Coronary collateral circulation

INTRODUCTION — This topic will discuss aspects of coronary artery collateral vessels, including their potential role in minimizing myocardial ischemia. The broader discussion on the management of myocardial ischemia is found elsewhere. (See "Silent myocardial ischemia: Epidemiology, diagnosis, treatment, and prognosis" and "Approach to the patient with suspected angina pectoris".)

DEFINITIONS — Anastomotic channels, known as collateral vessels, can develop in the heart as an adaptation to ischemia [1,2]. They serve as conduits that bridge severe stenoses or connect a territory supplied by one epicardial coronary artery with that of another [3]. Collaterals therefore provide an alternative source of blood supply to myocardium jeopardized by occlusive coronary artery disease, and they can help to preserve myocardial function in the setting of a chronic total coronary occlusion [4].

Two classes of collateral vessels have been recognized:

●Capillary size collaterals, in which smooth muscle cells are absent, may be observed throughout the myocardium, although they have a predilection for the subendocardium.

●Larger, muscular collaterals, which develop from pre-existing arterioles, are typically located epicardially [5].

DETERMINANTS OF RECRUITMENT — The clinical and pathophysiologic determinants of collateral recruitment are poorly understood. Although primarily thought to be initiated by ischemia, appreciable collateral perfusion is present in some patients who do not have coronary disease [6].

Studies of transient coronary occlusion during balloon angioplasty have found that the following independent clinical and angiographic variables are correlated with increases in collateral flow [4,7-9]:

●Longer duration of angina

●Greater level of long-term physical activity during leisure time

●Greater lesion severity

●Proximal lesion location

●Greater duration of lesion occlusion

Collateral recruitment may be diminished in older adults. In a study of over 1900 patients undergoing angiography within 72 hours after an acute myocardial infarction (MI), the prevalence of collaterals was 48 percent in patients <50 years of age and only 34 percent in patients ≥70 years of age [10].

Role of diabetes — Patients with coronary artery disease who have diabetes mellitus have a less favorable outcome than those without diabetes. In addition, they have a higher mortality and are more likely to experience a complication after MI, such as post-infarction angina or congestive heart failure. (See "Coronary artery revascularization in stable patients with diabetes mellitus" and "Acute myocardial infarction: Patients with diabetes mellitus".)

The less favorable outcomes among patients with diabetes may be attributable to fewer collaterals. One study of 306 patients with diabetes who underwent coronary angiography found that these patients had a lower mean collateral score compared with nondiabetics (2.4 versus 2.6), suggesting that patients with diabetes may have poorer development of collaterals [11]. Fewer collaterals were also observed among patients with type 2 diabetes in a contemporary study of percutaneous coronary intervention (PCI) performed for chronic total occlusion [12].

Role of cytokines — The underlying mechanisms by which ischemia induces collateral vessel formation are unknown. Several endothelial and smooth muscle cell mitogens have been implicated [13-16]. As an example, markedly elevated levels of basic fibroblast growth factor (bFGF) have been demonstrated in the pericardial fluid of patients undergoing coronary artery bypass grafting for unstable angina [13]. In patients with coronary disease undergoing angioplasty, intracoronary concentrations of bFGF and vascular endothelial growth factor (VEGF) are associated with a directly-measured index of collateral flow; the concentration of VEGF is influenced by the degree of coronary atherosclerosis [17]. In addition, increased expression of VEGF has been shown in animal models of acute myocardial ischemia [14] and in patients with an acute MI; levels are higher in those with, compared to those without, improvement of left ventricular function [18]. VEGF induces or stimulates collateral vessel formation and improves the impaired endothelium-dependent relaxation of collaterals [19]; the induction of coronary collateralization by VEGF requires the production of nitric oxide, which is an important regulator of collateral growth [20]. Newer studies continue to implicate new molecular signals possibly involved in coronary arteriogenesis, and this is an active area of investigation [15,16].

Although suggestive, these studies do not directly prove a causative role for these molecules in the induction of collateral vasculogenesis.

PHYSIOLOGY — In a canine model of collateral development, gradual occlusion of an epicardial coronary artery resulted in collateral vessel growth sufficient to provide approximately 80 percent of the blood flow to normally perfused myocardium [21]. Although collateral vessel blood flow after epicardial coronary occlusion may be adequate to meet myocardial needs at rest, collateral circulation is generally not sufficient to meet myocardial demands during exercise [22] and may not prevent myocardial ischemia during coronary occlusion. This was illustrated by one study of 450 patients, which found that 67 percent did not have enough collateral flow to prevent myocardial ischemia produced by balloon angioplasty [9].

