Circulation Journal Official Journal of the Japanese Circulation Society http://www. j-circ.or.jp
FOCUS REVIEWS ON FUNCTIONAL TESTINGS
Secondary Prevention Strategy of Cardiovascular Disease Using Endothelial Function Testing Yasushi Matsuzawa, MD, PhD; Raviteja R. Guddeti, MD; Taek-Geun Kwon, MD, PhD; Lilach O. Lerman, MD, PhD; Amir Lerman, MD
Over the past decades, secondary prevention of cardiovascular (CV) disease has improved and considerably reduced mortality rates. However, there remains a high-rate of new or recurrent CV events in those with established atherosclerotic vascular diseases. Although most of the prevailing therapies target the conventional risk factors, there is notable interindividual heterogeneity in adaptation to risk factors and response to therapies, which affects efficacy. It is desirable to have a methodology for directly assessing the functional significance of atherogenesis, and for managing individual patients based on their comprehensive vascular health. Endothelial function plays a pivotal role in all stages of atherosclerosis, from initiation to atherothrombotic complication. Endothelial function reflects the integrated effect of all the atherogenic and atheroprotective factors present in an individual, and is therefore regarded as an index of active disease process and a significant risk factor for future CV events. Moreover, improvement in endothelial function is associated with decreased risk of CV events, even in the secondary prevention setting. The introduction of endothelial function assessment into clinical practice may trigger the development of a more tailored and personalized medicine and improve patient outcomes. In this review, we summarize current knowledge on the contribution of endothelial dysfunction to atherosclerotic CV disease in the secondary prevention setting. Finally, we focus on the potential of an endothelial function-guided management strategy in secondary prevention. (Circ J 2015; 79: 685 – 694) Key Words: Coronary artery disease; Endothelial function; Secondary prevention
O
ver the past decades, secondary prevention measures have greatly improved and considerably reduced cardiovascular (CV) mortality.1 However, there remains a high-rate of new or recurrent CV events in those with established coronary and other atherosclerotic vascular diseases.2 The difficulty in identifying individual risk can be highlighted as a factor associated with the high prevalence of recurrent CV disease. Although a number of CV risk factors have been established and effective treatments for atherosclerotic CV disease (ASCVD) have been developed, there is a notable interindividual heterogeneity in response to risk factors and therapies, which affects efficacy. It is desirable to have a methodology for directly assessing the functional significance of atherogenesis at each stage. Endothelial function reflects the balance of all atherogenic and atheroprotective factors present in an individual. Dysfunctional endothelium is associated with unfavorable physiological vascular changes such as vasomotor tone alterations, thrombotic dysfunction, smooth muscle cell proliferation and migration, as well as leukocyte adhesion, and plays a pivotal role in the initial development and progression of atherosclerotic plaque, and subsequent atherosclerotic complications.3,4
Endothelial dysfunction can, therefore, be regarded as both an index of active disease process through the course of ASCVD, and a significant risk factor for future CV events.5 Thus, the introduction of endothelial function assessment into clinical practice may instigate the development of more tailored medicine. In this review, we update the evidence supporting the role of endothelial function assessment in patients with established ASCVD, and focus on the potential of an endothelial functionguided management strategy in the secondary prevention setting.
