- Email: [email protected]
Vascular endothelial dysfunction is associated with reversible myocardial perfusion defects in the absence of obstructive coronary artery disease
ORIGINAL ARTICLES Vascular endothelial dysfunction is associated with reversible myocardial perfusion defects in the absence of obstructive coronary artery disease Prem Soman, MD, PhD, MRCP(UK), Devang M. Dave, BS, James E. Udelson, MD, FACC, Hui Han, MD, Husam Z. Ouda, MD, Ayan R. Patel, MD, FACC, Richard H. Karas, MD, PhD, FACC, and Jeffrey T. Kuvin, MD, FACC Background. The purpose of this study was to investigate whether endothelial dysfunction contributes to abnormal myocardial perfusion imaging (MPI) observed in patients without obstructive coronary artery disease (CAD). It is unclear whether reversible MPI defects detected in the absence of obstructive CAD represent underlying vascular pathology or are false-positive MPI results. Recent evidence suggests that coronary endothelial dysfunction might play a role in the pathogenesis of these defects. Methods and Results. We prospectively recruited 36 patients with chest discomfort, reversible abnormalities on MPI, and nonobstructive or absent CAD (stenosis <50% on coronary angiography). The control group (n ⴝ 55) consisted of patients with chest discomfort and similar cardiac risk factors but with normal MPI findings. Vascular endothelial function was assessed in the brachial artery by ultrasound as the response to hyperemia and reported as percent flow-mediated dilation (FMD). Response to sublingual nitroglycerin was used as an indicator of endothelium-independent vasodilation. The patients with abnormal MPI findings and nonobstructive CAD had a significantly lower FMD (9.0% ⴞ 7.2%), indicating endothelial dysfunction, compared with those with similar risk factors and normal MPI findings (12% ⴞ 5.2%) (P ⴝ .03). Baseline brachial artery size and endothelium-independent dilation were similar between groups. On multivariate analysis, only endothelial dysfunction was predictive of reversible MPI defects. Conclusions. Patients with chest pain and reversible MPI defects but without obstructive CAD have lower FMD indicative of endothelial dysfunction, as compared with similar patients with normal MPI findings. The possibility of a causal link between reversible MPI defects and endothelial dysfunction needs further exploration. (J Nucl Cardiol 2006;13:756-60.) Key Words: Endothelial function • myocardial perfusion • coronary artery disease • brachial artery
See related article, p. 747 Reversible myocardial perfusion abnormalities can occur in the absence of obstructive coronary artery From the Division of Cardiology, Tufts-New England Medical Center, Boston, Mass. Dr Soman was funded by the Herbert J. Levine Foundation Fellowship in Cardiovascular Medicine at Tufts-New England Medical Center and the Kos Pharmaceutical Fellowship in Preventive Cardiology during the course of this project. Received for publication May 25, 2006; final revision accepted Aug 6, 2006. Reprint requests: Prem Soman, MD, PhD, Presbyterian Hospital, A 429, Scaife Hall, 200 Lothrop St, Pittsburgh, PA 15213; somanp@ upmc.edu. 1071-3581/$32.00 Copyright © 2006 by the American Society of Nuclear Cardiology. doi:10.1016/j.nuclcard.2006.08.018 756
disease (CAD) and have been ascribed to attenuation artifacts, false-negative coronary angiography (microvascular disease) findings, and vasospasm.1-3 Recently, coronary endothelial dysfunction has been postulated as a potential mechanism of perfusion abnormalities seen on single photon emission computed tomography (SPECT) imaging.4,5 Evaluating coronary endothelial function, however, is an invasive process and is not easily applicable to large populations. Peripheral endothelial function testing, on the other hand, may be performed noninvasively and has been shown to correlate well with coronary artery endothelial function.6,7 To determine the relationship between peripheral vascular endothelial function and reversible myocardial perfusion abnormalities, we compared brachial artery endothelial function, determined by flow-mediated dilation (FMD), in patients with chest discomfort, reversible perfusion defects on myocardial perfusion imaging (MPI),
Journal of Nuclear Cardiology Volume 13, Number 6;756-60
and angiographically nonobstructed coronary arteries with a matched cohort of subjects with normal MPI findings. METHODS Study Design and Patient Population Consecutive patients aged 18 years or older who were referred for coronary angiography to the cardiac catheterization laboratory at Tufts-New England Medical Center, Boston, Mass, between July 2003 and June 2004, on the basis of chest discomfort and a reversible MPI defect, and were found to have absent or nonobstructive epicardial CAD (defined as ⬍50% diameter stenosis) on coronary angiography were prospectively screened for study eligibility, and eligible patients were recruited if they provided informed consent for study participation. Exclusion criteria included left ventricular dysfunction (ejection fraction ⬍45% on gated SPECT), significant valvular heart disease, and unstable clinical states. The control population was drawn from an existing database of patients with chest pain and normal MPI findings, by matching with the control group by age, gender, and number of cardiac risk factors including smoking, hypertension, hyperlipidemia, diabetes mellitus, and family history of premature CAD (first-degree male and female relatives aged ⬍55 years and ⬍65 years, respectively). All patients from the database who could be matched to a control patient via the study criteria were included. The protocol was approved by the institutional review board. MPI was performed at the referring physician’s office or institution, according to standard local practices. Either exercise or pharmacologic stress testing was performed. All scans were performed by use of gated SPECT technology and technetium-labeled radiotracers (either Tc-99m sestamibi or Tc-99m tetrofosmin) and were not reread specifically for the purpose of this study. The number and location of reversible perfusion defects were determined from the report.
