Systolic Compression of Epicardial Coronary and Intramural Arteries

Abstract

It has been suggested that systolic compression of epicardial coronary arteries is an important cause of myocardial ischemia and sudden death in children with hypertrophic cardiomyopathy. We examined the associations between sudden death, systolic coronary compression of intra- and epicardial arteries, myocardial perfusion abnormalities, and severity of hypertrophy in children with hypertrophic cardiomyopathy.

We reviewed the angiograms from 57 children with hypertrophic cardiomyopathy for the presence of coronary and septal artery compression; coronary compression was present in 23 (40%). The left anterior descending artery was most often affected, and multiple sites were found in 4 children. Myocardial perfusion abnormalities were more frequently present in children with coronary compression than in those without (94% vs 47%, P = 0.002). Coronary compression was also associated with more severe septal hypertrophy and greater left ventricular outflow gradient. Septal branch compression was present in 65% of the children and was significantly associated with coronary compression, severity of septal hypertrophy, and outflow obstruction. Multivariate analysis showed that septal thickness and septal branch compression, but not coronary compression, were independent predictors of perfusion abnormalities. Coronary compression was not associated with symptom severity, ventricular tachycardia, or a worse prognosis.

We conclude that compression of coronary arteries and their septal branches is common in children with hypertrophic cardiomyopathy and is related to the magnitude of left ventricular hypertrophy. Our findings suggest that coronary compression does not make an important contribution to myocardial ischemia in hypertrophic cardiomyopathy; however, left ventricular hypertrophy and compression of intramural arteries may contribute significantly. (Tex Heart Inst J 2002;29:290–8)

Hypertrophic cardiomyopathy (HCM) is a genetic disease with a prevalence of 1 in 500–1,000 in the general U.S. population. ¹ In most cases, HCM is inherited in an autosomal dominant fashion and is characterized by left ventricular (LV) hypertrophy in the absence of another cause for the increased cardiac mass. ² Hypertrophic cardiomyopathy is often associated with disabling symptoms, arrhythmias, and sudden cardiac death. ^2–4 Several mechanisms, including myocardial ischemia, are thought to be important contributors to sudden death. ^3–11

Patients who have hypertrophic cardiomyopathy often have regional myocardial perfusion abnormalities, as shown by exercise thallium scintigraphy and positron emission tomography, and by net production of myocardial lactate. ^11–17 Sudden death is most common in pediatric populations (Fig. 1), and exercise thallium abnormalities are associated with increased risk. ^17–22 Several mechanisms are potentially responsible for ischemia in HCM, including the obstruction of epicardial and intramyocardial arteries by systolic compression. ^22–24

Fig. 1 Survival estimates in children diagnosed with hypertrophic cardiomyopathy in several studies. ^18–22 On the basis of these studies, the annual mortality rate from cardiac deaths in children is estimated at approximately 4% per year. It has been reported that in the absence of coronary bridging, survival is significantly improved. ²²

Yetman and colleagues ²² suggested that systolic compression of epicardial arteries by an overlying myocardial “bridge” in children with HCM persists into diastole, which results in coronary insufficiency and myocardial ischemia. In their study, compression was associated with reduced survival rates, ventricular tachycardia, chest pain, and impaired exercise performance. The magnitude of LV hypertrophy and the severity of obstruction were not associated with compression. In contrast, the prognosis in children without coronary compression was excellent (Fig. 1). Those authors suggested that surgical division of the bridge might improve prognosis. ²²

To validate the findings of Yetman's group, ²² we investigated relations between sudden death, systolic coronary compression of intra- and epicardial arteries, myocardial perfusion abnormalities, and severity of hypertrophy in a larger group of children with HCM who had been studied for a longer period and whose management represented more current practice. Our findings have been at odds with those of Yetman and colleagues, and have led to active discussion. ^25–27 Herein, we present the findings of our study and a literature review of coronary compression.

Methods

We studied children with HCM who had selective coronary angiograms available for review at the National Institutes of Health (NIH) from January 1989 through April 1999. Hypertrophic cardiomyopathy was defined echocardiographically as a hypertrophied, nondilated left ventricle in the absence of another cause for the increased cardiac mass. Cardiac events were defined as sudden death or resuscitated cardiac arrest. Survival was determined according to the age of the children, rather than years from diagnosis, because most cardiac events occur in the 2nd and 3rd decades of life. ²⁸

Right- and left-heart hemodynamic measurements and selective coronary angiograms were obtained during catheterization. All coronary segments showing evidence of compression were evaluated quantitatively to determine the severity of compression. Coronary compression was defined as a maximum systolic compression ≥50%. ²² We found identification of exact boundaries between diastole and systole from selective coronary angiograms to be inaccurate. It follows that angiographic determination of diastolic persistence of compression is not reliable; therefore, this variable was not measured. Septal perforator branches of the left anterior descending coronary artery (LAD) were considered compressed when they were angiographically obliterated in systole. Treadmill exercise tests had been performed with use of the Bruce protocol. Results of exercise thallium scintigraphy were evaluated when available, and images were analyzed qualitatively in the anterior, apical, inferior, septal, and lateral regions. We characterized a region of reduced thallium uptake as reversible or fixed, and determined the presence or absence of apparent cavity dilatation.

