Noninvasive diagnosis of coronary artery disease

Transcription

1 Transthoracic Doppler Echocardiography as a Noninvasive Tool to Assess Coronary Artery Stenoses A Comparison with Quantitative Coronary Angiography Markku Saraste, MD, Risto K. Vesalainen, MD, PhD, Antti Ylitalo, MD, PhD, Antti Saraste, MD, PhD, Juha W. Koskenvuo, MD, PhD, Jyri O. Toikka, MD, PhD, Mari-Anne Vaittinen, MD, Jaakko J. Hartiala, MD, PhD, and K. E. Juhani Airaksinen, MD, PhD, Turku and Pori, Finland We prospectively tested the diagnostic accuracy of Doppler transthoracic echocardiography in detection of coronary artery stenoses throughout the main coronary arterial tree. In all, 84 patients referred for diagnostic quantitative coronary angiography were studied. Coronary artery stenosis was identified with color Doppler as local spot of turbulence, and local flow velocity was measured using pulsed wave Doppler. Angiography showed significant stenoses (diameter reduction > 50%) in 33 patients. An abnormal maximal-to-prestenotic blood flow velocity ratio greater than 2.0 in subtotal stenoses, or the detection of collateral blood flow in the absence of normal antegrade flow in the case of total occlusion (N 6), resulted in overall sensitivity of 82% and specificity of 92%. The sensitivity and specificity were, respectively, 73% and 92% for left anterior descending coronary artery, 63% and 96% for right coronary artery, and 38% and 99% for left circumflex coronary artery stenoses. Transthoracic echocardiography is a promising noninvasive technique to diagnose significant coronary artery stenoses. (J Am Soc Echocardiogr 2005;18: ) Noninvasive diagnosis of coronary artery disease (CAD) remains an important challenge for clinical cardiology. The development of imaging technology has made it possible to study coronary artery flow noninvasively using Doppler echocardiography. Coronary artery stenoses can be identified by Doppler echocardiography because of turbulent and accelerated flow at the site of stenosis. 1 Stenoses can be assessed invasively by intravascular ultrasound and Doppler wire. 2 Transesophageal echocardiography has also been reported to diagnose stenoses semi-invasively with reasonable accuracy. 3 Previous studies have shown that Doppler transthoracic echocardiography (TTE) can be used to detect restenosis of the left anterior descending coronary artery (LAD) after coronary angioplasty and stenting with good sensitivity and specificity, 4,5 and to quantify the internal thoracic artery graft function after coronary bypass operation. 6 Moreover, total occlusions of the LAD have been detected by retrograde From the Departments of Clinical Physiology and Medicine (R.K.V., A.S., M-A.V., K.E.J.A.), Turku University Hospital, and Department of Cardiology, Satakunta Central Hospital (A.Y.). Reprint requests: Risto K. Vesalainen, MD, PhD, Department of Medicine, Turku University Hospital, Kiinamyllynkatu 4-8, 20520, Turku, Finland ( risto.vesalainen@tyks.fi) /$30.00 Copyright 2005 by the American Society of Echocardiography. doi: /j.echo flow in its middle segment. 7 We decided to prospectively assess the accuracy of TTE to detect significant coronary artery stenoses throughout the main coronary arterial tree using quantitative coronary angiography as a reference standard in patients with suggested significant CAD. METHODS Study Population We studied 84 consecutive patients (mean age 60 9 years; 50 men) who were referred for diagnostic coronary angiography because of a suggestion of significant CAD. Patients with unstable angina within a week before the study or with a previous myocardial infarction were excluded. The patients continued their usual medication during the study. Exercise testing was performed for 68 patients (81%) as a part of a clinical workup 6 3 months before coronary angiography, but exercise testing was not part of the study protocol. A positive exercise test for CAD according to the definition of American College of Cardiology/American Heart Association guidelines was found in 36 patients (52%). 8 Percutaneous coronary angioplasty was previously performed in two patients but no stents were used. The clinical characteristics and the medications of the study patients are summarized in Table 1. All patients gave written informed consent and the study 679

