Author + information
- Received June 26, 1998
- Revision received December 11, 1998
- Accepted January 21, 1999
- Published online May 1, 1999.
- Steven E Reis, MD, FACC∗,* (, )
- Richard Holubkov, PhD∗,
- Joon S Lee, MD, FACC∗,
- Barry Sharaf, MD, FACC†,
- Nathaniel Reichek, MD, FACC‡,
- William J Rogers, MD, FACC§,
- Edward G Walsh, PhD§,
- Anthon R Fuisz, MD§,
- Richard Kerensky, MD, FACC∥,
- Katherine M Detre, MD, DrPH, FACC¶,
- George Sopko, MD#,
- Carl J Pepine, MD, FACC∥,
- for the WISE Investigators
- ↵*Reprint requests and correspondence: Dr. Steven E. Reis, c/o WISE Coordinating Center, 127 Parran Hall, 130 DeSoto St., Pittsburgh, Pennsylvania 15261
We sought to develop and validate a definition of coronary microvascular dysfunction in women with chest pain and no significant epicardial obstruction based on adenosine-induced changes in coronary flow velocity (i.e., coronary velocity reserve).
Chest pain is frequently not caused by fixed obstructive coronary artery disease (CAD) of large vessels in women. Coronary microvascular dysfunction is an alternative mechanism of chest pain that is more prevalent in women and is associated with attenuated coronary volumetric flow augmentation in response to hyperemic stimuli (i.e., abnormal coronary flow reserve). However, traditional assessment of coronary volumetric flow reserve is time-consuming and not uniformly available.
As part of the Women’s Ischemia Syndrome Evaluation (WISE) study, 48 women with chest pain and normal coronary arteries or minimal coronary luminal irregularities (mean stenosis = 7%) underwent assessment of coronary blood flow reserve and coronary flow velocity reserve. Blood flow responses to intracoronary adenosine were measured using intracoronary Doppler ultrasonography and quantitative angiography.
Coronary volumetric flow reserve correlated with coronary velocity reserve (Pearson correlation = 0.87, p < 0.001). In 29 (60%) women with abnormal coronary microcirculation (mean coronary flow reserve = 1.84), adenosine increased coronary velocity by 89% (p < 0.001) but did not change coronary cross-sectional area. In 19 (40%) women with normal microcirculation (mean flow reserve = 3.24), adenosine increased coronary velocity and area by 179% (p < 0.001) and 17% (p < 0.001), respectively. A coronary velocity reserve threshold of 2.24 provided the best balance between sensitivity and specificity (90% and 89%, respectively) for the diagnosis of microvascular dysfunction. In addition, failure of the epicardial coronary to dilate at least 9% was found to be a sensitive (79%) and specific (79%) surrogate marker of microvascular dysfunction.
Coronary flow velocity response to intracoronary adenosine characterizes coronary microvascular function in women with chest pain in the absence of obstructive CAD. Attenuated epicardial coronary dilation response to adenosine may be a surrogate marker of microvascular dysfunction in women with chest pain and no obstructive CAD.
The clinical characteristics of angina pectoris are gender-specific: 1) angina is the predominant initial manifestation of coronary artery disease (CAD) in women but not in men, 2) the prevalence rate of angina is greater in women than men, and 3) chest pain both typical and atypical for angina in women is frequently not caused by fixed obstructive epicardial CAD (1–3). An alternative mechanism of recurrent chest pain in the absence of significant angiographically detected CAD is coronary microvascular dysfunction, which is defined as disordered function of the smaller coronary resistance vessels (<100 to 200 μm) that play a role in the regulation of coronary blood flow. Although these microvessels are usually spared from the gross morphologic changes of atherosclerosis, they may have abnormal motility, growth, inflammation and/or permeability resulting in disordered vascular function.
Microvascular angina is more prevalent in women than in men and may be manifested by chest pain, an abnormal stress test and abnormal myocardial perfusion and metabolism in the setting of angiographically normal appearing coronary arteries (4–8). However, from the clinical perspective, microvascular angina is a diagnosis of exclusion or one that is ignored after performance of noninvasive cardiac testing, coronary angiography and other costly evaluations of noncardiac chest pain. Often these patients continue to have pain leading to repeated evaluations. Therefore, early diagnosis would allow for initiation of appropriate palliative therapy and decrease utilization of medical resources.