Role of Endothelial Dysfunction in CVD Atherosclerosis begins early in life, and progresses over decades. Endothelial dysfunction contributes to atherosclerotic disease progression in all stages.3,6 Dysfunctional endothelium is also responsible for increased plaque vulnerability.7 Impaired endothelial function is associated with an increased inflammatory response, thrombogenicity, and enhanced local expression of matrix metalloproteinases, which are rendered prone to develop a fracture in the protective fibrous cap of plaques, and coronary
Received January 19, 2015; revised manuscript received February 4, 2015; accepted February 5, 2015; released online February 27, 2015 Division of Cardiovascular Diseases (Y.M., R.R.G., T.-G.K., A.L.), Division of Nephrology and Hypertension (L.O.L.), Mayo Clinic, Rochester, MN; Department of Internal Medicine, Marshfield Clinic/St Joseph’s Hospital, Marshfield, WI (R.R.G.), USA Mailing address: Amir Lerman, MD, Division of Cardiovascular Disease and Department of Internal Medicine, Mayo College of Medicine, 200 First Street SW, Rochester, MN 55905, USA. E-mail:
[email protected] ISSN-1346-9843 doi: 10.1253/circj.CJ-15-0068 All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail:
[email protected] Circulation Journal Vol.79, April 2015
686
MATSUZAWA Y et al. thrombosis.3,8 Moreover, apoptosis of endothelial cells could contribute to desquamation of endothelial cells in areas of superficial erosion, which can foster coronary thrombosis.9 In addition to its association with atherosclerosis, endothelial dysfunction has been implicated in other conditions that lead to CV events, such as coronary spasm,10 heart failure with preserved ejection fraction,11 cardiomyopathy,12 atrial fibrillation,13 and left atrial14 and venous15 thrombus formation. Thus, endothelial dysfunction is a systemic manifestation and represents comprehensive CV health (Figure 1). Effective identification of vulnerable patients with severe endothelial dysfunction is important to improve prognosis.
Local and Systemic Factors Damaging the Endothelium
Figure 1. Systemic manifestation of endothelial dysfunction in the vulnerable patient with vulnerable endothelium.
Atherosclerosis is a diffuse disease with focal complications in different vascular beds. The precise mechanisms by which a specific site is rendered more prone to the development of symptomatic disease and CV events are not known. Endothelial function status is not determined solely by an individual risk factor burden, but is rather an integrated index of all factors (Table 1).16 Although the entire systemic vasculature is exposed to the atherogenic effects of systemic risk factors, similarly, local risk factors, such as flow-generated endothelial shear stress,17 angioplasty,18 and local inflammation play a role in regional endothelial dysfunction and plaque formation. Iatrogenic vascular injury, including balloon angioplasty and stent implantation, disrupts endothelial cells, which promotes
Table 1. Local and Systemic Risk Factors for Atherosclerosis Known risk factors Local
Systemic
Hemodynamic forces (eg, shear stress)
Conventional risk factors (modifiable)
Vascular injury (eg, balloon angioplasty, stent implantation)
Smoking
Local inflammation
High LDL-C
Local oxidative stress
Low HDL-C
Impaired local endothelial repair
High triglycerides
Hypertension
Diabetes, metabolic syndrome, insulin resistance Non-conventional risk factors Male sex Older age Race Genetic factors Inflammation Lp-PLA2 Lipoprotein(a) Homocysteine Environmental exposures Depression, mental stress Low physical activity, sedentary behaviors Obesity Dietary factor Menopause, postmenopausal hormone therapy Noise Others Unknown risk factors HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; Lp-PLA2, lipoprotein-associated phospholipase A2. Circulation Journal Vol.79, April 2015
Endothelial Function in Secondary CVD Prevention
687
Table 2. Methods Used to Assess Endothelial Function Coronary or Peripheral artery
Vascular bed
Measurements
Stimulus
Invasive
Coronary epicardial vasoreactivity
Coronary
Conduit
Vessel diameter
Infusion of endothelial dependent vasodilator
+
Coronary microvascular vasoreactivity
Coronary
Resistance
Blood flow
Infusion of endothelial dependent vasodilator
+
Peripheral (brachial artery)
Conduit
Vessel diameter
Reactive hyperemia
−
Peripheral (finger microvasculature)
Resistance
Plethysmogram
Reactive hyperemia
−
Method
FMD RH-PAT
FMD, flow-mediated vasodilatation; RH-PAT, reactive hyperemia-peripheral arterial tonometry.