Quantitative radionuclide myocardial perfusion imaging with positron emission tomography has added to the understanding of blood flow regulation in collateral circulation in humans. Applying this technique to patients with single vessel coronary occlusion, blood flow to collateral-dependent myocardium was shown to increase by nearly 35 percent during atrial pacing-induced tachycardia, a condition that increases myocardial oxygen demand [23]. This degree of flow augmentation, however, was far less than that observed among normal controls. This method is not typically used for diagnostic assessment.

Endothelial release of nitric oxide [24] and activation of beta-adrenergic receptors, whose existence has been demonstrated in collateral vasculature [25], may be responsible for vasodilation of coronary collaterals. In contrast, alpha-adrenergic receptors have not been identified in isolated rings of collateral tissue [26].

Studies in animal models have shown that beta blockers induce vasoconstriction of coronary collaterals during exercise, thereby reducing collateral blood flow [27]. A similar effect can be induced by inhibitors of nitric oxide synthesis [28] and by a platelet activating factor that causes platelet aggregation and the release of the potent vasoconstrictors thromboxane A2 and serotonin [29]. These observations may have clinical implications in patients with collateral-dependent myocardium, especially during episodes of ischemia and subsequent platelet activation.

Changes in collateral vessel function over time — Increased antegrade flow through a collateralized coronary artery following successful percutaneous coronary intervention results in gradual regression of collateral function. In a study of 103 patients who underwent successful recanalization of chronically occluded coronary arteries, coronary pressure-derived collateral function dropped by 23 percent immediately following the intervention, and by another 23 percent during median five-month follow-up [30]. In this study, recruitable coronary collateral flow was best preserved among patients with larger collaterals.

One study compared temporal changes in coronary collateral function following coronary recanalization among 23 patients with acute occlusions and 74 patients with chronic occlusions [31]. Notably, recruitable coronary collateral flow was greater in successfully recanalized chronically occluded coronaries compared with acutely occluded coronaries. Furthermore, among patients who received successful recanalization of chronically (but not acutely) occluded coronaries, those who had greater collateral recruitability (as determined by the collateral pressure index) had fewer adverse cardiac events during the follow-up period.

Another study demonstrated that collateral recruitment as measured by coronary flow index improved in patients with stable coronary artery disease following four weeks of moderate or intense exercise [32].

These studies provide further evidence of a correlation between collateral vessel function and long-term clinical outcomes.

DIAGNOSIS — Coronary collaterals have been traditionally identified and assessed using selective coronary angiography; this technique offers direct visualization of collateral vessels exceeding 100 microns in diameter [33,34]. A classification system for collateral flow from patent vessels to the occluded vessel was developed by Rentrop [35]:

●Grade 0 – No visible filling of any collateral channel

●Grade 1 – Filling of the side branches of the occluded artery, with no dye reaching the epicardial segment

●Grade 2 – Partial filling of the epicardial vessel

●Grade 3 – Complete filling of the epicardial vessel by collaterals

This relatively simple angiographic method remains the mainstay for identification of the presence and extent of coronary collaterals.

Quantification of the extent of collateral flow can be performed by a number of methods, as described below. While of interest physiologically, the practical application of these methods in clinical care is not often done.

Intracoronary flow velocity, flow capacity, or pressure measurements obtained during routine catheterization and angioplasty can provide an accurate and quantitative method for assessing the collateral circulation, although these are not commonly used clinically [34,36]. (See "Clinical use of coronary artery pressure flow measurements".)

For patients undergoing percutaneous coronary intervention (PCI), the Doppler-derived or pressure-derived collateral flow index (CFI), which is defined as the ratio of flow during balloon inflation divided by resting flow, may be calculated. CFI is a useful measure for assessing the effects of various pharmacologic agents, such as adenosine and metoprolol, on coronary collateral perfusion [37]. Pressure-derived CFI was shown to predict left ventricular dilation one year following thrombolysis for acute MI [38].

Myocardial contrast echocardiography, which for this purpose involves the selective coronary injection of sonicated microbubbles, is an alternative diagnostic tool for the assessment of coronary collaterals. This technique may more precisely identify the spatial extent of perfusion via collateral vessels [39]. However, United States Food and Drug Administration (FDA) approved agents for cardiac imaging are contraindicated for intraarterial injection. (See "Contrast echocardiography: Contrast agents, safety, and imaging technique" and "Contrast echocardiography: Clinical applications".)

Myocardial contrast echocardiography and angiography appear to differ in their ability to detect the functional significance of collateral beds [33,40,41]. Surprisingly, there is a poor correlation between the presence of well-visualized collaterals at angiography and the preservation of segmental wall motion in patients with chronically occluded epicardial coronary arteries [33]. By comparison, myocardial contrast echocardiography more reliably reflects wall motion in these patients, perhaps due to more precise delineation of microvascular perfusion [40].