2 680 Saraste et al June 2005 Table 1 Clinical characteristics of the study patients (n 84) Age (y) 60 9 Body mass index (kg/m 2 ) Obese ( 30 kg/m 2 ) 15 (18%) Family history of coronary artery disease* 47 (56%) Hyperlipidemia 57 (68%) Hypertension 41 (49%) Diabetes 8 (10%) Current smoking 15 (18%) Peripheral or cerebral arterial disease 5 (6%) Chronic atrial fibrillation 3 (4%) Clinical heart failure 9 (11%) Medication -blocker 57 (68%) Nitrate 41 (49%) Calcium channel blocker 13 (15%) Digoxin 1 (1%) Angiotensin-converting enzyme 13 (15%) inhibitor Angiotensin-receptor blocker 8 (10%) Statin 36 (42%) Acetosalicylic acid 73 (87%) Anticoagulant therapy 3 (4%) *First-degree relatives. Serum total cholesterol 5 mmol/l or low-density lipoprotein cholesterol 3.5 mmol/l. Blood pressure 140/90 mm Hg. protocol was approved by the joint committee on ethics of Turku University and Turku University Central Hospital, Turku, Finland. Doppler TTE All patients had a TTE examination a day before the coronary angiography. The same physician (M. S.) carried out all the TTE examinations using ultrasound apparatus (Sequoia C 256, Acuson Inc, Mountain View, Calif) and a standard 3.5-MHz transducer. The mean duration of the TTE study was minutes and the mean heart rate during the study was 61 9/min. We studied anatomic course of the coronary arteries using color Doppler mapping with data postprocessing mix function, which makes the colors transparent. We used all possible standard and nonstandard windows and views to find coronary arteries. The velocity scale of color Doppler was primarily set to 0.24 m/s, but it was changed actively. A 2-dimensional mode image was used to facilitate identification of coronary arteries. During the TTE examination, we searched the left main coronary artery from the left parasternal short- and longaxis views focusing on area adjacent to the sinus Valsalva cranial to the aortic valve. The proximal LAD continuing from the left main coronary artery and turning slightly toward the transducer was best visualized in the same short-axis view using minor changes of imaging plane. Origin of the first septal branch of the LAD that marks border between the proximal and middle segments of the LAD could be visualized in most patients. We searched the middle and distal LAD from left parasternal windows at varying levels using modified short- and long-axis views focusing on the anterior interventricular sulcus. The segments of the LAD basal and apical to the papillary muscle level in the short-axis view were considered as its middle and distal segments, respectively. The proximal left circumflex coronary artery (LCX) was searched using the left parasternal short- and long-axis views focusing on the atrioventricular sulcus. The part of the LCX that was covered by the auricle of left atrium was considered to represent its proximal segment. We searched the distal LCX using the apical long-axis view focusing to the lateral mitral ring and the 4-chamber view focusing on the inferior mitral ring. The left posterolateral branch of the LCX was visualized using the apical 4-chamber view focusing on epicardial surface of the lateral wall of left ventricle. The ostium and first 2 cm of the right coronary artery (RCA) were seen from the left parasternal short-axis view in the area of right sinus Valsalva cranial to the aortic valve. The rest of the proximal RCA was searched from the right parasternal short- and long-axis views when patients were lying on their right side focusing on the anterior tricuspid ring. The part of the RCA passing anterior surface of the tricuspid ring until the inferior margin of the right ventricle was considered as the proximal segment of the RCA. The middle RCA was visualized from the subcostal short-axis view focusing to the medial tricuspid ring on the inferior surface of the heart. The distal RCA was visualized using the subcostal 4-chamber view focusing to the posterior tricuspid ring. Finally, the posterior descending artery in the posterior interventricular sulcus coursing toward the apex of the heart was visualized from the apical 2-chamber view. Coronary artery stenosis was identified as localized color aliasing indicating local acceleration and as turbulence of flow. Normal coronary artery flow is slow and laminar and causes a weak Doppler signal. In contrast, turbulent and accelerated flow at the site of stenosis causes strong signal. To assess severity of the stenoses we quantified flow acceleration as ratio of maximal