Adenosine stress myocardial perfusion tomographic imaging in patients with significant aortic stenosis

Adenosine stress myocardial perfusion tomographic imaging in patients with significant aortic stenosis Sotirios P. Patsilinakos, MD,a Stavros Spanodimos, MD,a Fivi Rontoyanni, MD,b Athanasios Kranidis, MD,c Ioannis P. Antonelis, MD,c Konstantinos Sotirellos, MD,a Dionysios Antonatos, MD,a Elias Tsaglis, MD,a Nikolaos Nikolaou, MD,a and Dimitris Tsigas, MDa Background. Myocardial perfusion scintigraphy has been used by some investigators for the diagnosis of coronary artery disease (CAD) in patients with mild, moderate, and moderate to severe aortic stenosis, with various results. The aim of this study was to assess the safety and diagnostic accuracy of adenosine stress myocardial perfusion scintigraphy (adenosine single photon emission computed tomography [Ad-SPECT]) for the detection of CAD in patients with significant aortic stenosis. Methods and Result. The study included 75 patients with significant aortic stenosis (maximal instantaneous aortic valve gradient >80 mm Hg [range, 81-149 mm Hg] and aortic valve area <0.75 cm2). All patients underwent Ad-SPECT after a 6-minute infusion of adenosine (140 ␮g/kg body weight per minute). At the third minute of adenosine infusion, a bolus of 3 mCi thallium 201 was injected, and SPECT acquisition was obtained immediately after completion of adenosine infusion. Coronary angiography was performed in all patients. No major complications during adenosine infusion were observed. All unpleasant symptoms lasted for only a few seconds and did not necessitate cessation of the test. Concerning the angiographically diagnosed CAD, we found that Ad-SPECT showed a sensitivity of 88.6%, a specificity of 72.5%, a positive predictive value of 73.8%, a negative predictive value of 87.8%, and a diagnostic accuracy of 80%. Conclusions. Ad-SPECT is a moderately accurate method for detecting the presence or absence of CAD in patients with severe aortic stenosis. However, further modification of this method is required before it can supplant cardiac catheterization in the preoperative evaluation of patients with severe aortic stenosis. (J Nucl Cardiol 2004;11:20-5.) Key Words: Adenosine • aortic stenosis • single photon emission computed tomography

Coexisting coronary artery disease (CAD) is observed in approximately 60% of patients with aortic stenosis.1-3 About two thirds of patients with significant aortic stenosis may have angina.1 Exercise electrocardiography (ECG) is generally discouraged in these patients, as hypotension, syncope, or even sudden cardiac death may be provoked by stress testing.4,5 Moreover, exercise ECG in patients with severe aortic stenosis is not used for the diagnosis of CAD; rather, it is mostly used for the stratification of asymptomatic patients with From the Cardiology Department, Konstantopoulio-Ag Olga General Hospital, N Ionia,a and Nuclear Medicine Department,b and Hemodynamic Department,c Evangelismos Hospital, Athens, Greece. Received for publication April 3, 2003; final revision accepted Oct 6, 2003. Reprint requests: Sotirios P. Patsilinakos, Ilia Mantouvalou 4-6E str, 15235 Vrilisia, Athens, Greece; [email protected]. Copyright © 2004 by the American Society of Nuclear Cardiology. 1071-3581/2004/$30.00 ⫹ 0 doi:10.1016/j.nuclcard.2003.10.003 20

severe aortic stenosis. This test may decide which of those patients will be guided to surgical therapy.5,6 Furthermore, the traditional exercise stress test and thallium 201 exercise scintigraphy have shown low specificity in detecting CAD in patients with aortic stenosis.6,7 Adenosine pharmacologic stress myocardial perfusion tomographic imaging with Tl-201 by use of single photon emission computed tomography (AdSPECT) in patients with moderate to severe aortic stenosis was found to be well tolerated and diagnostically accurate.8 The aim of this study was to evaluate the safety and diagnostic accuracy of Ad-SPECT imaging in detecting CAD in patients with significant aortic stenosis. METHODS The study included 75 patients with significant aortic stenosis (51 men and 24 women), with a mean age of 68.8 ⫾ 10.7 years. All patients were symptomatic. Their symptoms were dyspnea (57 patients), angina (49 patients), and syncope