Endothelial Function Testing Brachial artery ultrasound (BAUS) was performed according to our previously published method.6 In patients in the study group BAUS was performed at least 2 hours after coronary angiography and in the fasting state. Longitudinal brachial artery images were obtained with a high-resolution (10-MHz) linear-array vascular transducer (Philips Sonos 5500; Philips Medical Systems, Andover, Mass). Subjects were studied under quiet conditions while in the supine position in a temperature-controlled room. After a 10-minute equilibrium period, baseline 2-dimensional images of the right brachial artery were obtained approximately 2 cm above the antecubital fossa. A blood pressure cuff (Hokanson, Bellevue, Wash) placed proximal to the imaging transducer on the upper arm was inflated to suprasystolic pressure for exactly 5 minutes. The brachial artery was imaged continuously for 1 minute after release of occlusion, and reactive hyperemia was confirmed by
Soman et al Endothelial dysfunction and perfusion defects
757
pulse-wave Doppler interrogation. Repeat baseline resting brachial artery dimensions were obtained 10 minutes later. Subjects were then given sublingual nitroglycerin (400-g tablet), and the brachial artery was imaged for 5 minutes thereafter. Endothelium-dependent vasomotion was determined by the maximal brachial artery diameter after exactly 60 seconds of reactive hyperemia compared with the baseline diameter and was expressed as percent FMD. Endothelium-independent vasodilatation was defined as the maximal brachial artery diameter after administration of nitroglycerin compared with the baseline vessel diameter and was expressed as percent nitroglycerin-mediated diameter (NMD). Brachial artery measurements were performed with ultrasonic calipers, and maximal end-diastolic brachial arterial diameter was calculated within a 5-cm segment of the vessel as the mean of 5 evenly spaced measurements of the distance from the near to the far arterial wall along a line perpendicular to the long axis of the artery. Mean intraobserver variability and interobserver variability of brachial reactivity measurements in our laboratory in normal volunteers were 1.9% and 2.8%, respectively.6
Statistical Analysis Continuous variables are described as mean ⫾ SD and categoric variables as frequencies (percent). We used 2 statistics and t tests to detect differences between continuous and categoric variables, respectively. A multivariate regression analysis was used to determine predictors of an abnormal MPI scan. Those variables predictive of an abnormal MPI scan on univariate analysis were included in the multivariate model. A Pearson correlation coefficient was determined between every pair of predictor variables to assess for multicollinearity in the data. Significant correlation was considered to be present between variables if R ⬎ 0.9. SPSS (version 13.0; SPSS Inc, Chicago, Ill) was used for all statistical computations. P ⱕ .05 was considered significant for all analyses.
RESULTS Subject characteristics are shown in Table 1. The study group consisted of 36 patients, aged 56 ⫾ 10 years, of whom 18 (50%) were women. In 20 patients the chest pain was typical of angina.8 The distribution of the number of cardiac risk factors was as follows: 0 in 1 patient, 1 in 9 patients, 2 in 11 patients, 3 in 11 patients, and 4 in 4 patients. Stress testing was performed with treadmill exercise in 30 patients and pharmacologic agents in 6. The reversible perfusion abnormality was confined to the left anterior descending artery territory in 18 patients and the right coronary or left circumflex coronary artery territory in 14 patients, and it involved multiple vascular territories in 4 patients. The control group consisted of 55 patients, aged 55 ⫾ 11 years, of whom 27 (49%) were women (Table 1). The distribution of cardiac risk factors was not different from that in the study group. The number of risk factors
758
Soman et al Endothelial dysfunction and perfusion defects
Journal of Nuclear Cardiology November/December 2006
Table 1. Subject characteristics
No. of patients Age (y) Male gender [n (%)] Hypertension [n (%)] Hyperlipidemia [n (%)] Diabetes mellitus [n (%)] Smoking [n (%)] Family history of CAD [n (%)]
Study group
Control group
36 56 ⫾ 10 18 (50) 21 (58) 24 (67) 5 (14) 13 (35) 20 (56)
55 55 ⫾ 11 27 (49) 21 (38) 27 (49) 8 (15) 27 (49) 25 (46)
P ⫽ NS for all.
Figure 1. Scatter plot of individual percent FMD values of study and control groups with mean and SD.