Holter recordings were analyzed for the presence or absence of ventricular tachycardia. The QT and QTc intervals and QTc dispersion were measured from the 12-lead electrocardiogram. Electrophysiology studies were performed in a subset of the children with use of previously described methods. ⁶

Two-sample data were compared by use of the Student's t-test. Cardiac survival rates for patients with and without coronary compression were determined by Kaplan-Meier estimates and were compared by the log rank test. Multivariate logistic regression analysis was used to determine the independent contributions of LV wall thickness, LV outflow obstruction, septal compression, and bridging, to thallium perfusion abnormalities. A P value of <0.05 was considered statistically significant.

Results

Fifty-seven children with HCM who were admitted for evaluation to the NIH underwent selective coronary angiograms. These children had a high prevalence of risk factors associated with sudden death (Table I).

TABLE I. Clinical Findings in the 57 Children with HCM: Prevalence of Established Risk Factors

Angiographic Findings.

Using definitions of coronary compression identical to those of Yetman and colleagues ²² (systolic diameter ≤50% of diastolic diameter), we found varying degrees of coronary compression that occurred more frequently than had previously been reported. Coronary compression was detected at sites other than the mid LAD, which was the only site found to be compressed in Yetman's study. ²² Coronary compression (Fig. 2) affected 28 coronary segments in 23 of the 57 children (40%). Multiple coronary compression sites were identified in 4 children (7%): 2 coronary segments in 3 children, and 3 segments in 1 child. Compression was of the mid LAD in 13 of the 28 segments (46%), the proximal LAD in 1 (4%), the distal LAD in 2 (7%), a diagonal branch in 5 (18%), an obtuse marginal branch in 4 (14%), and the posterior descending branch of the right coronary artery in 3 (11%). The mean length of the compressed segment was 15 ± 7 (SD) mm (range, 6–39 mm), with a mean systolic narrowing of 76% ± 18%. The severity of maximum compression was between 75% and 100% in 13 of the segments (46%), and complete systolic occlusion was seen in 8 (29%). Coronary compression of severity less than that of the current definition was detected in several more children.

Fig. 2 Figure shows coronary angiograms in 3 children with hypertrophic cardiomyopathy. Diastolic frames are on the left and systolic frames on the right. Images are from A) a 6-year old boy with compression of the mid LAD and its septal branches, B) a 13-year old girl with compression of the proximal posterior descending branch of the right coronary artery and its septal branches, and C) an 8-year-old boy in whom there is complete compression of the septal branches of the LAD in the absence of LAD compression.

The broad arrows indicate sites of bridging of epicardial arteries, and the thinner arrows indicate compression of septal branches.

Compression of the intramural arteries of the septum (septal perforators) was detected in 37 of the 57 children (65%). Such compression was more frequent in children with coronary compression, affecting all 23 with a compressed epicardial artery but only 14 of the 34 without coronary compression (P < 0.001).

Clinical Features.

The ages at diagnosis of HCM and cardiac catheterization were similar in children with and without compression (Table II). The symptom status at presentation (chest pain and impaired consciousness, including cardiac arrest) was also similar in the 2 groups.

TABLE II. Clinical Findings in Children Who Had HCM with or without (A) Coronary and (B) Septal Compression

Cardiac events (defined as sudden death [n = 2] and cardiac arrest [n = 4]) occurred in 2 children with a compressed coronary artery and in 4 without. The cumulative survival rate at 20 years of age in children with coronary compression was 85% ± 10%; in those without, it was 82% ± 8% (P = 0.9). In addition, LV systolic dysfunction occurred in 3 children without coronary compression, one of whom experienced the dysfunction after transmural myocardial infarction (none occurred in children with coronary compression). One child in each group underwent cardiac transplantation for severe symptoms and exercise limitation.

Echocardiography.