Journal of Nuclear Cardiology Volume 11, Number 1;20-5

(29 patients). All patients underwent 2-dimensional echocardiography and Doppler analysis. The maximal instantaneous aortic transvalvular gradient was greater than 80 mm Hg (range, 81-149 mm Hg), and the aortic valve area was less than 0.75 cm2 (range, 0.74-0.33 cm2). For the 48 hours preceding Ad-SPECT, all xanthine derivatives and caffeine-containing products were discontinued. None of the patients had severe congestive heart failure, second- or third-degree atrioventricular block, or a history of asthma or significant chronic obstructive pulmonary disease. Adenosine was infused intravenously at a rate of 140 ␮g/kg body weight per minute for 6 minutes. At the third minute of adenosine infusion, a bolus of 3 mCi Tl-201 was injected intravenously and rapidly flushed with 10 mL of normal saline solution. SPECT acquisition was obtained immediately after completion of adenosine infusion. At baseline and every minute thereafter for at least 8 minutes, vital signs and a 12-lead electrocardiogram were monitored and recorded. ECG monitoring was continuous throughout the test for detecting the development of arrhythmias or ST-segment deviation. Redistribution images were obtained 4 hours later, while the patient was resting. After this, an additional 1 mCi Tl-201 was injected if the patient had shown evidence of a persisting perfusion defect on the redistribution recordings, and new images were obtained (reinjection imaging). Images were acquired by use of a circular orbit over a 180° range, starting at the 45° left posterior oblique projection and ending at the 45° right anterior oblique projection. Each of 32 projections was acquired with a 64 ⫻ 64 matrix at 40 seconds per image. A standard filtered backprojection technique was applied by use of a filter with a cutoff frequency of 0.83 cycles/cm to generate transaxial slices. From these transaxial images, the long axis of the left ventricle was identified and oblique angle images were generated in the short-axis, vertical long-axis, and horizontal long-axis orientations, divided into 16 slices each. In these slices the left ventricular wall was divided into 9 segments (ie, the middle and basal part of the anterior, lateral, inferior, and septal wall plus the apex). Thirty-two planar images were reviewed in a cine-loop mode to evaluate extracardiac activities and motion artifacts. An image was considered abnormal if, by visual inspection, it exhibited a decrease in Tl-201 uptake in any myocardial segment. The perfusion defects were graded by use of a 4-point scale: 0, normal; 1, equivocal; 2, moderately diminished uptake; and 3, severely diminished uptake. The presence or absence of redistribution was visually assessed in the 4-hour images. Myocardial segments with diminished uptake and complete redistribution or fill-in after reinjection were considered ischemic. The images were interpreted by 2 experienced, blinded, independent observers without knowledge of the results of angiography. The summed stress score (SSS), summed rest score (SRS), and reversibility index (RI) were estimated. Patients with an RI of 1 or greater were considered to have an abnormal scan for ischemia. Cardiac catheterization and angiography were performed in all patients with a standard technique. Narrowing of the luminal diameter of the vessel greater than 50% was considered as significant.

Patsilinakos et al SPECT in severe aortic stenosis

21

Table 1. Hemodynamic and angiographic data of 75 patients with severe aortic stenosis

Angiographic and hemodynamic results CAD (lesion ⬎50%) One-vessel disease Two-vessel disease Three-vessel disease Stenosis of left anterior descending artery Stenosis of right coronary artery Stenosis of circumflex artery Maximal left ventricular aortic gradient (mm Hg)* Ejection fraction*

Finding 35 patients (46.7%) 22 patients (29.3%) 11 patients (14.7%) 2 patients (2.7%) 24 patients (32%) 17 patients (22.7%) 9 patients (12%) 87.6 ⫾ 9.9 52.3% ⫾ 9.6%

*Left ventricular catheterization was achieved in 66 of 75 patients.

Informed Consent All patients included in the study had signed an informed consent form.

Statistics Sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of Ad-SPECT in detecting CAD were calculated in the conventional way. Numerical data are given as mean ⫾ SD, and comparisons between means were performed with the Student t test. P values ⬍ .05 were considered statistically significant.