Table 2. Endothelial function testing results
Baseline brachial size (mm) % FMD % NMD*
Study group
Control group
P value
3.7 ⫾ 0.7
3.6 ⫾ 0.7
.96
9.0 ⫾ 7.2 18.1 ⫾ 7.1
12.0 ⫾ 5.2 21.1 ⫾ 7.9
Table 3. Multivariate analysis for prediction of reversible SPECT abnormalities
Variable .03 .17
*Available in all study patients and 47 patients in control group.
was as follows: 0 in 4 patients, 1 in 13 patients, 2 in 19 patients, 3 in 9 patients, and 4 in 2 patients. There were no baseline characteristic differences between the study group and the control group. The results of BAUS testing are shown in Table 2. The study group had a significantly lower mean percent FMD (9.0% ⫾ 7.2%) compared with the control group (12.1% ⫾ 5.4%) (P ⫽ .03). There was no difference in the mean percent FMD between the group that underwent treadmill exercise testing and the group that underwent pharmacologic stress testing (9.2% ⫾ 7.5% and 7.8% ⫾ 5.4%, respectively; P ⫽ not significant [NS]). A scatter plot of the individual FMD values in both groups is shown in Figure 1. Sixteen patients in the study group had percent FMD values below 1 SD of the mean percent FMD of the control group (6.8%). The baseline brachial artery size was similar between the study and control groups (3.7 ⫾ 0.7 mm and 3.6 ⫾ 0.7 mm, respectively; P ⫽ NS). Similarly, percent NMD was also not significantly different between groups (18.1% ⫾ 7.1% and 21.1% ⫾ 7.9%, respectively; P ⫽ NS). Results of the multivariate logistic regression analysis are shown in Table 3. When the clinically significant predictor variables of age, gender, number of cardiac risk factors, and percent FMD were introduced into a multiple regression model, percent FMD was the only variable that was independently predictive of an abnormal
% FMD Age Male gender Hypertension Hyperlipidemia Diabetes mellitus Smoking Family history of CAD
Odds Confidence P ratio interval value 0.90 1.01 1.87 2.34 1.67 1.15 0.46 1.33
0.82–0.99 0.96–1.06 0.66–5.25 0.85–6.46 0.58–4.82 0.29–4.54 0.20–1.34 0.40–3.66
.03 .75 .24 .10 .35 .84 .16 .58
SPECT result (odds ratio, 0.9; 95% confidence interval, 0.82-0.99). DISCUSSION Chest pain without angiographic obstructive CAD is a well-defined clinical entity.1-3 When associated with electrocardiographic abnormalities indicative of myocardial ischemia on treadmill exercise testing, the constellation has been termed syndrome X.9 Myocardial perfusion abnormalities have been demonstrated in patients with angina and nonobstructed coronary arteries by use of radionuclide studies, and a reduction in coronary flow reserve has been demonstrated in invasive studies.10 Magnetic resonance imaging has demonstrated hypoperfusion of the subendocardial region during adenosine stress in patients with angina and normal coronary arteries.11 However, the pathogenesis of myocardial perfusion abnormalities in the absence of obstructive epicardial coronary disease and the vascular level at
Journal of Nuclear Cardiology Volume 13, Number 6;756-60
which these pathogenetic mechanisms operate remain a matter of contention.9 Whereas adenosine is the primary mediator of coronary arteriolar dilatation mediated by metabolic demand, the increased shear stress caused by adenosine-induced hyperemia results in endothelial cell release of nitric oxide, with further dilatation of the epicardial coronaries and microvasculature.12 Thus endothelial dysfunction might limit the full ability of the coronary vessels to dilate. The exploration of an association between endothelial dysfunction and myocardial perfusion abnormalities has been somewhat limited by the invasive nature of direct coronary endothelial function testing. Nevertheless, a few small studies have attempted to investigate this concept. In 27 patients with nonobstructed epicardial coronary arteries, Zeiher et al5 showed that those with exercise-induced myocardial perfusion abnormalities had blunted endothelium-mediated increases in coronary flow reserve, as tested by subselective coronary acetylcholine infusion, compared with patients with normal exercise myocardial perfusion. The coronary flow reserve response to the endotheliumindependent vasodilator papaverine was comparable in both groups. Hasdai et al4 demonstrated a decrease in the left anterior descending coronary artery diameter measured by quantitative coronary angiography and coronary flow reserve measured by Doppler flow wire in patients in whom a transient perfusion abnormality developed on Tc-99m sestamibi SPECT imaging during selective infusion of acetylcholine. Masoli et al13 compared the vasomotor response of the epicardial coronary artery to acetylcholine, as measured by quantitative coronary angiography, and simultaneous myocardial perfusion on planar thallium 201 imaging in 23 coronary vascular territories in 11 patients with suspected CAD. Perfusion abnormalities developed in all patients with a 20% reduction or greater in the coronary artery diameter in response to acetylcholine in vascular territories with absent, intermediate (50%-69%), and