in-stent thrombosis and restenosis. Restoration of healthy vascular endothelial cells is an important step in the prevention of subsequent coronary events. Locally derived endothelial cells and bone-marrow derived circulating endothelial progenitor cells (EPCs) have been suggested to participate in re-endothelialization.18,19 EPCs may have an important function as an endogenous repair mechanism by replacing denuded parts of the artery and regenerating low-grade endothelial damage. Dysfunctional EPCs are also considered part of endothelial dysfunction.20 Atherosclerotic plaque progression results from a complex interaction between local and systemic atherogenic and atheroprotective factors. Both local and systemic atherosclerotic disease manifestations vary depending on the stage, location, and other factors affecting the integrity of the vascular wall.
Assessment of Endothelial Function Development of clinical tests that evaluate endothelial function has paralleled the growing understanding of the biology of the vascular endothelium. Endothelial function can be measured by assessing various physiologic functions, which include regulation of vascular tone, expression of adhesion molecules and maintenance of an antithrombotic microenvironment. In general, loss of regulation and activity of vasoactive substances, in particular nitric oxide (NO), indicates a broadly dysfunctional phenotype across many properties of the endothelium; suppression of platelet aggregation, inflammation, oxidative stress, vascular smooth muscle cell migration and proliferation, and leukocyte adhesion.21 Thus, endothelium-dependent vasodilation is the most widely used clinical endpoint to assess and reflect the multiple aspects of endothelial function in humans. Coronary Endothelial Function Assessment The widely accepted method of evaluating coronary endothelial function involves intra-arterial administration of endothelium-dependent vasodilatory substances (eg, acetylcholine). The vasodilatory agent delivered into the coronary arteries results in measurable vasodilatation and an increase in coronary blood flow in normal subjects through activation of endothelial cells and stimulation of NO release, whereas in patients with endothelial dysfunction, it induces vasoconstriction and lack of increase in coronary blood flow via direct activation of muscarinic receptors on vascular smooth muscle cells. Changes in vessel diameter assessed by quantitative coronary angiography represent epicardial coronary endothelial function, whereas changes in coronary blood flow assessed by Doppler flow wire represent coronary microvascular endothelial function (Table 2). More recently, noninvasive methods of assess-
ing coronary endothelial function have been developed, such as transthoracic Doppler echocardiography, computed tomography imaging, magnetic resonance imaging, and positron emission tomography.22,23 Peripheral Endothelial Function: FMD and RH-PAT It has been reported that endothelial dysfunction in peripheral arteries had comparable prognostic value to coronary endothelial dysfunction,24 and several noninvasive methods for the assessment of peripheral endothelial function have been developed. Reactive hyperemia after artery occlusion is used as a trigger to detect endothelium-dependent vasodilation in most of the noninvasive methods. Brachial flow-mediated vasodilatation (FMD) and reactive hyperemia-peripheral arterial tonometry (RH-PAT) are some of the widely used noninvasive methods, and are based on the same principle of reactive hyperemia (Table 2). To evaluate the endothelium-dependent vasodilation capacity, the brachial artery diameter proximal to the antecubital fossa is measured in the FMD technique at rest and during reactive hyperemia, which is achieved by rapid release of a pneumatic pressure cuff after inflation to suprasystolic pressure for 5 min. In the RH-PAT technique, the pulse wave amplitude of the finger is measured. Thus, FMD assesses conduit artery vasodilation, and RH-PAT assesses microvessel vasodilation. Both of these techniques have been reported to correlate well with coronary artery endothelial function.10,25,26 However, the Framingham Heart Study reported that the relationship between RH-PAT and FMD is not statistically significant, and concluded that the 2 methods have differing relationships with CV risk factors.27 NO bioavailability has a substantial role in both,28,29 but other substances, such as prostaglandin, adenosine, and hydrogen peroxide, can also affect vasodilation in response to shear stress and ischemia.30 There are 2 techniques of measuring brachial FMD, using an occluding cuff placed distal or proximal to the imaged artery. FMD with distal occlusion is more NO-dependent than FMD with proximal occlusion or RH-PAT. Interestingly, FMD with proximal occlusion provides higher predictive value for CV events than FMD with distal occlusion.31 Furthermore, microvascular function measured by blood flow or shear stress response after ischemia possesses independent predictive value from endothelial function in conduit arteries, and such responses are not solely NO mediated.32–36 FMD and RH-PAT might reflect different and complementary aspects of vascular function. Other methods used for peripheral endothelial function assessment include laser Doppler flowmetry, biochemical markers (asymmetrical dimethylarginine, etc), endothelial microparticles, and EPCs.23
Circulation Journal Vol.79, April 2015
688
MATSUZAWA Y et al.