flow velocity at the site of aliasing to nearest upstream nonaccelerated prestenotic flow velocity (Figure 1). Blood flow velocity was measured at the beginning of diastole using pulsed wave Doppler with 2-MHz frequency in an average sample volume of 5 mm. Multiple consecutive cardiac cycles were analyzed to find average flow velocity. During measurements, the angle between flow and Doppler beam was kept as small as possible and angle correction was always used. In some coronary segments, such as the proximal RCA or distal LCX, it was sometimes impossible to optimize the angle because of horizontal course of the coronary artery. In these segments, stenotic flow velocity was approximated using rescaling of color Doppler. Prestenotic flow velocity was measured with pulsed wave Doppler from nearest possible upstream flow. A predefined maximal-to-prestenotic ratio more than 2 was used as a cut-off value for significant stenosis. 9 To detect total coronary occlusion by TTE, we analyzed flow in the septal branches of the LAD from left parasternal short-axis views using color Doppler. Normally, we

3 Volume 18 Number 6 Saraste et al 681 Coronary Angiography Coronary angiography was performed by the femoral artery by using the Judkins technique. An experienced interventional cardiologist being unaware of the results of TTE studies quantified the coronary stenoses in the coronary angiogram by using quantitative coronary angiography software (Quantcor, Siemens, Erlangen, Germany). A reduction greater than 50% in diameter of the stenosis was considered hemodynamically significant. Figure 1 Modified subcostal short-axis color Doppler view of distal right coronary artery stenosis (A). Open arrow, Normal slow prestenotic flow; open arrowheads, course of arterial lumen; white arrow, aliased and strong color Doppler signal caused by accelerated flow at site of stenosis. B, Same stenosis as seen in coronary angiography (white arrow). C, Measurement of peak flow velocity by pulsed wave Doppler immediately proximal to stenosis and at site of stenosis (D). Maximal stenotic flow velocity is 1.62 m/s and prestenotic flow velocity is 0.18 m/s, resulting in maximal-to-prestenotic flow velocity ratio of 9.0 and indicating presence of significant stenosis. Asterisk, Liver; RA, right atrium. Figure 2 Left parasternal short-axis color Doppler view of septal branches of left anterior descending artery in patient with total occlusion of right coronary artery (A) and in patient with healthy coronary arteries (B). Note increased coronary blood flow velocity in presence of coronary occlusion. Long segment of flow in single imaging plane can be seen in presence of occlusion, whereas normal slow blood flow can be seen only in short segment of septal arteries. observed only small spots of slow ( 0.35 m/s) flow signals in these vessels. However, during occlusion of either the left coronary artery or RCA there is enhanced collateral flow through these arteries that causes acceleration of flow and makes long continuous segments of flow signals visible by TTE. Thus, in those patients with nondetectable flow either in the LAD or RCA, signs of reversed enhanced flow from right to left in the septal branches of the LAD were considered as a marker of total occlusion in the LAD, whereas enhanced Doppler signals in normal direction were considered as a sign of total occlusion in the RCA (Figure 2). Statistical Analysis The normality of distribution of the variables was tested with Shapiro-Wilks W test. The results for the normally distributed variables are expressed as mean SD and for skewed variables as median (range). All statistical tests were performed using software (Statistica, Version 4.0, Statsoft Inc, Tulsa, Okla). RESULTS Quantitative Coronary Angiography Coronary angiography showed significant stenoses (diameter reduction 50%) or total occlusions in 33 (39%) of the patients. Single vessel disease was found in 19 (23%) and 2-vessel disease in 14 (17%) patients. There were no patients with 3-vessel disease. Of the 41 significant subtotal stenoses detected, 19 were located the in the LAD (12 proximal, 7 middle), 13 in the RCA (6 proximal, 4 middle, 3 distal), and 8 in the LCX (3 proximal, 5 distal). Only one patient had a significant stenosis in the left main coronary artery. A total of 6 total occlusions were detected, located in the LAD (one proximal, two middle) and RCA (two