significant (ⱖ70%) epicardial stenosis. No patient with a normal vasomotor response to acetylcholine had a perfusion abnormality develop. Finally, Fujita et al14 showed in 12 patients with angina and normal epicardial coronary arteries that the administration of L-arginine improved exercise time and the extent and severity of abnormalities on Tl-201 MPI. This study suggests that an improvement in nitric oxide availability, via arginine, improves vascular function and myocardial perfusion. Thus these important, albeit small, clinical studies have established that abnormal vasomotor responses in the coronary microvasculature resulting from endothelial dysfunction may cause myocardial perfusion abnormalities in the absence of significant epicardial coronary stenosis. The results of our study demonstrate that impaired peripheral vascular endothelial function, which has previously been shown to correlate with coronary endothelial
Soman et al Endothelial dysfunction and perfusion defects
759
function,6,7 is associated with reversible MPI defects in the absence of obstructive epicardial CAD. This finding adds to the existing data set linking coronary endothelial dysfunction to myocardial perfusion abnormalities by demonstrating this association noninvasively and thus sets the stage for further investigation of the relationship between endothelial dysfunction and myocardial ischemia. Furthermore, in this study brachial FMD was the only variable predictive of an abnormal MPI scan in a model that included traditional cardiovascular risk markers. The presence of comparable endothelium-independent vasomotion in patients with myocardial perfusion abnormalities and those without them suggests that the pathogenesis of these perfusion abnormalities is probably not related to a smooth muscle mechanism. These findings are concordant with those of Masci et al,15 who reported lower brachial FMD in syndrome X patients with myocardial perfusion abnormalities, as compared with those with normal myocardial perfusion. Another study has demonstrated impaired brachial FMD in patients with abnormal myocardial perfusion in more heterogeneous populations.6 A significant methodologic study limitation that must be noted is the fact that image interpretation was performed by multiple referring physicians and was not standardized at a central core laboratory. Quantitative assessment of perfusion defect extent and severity was not performed. However, all patients had chest pain, and coronary angiography was considered warranted by the MPI abnormality. Although the hypothesis of a causal relationship between endothelial dysfunction and the pathogenesis of myocardial perfusion is biologically plausible, several questions remain unanswered. For example, how does endothelial dysfunction, presumably a diffuse, systemic process, produce regional abnormalities on MPI? Are these perfusion abnormalities indicative of true myocardial ischemia, related to the chest pain syndrome in these patients? Is there a therapeutic implication to this finding? Answers to these and other questions can only be derived from the institution of large, prospective clinical trials designed to determine outcome data in patients with endothelial dysfunction and nonobstructed coronary arteries, as well as the results of repeat MPI after treatment of endothelial dysfunction. The establishment of noninvasive testing of brachial FMD as an accurate surrogate marker of coronary endothelial function certainly improves the feasibility of performing such trials.
Acknowledgment The authors have indicated they have no financial conflicts of interest.
760
Soman et al Endothelial dysfunction and perfusion defects
References 1. Kemp HG, Elliott WC, Gorlin R. The anginal syndrome with normal coronary arteriography. Trans Assoc Am Physicians 1967; 80:59-70. 2. Likoff W, Segal BL, Kasparian H. Paradox of normal selective coronary arteriograms in patients considered to have unmistakable coronary heart disease. N Engl J Med 1967;276:1063-6. 3. Egashira K, Inou T, Hirooka Y, Yamada A, Urabe Y, Takeshita A. Evidence of impaired endothelium-dependent coronary vasodilatation in patients with angina pectoris and normal coronary angiograms. N Engl J Med 1993;328:1659-64. 4. Hasdai D, Gibbons RJ, Holmes DR Jr, Higano ST, Lerman A. Coronary endothelial dysfunction in humans is associated with myocardial perfusion defects. Circulation 1997;96:3390-5. 5. Zeiher AM, Krause T, Schachinger V, Minners J, Moser E. Impaired endothelium-dependent vasodilation of coronary resistance vessels is associated with exercise-induced myocardial ischemia. Circulation 1995;91:2345-52. 6. Kuvin JT, Patel AR, Sliney KA, Pandian NG, Rand WM, Udelson JE, et al. Peripheral vascular endothelial function testing as a noninvasive indicator of coronary artery disease. J Am Coll Cardiol 2001;38:1843-9. 7. Teragawa H, Ueda K, Matsuda K, Masashi K, Higashi Y, Oshima T, et al. Relationship between endothelial function in the coronary and brachial arteries. Clin Cardiol 2005;28:460-6.