Echocardiography showed left ventricular wall thickness at the proximal interventricular septum to be significantly greater in the children with coronary compression than in those without (Table II). The mean septal thickness in children with coronary compression was 28 ± 8 mm, compared with 19 ± 8 mm in those without (P < 0.001), and 26 ± 9 mm in children with septal compression compared with 17 ± 5 mm in those without (P < 0.001). Children with coronary or septal compression also had higher LV outflow gradients than did children with no compression (Table II).

Hemodynamic Variables.

The mean resting LV outflow gradient was greater in patients with either coronary or septal compression than in those without compression (Table II).

Exercise Thallium Scintigraphy.

Reversible myocardial abnormalities were present in 31 of the 48 children (65%) who underwent exercise thallium scintigraphy (Fig. 3). Abnormalities in myocardial thallium uptake were much more frequent in children with coronary and septal compression (Table II). Left ventricular hypertrophy was greater in patients with reversible thallium abnormalities than in those with no reversible abnormalities (septum, 26 ± 9 mm vs 16 ± 3 mm, respectively [P < 0.001]), and LV outflow gradients at cardiac catheterization were higher (26 ± 33 mmHg vs 7 ± 12 mmHg, respectively [P < 0.05]).

Fig. 3 Myocardial thallium uptake is shown during exercise (upper row), and after rest and re-injection (below). Reversible apical, subendocardial, posterior, and septal defects can be seen in the results from this child with mid LAD compression. We suggest that these defects cannot all be due to LAD compression.

The distribution of thallium abnormalities was often unrelated to the coronary artery compressed, and was similar in children with LAD compression to that in children with compression of another coronary artery. Stress-induced apparent LV cavity dilatation, thought to be caused by subendocardial ischemia and not easily attributed to occlusion of a single coronary artery, was present in over a third of the children. Almost half of the children without coronary compression had reversible abnormalities. Multiple regression analysis identified LV septal thickness and septal perforator compression, and not LV obstruction or coronary compression, as independent predictors of reversible myocardial thallium uptake abnormalities (Table III).

TABLE III. Relation of Thallium Perfusion Abnormalities to Clinical Parameters: Logistic Regression Model for Predicting Thallium Perfusion Abnormalities

These results show that factors other than coronary compression are important in causing perfusion abnormalities, and that compression of septal perforators may contribute to myocardial perfusion abnormalities.

Exercise and Electrophysiology.

There were no significant differences between children with and without coronary or septal compression with regard to mean exercise duration and resting and peak systolic blood pressure. Nor were the mean QT and QTc intervals or the QTc dispersion significantly different in those children. Nonsustained ventricular tachycardia was recorded in 17% of the children during ambulatory electrocardiographic monitoring. Electrophysiology studies were performed in 36 of the 57 children (63%); sustained ventricular tachycardia was induced in 28% of the patients. Differences between ventricular tachycardia on Holter monitoring and ventricular tachycardia induced at electrophysiology study in children with or without coronary compression were not statistically significant.

Discussion

Myocardial Ischemia in HCM.

Evidence for myocardial ischemia in HCM is provided by stress-induced 1) chest pain and electrocardiographic changes; 2) reversible myocardial perfusion defects at thallium scintigraphy; 3) limitations in coronary flow reserve; and 4) anaerobic metabolism in patients with HCM who have a history of chest pain. ^11–17 Myocardial ischemia results from true hypoxia (mismatch of flow supply to myocardial demand), or from abnormalities in production and use of high-energy phosphates. ^29–31

Coronary flow reserve limitation in HCM, like thallium perfusion defects in HCM, is often most striking in the subendocardium and does not typically show single coronary artery distribution. ¹¹ Coronary flow reserve is lowest in HCM patients with microvascular abnormalities detected after biopsy. ^14–17,24 Myocardial perfusion may also be limited by elevated LV diastolic pressure, abbreviated diastolic intervals, and systolic arterial compression, as illustrated in Figure 4. ^23,32

Fig. 4 Simultaneous recordings of arterial pressure and left ventricular (LV) pressure in a young patient with severe LV hypertrophy, no outflow obstruction, and septal artery compression, but no coronary compression. A) The gradient between the arterial and LV pressure in diastole, as indicated by the shaded area, drives myocardial perfusion. B) After an infusion of isoproterenol, the LV diastolic pressures have risen dramatically, the aortic diastolic pressure has fallen, and the diastolic period is abbreviated. Myocardial perfusion pressure and duration are severely diminished, as shown by the dramatic reduction in the shaded area.

Definitions and Prevalence of Coronary Compression.

Coronary angiography reveals systolic compression of epicardial coronary arteries in a proportion of otherwise normal hearts. A segment of an epicardial artery is crossed by a myocardial band or “bridge” that contracts and compresses the underlying artery in systole. ^33–38 In this article, bridging refers to the muscle bridge, and the angiographic findings are termed coronary compression.