Table 3. Clinical Evidence of the Association of Endothelial Function With Coronary Plaque Endothelial function assessment
Plaque assessment
30
Coronary epicardial vasoreactivity
VH-IVUS
Nonobstructive CAD
32
Coronary epicardial vasoreactivity
NIRS
Coronary epicardial endothelial dysfunction associated with lipid core plaques
CAD
111
FMD
VH-IVUS
Lower FMD associated with more necrotic core and higher prevalence of TCFA
Nonobstructive CAD
40
Coronary epicardial vasoreactivity
OCT
Coronary epicardial endothelial dysfunction associated with macrophages and microchannels in coronary plaques
Yoon et al41 (2013)
Nonobstructive CAD
35
Coronary epicardial vasoreactivity
IVUS
In segment with endothelial dysfunction, coronary plaque progressed
Gössl et al4 (2014)
Nonobstructive CAD
22
Coronary epicardial vasoreactivity
IVUS
In segment with endothelial dysfunction, coronary plaque progressed
Study
Population
n
Lavi et al37 (2009)
Nonobstructive CAD
Choi et al38 (2013)
Results
Coronary plaque vulnerability
Sawada et al40 (2013) Choi et al39 (2014)
Local coronary endothelial dysfunction associated with greater necrotic core
Coronary plaque progression
CAD, coronary artery disease; FMD, flow-mediated vasodilatation; NIRS, near-infrared spectroscopy; OCT, optical coherence tomography; TCFA, thin-cap fibroatheroma; VH-IVUS, virtual histology-intravascular ultrasound.
Clinical Evidence of the Association of Endothelial Function With Atherosclerotic Disease Plaque Vulnerability Atherosclerotic lesions prone to acute thrombotic complications because of plaque rupture or superficial endothelial erosion are known as “vulnerable plaques”. Recent clinical studies demonstrated that coronary endothelial dysfunction was associated with vulnerable plaque characteristics than those with normal endothelial function (Table 3). Lavi et al reported that coronary segments with attenuated endothelial function were associated with a larger necrotic core of plaque.37 Choi et al reported that epicardial coronary artery segments with endothelial dysfunction had more lipid deposition,38 macrophages and microchannels of plaque, consistent with vasa vasorum proliferation.39 In addition to invasive coronary endothelial function assessment, peripheral endothelial dysfunction as assessed by FMD was reported to be associated with larger necrotic core content of plaque and higher frequency of thincap fibroatheroma.40 Plaque Progression Endothelial dysfunction is not only a marker for CV risk but also a contributor to the progression of atherosclerosis. A randomized study of endothelin-A receptor antagonist was reported recently by Yoon et al.41 In addition to its vasoconstrictive properties, endothelin has mitogenic properties, and plays an important role in the development of endothelial dysfunction and progression of atherosclerosis.42 Plaque volume change was evaluated by intravascular ultrasound in patients with nonobstructive coronary artery disease (CAD) at baseline and 6-month follow-up, and in the coronary artery segments with endothelial dysfunction, significant plaque progression had occurred at the 6-month follow-up, but not in segments with normal endothelial function. Moreover, plaque progression was attenuated by endothelin-A receptor antagonist. Those results indicate the important role of coronary endothelial dysfunction in CAD progression. Similarly, Gossel et al reported that coronary plaque progressed more in segments with endothelial dysfunction than in segments with normal endothelial function.4 Thus, coronary segments with