Journal of Nuclear Cardiology November/December 2006
8. Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary-artery disease. N Engl J Med 1979;300:1350-8. 9. Cannon RO, Camici PG, Epstein SE. Pathophysiological dilemma of syndrome X. Circulation 1992;85:883-92. 10. Kaski JC, Aldama G, Cosin-Sales J. Cardiac syndrome X. Diagnosis, pathogenesis and management. Am J Cardiovasc Drugs 2004;4:179-94. 11. Panting JR, Gatehouse PD, Yang GZ, Grothues F, Firmin DN, Collins P, et al. Abnormal subendocardial perfusion in cardiac syndrome X detected by cardiovascular magnetic resonance imaging. N Engl J Med 2002;346:1948-53. 12. Cannon RO. Does coronary endothelial dysfunction cause myocardial ischemia in the absence of obstructive coronary artery disease? Circulation 1997;96:3251-4. 13. Masoli O, Balino NP, Sabate D, Jalon J, Meretta A, Cragnolino D, et al. Effect of endothelial dysfunction on regional perfusion in myocardial territories supplied by normal and diseased vessels in patients with coronary artery disease. J Nucl Cardiol 2000;7:199204. 14. Fujita H, Yamabe H, Yokoyama M. Effect of L-arginine administration on myocardial thallium-201 perfusion during exercise in patients with angina pectoris and normal coronary angiograms. J Nucl Cardiol 2000;7:97-102. 15. Masci PG, Laclaustra M, Lara JG, Kaski JC. Brachial artery flow-mediated dilation and myocardial perfusion in patients with cardiac syndrome X. Am J Cardiol 2005;95:1478-80.
See related article, p. 747 Reversible myocardial perfusion abnormalities can occur in the absence of obstructive coronary artery From the Division of Cardiology, Tufts-New England Medical Center, Boston, Mass. Dr Soman was funded by the Herbert J. Levine Foundation Fellowship in Cardiovascular Medicine at Tufts-New England Medical Center and the Kos Pharmaceutical Fellowship in Preventive Cardiology during the course of this project. Received for publication May 25, 2006; final revision accepted Aug 6, 2006. Reprint requests: Prem Soman, MD, PhD, Presbyterian Hospital, A 429, Scaife Hall, 200 Lothrop St, Pittsburgh, PA 15213; somanp@ upmc.edu. 1071-3581/$32.00 Copyright © 2006 by the American Society of Nuclear Cardiology. doi:10.1016/j.nuclcard.2006.08.018 756
disease (CAD) and have been ascribed to attenuation artifacts, false-negative coronary angiography (microvascular disease) findings, and vasospasm.1-3 Recently, coronary endothelial dysfunction has been postulated as a potential mechanism of perfusion abnormalities seen on single photon emission computed tomography (SPECT) imaging.4,5 Evaluating coronary endothelial function, however, is an invasive process and is not easily applicable to large populations. Peripheral endothelial function testing, on the other hand, may be performed noninvasively and has been shown to correlate well with coronary artery endothelial function.6,7 To determine the relationship between peripheral vascular endothelial function and reversible myocardial perfusion abnormalities, we compared brachial artery endothelial function, determined by flow-mediated dilation (FMD), in patients with chest discomfort, reversible perfusion defects on myocardial perfusion imaging (MPI),
Journal of Nuclear Cardiology Volume 13, Number 6;756-60
and angiographically nonobstructed coronary arteries with a matched cohort of subjects with normal MPI findings. METHODS Study Design and Patient Population Consecutive patients aged 18 years or older who were referred for coronary angiography to the cardiac catheterization laboratory at Tufts-New England Medical Center, Boston, Mass, between July 2003 and June 2004, on the basis of chest discomfort and a reversible MPI defect, and were found to have absent or nonobstructive epicardial CAD (defined as ⬍50% diameter stenosis) on coronary angiography were prospectively screened for study eligibility, and eligible patients were recruited if they provided informed consent for study participation. Exclusion criteria included left ventricular dysfunction (ejection fraction ⬍45% on gated SPECT), significant valvular heart disease, and unstable clinical states. The control population was drawn from an existing database of patients with chest pain and normal MPI findings, by matching with the control group by age, gender, and number of cardiac risk factors including smoking, hypertension, hyperlipidemia, diabetes mellitus, and family history of premature CAD (first-degree male and female relatives aged ⬍55 years and ⬍65 years, respectively). All patients from the database who could be matched to a control patient via the study criteria were included. The protocol was approved by the institutional review board. MPI was performed at the referring physician’s office or institution, according to standard local practices. Either exercise or pharmacologic stress testing was performed. All scans were performed by use of gated SPECT technology and technetium-labeled radiotracers (either Tc-99m sestamibi or Tc-99m tetrofosmin) and were not reread specifically for the purpose of this study. The number and location of reversible perfusion defects were determined from the report.