RESULTS Echocardiography The mean value of maximal instantaneous aortic valve gradient was 90.5 ⫾ 11.5 mm Hg, and the mean aortic valve area was 0.56 ⫾ 0.13 cm2. Left ventricular hypertrophy was present in all patients. Hemodynamic Study and Angiography Of the 75 patients with significant aortic stenosis, 35 (47%) were found to have CAD. Of these, 22 had 1-vessel disease, whereas 11 had 2-vessel disease and 2 had 3-vessel disease (Table 1). Effects of Adenosine Infusion A reduction in systolic pressure (from 142.87 ⫾ 18.7 mm Hg to 139.4 ⫾ 15.5 mm Hg, P ⫽ .22) and an increase in heart rate (from 78.5 ⫾ 11.2 beats/min to 81.5

22

Patsilinakos et al SPECT in severe aortic stenosis

Journal of Nuclear Cardiology January/February 2004

Table 2. Symptoms experienced during Ad-SPECT

Symptoms

Patients with aortic stenosis (n ⴝ 75)

Flushing Chest pain Dyspnea Dizziness Headache Nausea

31 (41.3%) 26 (34.7%) 26 (34.7%) 23 (30.7%) 17 (22.7%) 6 (8%)

SSS SRS RI

Table 3. Ad-SPECT data in patients with or without angiographically significant CAD

Normal Ad-SPECT Abnormal Ad-SPECT Total No. of patients

Table 4. Comparison of SSS, SRS, and RI between patients with CAD and patients without CAD

With CAD

Without CAD

Total No. of patients

4 31 35

29 11 40

33 42 75

⫾ 9.4 beats/min, P ⫽ .08) were observed after the infusion of adenosine. The mean rate-pressure product change was not significant (from 11,294 ⫾ 2,011 to 10,778 ⫾ 1,957, P ⫽ .11). The symptoms patients had during the infusion of adenosine are shown in Table 2. We did not observe any major complications. All of the symptoms lasted only a few seconds, and thus we did not have to stop the test or inject aminophylline intravenously. Second-degree heart block developed in 7 patients and third-degree heart block in 2. All of the episodes of heart block were temporary and terminated spontaneously. In 6 patients a temporary ST-segment depression greater than 1 mm during the infusion of adenosine was observed. Ad-SPECT Table 3 compares the Ad-SPECT data in patients with and without CAD. We analyzed 675 left ventricular segments (for all 75 patients). Identical results were given by the 2 observers for 658 segments (97.5%). The disagreement that occurred for the remaining 17 segments was resolved by discussion and consensus between the 2 observers. Completely or partially reversible perfusion defects after adenosine infusion occurred in 57 segments (18 of which were completely reversible), fixed defects in 16 segments, and normal perfusion images in the remaining 602 segments. The mean SSS, SRS, and RI for the total population

CAD (35 patients)

No CAD (40 patients)

P

3.89 ⫾ 2.23 2.91 ⫾ 1.84 0.97 ⫾ 1.46

0.60 ⫾ 1.03 0.43 ⫾ 0.87 0.18 ⫾ 0.59

.000 .000 .002

were 2.13 ⫾ 2.36, 0.68 ⫾ 1.21, and 1.45 ⫾ 1.91, respectively. SSS, SRS, and RI were significantly higher in patients with CAD compared with those without CAD (Table 4). Clinical, echocardiographic, angiographic, and perfusion data of patients with false-positive and falsenegative Tl-201 scans are shown in Tables 5 and 6, respectively. Compared with the results of coronary angiography, Ad-SPECT demonstrated the following: sensitivity, 88.6% (31/35); specificity, 72.5% (29/40); positive predictive value, 73.8%; negative predictive value, 87.8%; and diagnostic accuracy, 80% (Figure 1). DISCUSSION Patients with severe aortic stenosis are usually symptomatic. We very often have to investigate whether there is concomitant CAD, especially if surgery is recommended. This distinction is of particular significance if it can be made by a noninvasive and safe diagnostic modality. Such a possibility is of even greater value, because about two thirds of patients with significant aortic stenosis may have angina.1 The traditional ECG exercise test cannot be used to assess the presence or absence of CAD, as it has low diagnostic accuracy. Certainly, exercise testing should not be performed in symptomatic patients. However, in asymptomatic patients, exercise testing is of some value because it can identify those patients who are at a higher risk of death or symptoms.4,5,7-9 Pharmacologic stress tests seem to be safer and may offer higher diagnostic accuracy than ECG exercise tests. They can contribute substantially to the detection of CAD in patients with severe aortic stenosis. Planar Tl-201 myocardial imaging after combined intravenous dipyridamole infusion and isometric handgrip testing was found to be well tolerated and diagnostically adequate in patients with aortic stenosis.10 Adenosine is a more potent but safer vasodilator than dipyridamole and has a high diagnostic accuracy in detecting CAD.11-14 Samuels et al8 evaluated adenosine pharmacologic stress myocardial perfusion tomographic imaging in patients with moderate to severe aortic stenosis. All investigators found that adenosine infusion is safe, as they did not