Microvascular angina can be diagnosed in women without obstructive epicardial CAD by demonstrating microvascular dysfunction manifested by an attenuated increase in coronary flow in response to hyperemic stimuli (i.e., abnormal coronary flow reserve). The traditional definition of microvascular dysfunction as an explanation for microvascular angina requires that maximal hyperemic stimuli (e.g., adenosine) increase coronary volumetric blood flow less than 2.5-fold, which is the lower limit of normal flow reserve in coronaries free of significant obstructive CAD (9–11). Coronary volumetric flow reserve is most commonly measured invasively by quantitative coronary angiography and intracoronary Doppler ultrasonography, which is time-consuming, may require off-line analysis and is performed in few catheterization laboratories. Recent studies, which have primarily been performed in men, have assessed the functional integrity of the coronary microcirculation by using intracoronary Doppler ultrasound alone to measure coronary flow velocity responses to adenosine. However, a definition of microvascular dysfunction based on coronary flow velocity reserve in women with chest pain in the absence of obstructive CAD has not been validated.
As part of the pilot phase of the Women’s Ischemia Syndrome Evaluation (WISE) study, a cohort of women with chest pain and angiographically normal coronary arteries or minimal coronary luminal irregularities were identified. They underwent invasive assessment of both coronary volumetric flow reserve and coronary flow velocity reserve using quantitative coronary angiography and intracoronary Doppler ultrasonography (12). The purpose of this report is to validate a definition of microvascular dysfunction based on coronary velocity reserve. The results of our study demonstrate that coronary velocity reserve characterizes coronary microvascular function in women with chest pain in the absence of obstructive epicardial CAD, with the threshold for normal coronary velocity reserve being lower than that of normal coronary volumetric flow reserve.
The WISE study is a National Heart, Lung and Blood Institute–sponsored four-center study assessing cardiovascular function with state-of-the-art techniques in women referred for coronary angiography to evaluate chest pain or suspected ischemia (12). As part of the WISE pilot phase protocol, the University of Florida (Gainesville, Florida) and University of Pittsburgh (Pittsburgh, Pennsylvania) sites used intracoronary Doppler ultrasonography and quantitative coronary angiography to evaluate the functional integrity of the coronary microcirculation in women with chest pain in the absence of significant fixed obstructive CAD.
Evaluation of coronary volumetric flow and flow velocity
Diagnostic coronary angiography was performed using standard techniques. Women with epicardial coronaries that were angiographically normal or had minimal luminal irregularities, defined by the absence of ≥50% diameter stenoses in all coronary arteries, underwent functional assessment of their coronary microcirculation. In these subjects, a 0.014-in. or 0.018-in., Doppler-tipped guidewire (Cardiometrics, Mountain View, California) was advanced through a coronary catheter engaged in the left main or right coronary artery and positioned in the left anterior descending artery (n = 34), left circumflex (n = 9), first obtuse marginal (n = 4) or posterior descending artery (n = 1). Pulsed-wave Doppler ultrasonography was used to assess the time-averaged peak coronary flow velocity calculated on-line over two cardiac cycles through a 2-mm sample volume at a location approximately 5 mm distal to the tip of the guidewire.
Quantitative coronary angiographic analysis was performed off-line and masked to all other patient data at the WISE central angiographic core laboratory (Rhode Island Hospital, Providence, Rhode Island). Luminal diameter was measured at the location corresponding to the site of flow velocity assessment (5 mm distal to the guidewire tip) using an electronic cine projector-based “crosshair” technique (Vanguard Instrument Corporation, Melville, New York). This technique is analogous to electronic calipers built into the hood of the cine projector, thus removing viewer parallax, and compares favorably with methods using automated edge detection (13). In the WISE core laboratory, the standard deviation of the mean difference in measured percent stenosis between the two methods was <8%. Although multiple views were filmed to determine that no severe stenoses were present, the cine view used for quantitative analysis was chosen to minimize vessel foreshortening and overlap.