Endothelial Function Testing Brachial artery ultrasound (BAUS) was performed according to our previously published method.6 In patients in the study group BAUS was performed at least 2 hours after coronary angiography and in the fasting state. Longitudinal brachial artery images were obtained with a high-resolution (10-MHz) linear-array vascular transducer (Philips Sonos 5500; Philips Medical Systems, Andover, Mass). Subjects were studied under quiet conditions while in the supine position in a temperature-controlled room. After a 10-minute equilibrium period, baseline 2-dimensional images of the right brachial artery were obtained approximately 2 cm above the antecubital fossa. A blood pressure cuff (Hokanson, Bellevue, Wash) placed proximal to the imaging transducer on the upper arm was inflated to suprasystolic pressure for exactly 5 minutes. The brachial artery was imaged continuously for 1 minute after release of occlusion, and reactive hyperemia was confirmed by
Soman et al Endothelial dysfunction and perfusion defects
757
pulse-wave Doppler interrogation. Repeat baseline resting brachial artery dimensions were obtained 10 minutes later. Subjects were then given sublingual nitroglycerin (400-g tablet), and the brachial artery was imaged for 5 minutes thereafter. Endothelium-dependent vasomotion was determined by the maximal brachial artery diameter after exactly 60 seconds of reactive hyperemia compared with the baseline diameter and was expressed as percent FMD. Endothelium-independent vasodilatation was defined as the maximal brachial artery diameter after administration of nitroglycerin compared with the baseline vessel diameter and was expressed as percent nitroglycerin-mediated diameter (NMD). Brachial artery measurements were performed with ultrasonic calipers, and maximal end-diastolic brachial arterial diameter was calculated within a 5-cm segment of the vessel as the mean of 5 evenly spaced measurements of the distance from the near to the far arterial wall along a line perpendicular to the long axis of the artery. Mean intraobserver variability and interobserver variability of brachial reactivity measurements in our laboratory in normal volunteers were 1.9% and 2.8%, respectively.6
Statistical Analysis Continuous variables are described as mean ⫾ SD and categoric variables as frequencies (percent). We used 2 statistics and t tests to detect differences between continuous and categoric variables, respectively. A multivariate regression analysis was used to determine predictors of an abnormal MPI scan. Those variables predictive of an abnormal MPI scan on univariate analysis were included in the multivariate model. A Pearson correlation coefficient was determined between every pair of predictor variables to assess for multicollinearity in the data. Significant correlation was considered to be present between variables if R ⬎ 0.9. SPSS (version 13.0; SPSS Inc, Chicago, Ill) was used for all statistical computations. P ⱕ .05 was considered significant for all analyses.
RESULTS Subject characteristics are shown in Table 1. The study group consisted of 36 patients, aged 56 ⫾ 10 years, of whom 18 (50%) were women. In 20 patients the chest pain was typical of angina.8 The distribution of the number of cardiac risk factors was as follows: 0 in 1 patient, 1 in 9 patients, 2 in 11 patients, 3 in 11 patients, and 4 in 4 patients. Stress testing was performed with treadmill exercise in 30 patients and pharmacologic agents in 6. The reversible perfusion abnormality was confined to the left anterior descending artery territory in 18 patients and the right coronary or left circumflex coronary artery territory in 14 patients, and it involved multiple vascular territories in 4 patients. The control group consisted of 55 patients, aged 55 ⫾ 11 years, of whom 27 (49%) were women (Table 1). The distribution of cardiac risk factors was not different from that in the study group. The number of risk factors
758
Soman et al Endothelial dysfunction and perfusion defects
Journal of Nuclear Cardiology November/December 2006
Table 1. Subject characteristics
No. of patients Age (y) Male gender [n (%)] Hypertension [n (%)] Hyperlipidemia [n (%)] Diabetes mellitus [n (%)] Smoking [n (%)] Family history of CAD [n (%)]
Study group
Control group
36 56 ⫾ 10 18 (50) 21 (58) 24 (67) 5 (14) 13 (35) 20 (56)
55 55 ⫾ 11 27 (49) 21 (38) 27 (49) 8 (15) 27 (49) 25 (46)
P ⫽ NS for all.
Figure 1. Scatter plot of individual percent FMD values of study and control groups with mean and SD.
Table 2. Endothelial function testing results
Baseline brachial size (mm) % FMD % NMD*
Study group
Control group
P value
3.7 ⫾ 0.7
3.6 ⫾ 0.7
.96
9.0 ⫾ 7.2 18.1 ⫾ 7.1
12.0 ⫾ 5.2 21.1 ⫾ 7.9
Table 3. Multivariate analysis for prediction of reversible SPECT abnormalities
Variable .03 .17
*Available in all study patients and 47 patients in control group.