Journal of Nuclear Cardiology Volume 11, Number 1;20-5

Patsilinakos et al SPECT in severe aortic stenosis

23

Table 5. Clinical, echocardiographic, and perfusion data of 11 patients with false-positive TI-201 scans (ie, patients with no significant CAD)

Echocardiography Patient No.

Sex

Symptoms

Max grad

1 2 3 4 5 6 7

F M M M M M F

110 mm Hg 86 mm Hg 83 mm Hg 92 mm Hg 133 mm Hg 149 mm Hg 92 mm Hg

15 16 14 15 21 19 17

8 9 10 11

M F M M

Dyspnea, angina Dyspnea Angina Syncope, angina Dyspnea Angina Dyspnea, syncope Dyspnea Dyspnea, angina Angina Dyspnea

85 mm Hg 98 mm Hg 103 mm Hg 95 mm Hg

19 18 16 16

EF

SSS

SRS

RI

Segments considered as ischemic

mm mm mm mm mm mm mm

0.35 0.50 0.70 0.70 0.55 0.65 0.40

2 3 2 2 2 1 3

2 0 2 2 2 0 3

0 3 0 0 0 1 0

Apex Basal and middle inferior Basal inferior Basal inferior Basal septal Basal inferior Middle anterior, apex

mm mm mm mm

0.65 0.65 0.55 0.40

2 3 2 2

2 2 0 2

0 1 2 0

Basal Basal Basal Basal

IVS

septal inferior inferior inferior

Max grad, Maximum gradient through aortic valve; IVS, intraventricular septum; EF, ejection fraction.

Table 6. Clinical, echocardiographic, angiographic, and perfusion data of 4 patients with false-positive TI-201 scans

Echocardiography Patient No.

Sex

Symptoms

1 2 3 4

F M M M

Dyspnea, angina Dyspnea Angina Dyspnea

Max grad 87 mm Hg 86 mm Hg 103 mm Hg 116 mm Hg

IVS 14 18 16 18

mm mm mm mm

EF 0.65 0.70 0.50 0.65

SSS 0 0 0 0

SRS 0 0 0 0

RI

Diseased artery

0 0 0 0

LAD (50%) LAD (80%) LAD (60%) CX (60%)

Max grad, Maximum gradient through aortic valve; IVS, intraventricular septum; EF, ejection fraction; LAD, left anterior descending; CX, circumflex.

observe significant differences between patients with aortic stenosis and control subjects with regard to hemodynamic manifestations and the frequency of transient atrioventricular block. Adenosine stress myocardial perfusion imaging was not found to be adequately accurate in detecting CAD in these patients. The diagnostic accuracy of 80% to 85% is certainly not low; however, currently, when CAD can be diagnosed safely enough with coronary angiography, missing about 15% to 20% of hemodynamically important coronary artery lesions with this modality makes this accuracy less satisfactory.15 As an extension to these data, we investigated whether pharmacologic stress testing with adenosine was safe in a larger group of patients (75 patients) with severe aortic stenosis and left ventricular hypertrophy, and we assessed the diagnostic accuracy of myocardial perfusion tomographic imaging in these patients. In our study group with severe aortic stenosis (mean