Quantitative angiographic measurements of luminal diameter were used to calculate coronary cross-sectional area. Coronary volumetric flow was then calculated by the product of epicardial coronary cross-sectional area (cm2), a time constant (60 s/min) and mean coronary flow velocity, which is one half of the average peak velocity assuming a parabolic flow velocity profile (14).
Assessment of the functional integrity of the coronary microcirculation
The functional integrity of the coronary microcirculation (i.e., vasodilator reserve) was assessed by measuring coronary volumetric flow and velocity responses to intracoronary adenosine. Coronary velocity and diameter were measured at baseline and after a hand-injected intracoronary bolus of adenosine (18 μg in left main or 12 μg in right coronary artery; Adenocard, Fujisawa USA, Deerfield, Illinois) diluted in 2 ml of normal saline and followed by a 5-ml saline flush. Coronary volumetric flow reserve was calculated by the ratio of average peak coronary volumetric flow after adenosine administration to baseline flow. Volumetric coronary flow reserves of ≥2.5 were considered normal (9–11). Coronary velocity reserve was determined by the ratio of average peak coronary velocity after adenosine to baseline velocity.
The data are summarized as means ± standard deviations or percent changes when appropriate. P values of ≤0.05 are considered statistically significant. Associations between continuous measures were assessed using the Pearson correlation coefficient. Significance of differences between baseline and adenosine measures was assessed using the Wilcoxon signed rank test. Comparisons between subgroups were performed by the Wilcoxon rank-sum test for continuous measures and by the chi-squared test or Fisher exact test (when expected cell sizes were <5) for discrete measures.
Forty-eight women, 54.2 ± 9.7 years of age, with chest pain and coronary arteries that were angiographically normal or had minimal luminal irregularities underwent functional assessment of their coronary microcirculation. Patient characteristics are presented in Table 1. The mean stenosis in the study artery was 7% (range 0% to 48%), with 38 women (79%) having 0% stenosis. Of the 30 women who had a prior noninvasive test to evaluate ischemia, 18 (60%) had ischemic or indeterminate electrocardiographic responses to stress; 10 women underwent stress scintigraphy, with 5 (50%) exhibiting perfusion defects.
Coronary volumetric flow and diameter reserve
Adenosine significantly increased coronary volumetric flow by 139.1 ± 85.6% from 39.1 ± 23.0 to 91.0 ± 56.9 ml/min (p < 0.001). Mean coronary volumetric flow reserve was 2.39 ± 0.86; 29 women (60%) had coronary microvascular dysfunction defined as a volumetric flow reserve <2.5. In this subgroup, mean volumetric flow reserve was 1.84 ± 0.41 (Table 2). The 19 (40%) women who exhibited normal coronary microcirculation vasodilator function (i.e., volumetric flow reserve ≥2.5) had a mean volumetric flow reserve of 3.24 ± 0.63. A comparison of clinical characteristics between these groups found that several risk factors, including history of diabetes and hypertension, were present more frequently in women with coronary microvascular dysfunction (Table 3).
The adenosine-induced increase in coronary flow observed in the 19 women with normal coronary microcirculation resulted from significant 178.8 ± 55.3% (p < 0.001) and 17.2 ± 14.3% (p < 0.001) adenosine-induced increases in coronary flow velocity and epicardial cross-sectional area, respectively. In contrast, the adenosine-induced increase in coronary volumetric flow seen in the 29 women with microvascular dysfunction resulted solely from an adenosine-induced 89.2 ± 37.4% (p < 0.001) increase in coronary velocity, since baseline and postadenosine cross-sectional areas were not significantly different in this group (Table 2). Indeed, the lack of adenosine-induced epicardial dilation, defined as less than 9% dilation in response to an 18-μg bolus of intracoronary adenosine, has a 79% sensitivity and 79% specificity for the diagnosis of microvascular dysfunction. Therefore, epicardial coronary responses to adenosine may be a surrogate marker for the diagnosis of coronary microvascular dysfunction in women with chest pain in the absence of obstructive CAD.