was as follows: 0 in 4 patients, 1 in 13 patients, 2 in 19 patients, 3 in 9 patients, and 4 in 2 patients. There were no baseline characteristic differences between the study group and the control group. The results of BAUS testing are shown in Table 2. The study group had a significantly lower mean percent FMD (9.0% ⫾ 7.2%) compared with the control group (12.1% ⫾ 5.4%) (P ⫽ .03). There was no difference in the mean percent FMD between the group that underwent treadmill exercise testing and the group that underwent pharmacologic stress testing (9.2% ⫾ 7.5% and 7.8% ⫾ 5.4%, respectively; P ⫽ not significant [NS]). A scatter plot of the individual FMD values in both groups is shown in Figure 1. Sixteen patients in the study group had percent FMD values below 1 SD of the mean percent FMD of the control group (6.8%). The baseline brachial artery size was similar between the study and control groups (3.7 ⫾ 0.7 mm and 3.6 ⫾ 0.7 mm, respectively; P ⫽ NS). Similarly, percent NMD was also not significantly different between groups (18.1% ⫾ 7.1% and 21.1% ⫾ 7.9%, respectively; P ⫽ NS). Results of the multivariate logistic regression analysis are shown in Table 3. When the clinically significant predictor variables of age, gender, number of cardiac risk factors, and percent FMD were introduced into a multiple regression model, percent FMD was the only variable that was independently predictive of an abnormal
% FMD Age Male gender Hypertension Hyperlipidemia Diabetes mellitus Smoking Family history of CAD
Odds Confidence P ratio interval value 0.90 1.01 1.87 2.34 1.67 1.15 0.46 1.33
0.82–0.99 0.96–1.06 0.66–5.25 0.85–6.46 0.58–4.82 0.29–4.54 0.20–1.34 0.40–3.66
.03 .75 .24 .10 .35 .84 .16 .58
SPECT result (odds ratio, 0.9; 95% confidence interval, 0.82-0.99). DISCUSSION Chest pain without angiographic obstructive CAD is a well-defined clinical entity.1-3 When associated with electrocardiographic abnormalities indicative of myocardial ischemia on treadmill exercise testing, the constellation has been termed syndrome X.9 Myocardial perfusion abnormalities have been demonstrated in patients with angina and nonobstructed coronary arteries by use of radionuclide studies, and a reduction in coronary flow reserve has been demonstrated in invasive studies.10 Magnetic resonance imaging has demonstrated hypoperfusion of the subendocardial region during adenosine stress in patients with angina and normal coronary arteries.11 However, the pathogenesis of myocardial perfusion abnormalities in the absence of obstructive epicardial coronary disease and the vascular level at
Journal of Nuclear Cardiology Volume 13, Number 6;756-60
which these pathogenetic mechanisms operate remain a matter of contention.9 Whereas adenosine is the primary mediator of coronary arteriolar dilatation mediated by metabolic demand, the increased shear stress caused by adenosine-induced hyperemia results in endothelial cell release of nitric oxide, with further dilatation of the epicardial coronaries and microvasculature.12 Thus endothelial dysfunction might limit the full ability of the coronary vessels to dilate. The exploration of an association between endothelial dysfunction and myocardial perfusion abnormalities has been somewhat limited by the invasive nature of direct coronary endothelial function testing. Nevertheless, a few small studies have attempted to investigate this concept. In 27 patients with nonobstructed epicardial coronary arteries, Zeiher et al5 showed that those with exercise-induced myocardial perfusion abnormalities had blunted endothelium-mediated increases in coronary flow reserve, as tested by subselective coronary acetylcholine infusion, compared with patients with normal exercise myocardial perfusion. The coronary flow reserve response to the endotheliumindependent vasodilator papaverine was comparable in both groups. Hasdai et al4 demonstrated a decrease in the left anterior descending coronary artery diameter measured by quantitative coronary angiography and coronary flow reserve measured by Doppler flow wire in patients in whom a transient perfusion abnormality developed on Tc-99m sestamibi SPECT imaging during selective infusion of acetylcholine. Masoli et al13 compared the vasomotor response of the epicardial coronary artery to acetylcholine, as measured by quantitative coronary angiography, and simultaneous myocardial perfusion on planar thallium 201 imaging in 23 coronary vascular territories in 11 patients with suspected CAD. Perfusion abnormalities developed in all patients with a 20% reduction or greater in the coronary artery diameter in response to acetylcholine in vascular territories with absent, intermediate (50%-69%), and significant (ⱖ70%) epicardial stenosis. No patient with a normal vasomotor response to acetylcholine had a perfusion abnormality develop. Finally, Fujita et al14 showed in 12 patients with angina and normal epicardial coronary arteries that the administration of L-arginine improved exercise time and the extent and severity of abnormalities on Tl-201 MPI. This study suggests that an improvement in nitric oxide availability, via arginine, improves vascular function and myocardial perfusion. Thus these important, albeit small, clinical studies have established that abnormal vasomotor responses in the coronary microvasculature resulting from endothelial dysfunction may cause myocardial perfusion abnormalities in the absence of significant epicardial coronary stenosis. The results of our study demonstrate that impaired peripheral vascular endothelial function, which has previously been shown to correlate with coronary endothelial