aortic valve area, 0.56 ⫾ 0.13 cm2), pharmacologic stress testing with adenosine was indeed safe. In this study group we observed no significant decrease in systolic blood pressure or increase in heart rate and, consequently, no significant difference in the double product. We did not encounter any episode of major hypotension. The symptoms and disturbances of atrioventricular conduction that we observed in our subjects during adenosine infusion were transient. The proportion of patients in whom these untoward effects developed was not significantly different from that observed during the infusion of adenosine in patients with moderate to severe aortic stenosis8 or from that in patients with no aortic stenosis.12 Chest pain after adenosine infusion was not related to the presence of concomitant CAD in our study group. This could probably mean that this symptom was not caused by ischemia. The sensitivity of 88.6% in detecting CAD in

24

Patsilinakos et al SPECT in severe aortic stenosis

Journal of Nuclear Cardiology January/February 2004

Figure 1. Sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of Ad-SPECT in detecting CAD.

patients with severe aortic stenosis and left ventricular hypertrophy that we found with Ad-SPECT was lower but comparable to that found (92%) by Samuels et al8 in patients with moderate aortic stenosis. It is well known that the principal underlying mechanism of myocardial perfusion defects during adenosine-induced coronary vasodilatation in patients with CAD is the disparity in regional myocardial perfusion between vascular territories supplied by normal coronary vessels and those supplied by stenotic arteries. Adenosine infusion produces coronary arteriolar vasodilatation, which produces a 3- to 5-fold increase in blood flow in normal coronary arteries with little or no increase in stenotic arteries. Concomitant left ventricular hypertrophy that was present in all of our patients with severe aortic stenosis does not seem to decrease the sensitivity of Ad-SPECT in detecting CAD. We should mention that all 4 patients with a false-negative scan had 1-vessel disease with a mean stenosis of 62.5% (50%, 60%, 60%, and 80%). This mean stenosis, as compared with the mean stenosis of 84.2% for the remaining 18 patients with 1-vessel disease with true-positive scans, was not significantly different (P ⫽ .84). Nevertheless, because of the small number of patients, we cannot safely conclude whether the degree of stenosis plays any significant role in the interpretation of the scan tests. The specificity found in this study was 72.5% and was also similar to that found (71%) in the group of

patients with moderate to severe aortic stenosis in the study of Samuels et al8. However, Verani et al12 found a much higher specificity (94%) with adenosine and Tl201 scintigraphy in detecting CAD in patients without aortic stenosis. This may be explained by the fact that patients with significant aortic stenosis may have ischemia and thus scintigraphic perfusion defects develop, despite the absence of obstructive CAD.8 Some possible mechanisms (eg, impaired coronary vasodilator reserve due to myocardial hypertrophy, significant systolic reduction or inversion of coronary flow caused by compression of intramyocardial vessels, and so on16) may produce perfusion patterns of heterogeneous flow despite patent coronary arteries. The mean intraventricular septum thickness of the 11 patients with false-positive scans did not differ significantly from the intraventricular septum thickness of the 29 patients with true-negative scans (16.9 ⫾ 2.1 mm vs 16.3 ⫾ 2.2 mm, P ⫽ .44). This may imply that ventricular hypertrophy is not the only cause for the false-positive scans and probably other conditions, such as fibrosis or scar formation in the myocardium and dilation of the left ventricle, may contribute to producing false-positive images in patients with no CAD.17,18 Recently, Van Tosh19 has reviewed and compared all of these similar studies (which, however, included patients with different degrees of aortic stenosis) and calculated a mean sensitivity of 86% and a mean specificity of 72%.

Journal of Nuclear Cardiology Volume 11, Number 1;20-5

In conclusion, the administration of adenosine was well tolerated by all patients with severe aortic stenosis. Ad-SPECT, a noninvasive and safe method, was found to have a high sensitivity for detecting CAD in patients with severe aortic stenosis and left ventricular hypertrophy. Ad-SPECT is a moderately accurate method for detecting the presence or absence of CAD in patients with severe aortic stenosis. However, the relative lack of predictive accuracy of a negative test currently renders this method inadequate for clinical decision making. The method remains an interesting research tool and requires further modification before it can supplant cardiac catheterization in the preoperative evaluation of patients with severe aortic stenosis. Acknowledgment The authors have indicated they have no financial conflicts of interest.

References 1. Lombart JT, Selzer A. Valvular aortic stenosis. A clinical and hemodynamic profile of patients. Ann Intern Med 1987;106:292-8. 2. Selzer A. Changing aspects of the natural history of valvular aortic stenosis. N Engl J Med 1987;317:91-8. 3. Hakki AH, Kimbiris D, Iskandrian AS, et al. Angina pectoris and coronary artery disease in patients with severe aortic valvular disease. Am Heart J 1980;100:441-9. 4. Atwood JE, Kawaniski S, Myers J, Froehicher VF. Exercise testing in patients with aortic stenosis. Chest 1988;93:1083-7. 5. Clyne CA, Arrighi JA, Maron BJ, Bonow RO, Cannon RO. Systemic and left ventricular responses to exercise stress of asymptomatic patients with valvular aortic stenosis. Am J Cardiol 1991;68:1469-76. 6. Aronow WS, Harris CN. Treadmill exercise test in aortic stenosis and mitral stenosis. Chest 1975;68:507-9. 7. Baily IK, Come PC, Kelly DT, et al. Thallium-201 myocardial perfusion imaging in aortic valve stenosis. Am J Cardiol 1977;40: 889-98.

Patsilinakos et al SPECT in severe aortic stenosis

25

8. Samuels B, Kiat H, Friedman JD, Berman DS. Adenosine pharmacologic stress myocardial perfusion tomographic imaging in patients with significant aortic stenosis. J Am Coll Cardiol 1995; 25:99-106. 9. Kupari M, Virtanen KS, Turto H, et al. Exclusion of coronary artery disease by exercise thallium-201 tomography in patients with aortic valve stenosis. Am J Cardiol 1992;70:635-40. 10. Huikuri HV, Korhonen UR, Ivaheimo MJ, Heikkila J, Takkunen JJ. Detection of coronary artery disease by thallium imaging using a combined intravenous dipyridamole and isometric handgrip test in patients with aortic valve stenosis. Am J Cardiol 1987;59:33640. 11. 〈nthopoulos LP, Bonou SM, Kardaras GF, et al. Stress echocardiography in elderly patients with coronary artery disease. Applicability, safety and prognostic value of dobutamine and adenosine echocardiography in elderly patients. J Am Coll Cardiol 1996;28: 52-9. 12. Verani MS, Mahmarian JJ, Hixcon JB, Bayce TM, Staudacher RA. Diagnosis of coronary artery disease by controlled coronary vasodilation with adenosine and thallium-201 scintigraphy in patients unable to exercise. Circulation 1990;82:80-7. 13. Rask LP, Karp K, Edlund B, et al. Computer-assisted evaluation of dipyridamole thallium-201 SPECT in patients with aortic stenosis. J Nucl Med 1994;35:983-8. 14. Rask LP, Karp KH, Eriksson P, Mooe T. Dipyridamole thallium201 single-photon emission tomography in aortic stenosis. Eur J Nucl Med 1995;22:1155-62. 15. Patsilinakos SP, Kranidis AI, Antonelis IP, et al. Detection of coronary artery disease in patients with severe aortic stenosis with noninvasive methods. Angiology 1999;50:309-17. 16. Lazzeroni E, Picano E, Dodi C, et al, on behalf of the EPIC Study Group. Dipyridamole echocardiography for diagnosis of coexistent coronary artery disease in hypertrophic cardiomyopathy. Am J Cardiol 1995;75:810-3. 17. Salcedo EE, Marwick TH, Korzick DH, Goormastic M, Go RT. Left ventricular hypertrophy sensitizes the myocardium to the development of ischaemia. Eur Heart J 1990;11(Suppl):72-8. 18. Omori Y, Kajiya T, Shiotani H, et al. Tl-201 single-photon emission computed tomography for clinical assessment of dilated cardiomyopathy. J Cardiogr 1984;14:707-18. 19. Van Tosh A. The value of myocardial perfusion imaging for diagnosing coronary artery disease with aortic valve stenosis. In: Borer JS, Isom OW, editors. Pathophysiology, evaluation and management of valvular heart diseases. Advances in cardiology, volume 39. Basel: Karger; 2002. p. 61-9.