Subgroup analysis of the 38 women without any angiographic evidence of coronary stenoses in the study artery demonstrated that, as in the entire study population, 40% and 60% exhibited normal and abnormal microvascular function, respectively. The sample size from the pilot phase of WISE is insufficient for evaluation of clinical predictors of microvascular dysfunction in this cohort. However, the relationships observed in Table 3will be explored in the final WISE population using the methods validated in the present study.
Coronary velocity reserve
Coronary microvascular functional integrity was also classified in each woman by the assessment of her maximum coronary flow velocity response to adenosine. On average, velocity increased from 22.4 ± 8.7 cm/s at baseline to 49.6 ± 20.2 cm/s (p < 0.001) during adenosine. Mean coronary velocity reserve was 2.25 ± 0.63. Figure 1demonstrates a significant correlation between coronary velocity reserve and coronary volumetric flow reserve (Pearson correlation = 0.87, p < 0.001). However, examination of the line fitted to this scatterplot indicates that its slope deviates from identity, with coronary velocity reserve being less than flow reserve in the majority of subjects. Interestingly, this observation is primarily due to the finding that coronary velocity reserve underestimated volumetric flow reserve in 16 of the 19 women (84%) with normal microvascular function. In contrast, coronary velocity reserve underestimated volumetric flow reserve in only 13 of the 29 women (45%) with microvascular dysfunction. Subgroup analysis of the 38 women without any angiographic evidence of CAD demonstrated similar results (analysis not shown).
As noted above, microvascular dysfunction, defined by coronary volumetric flow reserve <2.5, was present in 29 of the 48 studied women. Using this classification as the reference standard, we assessed the accuracy of a range of coronary velocity reserve thresholds for microvascular dysfunction (Fig. 2). We found that a coronary velocity reserve threshold of 2.24 provided optimum sensitivity (90%) and specificity (89%) for the diagnosis of microvascular dysfunction. This velocity reserve threshold correctly classified coronary microvascular function in 43 of 48 women (90%). Specifically, 26 of the 29 (90%) women with microvascular dysfunction (i.e., coronary volumetric flow reserve <2.5) had a coronary velocity reserve of <2.24, and 17 of 19 (89%) with normal microvascular function (i.e., coronary volumetric flow reserve ≥2.5) had a coronary velocity reserve of ≥2.24. Of note, this cutpoint also provided nearly optimal sensitivity and specificity when the analysis was limited to the subgroup of 38 women without angiographic evidence of coronary atherosclerosis in the study artery.
The results of this WISE pilot phase study suggest that the coronary blood flow velocity response to intracoronary adenosine characterizes the functional integrity of the coronary microcirculation in women with chest pain and angiographically normal or only minimally irregular coronary arteries. Coronary velocity reserve can be conveniently measured by an intracoronary Doppler-tipped guidewire and correlates significantly with coronary volumetric flow reserve, which is currently the standard but more labor-intensive method used to objectively diagnose microvascular dysfunction. However, we found that coronary velocity reserve is lower than coronary volumetric flow reserve in nearly all women with normal microvascular function and is consistently lower than coronary volumetric flow reserve in women with microvascular dysfunction. This suggests that the threshold for normal coronary velocity reserve should be less than that of coronary volumetric flow reserve. Our results indicate that microvascular dysfunction is present in nearly all women with chest pain and insignificant CAD when coronary flow velocity increases less than 2.24-fold in response to an intracoronary bolus of adenosine. Furthermore, our data suggest that attenuated adenosine-induced epicardial coronary dilation may be a surrogate marker for coronary microvascular dysfunction.
Coronary flow and velocity responses to adenosine
The functional integrity for flow regulation of the coronary microcirculation is typically assessed by measuring maximal changes in coronary volumetric flow in response to vasodilators such as adenosine, which activates adenylate cyclase and induces cell membrane hyperpolarization in arterial myocytes present within the coronary microcirculation. As a result, adenosine causes microvascular dilation and maximally increases both coronary blood velocity and volumetric flow. Accordingly, recent studies have used coronary velocity reserve to assess coronary microvascular function. In addition, coronary velocity reserve is now commonly used to evaluate the functional significance of “intermediate” coronary stenoses (15,16). However, the results of these studies, which suggest that coronary velocity reserves of at least 2.0 are normal, are not applicable to the evaluation of the functional integrity of the coronary microcirculation in the cohort of women with chest pain in the absence of obstructive CAD.