proximal, one middle). Proximal CAD was found in 23 patients, defined as at least one significant stenosis or total occlusion in the left main coronary artery, or in the proximal segment of the LAD, LCX, or RCA. A single stenosis greater than 50% in a small side branch of less than 2-mm diameter was found in 3 patients (two in the first diagonal branch of the LAD and one in the left posterolateral branch of the LCX). These stenoses were considered hemodynamically insignificant and excluded from further analysis. Doppler TTE The diagnostic accuracy of TTE compared with quantitative coronary angiography is summarized in Tables 2and 3. For subtotal stenoses, accelerated coronary flow was detected in 43 coronary segments (median maximal-to-prestenotic ratio 3.0 [ ]). Maximal-to-prestenotic ratio exceeded the cutoff greater than 2 (median 3.1 [ ]) in 34 segments. None of the patients were excluded from the study even if the visibility was suboptimal. There were 9 false-positive and 16 false-negative results

4 682 Saraste et al June 2005 Table 2 Diagnostic accuracy of combined criteria (maximal-to-prestenotic flow velocity ratio 2 or total occlusion) in Doppler transthoracic echocardiography to detect significant ( 50%) coronary artery stenosis in quantitative coronary angiography Coronary segment Stenoses in coronary angiography Sensitivity (%) Specificity (%) Positive predictive value (%) Negative predictive value (%) LAD LCX RCA Any significant stenosis* Significant disease Proximal disease LAD, Left anterior descending coronary artery; LCX, left circumflex coronary artery; RCA, right coronary artery. *Stenosis or total occlusion in the left main coronary artery or in LAD (considered as a single vessel), LCX, or RCA; calculated for number of coronary arteries analyzed. At least one significant stenosis or total occlusion in any coronary artery; calculated for number of patients analyzed. At least one significant stenosis or total occlusion in the left main coronary artery or in the proximal segment of the LAD, LCX, or RCA; calculated for number of patients analyzed. Table 3 Diagnostic accuracy of transthoracic Doppler echocardiography to detect significant subtotal ( 50%) stenoses or total occlusions by coronary segments when compared with quantitative coronary angiography Echocardiography Coronary angiography Stenoses Occlusions Coronary segment Stenoses (n) Occlusions (n) True negative (n) True negative (n) Any segment Left main Any LAD Proximal LAD Mid-LAD Distal LAD Any LCX Proximal LCX Distal LCX Any RCA Proximal RCA Mid-RCA Distal RCA LAD, Left anterior descending coronary artery; LCX, left circumflex coronary artery; RCA, right coronary artery. according to the predefined cutoff. Of the falsepositive results (mean maximal-to-prestenotic ratio ), 6 (67%) were associated with subclinical stenoses in coronary angiography (median 20% [20%-49%]), whereas only 3 (mean maximal-to-prestenotic ratio ) were associated with normal coronary segment in coronary angiography. In contrast, in 13 of the false-negative results (81%) no accelerated flow at all was detected with Doppler in the stenosed segment (mean stenosis 57 7%), suggesting poor visibility of the stenosed coronary segment in TTE. Only 3 of the false-negative results were associated with maximal-to-prestenotic ratio below the cutoff. Accelerated septal collateral flow identified a total occlusion in 5 patients. The sensitivity, specificity, and positive and negative predicting power for identifying a patient with a total occlusion alone were 80%, 100%, 100%, and 99%, respectively. For patients with total occlusion of the RCA, flow velocities in the septal branches of the LAD were 0.4 m/s and 0.5 m/s, whereas normal velocity was always lower than 0.35 m/s. Combined criteria of maximal-to-prestenotic ratio greater than 2.0 or total occlusion correctly identified 27 of 33 patients with significant CAD defined as at least one stenosis greater than 50% or total occlusion in any coronary artery. There were 4 false-positive and 6 false-negative cases resulting in sensitivity of 82% and specificity of 92%. Combined criteria of maximal-to-prestenotic ratio greater than 2.0 or total occlusion also correctly identified 17 of 23 patients with proximal CAD. In all, 6 falsepositive and 6 false-negative results were found,