Soman et al Endothelial dysfunction and perfusion defects
759
function,6,7 is associated with reversible MPI defects in the absence of obstructive epicardial CAD. This finding adds to the existing data set linking coronary endothelial dysfunction to myocardial perfusion abnormalities by demonstrating this association noninvasively and thus sets the stage for further investigation of the relationship between endothelial dysfunction and myocardial ischemia. Furthermore, in this study brachial FMD was the only variable predictive of an abnormal MPI scan in a model that included traditional cardiovascular risk markers. The presence of comparable endothelium-independent vasomotion in patients with myocardial perfusion abnormalities and those without them suggests that the pathogenesis of these perfusion abnormalities is probably not related to a smooth muscle mechanism. These findings are concordant with those of Masci et al,15 who reported lower brachial FMD in syndrome X patients with myocardial perfusion abnormalities, as compared with those with normal myocardial perfusion. Another study has demonstrated impaired brachial FMD in patients with abnormal myocardial perfusion in more heterogeneous populations.6 A significant methodologic study limitation that must be noted is the fact that image interpretation was performed by multiple referring physicians and was not standardized at a central core laboratory. Quantitative assessment of perfusion defect extent and severity was not performed. However, all patients had chest pain, and coronary angiography was considered warranted by the MPI abnormality. Although the hypothesis of a causal relationship between endothelial dysfunction and the pathogenesis of myocardial perfusion is biologically plausible, several questions remain unanswered. For example, how does endothelial dysfunction, presumably a diffuse, systemic process, produce regional abnormalities on MPI? Are these perfusion abnormalities indicative of true myocardial ischemia, related to the chest pain syndrome in these patients? Is there a therapeutic implication to this finding? Answers to these and other questions can only be derived from the institution of large, prospective clinical trials designed to determine outcome data in patients with endothelial dysfunction and nonobstructed coronary arteries, as well as the results of repeat MPI after treatment of endothelial dysfunction. The establishment of noninvasive testing of brachial FMD as an accurate surrogate marker of coronary endothelial function certainly improves the feasibility of performing such trials.
Acknowledgment The authors have indicated they have no financial conflicts of interest.
760
Soman et al Endothelial dysfunction and perfusion defects
References 1. Kemp HG, Elliott WC, Gorlin R. The anginal syndrome with normal coronary arteriography. Trans Assoc Am Physicians 1967; 80:59-70. 2. Likoff W, Segal BL, Kasparian H. Paradox of normal selective coronary arteriograms in patients considered to have unmistakable coronary heart disease. N Engl J Med 1967;276:1063-6. 3. Egashira K, Inou T, Hirooka Y, Yamada A, Urabe Y, Takeshita A. Evidence of impaired endothelium-dependent coronary vasodilatation in patients with angina pectoris and normal coronary angiograms. N Engl J Med 1993;328:1659-64. 4. Hasdai D, Gibbons RJ, Holmes DR Jr, Higano ST, Lerman A. Coronary endothelial dysfunction in humans is associated with myocardial perfusion defects. Circulation 1997;96:3390-5. 5. Zeiher AM, Krause T, Schachinger V, Minners J, Moser E. Impaired endothelium-dependent vasodilation of coronary resistance vessels is associated with exercise-induced myocardial ischemia. Circulation 1995;91:2345-52. 6. Kuvin JT, Patel AR, Sliney KA, Pandian NG, Rand WM, Udelson JE, et al. Peripheral vascular endothelial function testing as a noninvasive indicator of coronary artery disease. J Am Coll Cardiol 2001;38:1843-9. 7. Teragawa H, Ueda K, Matsuda K, Masashi K, Higashi Y, Oshima T, et al. Relationship between endothelial function in the coronary and brachial arteries. Clin Cardiol 2005;28:460-6.
Journal of Nuclear Cardiology November/December 2006
8. Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary-artery disease. N Engl J Med 1979;300:1350-8. 9. Cannon RO, Camici PG, Epstein SE. Pathophysiological dilemma of syndrome X. Circulation 1992;85:883-92. 10. Kaski JC, Aldama G, Cosin-Sales J. Cardiac syndrome X. Diagnosis, pathogenesis and management. Am J Cardiovasc Drugs 2004;4:179-94. 11. Panting JR, Gatehouse PD, Yang GZ, Grothues F, Firmin DN, Collins P, et al. Abnormal subendocardial perfusion in cardiac syndrome X detected by cardiovascular magnetic resonance imaging. N Engl J Med 2002;346:1948-53. 12. Cannon RO. Does coronary endothelial dysfunction cause myocardial ischemia in the absence of obstructive coronary artery disease? Circulation 1997;96:3251-4. 13. Masoli O, Balino NP, Sabate D, Jalon J, Meretta A, Cragnolino D, et al. Effect of endothelial dysfunction on regional perfusion in myocardial territories supplied by normal and diseased vessels in patients with coronary artery disease. J Nucl Cardiol 2000;7:199204. 14. Fujita H, Yamabe H, Yokoyama M. Effect of L-arginine administration on myocardial thallium-201 perfusion during exercise in patients with angina pectoris and normal coronary angiograms. J Nucl Cardiol 2000;7:97-102. 15. Masci PG, Laclaustra M, Lara JG, Kaski JC. Brachial artery flow-mediated dilation and myocardial perfusion in patients with cardiac syndrome X. Am J Cardiol 2005;95:1478-80.