Coronary volumetric flow cannot be directly measured in patients undergoing coronary angiography. Instead, coronary volumetric flow can be calculated by the use of validated equations that incorporate measurements of mean coronary flow velocity and epicardial cross-sectional area (14,17). Adenosine-induced changes in coronary flow velocity are expected to approximate adenosine-induced changes in coronary volumetric blood flow only if epicardial coronary caliber remains constant. However, adenosine’s vasodilatory effect is not limited to the coronary microcirculation; when administered in an epicardial coronary segment, adenosine induces endothelium-independent vasodilation distal to the site of infusion (18,19). In addition, adenosine-induced increases in coronary blood flow result in endothelium-dependent dilation of epicardial coronaries with intact endothelial function (18,19). Therefore, coronary volumetric flow and velocity responses to adenosine may not be equivalent. Indeed, although coronary velocity responses to adenosine correlated significantly with volumetric flow responses in our study, the threshold for velocity reserve that provided the best combination of sensitivity and specificity for the diagnosis of microvascular dysfunction (i.e., 2.24) was lower than the threshold for normal coronary volumetric flow reserve (i.e., 2.5).
Epicardial coronary artery responses to adenosine
The finding that mean coronary volumetric flow reserve is generally greater than mean coronary velocity reserve is consistent with those of a previous report (20)and may be explained by our observation that adenosine increases both of the independent variables in the equation used to calculate coronary flow: epicardial coronary cross-sectional area and coronary flow velocity. Interestingly, we found distinct large vessel responses to adenosine: 1) dilation of epicardial coronaries in women with normal microvascular flow reserve function, and 2) lack of epicardial dilation in those with microvascular dysfunction. Because adenosine-induced epicardial coronary dilation results partly from endothelium-dependent flow-mediated dilation, it appears that women with chest pain and microvascular dysfunction may have endothelial dysfunction in their epicardial coronaries. This observation is supported by previous reports suggesting that microvascular dysfunction is caused by abnormal endothelial function in the coronary microvasculature (21–23). Alternatively, our findings may be explained by the attenuated adenosine-induced increase in coronary flow in those with microvascular dysfunction because a suboptimal increase in flow for any reason may provide an inadequate increase in shear stress that is sensed by the endothelial cell to result in epicardial coronary dilation (24).
The association identified between epicardial coronary dilation and coronary flow velocity augmentation in response to an intracoronary bolus of adenosine suggests that the conductance coronary response to adenosine may be used as a low risk inexpensive method to assess microvascular function in women without obstructive CAD. Indeed, the sensitivity and specificity of attenuated epicardial coronary dilation in response to adenosine (i.e., <9% diameter increase) are both 79% for the diagnosis of microvascular dysfunction. However, further validation studies are required before this observation can be applied to the large cohort of women with chest pain in the absence of obstructive CAD.
Our study is limited by the use of an intracoronary bolus of adenosine as the hyperemic stimulus to assess coronary reserve and its reliance on intracoronary Doppler ultrasonography and quantitative coronary angiography to estimate coronary blood flow. However, we chose intracoronary adenosine instead of intravenous adenosine to avoid confounding effects associated with intravenous use such as transient lowering of systemic blood pressure and changes in heart rate, both of which can decrease coronary perfusion pressure and alter coronary flow independent of the functional integrity of the coronary microcirculation. In addition, direct measurement of intracoronary blood flow is impractical in humans. Therefore, we used a validated method (14)to indirectly quantify coronary blood flow using an intracoronary Doppler-tipped guidewire because of its ability to assess coronary blood flow when used in conjunction with quantitative coronary angiography and because of its low profile which is not expected to impede coronary flow.