Myocardial contrast echocardiography has also been used to predict functional recovery after revascularization of collateralized myocardium. In one series, for example, improvement in left ventricular function correlated with successful angioplasty of the infarct-related artery [41]. Among patients in whom angioplasty was successful, those with >50 percent of the infarct bed perfused by collaterals, as demonstrated by myocardial contrast echocardiography, had superior baseline wall motion and wall motion one month following revascularization compared with those whose infarct beds were ≤50 percent perfused by collateral flow. This suggests that myocardial viability may be associated with the presence of collateral blood flow within the infarct bed.

Other noninvasive techniques can assess the collateral circulation. The severity of 99m-technetium-sestamibi tomographic perfusion defects, for example, has been shown to correlate closely with absolute collateral flow as measured by radiolabeled microspheres in animal models of coronary occlusion [42] and with collateral fractional flow reserve, an index of collateral flow in patients [43]. (See "Assessment of myocardial viability by nuclear imaging in coronary heart disease".) Spiral computerized tomography (CT) [44], magnetic resonance imaging [45], and positron emission tomography [23,46] have also been validated and may ultimately prove clinically useful in the assessment of myocardial collateralization. (See "Clinical utility of cardiovascular magnetic resonance imaging".)

CLINICAL SIGNIFICANCE — In the setting of acute MI, thrombotic occlusion of an epicardial coronary artery results in downstream myonecrosis and cell death, with resultant segmental abnormalities of regional wall motion. More myocardial injury leads to greater decrease in left ventricular ejection fraction and a worse prognosis. Pre-existing collateral vessels have the potential to reduce the severity of injury by providing blood flow to the threatened segments. Indeed, this concept has been shown during brief episodes of coronary occlusion produced by angioplasty, where the presence of recruitable collaterals reduces the severity of regional myocardial hypoperfusion in some patients, as evaluated with radionuclide myocardial perfusion imaging [47].

Specific beneficial effects observed in the presence of collateral vessels include:

●Reduced infarction size – In a canine model of electrically-induced left circumflex coronary thrombosis, infarct size as determined histochemically was smaller among dogs with contrast echo-detected collateral perfusion than among those without collaterals (4 versus 11 percent) [48]. In patients in the TIMI trial, infarct size, as determined by serum creatine kinase, was smaller among patients with collateral vessels compared to those without such vessels (20.6 versus 31.4 CK gram equivalents, p = 0.001) [49].

●Greater postinfarction ejection fraction – Smaller infarct size should translate into a greater postinfarction left ventricular ejection fraction (LVEF) [50,51]. One study found that, in patients with sufficient collaterals, left ventricular recovery after reperfused MI is primarily determined by the amount of collateral blood flow and is less dependent upon the time to reperfusion [52]. In one series of patients with occluded culprit arteries after fibrinolytic therapy, the presence of significant collateral flow (as determined angiographically) was associated with a significantly higher postinfarction LVEF (60 versus 54 percent for those with lesser collateral flow) [51]. Preservation of left ventricular function has also been observed following the late administration of fibrinolytic therapy to patients suffering from acute MI who had a collateralized infarct-related artery [53].

●Reduced risk for rupture – The presence of collaterals may reduce the risk for rupture of papillary muscle, myocardial free wall, or interventricular septum. (See "Acute myocardial infarction: Mechanical complications".)

●Decreased aneurysmal dilatation – In one series of 47 patients given fibrinolytic therapy within six hours of a first MI, the incidence of left ventricular aneurysm was 4 percent in those with successful reperfusion and, among those with unsuccessful reperfusion, was much lower in patients with significant collaterals to the infarct-related artery (10 versus 58 percent in those without collaterals) [54]. This seems consistent with the findings (as noted above) of reduced infarction size in the presence of collaterals.

PROGNOSTIC SIGNIFICANCE

Acute myocardial infarction — The absence of collateral circulation to the infarct-related artery may be an independent predictor of short-term morbidity and mortality in patients with an acute MI although association with long-term outcomes is less certain.

●Short-term outcomes – In a report of 1059 patients undergoing primary percutaneous coronary intervention (PCI) for ST elevation MI (STEMI), increased collateral flow was associated with a lower incidence of Killip class ≥2 at presentation, less need for intraaortic balloon pumping after PCI, better myocardial blush grade after intervention, and smaller enzymatic infarct size [55]. In another series of 180 patients undergoing primary PCI for anterior STEMI, the in-hospital mortality rate was substantially lower in those with demonstrable collaterals as determined by angiography (8 versus 23 percent) [56]. This difference was primarily due to a lower frequency of cardiogenic shock among the patients with pre-existing collaterals.

●Long-term outcomes – Studies are mixed as to whether collateral circulation has a long-term protective effect after an acute MI [57-61]. A smaller study with longer follow-up has shown benefit. Among 227 patients with STEMI (median follow up 797 days), the presence of extensive coronary collateral flow before primary PCI was associated with a significantly lower mortality (4.4 versus 13.8 percent) [60]. Similarly, a retrospective study of 3340 patients with STEMI treated with PCI found that patients with Rentrop class 1 and 2 collaterals had better in-hospital and five-year survival as compared with those without collateral flow (class 0). Persons with class 3 collaterals in this study did not have lower mortality, perhaps because they had a worse overall clinical profile [61].