5 Volume 18 Number 6 Saraste et al 683 resulting in sensitivity of 74% and specificity of 90%. The specificity of the method to diagnose a significant stenosis in all of the 3 major coronary arteries was systematically high (92%-99%) (Table 2). The sensitivity of the method was highest in the LAD (73%), moderate in the RCA (63%), and lowest in the LCX (38%). DISCUSSION In this study we have shown that Doppler TTE is a feasible and promising method to diagnose significant coronary artery stenoses. Notably, in this typical population admitted to diagnostic coronary angiography, TTE gained good sensitivity and specificity, and high negative predictive value to identify a patient with significant CAD (82%, 92%, and 89%, respectively) or with a proximal coronary artery stenosis (74%, 90%, and 90%, respectively). Coronary peak flow velocity measured by TTE correlates closely with diameter of the LAD stenosis measured by quantitative coronary angiography. 4 In addition, flow velocity measured with TTE has been shown to be closely associated with invasively measured flow using a Doppler wire and to correlate closely with coronary flow reserve measured with 15 H 2 0 positron emission tomography, a current noninvasive standard for measuring myocardial blood flow. 10,11 Furthermore, the coronary blood flow measurements by TTE have shown low intraobserver and interobserver variability and acceptable reproducibility with coefficient of variation of 8.1%. 4,5 The absolute coronary blood flow velocity is complexly defined by perfusion pressure, heart rate, myocardial mass, vasoregulatory mechanisms, and rheologic properties of the blood In the presence of coronary artery stenosis the flow is further affected by the severity, length, and shape of the stenosis and collateral flow Furthermore, the blood flow velocity measured by pulsed wave Doppler is directly proportional to the cosine of the angle between the blood flow and the ultrasonic beam. The optimal imaging angle of 0 degrees is almost impossible to achieve systematically, given the anatomic course and tortuous anatomy of the epicardial coronary arteries. We always used angle correction to minimize the effect of angle error. To normalize the effects of various factors affecting the blood flow velocity independent of coronary artery stenosis, we used maximal-to-prestenotic flow velocity ratio to assess the severity of the stenosis. This has been shown to correlate more closely to the severity of the stenosis in quantitative coronary angiography than detecting localized aliasing or measuring peak blood flow velocity alone. 4,5 Previous studies using either TTE or transesophageal echocardiography have shown that at least doubling of the maximal blood flow velocity, mean diastolic velocity, or time-velocity integral at the site of stenosis can reliably be used as a marker of a significant coronary artery stenosis with good sensitivity and specificity. 5,9,16 Therefore, we decided to use maximal-toprestenotic ratio greater than 2, an equivalent of doubling of the blood flow velocity, as a predefined cutoff in this study. In our previous small and nonblinded study, maximal-to-prestenotic ratio greater than 3 was suggested as an optimal cut-off value for detecting LAD stenoses. 4 However, in a retrospective analysis of the current data this higher cutoff was found to be less accurate to diagnose significant stenoses in all major coronary arteries (data not shown). In a previous study total coronary occlusions have been successfully detected in the middle segment of the LAD by detecting retrograde blood flow, 7 but the detection of total occlusions in other coronary arteries has not been successful. 9 In the current study, a nondetectable flow either in the LAD or RCA together with abnormal septal collateral flow (reversed flow from right to left in LAD occlusions and increased flow from left to right in RCA occlusions) was used as criteria for a total occlusion. We were able to correctly diagnose 5 of 6 total occlusions (3/3 in the LAD and 2/3 in the RCA), suggesting high accuracy for detecting total occlusions with this technique. However, none of our patients had a total occlusion in the LCX, distal LAD, or distal RCA. Therefore, larger prospective trials are needed to reliably evaluate the value of Doppler TTE in the systematic detection of total occlusions. Krzanowski et al 9 were the first to report the accuracy of coronary Doppler TTE in detecting coronary artery stenoses not only in the