This study developed and validated a definition of coronary microvascular dysfunction in women with chest pain and no obstructive CAD based on coronary flow velocity reserve. It also identified a high (60%) prevalence of microvascular dysfunction in this cohort. However, we may have overestimated the role of this disease entity because 1) the relationship between microvascular dysfunction and chest pain is uncertain (25), 2) no association has been demonstrated between microvascular dysfunction and myocardial ischemia, and 3) the limited sample size of this WISE pilot study does not allow us to control for clinical variables that may influence coronary reserve including atherosclerotic risk factors, blood pressure and heart rate. Nevertheless, the methods developed by the present study will be used to assess the prevalence of microvascular dysfunction and its relationship to myocardial ischemia and clinical factors in the final WISE cohort.
In summary, chest pain in women with coronary arteries that are angiographically normal or have minimal luminal irregularities is frequently attributed to a “noncardiac” etiology. Although microvascular dysfunction in the absence of epicardial CAD may be associated with a relative freedom from death or myocardial infarction (26,27), it may result in debilitating symptoms, frequent hospitalizations, extensive evaluations and loss of productivity and wages. Therefore, early diagnosis of microvascular dysfunction may facilitate initiation of appropriate therapy and may decrease health care costs in women with chest pain. Our results indicate that the functional integrity of the coronary microcirculation can be accurately assessed during coronary angiography by measuring coronary flow velocity responses to an intracoronary bolus of adenosine. Although this diagnostic approach requires the use of an intracoronary Doppler-tipped guidewire, it is simpler and less time-consuming than the traditional approach, which necessitates both intracoronary Doppler ultrasonography and quantitative coronary angiography to calculate coronary volumetric flow reserve. In addition, we observed that changes in epicardial coronary diameter in response to adenosine may provide an even less invasive assessment of microvascular function. As a result, women with chest pain not attributed to CAD may undergo a rapid diagnostic evaluation to accurately assess the functional integrity of their coronary microcirculation and to diagnose the etiology of their symptoms. These diagnostic approaches are being used in the second phase of the WISE study to further characterize the pathophysiologic mechanisms, ischemic and clinical sequelae and associated demographics of chest pain in a large cohort of women.
☆ This work was supported by contracts from the National Heart, Lung and Blood Institute, contracts N01-HV-68161, N01-HV-68162, NO1-HV-68163 and NO1-HV-68164 and grants from the Gustavis and Louis Pfeiffer Research Foundation, The Women’s Guild, Cedars-Sinai Medical Center, the Ladies Hospital Aid Society of Western Pennsylvania, University of Pittsburgh and qmed, Inc. (Laurence Harbor, NJ)
- coronary artery disease
- Women’s Ischemia Syndrome Evaluation
- Received June 26, 1998.
- Revision received December 11, 1998.
- Accepted January 21, 1999.
- American College of Cardiology
- Murabito J.M,
- Evans J.C,
- Larson M.G,
- Levy D
- (1998) 1998 Heart and Stroke Statistical Update (American Heart Association, Dallas (TX)).
- Cannon R.O,
- Bonow R.O,
- Bacharach S.L,
- et al.
- Opherk D,
- Zebe H,
- Weihe E,
- et al.
- Camici P.G,
- Marraccini P,
- Lorenzoni R,
- et al.
- Bugiardini R,
- Pozzati A,
- Ottani F,
- et al.
- Klocke F.J
- Bergmann S.R,
- Herrero P,
- Markham J,
- et al.
- Geltman E.M,
- Henes C.G,
- Senneff M.J,
- et al.
- Merz C.N.B,
- Kelsey S.F,
- Pepine C.J,
- et al.
- Doucette J.W,
- Corl D,
- Payne H.M,
- et al.
- Miller D.D,
- Donahue T.J,
- Younis L.T,
- et al.
- Heller L.I,
- Cates C,
- Popma J,
- et al.
- Vita J.A,
- Treasure C.B,
- Ganz P,
- et al.
- Cox D.A,
- Vita J.A,
- Treasure C.B,
- et al.
- Zeiher A.M,
- Krause T,
- Schachinger V,
- et al.
- Egashira K,
- Hirooka Y,
- Kuga T,
- et al.
- Liao J.C,
- Kuo L
- Rosen S.D,
- Uren N.G,
- Kaski J.-C,
- et al.
- Kaski J.C,
- Rosano G.M,
- Collins P,
- et al.