However, an analysis from the large ACUITY trial of 5412 patients with acute coronary syndrome had collateral circulation assessed by a core laboratory using the Rentrop score and did not show benefit of collateral circulation [60]. Patients with collateral circulation had no difference in one-year major adverse cardiac events (hazard ratio [HR] 0.94 95% CI, 0.76-1.16), mortality (HR 1.03 95% CI, 0.65-1.62), MI (HR 1.07 95% CI 0.83-1.38), and unplanned target vessel revascularization (HR 0.95 95% CI 0.71-1.28) [60]. Furthermore, among patients with collateral circulation who underwent PCI, the risk of unplanned target vessel revascularization was increased (HR 2.74 95% CI 1.48-5.10).

Longer-term outcome studies with core laboratory collateral flow measurements are needed.

Chronic coronary syndrome — Patients with chronic coronary syndrome may have more favorable outcomes if they have collateral circulation. In a prior investigation, 845 patients (of whom 739 had chronic coronary syndrome and 106 patients had no obstructive coronary artery disease) underwent quantitative, coronary pressure-derived collateral measurements and were followed for a mean of five years [62]. Persons with a high collateral flow index, compared with those with a low index, had significantly higher cumulative rates of 10-year survival (88 versus 71 percent) and survival free of cardiac death (97 versus 89 percent).

The mechanism for impaired collateral formation in diabetics was evaluated in a series of 16 diabetics and 14 healthy volunteers [63]. The cellular response of monocytes to VEGF-A, an important step during collateral formation due to regional myocardial ischemia, was attenuated in the diabetic patients. The expression of vascular endothelial growth factor (VEGF) and hypoxia inducible factor (HIF-1alpha), a transcriptional activator of VEGF, was measured in ventricular biopsy specimens from patients undergoing coronary bypass surgery. Patients without diabetes had statistically significantly higher levels of VEGF and HIF-1alpha than diabetic patients, suggesting reduced expression of these angiogenesis factors [64].

THERAPEUTIC POTENTIAL — Attention has been directed toward therapies that might augment the natural supply of collaterals through myocardial neovascularization. This topic is discussed in detail separately (see "Therapeutic angiogenesis for management of refractory angina"). At present, there are no specific pharmacologic therapies that are approved for augmenting collateral flow, although there is some evidence supporting exercise training:

●Serotonin receptor antagonists may ultimately be shown to be useful given that serotonin can reduce coronary artery collateral flow by vasoconstriction, possibly via platelet activation [65]. (See "The role of platelets in coronary heart disease".)

●Angiotensin converting enzyme (ACE) inhibitors improve the endothelial dysfunction seen in atherosclerosis, suggesting that they may have a role in angiogenesis [66].

●There is no strong evidence supporting the angiogenic potential of statin therapy [67].

●Exercise training has been proposed as a means to stimulate collateral vessel formation. A prospective trial that assigned 60 subjects with clinically stable, hemodynamically significant coronary stenosis randomized participants to moderate- or high-intensity supervised exercise training for 10 hours per week over four weeks versus usual care [32]. Moderate- and high-intensity training led to 41.3 and 39.4 percent increases, respectively, in coronary collateral flow index (CFI), whereas CFI in the control group remained unchanged. Recruitment of preexisting collateral vessels was proposed as a potential mechanism underlying these observations.

SUMMARY

●Anastomotic channels known as collateral vessels can develop in the heart as an adaptation to ischemia. They serve as conduits that bridge severe stenoses or connect a territory supplied by one epicardial coronary artery with that of another. They are diagnosed by standard coronary angiographic observation. (See 'Introduction' above.)

●Two classes of collateral vessels have been recognized: capillary size collaterals, in which smooth muscle cells are absent, may be observed throughout the myocardium, although they have a predilection for the subendocardium; and larger, muscular collaterals, which develop from pre-existing arterioles, are typically located epicardially. (See 'Introduction' above.)

●Coronary collaterals can provide substantial blood flow to resting myocardium, but are generally insufficient during increased demand (exercise). (See 'Physiology' above.)

●Collaterals may: reduce infarct size after myocardial infarction (MI); reduce post-MI complications such as rupture of a papillary muscle, myocardial free wall, or interventricular septum; and reduce aneurysmal dilatation. (See 'Clinical significance' above.)

●The absence of collateral circulation to the infarct-related artery may be an independent predictor of mortality in patients with an acute MI, although evidence may be mixed. (See 'Prognostic significance' above.)

●Patients with diabetes have less well-developed collateral vessels. (See 'Role of diabetes' above.)