LAD but also in the LCX and RCA. They reported similar sensitivity and specificity values compared with our study for those coronary artery segments that could be visualized. However, because of a large proportion of nondiagnostic examinations they were only able to correctly diagnose 48% of the LAD stenoses, 17% of the RCA stenoses, and 37% of the LCX stenoses confirmed in quantitative coronary angiography. We were able to identify 16/22 (72%) of the angiographically confirmed LAD stenoses, 10/16 (62%) of RCA stenoses, and 3/8 (38%) of the LCX stenoses. This improvement can be partly explained by our ability to detect total occlusions and, on the other hand, by our better accuracy to detect subtotal stenoses in the LAD and in the middle segment of the RCA. Given the low sensitivity for diagnosing LCX stenoses in the current (38%) and previous studies, 9 imaging of the LCX remains an obvious challenge for coronary Doppler TTE. The use of echocardiographic contrast media has been shown to enhance the pulsed wave Doppler TTE scan

6 684 Saraste et al June 2005 quality during coronary flow reserve measurement in the LAD. 17 Therefore, systematic use of ultrasonographic contrast media could improve the visualization of coronary arteries and detection of stenoses or total occlusions with Doppler TTE. We recognize several limitations in our study. First, we did not perform a detailed analysis on the visibility of the different coronary artery segments by TTE. The reason for this is that normal coronary flow causes weak Doppler signal and often only a small length of a coronary artery can be viewed simultaneously in any imaging plane. This makes it difficult to assess whether or not the evaluation of flow by TTE has been comprehensive throughout the course of a coronary artery. In contrast to normal flow, turbulent and accelerated flow at the site of stenosis causes strong Doppler signal that can be easily recognized by TTE. Stenoses were visible even in those coronary segments in which the angle between flow and ultrasound beam was extremely suboptimal. This is likely to be explained by the turbulence of flow at the site of stenoses. Thus, we designed our study to specifically assess the value of TTE in detection of stenosed coronary flow using normal flow signals and 2-dimensional mode as guides of the anatomic course of coronary arteries and normal flow velocity as a reference in quantifying the stenosis severity. Second, the visibility of different coronary artery segments by TTE is affected by conventional patient-related determinants of the echocardiographic image quality, such as the anatomic structure of the bony thorax, presence of obesity or pulmonary disease, and the position and rotational axis of the heart. In addition, fast heart rate decreases the time to observe coronary flow by Doppler, because coronary flow predominantly occurs during diastole. Third, there is no prognostic information available on coronary Doppler TTE in CAD. The good accuracy of the technique to detect significant proximal coronary stenoses suggests that TTE might have prognostic implications, because proximal stenoses expose large myocardial areas to ischemia. On the other hand, quantitative coronary angiography does not always correctly reflect the hemodynamic significance of the stenosis. Studies on intravascular ultrasound have shown that quantifying the stenoses with measuring coronary blood flow by using Doppler techniques may help to recognize lesions with hemodynamic significance. 2 Finally, although this was a prospective study on patients admitted for coronary angiography because of suggested significant CAD, we cannot rule out the effect of a selection bias on our results, because a majority of the patients had undergone exercise testing as a part of clinical workup before entering the study. The sensitivity and specificity, respectively, of TTE to detect CAD reported in our study is comparable with those previously reported in meta-analyses for exercise testing (sensitivity 67% and specificity 72%), stress echocardiography (86% and 81%), and exercise myocardial perfusion single photon emission tomography imaging (87% and 73%). 