endothelial dysfunction represent vulnerable segments. Stent Thrombosis and In-Stent Restenosis To our knowledge, no direct clinical evidence of the association between endothelial dysfunction and in-stent thrombosis has been reported. Compared with bare-metal stents, drugeluting stents reduce the incidence of in-stent restenosis, but also increase the risk of in-stent thrombosis, possibly mediated by their effects on the endothelium. It has been reported that drug-eluting stents are associated with a hypersensitivity reaction, delayed healing, and incomplete endothelialization, which may contribute to the increased risk of late and very late stent thrombosis compared with bare-metal stents.43 Moreover, in patients with CAD on dual antiplatelet therapy, peripheral endothelial function as assessed by RH-PAT is associated with residual platelet reactivity that may also contribute to an increased risk of in-stent thrombosis.44 Both systemic and local endothelial dysfunction may be modifiable factors in the prevention of in-stent thrombosis, and several stent technologies are being developed in an attempt to decrease the risk of late thrombotic events, including bioabsorbable polymers, nonpolymeric stent surfaces, bioabsorbable stents, and an EPC capture stent. Possible mechanisms involved in the pathogenesis of instent restenosis include platelets and inflammatory cell activation by procedural vascular injury with subsequent local release of cytokines and growth factors, leukocyte adherence, smooth muscle cell proliferation, and extracellular matrix synthesis. Dysfunctional endothelium may be partly responsible for instent restenosis, and several prospective studies using FMD have reported that endothelial dysfunction is an independent predictor of in-stent restenosis.45,46
Validity of Endothelial Function Testing for Event Prediction In 2000, we reported the first evidence of the long-term prognostic significance of coronary endothelial vasodilator dysfunction on atherosclerotic disease progression and CV events.47 In addition, the independent association between coronary microvascular endothelial dysfunction and the risk of future
Circulation Journal Vol.79, April 2015
Endothelial Function in Secondary CVD Prevention
689
Table 4. Clinical Evidence of the Association of Noninvasive Endothelial Function Assessment With CV Events in the Secondary Prevention Setting Study
Population
Follow-up
n
Method
Clinical endpoints
Events
rick et al53 F (2005)
Patients with chest pain (CAD 79%)
39 months (mean)
398
FMD
Cardiac death+MI+angina+coronary revascularization+progression of coronary plaque
44
FMD predicts CV events
Rubinshtein et al70 (2010)
Patients with chest pain
70 months (median)
270
PAT
CV death+MI+coronary revascularization+cardiac hospitalization
98
PAT predicts CV events
Matsuzawa et al64 (2013)
Patients with chest pain (CAD 84%)
34 months (mean)
528
PAT
CV death+MI+UA+coronary revascularization+HF+peripheral vascular events
105
PAT predicts CV events
Chan et al52 (2003)
CAD
34 months (mean)
152
FMD
CV death+MI+UA+coronary revas cularization+stroke+TIA+carotid endarterectomy+peripheral vascular events
22
FMD predicts CV events
Bosevski et al55 (2007)
DM+CAD
12 months (mean)
82
FMD
CV death+MI+angina+stroke+ worsening HF
46
FMD predicts CV events
Akcakoyun et al46 (2008)
CAD
12 months (mean)
135
FMD
CV death+MI+UA+stroke
30
FMD predicts CV events
In-stent restenosis
16
FMD predicts in-stent restenosis
Corrado et al56 (2009)
CAD
12 months (mean)
58
FMD
CV death+MI+UA+angina
itta et al58 K (2009)
CAD
31 months (mean)
251
FMD
Cardiac death+MI+coronary revascularization+stroke
42
Persistent low FMD over 6 months predicts CV events
atsue et al63 M (2014)
CAD (LDL