, not in circulation. Interestingly, our study found that plasma HA levels decreased in STEMI patients, particularly in plaque erosion. This may indicate that regional disturbed flow and eroded plaques might influence the systemic change in HA levels in STEMI patients. HA experiences many pathophysiological processes in plaque erosion, including Toll-like receptor 2 (TLR2) stimulation, endothelial activation and neutrophil accumulation. Based on a previous study, we hypothesized that plasma HA depletion is induced by plaque erosion in three phases. The first hit is mediated by TLR2. Previous studies found that TLR2 is widely expressed on the surface of eroded plaques in the zone of flow perturbation.20 As one of the endogenous ligands, HA activate TLR2 which contributes to endothelial cell detachment and apoptosis.21 The second hit of HA is binding to immune cells, such as macrophages and neutrophils, and then eventually being degraded via the HYAL2 and CD44 pathways. In plaque erosion, overexpression of CD44 induces the adhesion of neutrophils to HA.7 HA fragmentation can result from degradation by the reactive oxygen species (ROS) that are produced by neutrophils.22 On the other hand, there is evidence that macrophages are involved in HA uptake and the removal of HA fragments in inflammation.23 The third hit is the degradation of HA by platelets. In the condition of platelet aggregation, HYAL2 becomes expressed on the cell surface but is stored into α-granules in rest.24 In particular, activated platelets have higher hyaluronidase activity than nonactivated platelets and can reduce the concentration of free HA in plasma. This phenomenon is more obvious in plaque erosion than plaque rupture because platelet-rich thrombi usually occur in eroded plaques. Furthermore, platelets also express CD44, which can bind to free HA in plasma when it is activated.25
Furthermore, in our study, the plasma level of CD44 was also decreased in STEMI patients compared with healthy controls. This result mirrors the combination of HA and soluble CD44 in circulation. In reference with a previous study, we deduced that plasma HA binds to the N-terminal hyaluronan binding domain (HABD) of CD44 in an inflammatory state to consume free CD44 in plasma.26 Although the mRNA expression of CD44v1 and CD44v6 in PBMCs has been reported to be different between plaque rupture and erosion in the past,7 there was no difference in soluble CD44 between plaque erosion and rupture in our study. Soluble CD44 is not only regulated by HA but also other factors, such as cytokines and shedding from immune cells.27 The metabolism of soluble CD44 in STEMI patients needs further investigation.
It has been reported that the inhibition of HA synthesis accelerates the process of atherosclerosis because HA provides a protective effect on blood vessels.28 Therefore, the relationship between HA reduction and atherosclerosis may exist as a circle of positive feedback. However, the continuous production of HA has a compensatory function. HA is produced by stromal cells via hyaluronan synthases (HAS1-3).4 A recently published study found that the gene expression of CD44 and HYAL2 are different between plaque rupture and erosion, but they can return to baseline after 1 year of follow-up.7 This finding implies that HA has a compensatory effect but that this effect does not occur rapidly. In the acute phase of ischemia and inflammation, the level of HA decreased, but whether it can recover to normal needs further investigation.
Taken together, these findings suggest that plasma HA is consumed by an enhanced expression of HYAL2 in monocytes and platelets, a combination of TLR2 on endothelial cells and the uptake by macrophages mediated with the CD44 receptor (central illustration). This model derives from both the existing experimental and clinical investigations about plaque erosion and the data emerging from our study.
Limitation: This study has several potential limitations. First, patients with cardiac shock, congestive heart failure, a history of coronary artery bypass graft, left main diseases, extremely tortuous or heavily calcified vessels, or chronic total occlusion were not enrolled in our study. In addition, patients with massive thrombi have poor image quality. Therefore, selection bias cannot be excluded. Second, there is no independent cohort to validate the predictive value of HA in discriminating between plaque morphologies, which we aim to include in future studies. Third, hyaluronan ELISA kit (DHYAL0, R&D Systems, Abingdon, UK) is only able to test the hyaluronan > 35 kDa, and the circulating level of LMW-HA which is lower than 35 kDa is unknown. In the future, we hope to use mass spectrometer to separate different molecular weight of HA. Finally, the sensitivity and specificity of plasma HA levels to predict erosion is not very high, which may be related to the small sample size. We hope to expand the number of enrolled patients in our further investigation.