In populations of adults without cardiomyopathy, systolic coronary compression is seen at angiography in 0.5% to 12%. The compression most often affects the mid LAD. ^33–36 Muscle bridges are found much more frequently at autopsy or aortocoronary bypass (5%–86%). ³⁷ These bridges most often cross the mid LAD but have been found over most epicardial arteries. Therefore, anatomic muscular bridges do not always lead to coronary compression. An intramyocardial segment of an epicardial artery, most often of the mid LAD, should possibly be considered a (frequent) coronary variant rather than a coronary anomaly. Pharmacologic agents such as nitroglycerin and isoproterenol increase the angiographic severity and incidence of compression, with detection rates as high as 40% in patients undergoing angiography. ³⁸

In hypertrophic hearts, coronary compression is more common; myocardial hyperdynamic and hypertrophic adaptations may extend to previously “silent” bridges that become compressive. Coronary compression is found in 30% to 80% of adults who have HCM or aortic stenosis. ^35,39,40 Studies in adults with HCM have described no relationship between coronary compression and ischemia. ¹¹

Coronary Compression and Myocardial Ischemia.

It has previously been suggested that coronary compression 1) results in coronary insufficiency and myocardial ischemia and 2) might provide protection against the development of coronary atheroma. The latter, the subject of several reviews, is not considered here. ^35,37

Coronary flow occurs predominantly in diastole; only 15% of myocardial flow occurs during systole. Systolic compression may result in ischemia if the contribution of systolic flow becomes critical, or if a sufficient severity of compression persists into diastole. Compression may result in elliptical narrowing of the coronary artery and not the concentric narrowing typical of coronary disease. Relations between the minimum diameter and luminal diameter in an ellipse are different from those of a circle: a 50% concentric decrease in arterial diameter decreases luminal area by 75%; the area will only be reduced by about 30% if the artery is distorted into an ellipse. ³⁵ To reduce flow substantially, diastolic coronary artery compression must remain at maximum severity (Fig. 5).

Fig. 5 A) The reduction of luminal area of a compressed coronary artery follows elliptical geometry rather than concentric geometry. B) The minimum systolic diameter of a compressed coronary artery results in substantially less luminal area reduction than does concentric narrowing.

Multiple case reports implicate compression as the cause for ischemic syndromes in patients with otherwise normal coronary arteries. ^41–46 Both coronary stenting and surgery to “de-roof” or bypass the bridge have been associated with symptomatic improvement in case reports and small uncontrolled series. ^47–51 Noble and associates ⁴¹ found increased myocardial lactate production after rapid atrial pacing in symptomatic patients with coronary compression; however, approximately 25% of the patients may have had myocardial hypertrophy, which is itself associated with myocardial lactate production and symptoms. ¹¹ Quantitative angiography and coronary Doppler flow studies detect some diastolic persistence of compression, but its hemodynamic significance has not been established. ^52–54 No controlled study has associated increases in morbidity or mortality with coronary compression, and the 5-year survival rate is similar to that of an unselected population. ³⁵ In several of the case reports describing ischemic syndromes (chest pain, myocardial infarction, sudden death, or ventricular tachycardia), patients also had myocardial hypertrophy, which might have been due to HCM. The literature remains equivocal.

Finally, in cases of HCM and other hypertrophic conditions, angiographic obliteration in systole of the septal perforator branches of the LAD and posterior descending artery (often called septal blanching) occurs commonly. This obliteration is present in about 70% of patients with aortic stenosis or HCM and has previously been associated with the degree of LV hypertrophy and with thallium perfusion abnormalities. ^55–57 Doppler studies show the reversal of coronary flow in systole, an abnormality that correlates with the severity of septal compression. ^58–60 Therefore, in the presence of systolic septal compression, epicardial compression may not be of additional importance in limiting coronary flow, in which case, surgical release of the compressed coronary will not improve myocardial perfusion.

Conclusions

There are several mechanisms for sudden death in HCM other than ischemia, and ischemia is multifactorial in origin. This study finds no evidence that epicardial compression is an important cause of myocardial perfusion abnormalities in children who have HCM. Previously described associations may have been confounded by associated features, such as the magnitude of myocardial hypertrophy. Rather, our data support the hypothesis that LV hypertrophy and septal artery compression are important determinants of the severity of perfusion abnormalities. It remains possible that, in a few patients, a severe degree of coronary compression contributes to ischemia. However, in the children studied here, there was no evidence that this contribution was significant, even in those with the most severe compression. Currently, there is little evidence to suggest that relieving coronary compression in children with HCM will improve their symptoms or prognosis.