8,18,19 Further studies are needed to directly compare the advantages and disadvantages of TTE over these clinically widely used noninvasive diagnostic tests for CAD. REFERENCES 1. Kenny A, Shapiro LM. Transthoracic high-frequency twodimensional echocardiography, Doppler and color flow mapping to determine anatomy and blood flow patterns in the distal left anterior descending coronary artery. Am J Cardiol 1992;69: Newby DE, Fox KA. Invasive assessment of the coronary circulation: intravascular ultrasound and Doppler. Br J Clin Pharmacol 2002;53: Vrublevsky AV, Boschenko AA, Karpov RS. Diagnosis of main coronary artery stenoses and occlusions: multiplane transoesophageal Doppler echocardiographic assessment. Eur J Echocardiogr 2001;2: Saraste M, Koskenvuo JW, Mikkola J, Pelttari L, Toikka JO, Hartiala JJ. Technical achievement: transthoracic Doppler echocardiography can be used to detect LAD restenosis after coronary angioplasty. Clin Physiol 2000;6: Hozumi T, Yoshida K, Akasaka T, Asami Y, Kanzaki Y, Ueda Y, et al. Value of acceleration flow and the pre-stenotic to stenotic coronary flow velocity ratio by transthoracic color Doppler echocardiography in noninvasive diagnosis of restenosis after percutaneous transluminal coronary angioplasty. J Am Coll Cardiol 2000;35: Takemura H, Kawasuji M, Sakakibara N, Tedoriya T, Ushijima T, Watanabe Y. Internal thoracic artery graft function during exercise assessed by transthoracic Doppler echocardiography. Ann Thorac Surg 1996;61: Watanabe N, Akasaka T, Yamaura Y, Akiyama M, Koyama Y, Kamiyama N, et al. Noninvasive detection of total occlusion of the left anterior descending coronary artery with transthoracic Doppler echocardiography. J Am Coll Cardiol 2001;38: American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Committee to update the 1997 exercise testing guidelines: ACC/AHA 2002 guideline update for exercise testing; summary article a report of the American College of Cardiology/American Heart Association task force on practice guidelines (committee to update the 1997 exercise testing guidelines). J Am Coll Cardiol 2002;40: Krzanowski M, Bodzon W, Brzostek T, Nizankowski R, Szczeklik A. Value of transthoracic echocardiography for the detection of high-grade coronary artery stenosis: prospective evaluation in 50 consecutive patients scheduled for coronary angiography. J Am Soc Echocardiogr 2000;13: Hozumi T, Akasaka T, Yoshida K, Yoshikawa J. Noninvasive estimation of coronary flow reserve by transthoracic Doppler echocardiography with high-frequency transducer. J Cardiol 2001;37:43-50.

7 Volume 18 Number 6 Saraste et al Saraste M, Koskenvuo JW, Knuuti J, Toikka JO, Laine H, Niemi P, et al. Coronary flow reserve: measurement with transthoracic Doppler echocardiography is reproducible and comparable with positron emission tomography. Clin Physiol 2001;1: Anderson HV, Stokes MJ, Leon M, Abu-Halawa SA, Stuart Y, Kirkeeide RL. Coronary artery flow velocity is related to lumen area and regional left ventricular mass. Circulation 2000;102: Kern MJ. Coronary physiology revisited: practical insights from the cardiac catheterization laboratory. Circulation 2000; 101: Nichols WW, O Rourke MF. The coronary circulation. In: Nichols WW, O Rourke MF, editors. McDonald s blood flow in arteries. London: Arnold; p Gould KL, Libscomb K. Effects of coronary stenoses on coronary flow reserve and resistance. Am J Cardiol 1974;34: Isaaz K, Da Costa A, De Pasquale JP, Cerisier A, Lamaud M. Use of the continuity equation for transesophageal Doppler assessment of severity of proximal left coronary artery stenosis: a quantitative coronary angiography validation study. J Am Coll Cardiol 1998;32: Okoyama H, Sumimoto T, Hiasa G, Morioka N, Yamamoto K, Kawada H. Usefulness of an echo-contrast agent for assessment of coronary flow velocity and coronary flow velocity reserve in the left anterior descending coronary artery with transthoracic Doppler scan echocardiography. Am Heart J 2002;143: ACC/AHA/ASE 2003 guideline update for the clinical application of echocardiography. Summary article: a report of the American College of Cardiology/American Heart Association task force on practice guidelines (ACC/AHA/ASE committee to update the 1997 guidelines for the clinical application of echocardiography). J Am Coll Cardiol 2003;42: ACC/AHA/ASNC guidelines for the clinical use of cardiac radionuclide imaging. Executive summary: a report of the American College of Cardiology/American Heart Association task force on practice guidelines (ACC/AHA/ASNC committee to revise the 1995 guidelines for the clinical use of cardiac radionuclide imaging). Circulation 2003;108: